Molecular interactions controlling immune

homeostasis in the marginal zone

Matthias Vanderkerken

Thesis submitted to fulfill the requirements for the degree of

“Doctor in Health Sciences”

Promotors: Prof. dr. Bart Lambrecht and Prof. dr. Tessa Kerre

Faculty of Medicine and Health Sciences, Ghent University, 2019

© Matthias Vanderkerken, 2019
No part of this thesis may be reproduced or used in any form without prior written permission
of the author. Please treat these results as confidential and do not perform any actions that may
obstruct patent approval or publication.
Molecular interactions controlling immune homeostasis in the
marginal zone

Academic year 2018-2019

Matthias Vanderkerken

Promotors: Prof. dr. Bart Lambrecht

Department of Internal Medicine and Pediatrics, Ghent University
VIB-UGent Center for Inflammation Research, Ghent, Belgium
Department of pulmonary medicine, Erasmus Medical Center,
Rotterdam, The Netherlands

Prof. dr. Tessa Kerre

Department of Internal Medicine and Pediatrics, Ghent University
Cancer Research Institute Ghent, Ghent, Belgium

Matthias Vanderkerken was supported by an Aspirant

grant from the Flanders Research Foundation (FWO)
Examination Committee:

Chair: Prof. dr. Bruno Verhasselt

Department of Medical Microbiology, Ghent University Hospital, Ghent, Belgium

Secretary: Prof. dr. Rudi Beyaert

Unit of Molecular Signal Transduction in Inflammation, VIB-UGent Center for
Inflammation Research, Ghent, Belgium

Prof. dr. Dirk Elewaut

VIB-UGent Center for Inflammation Research, Ghent, Belgium
Laboratory for Molecular Immunology and Inflammation, Department of Rheumatology,
University Hospital Ghent, Ghent, Belgium

Prof. dr. Andrew MacDonald

Manchester Collaborative Centre for Inflammation Research, University of Manchester,
Manchester, UK

Prof. dr. Muriel Moser

Department of Molecular Biology, ULB Cancer Research Center, Université Libre de
Bruxelles, Gosselies, Belgium

Prof. dr. Charlotte Scott

Laboratory of Myeloid Cell Ontogeny and Functional Specialisation
VIB-UGent Center for Inflammation Research, Ghent, Belgium
Department of Biomedical Molecular Biology, Ghent University, Ghent, Belgium

Prof. dr. Pieter Van Vlierberghe

Center for Medical Genetics, Ghent University Hospital, Ghent, Belgium
Cancer Research Institute Ghent, Ghent, Belgium
Voor mijn ouders, Nele en Lukas
 Table of contents

List of abbreviations 3

Summary 7

Samenvatting 9

I. Introduction 11

1. The marginal zone and its constituents 12

2. Notch signaling and function of adam proteinases 27
3. Thousand-and-one kinase 3 44
4. Innate lymphoid cells 46
5. Lymphotoxin 48

II. Aims and outline 67

III. Results part one 69

Transitional B cells commit to marginal zone B cell fate by Taok3-mediated
surface expression of ADAM10

IV. Results part two 95

TAO-kinase 3 governs the terminal differentiation of conventional dendritic
cells through Notch2 signaling

V. Results part three 129

Innate lymphoid cells control the homeostasis of ESAM+ type 2 splenic
dendritic cells through the lymphotoxin β receptor

VI. Results part four 159

IRF8-dependent dendritic cells are required to control myeloproliferation

VII. General discussion 181

1
2
 List of abbreviations

ADAM A disintegrin and metalloproteinase

AHR Aryl hydrocarbon receptor
APP Amyloid precursor protein
ARHGEF Rho guanine nucleotide exchange factor
ATM Ataxia telangiectasia mutated
BAFF(R) B cell activating factor (receptor)
BATF Basic leucine zipper transcriptional factor ATF-like 3
BCL10 B cell lymphoma/leukemia 10
BCR B cell receptor
BM Bone marrow
BMDC Bone marrow derived dendritic cell
BTK Bruton's tyrosine kinase
CARMA1 Caspase recruitment domain-containing membrane-associated
guanylate kinase protein-1
CB2 Cannabinoid receptor type 2
CCL C-C motif chemokine ligand
CCR C-C motif chemokine receptor
CD Cluster of differentiation
cDC Conventional dendritic cell
CDC42 Cell division control protein 42 homolog
CDP Common dendritic cell progenitor
cMoP Common monocyte/macrophage precursor
CSL CBF-1, Suppressor of Hairless, Lag-2
CSR Class switch recombination
CTL Cytotoxic T lymphocyte
CXCL C-X-C motif chemokine ligand
CXCR C-X-C mtof chemokine receptor
DC Dendritic cell
DCR3 Decoy receptor 3
DLL Delta-like canonical Notch ligand
DNA Desoxyribonucleic acid
DOCK Dedicator of cytokinesis
DPK Dendritic cell derived protein kinase
DPP4 Dipeptidyl peptidase 4
EBI2 Epstein-Barr virus-induced G-protein coupled receptor 2
EGF Epidermal growth factor
ER Endoplasmic reticulum
ERK Extracellular signal regulated kinase
ESAM Endothelial cell selective adhesion molecule
FDC Follicular dendritic cell
FLT3 fms like tyrosine kinase 3
FoB cell Follicular B cell

3
FRC Fibroblastic reticular cell
GALT Gut-associated lymphoid tissue
GATA3 GATA binding protein 3
GC Germinal center
GCK Germinal center kinase
GPCR G-protein coupled receptor
GPR183 G-protein coupled receptor 183
GTP Guanosine triphosphate
Hes Hairy and enhancer of split
Hey Hairy/enhancer of split-related with YRPW motif
HIF1 Hypoxia induced factor 1
HIV Human immunodeficiency virus
HVEM Herpesvirus entry mediator
ICAM Intercellular adhesion molecule
ID Inhibitor of DNA binding
IFNγ Interferon gamma
Ig Immunoglobulin
IκB Inhibitor of kappa B
IKKα/β/γ Inhibitor of kappa B kinase α/β/γ
IL Interleukin
ILC Innate lymphoid cell
ILF Isolated lymphoid follice
iNOS Inducible nitric oxide synthase
IRF Interferon regulatory factor
JIK Jnk-inhibitory kinase
JNK c-Jun N-terminal kinase
KLF4 Krüppel-like factor 4
LIGHT homologous to lymphotoxin, exhibits inducible expression and
competes with HSV glycoprotein D for binding to herpesvirus entry
mediator, a receptor expressed on T lymphocytes
LPP3 Lipid phosphate phosphatase
LT Lymphotoxin
LTβR Lymphotoxin β receptor
LXRα Liver X receptor α
M-CSF Macrophage colony-stimulating factor
MadCAM Mucosal vascular addressin cell adhesion molecule
MAIT Mucosal associated invariant T cell
MALT1 Mucosa-associated lymphoid tissue lymphoma translocation protein 1
MAML Mastermind-like protein
MAPK Mitogen activated protein kinase
MARCO Macrophage receptor with collagenous structure
MC Monocyte-derived cell
MEF Mouse embryonic fibroblasts
MEKK MAP/ERK kinase kinase
MERTK MER tyrosine kinase
MHC Major histocompatibility complex

4
MINT Msx2-interacting nuclear target protein
MMM Marginal metallophilic macrophages
MPD Myeloproliferative disorder
MRC Marginal reticular cell
MZB cell Marginal zone B cell
MZM Marginal zone macrophage
NFIL3 Nuclear factor, interleukin-3 regulated
NFκB Nuclear factor kappa-light-chain-enhancer of activated B cells
NICD Notch intracellular domain
NIK NFκB inducing kinase
NK cell Natural killer cell
NKT cell Natural killer T cell
PAK p21-activated kinase
PAMP Pathogen-associated molecular pattern
PALS Periarteriolar lymphoid sheath
pDC Plasmacytoid dendritic cell
PDPN Podoplanin
RAC Ras-related C3 botulinum toxin substrate 1
RAG Recombination activating gene
RBC Red blood cell
RBPJκ Recombination signal binding protein for immunoglobulin kappa J
region
REL Reticuloendotheliosis oncogene
RhoA Ras homolog family member A
RhoGEF Rho guanine exchange factor
RORγt Retinoic acid receptor-related orphan receptor gamma
RUNX Runt Related Transcription Factor
S1P Sphingosine-1-phosphate
SAPK Stress-activated protein kinase
SHM Somatic hypermutation
SHP src-homology region 2 (SH2) domain-containing phosphatase
SIGLEC Sialic acid-binding immunoglobulin-type lectins
SIGN(R) Dendritic cell-specific intercellular adhesion molecule-3-grabbing
non-integrin (-related)
SIRPα Signal regulatory protein alpha
SLO Secondary lymphoid organ
Ste20 Sterile20
T1/T2 Transitional B cell
TAOK Thousand-and-one kinase
TBX21 T-box transcription factor 21
TCF4 Transcription factor 4
TCR T cell receptor
TEC Thymic endothelial cell
Th T helper cell
TipDC TNF/iNOS producing dendritic cell
TKI Tyrosine kinase inhibitor

5
TLO Tertiary lymphoid organ
TLR Toll-like receptor
TNF Tumor necrosis factor
TNFR Tumor necrosis factor receptor
TNFSF Tumor necrosis factor superfamily
TSLP Thymic stromal lymphopoietin
TSPAN Tetraspanin
UPR Unfolded protein response
VCAM Vascular cell adhesion molecule
WASP Wiskott-Aldrich syndrome protein
XCR1 X-C motif chemokine receptor 1
ZEB Zinc finger E-Box binding homeobox

6
 Summary

The splenic marginal zone constitutes a unique micro-anatomical region that segregates
the white pulp from the red pulp. Here, circulating antigens are recognized by
specialized immune cells such as marginal zone B (MZB) cells, conventional dendritic
cells (cDCs) and distinct macrophage subsets. These cells can exert innate effector
functions or initiate adaptive immune responses in the white pulp, and are crucial for
the protection against blood-borne pathogens. While B cells and dendritic cells in the
marginal zone are highly adapted to their microenvironment and differ considerably
from other B cells or cDCs, the molecular interactions that control their fate
commitment and functional specialization remain poorly characterized. As a
serendipitous finding, we discovered that mice deficient for Thousand-and-one kinase 3
(TAOK3) lack MZB cells. We investigated the underlying mechanism and found that
Taok3 controlled the ability of transitional B cell precursors to respond to ligation of the
Notch2 receptor. Activation of the Notch signaling pathway, a key event in the MZB cell
fate decision, is regulated by the proteolytic cleavage of Notch2 by the
metalloproteinase ADAM10. We demonstrate that in absence of Taok3, signals through
the B cell receptor (BCR) fail to trigger surface expression of ADAM10 in transitional B
cells, thus uncovering how BCR and Notch2 signals cooperate to instruct MZB cell
development.

At the same time, Notch2-instruction also determines the terminal differentiation of a

specific subset of type 2 dendritic cells (cDC2s) in the spleen that expresses ESAM and
locates at marginal zone bridging channels. Remarkably, Taok3 also controlled the
development of ESAM+ cDC2s in a cell-intrinsic manner. Using both in vivo and in vitro
approaches, we confirmed that Notch signaling was defective in Taok3-deficient DCs. In
parallel with our observations in MZB cells, conditional ablation of ADAM10 equally
targeted ESAM+ cDC2s, revealing ADAM10 as the relevant Notch sheddase in dendritic
cells. Taken together, TAOK3 appears to control the development and differentiation of
B cells and dendritic cells in the splenic marginal zone by regulating the receptiveness
to Notch2 ligation. Apart from Notch signaling, the differentiation of ESAM+ cDC2s in the
spleen also relies on homeostatic signals through the lymphotoxin β receptor (LTβR).
While it was long thought that interactions between lymphotoxin α1β2 (LT)-expressing
B cells and LTβR-expressing DCs were indispensable, we discovered that ESAM+ cDC2s
still develop in B cell-deficient mice. Rather, we demonstrate a hitherto unappreciated
role of type 3 innate lymphoid cells in the maintenance of splenic dendritic cells through
the LT-LTβR axis.

During our study of cDC development, we observed the onset of neutrophilia and
splenomegaly in various mouse models that lack (a subset of) dendritic cells. In line
with these findings, it has been suggested that absence of cDCs might trigger the

7
development of a myeloproliferative disorder (MPD). To address this hypothesis more
directly, we used mice that are genetically deficient for the transcription factor
interferon regulatory factor 8 (IRF8). We found that the development of MPD in Irf8-/-
mice requires two hits, whereby loss of Irf8 in DCs disrupts an extrinsic
myelosuppressive effect that restricts neutrophil production by Irf8-deficient myeloid
progenitors. This regulatory action seems to depend on type 2 cDCs or CD64+
monocyte-derived dendritic cells. A better understanding of the precise cellular
interactions by which the immune system surveils myelopoiesis could open up new
therapeutic avenues for human myeloproliferative diseases.

In conclusion, we identified novel molecular interactions that govern the development

of immune cells in the splenic marginal zone, and reveal an unconventional role of
dendritic cells in the regulation of hematopoiesis. These insights might contribute to the
design of novel therapeutic strategies to modulate immune responses.

8
 Samenvatting

De marginale zone in de milt is een unieke micro-anatomische structuur die de witte

pulpa van de rode pulpa scheidt. Antigenen in de bloedcirculatie worden hier herkend
en opgenomen door gespecialiseerde immuuncellen zoals marginale zone B (MZB)-
cellen, conventionele dendritische cellen (cDC) en macrofagen. De immuunreacties die
door deze cellen worden geïnitieerd, zijn onmisbaar in de bescherming tegen
pathogenen in de bloedbaan. Dendritische cellen en B-cellen in de marginale zone zijn
uitermate aangepast aan hun omgeving en verschillen daardoor erg van andere cDCs en
B-cellen. Desondanks is weinig bekend over de moleculaire interacties die hun
ontwikkeling en functionele specialisatie sturen. We ontdekten dat muizen met een
deletie van het serine/threonine kinase TAOK3 geen MZB-cellen ontwikkelen en
ontrafelden het moleculaire mechanisme hierachter. Activatie van de Notch
signaalcascade is cruciaal voor de differentiatie van MZB-cellen en wordt gecontroleerd
door proteolyse van de Notch2 receptor door het metalloproteinase ADAM10. In
tegenstelling tot wild type cellen konden Taok3-/- transitionele B-cellen ADAM10 niet
tot expressie brengen op het celmembraan na activatie van de B-cel receptor,
resulterend in verminderde Notch activiteit.

Notch2 signalen zijn eveneens vereist voor de differentiatie van een subpopulatie van
type 2 cDCs (cDC2s) in de marginale zone die de molecule ESAM tot expressie brengt.
Naar analogie met onze bevindingen in B-cellen observeerden we een reductie in het
aantal ESAM+ cDC2s in de milt van Taok3-/- muizen. Door middel van in vivo en in vitro
experimenten hebben we aangetoond dat de Notch signaalactiviteit in Taok3-deficiënte
cDCs verminderd was. Conditionele ablatie van ADAM10 in dendritische cellen had een
gelijkaardige impact op de frequentie van ESAM+ cDC2s in de milt, wat de essentiële rol
van ADAM10 als regulator van Notch2 activiteit in cDCs onderstreept. Kortom, TAOK3
en ADAM10 controleren de ontwikkeling en differentiatie van B-cellen en dendritische
cellen in de marginale zone door de ontvankelijkheid voor Notch2 signalen te
beïnvloeden.

De differentiatie van ESAM+ cDC2s in de milt staat naast Notch-instructie ook onder
invloed van trofische signalen via de lymfotoxine β receptor (LTβR). Hoewel het lang
werd aangenomen dat interacties tussen B-cellen (een copieuze bron van lymfotoxine)
en dendritische cellen onmisbaar waren, ontdekten we dat de ontwikkeling van ESAM+
cDC2s intact was in B cel-deficiënte muizen. We identificeerden echter een onverwachte
rol voor type 3 ‘innate’ lymfocyten, gemedieerd via LTβR signalen, in de homeostase van
dendritische cellen.

Tijdens onze studie van dendritische cellen viel op dat verschillende muizen met een
verstoord cDC compartiment neutrofilie en splenomegalie ontwikkelden. We opperden

9
dat dendritische cellen een onconventionele rol hadden in de regulatie van
hematopoiese, en testten deze hypothese in Irf8-/- muizen. Hieruit bleek dat
dendritische cellen het ontstaan van een myeloproliferatief syndroom kunnen
onderdrukken, maar dat het ontbreken van Irf8-afhankelijke dendritische cellen alleen
niet volstaat voor het ontwikkelen van een myeloproliferatief syndroom. Verder
onderzoek naar het precieze mechanisme waarmee dendritische cellen myelopoiese
beïnvloeden, kan leiden tot alternatieve therapeutische strategieën voor
myeloproliferatieve ziekten zoals chronische myeloïde leukemie.

Samengevat hebben we een deel van de moleculaire interacties ontrafeld die de

ontwikkeling van immuuncellen in de marginale zone sturen. Eveneens toonden we een
belangrijke rol aan van dendritische cellen in de regulatie van hematopoiese. Deze
inzichten kunnen bijdragen tot de ontwikkeling van innovatieve behandelingen door
het moduleren van ons immuunsysteem.

10
I. Introduction

11
1. THE MARGINAL ZONE AND ITS CONSTITUENTS

1.1 Anatomy of the spleen

Aside from its roles in hematopoiesis and red blood cell clearance, the spleen exerts
unique immune functions. Owing to its exceptional exposure to circulating antigens, the
spleen as a secondary lymphoid organ primarily serves to generate protective
responses to blood-borne pathogens1. To accommodate these distinct functions, it is
divided structurally into white pulp and red pulp (Fig. 1a). In the red pulp, afferent
blood arrives in cords that consist of fibroblasts and reticular fibres without an
endothelial lining, creating an open circulation2. Red blood cells (RBCs) must pass into
venous sinuses, eventually drained by the vena lienalis, to re-enter the circulation3, 4.
Aged or damaged RBCs that are not able to traverse the endothelial slits in these
sinusoids are phagocytosed by red pulp macrophages. These macrophages clear the
blood from senescent cells or non-self particulates, but also play crucial roles in iron
homeostasis5. The red pulp also contains innate immune cells such as neutrophils,
monocytes, γδ T cells and dendritic cells that mediate early responses to circulating
pathogens. In addition, plasmablasts and effector CD8+ T cells can exert their function in
the red pulp after leaving the white pulp, following a CXCL12 gradient6.

In contrast, the white pulp is made up of lymph node like structures that contain B cell
follicles and T cell-rich areas. These structures locate around the central arterioles,
explaining why the inner T cell zone is often denoted as the periarteriolar lymphoid
sheath (PALS). Circulating B and T cells enter the white pulp in search of their cognate
antigen and initiate adaptive immune responses after being primed by antigen-
presenting cells. The formation of the white pulp and segregation of B- and T-cell zones
relies on cytokines produced by stromal cells, in a manner that depends on the
lymphotoxin β receptor. CXCL13, produced by follicular dendritic cells (FDCs), attracts
CXCR5-expressing B cells, while CCL19 and CCL21 derived from fibroblastic reticular
cells (FRCs) are required for the migration of T cells through CCR71, 7, 8. The splenic
white pulp can be thought of as lymph node equivalents that monitor circulating
antigens. However, in contrast to lymph nodes, the white pulp is not confined by a
fibrous capsule and does not receive afferent lymph. Rather, a specialized structure,
denoted as the marginal zone, separates white pulp from red pulp and directly receives
most of the afferent blood.

1.2 Structure and function of the marginal zone

While some terminal arterioles end directly in the red pulp cords, many branches of the
central arteriole drain in the marginal sinus, a structure lined by endothelial cells that
surrounds the white pulp. The endothelium of the marginal sinus is fenestrated,
allowing further passage of the blood from the marginal sinus into the red pulp9. Around
the marginal sinus, specific subtypes of macrophages, B cells and dendritic cells can be
found. Together these cells form the marginal zone, a transitional site between the

12
circulation and peripheral lymphoid structures (Fig. 1b). The marginal zone serves to
monitor the blood for antigens, activate early effector functions and initiate adaptive
responses.

Figure 1 | Structure of the spleen and the marginal zone.

a. Micro-anatomy of the spleen. The splenic artery branches into central arterioles, ending in cords in the
red pulp that lack an endothelial lining. Blood then passes into venous sinuses, eventually draining into
the circulation. Central arterioles are sheathed by white pulp areas made up of T cell zones and, more
peripherally, B cell follicles. The outer white pulp is segregated from red pulp by a specialized structure
called the marginal zone. Larger arteries and veins run in connective-tissue trabeculae that are
continuous with the splenic capsule. b. Organization of the marginal zone. Afferent blood arrives in the
marginal sinus, which is lined by fenestrated endothelium and surrounds the white pulp. Marginal zone
metallophilic macrophages (MMMs) lie between the endothelial cells and the white pulp, while marginal
zone macrophages (MZMs), marginal zone B cells and dendritic cells monitor blood that passes from the
marginal zone into the red pulp. Reprinted from Mebius and Kraal8, with permission.

13
At the edge of the white pulp lies a layer of marginal reticular cells (MRC), fibroblasts
that are characterized by a TNFSF11+ MadCAM+ VCAM1+ ICAM1+ PDPN+ phenotype.
These cells are contiguous with the network of stromal cells in the B cell follicles, but
have a distinct transcriptional profile10, 11. Between MRCs and the marginal sinus, a
unique subtype of macrophages can be identified by their expression of sialic-acid
binding immunoglobulin like lectin 1 (SIGLEC1, also known as sialoadhesin or
CD169)12. These macrophages were named marginal metallophilic macrophages
(MMMs) after their appearance upon silver staining. They closely resemble a population
of CD169+ macrophages in lymph nodes that are interposed between the subcapsular
sinus and the MRC network13. MMMs scavenge circulating antigens such as
polysaccharides and immune complexes and contribute substantially to the generation
of adaptive immune responses14. In particular, their ability to trap antigen is crucial for
the priming of CD8+ T cells and for activation of memory cytotoxic T cells against
intracellular pathogens13-16. Indeed, ablation of MMMs was shown to abrogate cross-
presentation of viral and tumor antigens, either in a direct manner or via antigen
transfer to CD8α+ dendritic cells16, 17. Whether MMMs themselves migrate to the T cell
zone, or extend their processes (and captured antigen) into the white pulp, requires
further investigation18.

Another population of marginal zone macrophages (MZMs) resides on the outer side
(i.e. red pulp side) of the marginal sinus, in close association with marginal zone B cells.
The expression of the C-type lectin specific intercellular adhesion molecule-grabbing
nonintegrin receptor 1 (SIGNR1) and the type 1 scavenger receptor macrophage
receptor with collagenous structure (MARCO) allows discrimination of MZMs from
other macrophages in the spleen19, 20. Using these receptors, MZMs recognize molecular
patterns on a variety of pathogens such as Streptococcus pneumonia, Escherichia coli,
Mycobacterium tuberculosis, Candida albicans and HIV21. Polysaccharide antigens
trapped by SIGNR1 can subsequently be captured by marginal zone B cells22.
Accordingly, mice that lack MZMs are more susceptible to infection and display altered
T-cell responses to antigenic challenge23-25. MARCO is not only important for capture of
antigens, as MARCO-deficient mice display an abnormal marginal zone architecture and
defective responses to TI-2 independent antigens (polysaccharide antigens that can
activate B cells independently of T cell help through B cell receptor cross-linking).
Interactions between MZMs and marginal zone B (MZB) cells via MARCO might control
MZB migration to the follicle 26, 27. Moreover, depletion of MZMs led to loss of MZB cells.
Inversely, the expression of SIGNR1 is induced by MZB cells22. MZMs have also been
demonstrated to capture circulating apoptotic cells and induce tolerogenic responses to
autoantigens28, 29. In line, defects in marginal zone macrophages have been associated
with the development of autoimmunity. Both macrophage subtypes develop from bone
marrow (BM) monocytes in a manner that depends on macrophage colony-stimulating
factor (M-CSF) and liver X receptor α (LXRα). During inflammation, they rapidly turn
over and are replaced by BM-derived monocytes. Both are important for responses

14
against polysaccharide antigens and for the induction of tolerance to self-antigens, but
the precise division of labor requires further investigation30.

The marginal zone is not continuous, but displays interruptions. These so-called
bridging channels are direct conduits between the T cell zone and the red pulp. B and T-
lymphocytes use this conduit to enter the white pulp from the circulation, or to exit the
white pulp to the red pulp31. However, follicular B cells might cross the marginal sinus
directly32. Furthermore, the bridging channel is home to a specific population of
dendritic cells that are essential for the priming of CD4+ T cells. To guide migration,
fibroblastic reticular cells (FRCs) extend a tubular network from the T cell zone to the
red pulp. While small antigens can flow freely into the white pulp through this conduit
system, larger antigens (> 60kDa) require transport by cells in the marginal zone33.
Whereas it is not known whether MMMs or MZMs are able to migrate to the white pulp,
important roles have been demonstrated for marginal zone B cells and dendritic cells in
transporting antigens into the follicle8, 18. The function and development of these cell
types will be discussed in more detail below.

1.3 Marginal zone B cells

Located in between the previously described macrophage subsets, a unique population
of B cells can be identified in the splenic marginal zone. MZB cells are an innate-like
subtype of B lymphocytes that express a limited repertoire of germline encoded antigen
receptors. In contrast to follicular B cells, they do not recirculate and are long-lived34.
While their role in the early defense against blood-borne pathogens is well
characterized, accumulating evidence suggests a broader role involving both T-
independent and T-dependent immune responses. Phenotypically, marginal zone B
(MZB) cells are characterized by high levels of IgM, CD21/35, CD1d and CD9, while
expressing lower levels of IgD, B220 and CD23 than follicular B (FoB) cells. Thus, the
MZB versus FoB fate choice of immature B cells will determine their eventual
phenotype, location, lifespan, and functional properties21, 35-38. How exactly this lineage
commitment is controlled, has been the subject of intensive investigation, and will be
discussed in more detail below.

1.3.1 Development
Hematopoietic progenitor cells in the bone marrow that have committed to the B cell
fate will rearrange their immunoglobulin (Ig) heavy chain loci. Productive
rearrangements, leading to expression of the pre-B cell receptor (BCR), allow further
proliferation and rearrangement of light chain loci. Successful rearrangements of heavy
and light chain genes eventually give rise to expression of IgM on the B cell surface. As a
part of central tolerance, B cells that react strongly to self will undergo receptor editing
or clonal deletion. Along the development of these immature B cells into mature splenic
follicular or marginal zone B cells, the cells go through intermediate, transitional (T1
and T2) stages. T1 cells do not recirculate yet, are IgD- CD23- but express CD93
(recognized by the AA4.1 antibody) and IgM. After passing through follicles, they can

15
develop into CD93+ IgD+ IgM+ CD23+ T2B cells that are able to recirculate. Before the T2
stage, most B cells have homed to the spleen. There, a T2 B cell has to make a fate
decision between the MZB or FoB lineage (Fig. 2). Notably, a subset of follicular B cells
has an IgDhi IgMhi CD21int phenotype and might serve as a reservoir that can replete
MZB cells after infections39. These cells are referred to as follicular type II B cells.
Analysis of transgenic mice has revealed that the pathways governing the MZB fate
decision largely fall into four categories: BCR signaling, the Notch pathway, the B-cell
activating factor receptor (BAFFR) and its downstream NFκB pathway, and
integrin/chemokine activity35.

BCR signaling
The concept that differences in BCR signaling strength affect the cell fate of developing
B cells, whereby strong signals favor FoB generation, is mainly based on analysis of
mice that lack negative regulators of BCR signaling such as Aiolos or CD2240-42. In these
mice, increased BCR signaling was associated with a decreased MZB pool. In contrast, a
follicular B cell deficit dominated in mice that lack components of the BCR-Bruton’s
tyrosine kinase (BTK) pathway 42-45. However, some findings have challenged this view.
In BCR transgenic mice, binding of weak self-antigens was shown to be required for
MZB development in a BTK and CD19 dependent manner. Thus, positive selection is
also important for MZB cells46-48. Probably the effect of BCR activation is modulated by
as yet unidentified mechanisms, whereby the context might dictate the impact of BCR
signals on this fate decision. Aside from the signal intensity, the nature of the antigen
receptor might also affect the follicular versus marginal zone B cell fate specification37,
46. This explains why B cells in different anatomical regions harbor distinct antigen

specificities, and ensures the location of B cells with broad specificities close to
circulating pathogens in the marginal zone.

Notch signaling
Notch is an evolutionary conserved pathway that regulates many binary fate decisions
(see below). Several lines of evidence have demonstrated a crucial requirement for the
Notch pathway in MZB cell development. Mice lacking the transcription factor RBPjκ,
the Notch2 receptor or Mastermind-like 1 (MAML1) in B cells display a specific MZB cell
deficit49-51. In follicular B cells, Msx2-interacting nuclear target protein (MINT)
negatively regulates Notch target genes by interfering with the interaction between
NICD and factor RBPjκ52. Cleavage of the Notch2 receptor by A disintegrin and
metalloprotease 10 (ADAM10) was shown to be essential for activation of the Notch2
pathway upon ligand binding53. Delta-like 1 (DLL1), expressed by fibroblastic reticular
cells in the white pulp, was identified as the relevant Notch ligand for MZB cell
development54-56. How exactly Notch controls the MZB fate, is not well known.
Interestingly, canonical Notch signals stimulated degradation of both isoforms of the
helix-loop-helix transcription factor E2A. Decreased levels of E2A, or increased levels of
the E2A inhibitors ID2 and ID3, were shown to promote MZB cell development.
Inversely, Id3 null mice harbor more follicular B cells57. The levels of ID3, in turn, were

16
regulated by Pyk-2 kinase, which is crucially required for MZB cell formation58.
Together, these findings suggest that Notch signals can trigger a transcriptional
program that determines MZB cell identity through regulation of helix-loop-helix
transcription factors.

BAFFR and NFκB

BAFF, a member of the tumor necrosis factor family, stimulates the survival of T2 B cells
and follicular B cells. In contrast, BAFF appears to play a different, developmental role
in MZB cells59-61. Ligation of the BAFFR is able to activate both canonical and non-
canonical NFκB pathways. While the survival signals in transitional and follicular B cells
are mainly mediated by the non-canonical cleavage of p100 into p52 (NFκB2) by IκB
kinase-α (IKKα), activation of the canonical pathway was shown to be essential and
sufficient for the development of marginal zone B cells. This is exemplified by the deficit
of MZB cells in mice that lacked p50 (NFκB1), REL (also known as c-REL) and RELA
(also known as p65), and the restoral of MZB cells by constitutively active IKKβ in mice
that lack the BAFFR62-67. Similarly, mutations in the BCL10-MALT1-CARMA1 complex
upstream of IKKβ were essential for the generation of MZB cells. However, whether this
complex mediates BAFFR signals or functions downstream the BCR in MZB cells,
remains uncertain. How BAFF signals influence the MZB fate, is not known. Intriguingly,
the factor p50 activated a transcriptional program in synergy with Notch transcription
factors, highlighting how these pathways may cooperatively instruct lineage
determination68.

Adhesion/migration
It was long believed that marginal zone B cells were a sessile population. In two
landmark reports, Cyster and colleagues have demonstrated that the very opposite is
true32, 69. Marginal zone B cells were found to shuttle continuously between the
marginal zone and B cell follicles, with 20% of all MZB cells switching compartments
every hour. Migration to the B cell follicle allows the deposition of captured antigens on
follicular dendritic cells, and depends on the CXCR5-CXCL13 axis. Return to the
marginal zone depends on the chemokine sphingosine-1-phosphate (S1P) and its
receptors S1P1 and S1P3 in MZB cells. In absence of functional S1P1, B cells with a MZB
phenotype mostly reside in B follicles32, 69. The expression of the S1P degrading enzyme
lipid phosphate phosphatase 3 (LPP3) in the red pulp ensures a high local gradient of
S1P in the marginal zone and enables efficient shuttling70. In mice lacking both S1P1
and CXCL13, MZB cells appear to be correctly positioned in the marginal zone again,
probably by retention via integrins and other receptors such as the cannabinoid
receptor 2 (CB2). CB2, as well as ICAM1-αLβ2 and VCAM1-α4β1 interactions, retains MZB
cells in the marginal zone, preventing them from being washed into the red pulp and
circulation by the blood flow71-74. Many mutations in intracellular signal mediators that
affect MZB cells can be linked to chemokine/integrin signaling and migration. In
absence of the Wiskott-Aldrich syndrome protein (WASP), actin polymerization is
abrogated and MZB cells fail to migrate towards S1P75. The Rho family GTPases RAC1

17
and RAC2, as well as ARHGEF1, a guanine nucleotide exchange factor for RhoA, also
control cytoskeletal rearrangements necessary for cell migration and are required for
normal MZB cell homeostasis76-79. The S1P receptors are coupled to G proteins, and
intact Gα12/13 proteins were equally required for a normal MZB cell pool80, 81. To guide
MZB directionally towards the white pulp, differential expression or relocalization of
the integrins αLβ2 and α4β1 might be triggered by shear flow82.

Figure 2. | The follicular versus marginal zone B cell fate decision. Transitional (T1 and T2) B cells
can mature into either a follicular type I (FO-I), a follicular type II (FO-II) or a marginal zone B (MZB) cell.
This lineage decision is governed, among other factors, by signals through the B cell receptor (BCR), the B
cell-activating factor receptor (BAFFR) and Notch2. Reprinted from Pillai et al.35, with permission.

1.3.2 Function
Unlike follicular B cells, MZB cells carry polyreactive B cell receptors, as well as Toll-like
receptors (TLRs), placing them between innate and adaptive immunity38. They are in a
‘pre-activated’ state, allowing rapid low-affinity responses to conserved microbial
molecules. MZB cells can differentiate rapidly into IgM-secreting plasmablasts and
bridge the temporal gap that is inherent to the T-cell dependent production of high-
affinity antibodies. It is widely accepted that MZB cells play an essential role in the
defense against encapsulated pathogens. Defective function or absence of MZB cells is
associated with impaired antibody responses to T-independent type 2 multivalent
antigens such as capsular polysaccharides, and increased sensitivity to infections with
encapsulated bacteria such as Streptococcus pneumonia, Neisseria meningitides or
Haemophilus influenzae58, 83-89. A mature marginal zone, containing MZB cells and
specialized macrophage subsets, is only seen from 3 weeks after birth in rodents and
from 1-2 years of life in humans, explaining why infants fail to mount effective
responses to polysaccharide antigens38, 90-94. More direct proof comes from studies in
Pyk-2- and p50-deficient mice, where absence of MZB cells impaired the generation of
antibody responses to polysaccharide antigens and pneumococcal infection58, 95.
Expression of the complement receptor CD21 may assist in capturing complement

18
coated polysaccharides58. In addition, MZB cells may also generate antibody responses
against blood-borne viruses and RBC-derived alloantigens, at least in mice96, 97. Under
homeostatic conditions, MZB cells produce natural antibodies that recognize self and
foreign molecules, and may provide a first line of defense to intruding microorganisms.
Upon activation, MZB cells rapidly differentiate into plasmablasts that can be found in
the bridging channels and at the T:B cell border, and are thought to mature into plasma
cells in the red pulp98, 99. The function of MZB cells can be enhanced by other innate cells
in the marginal zone, such as neutrophils and natural killer T (NKT) cells100-104.

Increasing evidence also points to a role for MZB cells in T-dependent immune
responses. MZB cells express higher levels of major histocompatibility class II (MHCII)
and costimulatory molecules than FoB cells. Both in vitro and in vivo, MZB cells activated
CD4+ T cells more potently than FoB when exposed to soluble antigen105. Cell transfer
studies have revealed that MZB cells are fully capable of initiating germinal center
reactions and undergoing somatic hypermutation (SHM)106. Remarkably, extrafollicular
B cell responses to proteins have been described that may also include class-switch
recombination (CSR) and SHM107, 108. Still, most MZB cells in mice remain IgM+ and
display little evidence of SHM. To what extent CSR and SHM occur in activated MZB
cells, and whether this requires follicular germinal centers, awaits further elucidation.

1.4 Dendritic cells

In 1973, Steinman and Cohn discovered a novel cell type in mouse spleen, which they
named dendritic cells (DCs) after their stellate morphology109. Only later, the
exceptional ability of these cells to process antigens and prime T cells was recognized.
Because of their importance for the induction of both protective and tolerogenic
adaptive immune responses, DCs have since represented a very attractive target for
vaccination strategies and novel therapies targeting tumors and autoimmune
diseases110.

Dendritic cells reside in all lymphoid and most non-lymphoid tissues, where they
continuously sample antigens from their environment. Antigen-presenting DCs from
peripheral tissues migrate through the lymphatics in a CCR7-dependent manner and
present these antigens to cognate T cells in draining lymph nodes. In the spleen,
dendritic cells capture circulating antigens in the marginal zone and migrate to the T
cell zone in the white pulp. Under non-inflammatory conditions, antigen presentation
contributes to the maintenance of peripheral tolerance to harmless antigens. In
contrast, the recognition of pathogen-associated molecular patterns (PAMPs) through
germline-encoded receptors such as Toll-like receptors (TLRs) and C-type lectins
(CTLs) activates a maturation program in DCs that triggers upregulation of MHCII, CCR7
and co-stimulatory molecules. The ability of these mature DCs to prime antigen-specific
T cells is unrivalled in the mammalian immune system110-112. Since the discovery that
CD8α expression discriminates two functionally distinct DC subtypes in the spleen, it
has increasingly become apparent that DCs are phenotypically and functionally

19
heterogeneous111, 113, 114. A new nomenclature system based on ontogeny was recently
proposed to classify dendritic cells, and will be used throughout this thesis115.

1.4.1 Plasmacytoid dendritic cells

Under homeostatic conditions, dendritic cells can be divided into three distinct classes.
Plasmacytoid dendritic cells (pDCs) constitute a distinct cell type, characterized by a
rounded morphology that resembles plasma cells. This is due to their elaborate
endoplasmic reticulum, which enables pDCs to secrete large amounts of type I
interferons during viral infections. In the mouse, pDCs are identified as CD11cint
SiglecHhi CD317hi B220+ CD11blo Ly6C+ cells and lack MHCII expression. However,
activated pDCs can upregulate MHCII and costimulatory markers, and acquire antigen-
presenting capacities116, 117. Plasmacytoid dendritic cells share a common progenitor
with cDCs, the common dendritic cell precursor (CDP), although several reports have
demonstrated that at least a part of pDCs may derive from a lymphoid precursor111, 118-
123.

1.4.2 Conventional dendritic cells

To distinguish them from plasmacytoid dendritic cells, the antigen presenting cells
described by Steinman are often designated as ‘classical’ or ‘conventional’ dendritic
cells (cDCs). As described above, their primordial function is to sample and process
environmental and cell-associated antigens, and present them on major
histocompatibility molecules (MHC) I and II to cognate T cells. Conventional DCs are
tissue resident and short-lived, although they can proliferate locally to some extent124-
126. Their development and homeostasis requires signals through the fms-related

tyrosine kinase 3 (Flt3) receptor. cDCs are continuously replenished by committed

precursor cells, called pre-cDCs, that originate from the CDP in the bone marrow. Pre-
cDCs migrate via the blood and seed peripheral tissues and SLOs, where they will
differentiate into mature dendritic cells. Within cDCs, two distinct subtypes are
conserved across species that differ in their ontogeny, transcriptional profile and
function127.

Type 1 dendritic cells (cDC1s) can be distinguished by their XCR1+ CD24+ CLEC9A+
CD11b- CD172a- surface phenotype. Resident cDC1s in lymphoid tissues also express
CD8α, while migratory cDC1s or cDC1s in non-lymphoid are CD8α- but express the
integrin αE (CD103). In contrast, type 2 dendritic cells (cDC2s) are XCR1- CD11b+
CD172a+ across tissues and species. While both cDC subsets function as sentinel
antigen-presenting cells (APCs), some functional specialization is clearly present.
Typically, cDC1s are implicated in the activation of cytotoxic T cells and polarization
towards Th1 responses against intracellular pathogens and tumoral antigens. Through a
process called cross-presentation, cDC1s can present exogenous antigens on MHCI
molecules. On the other hand, cDC2s preferentially present antigen in an MHCII context
and prime CD4+ T cells, often inducing type 2 or type 3 immune responses110-112.
However, this distinction is not absolute. For example, cDC2 can cross-present antigen

20
to CD8+ T cells after TLR stimulation, and cDC1s are perfectly capable of CD4+ T cells
activation in vitro112, 128. Similarly, both subsets have been implicated in the induction of
tolerance against self-antigens, although the role of cDC1s appears to be dominant129-
134.

1.4.3 Monocyte-derived dendritic cells

During inflammation, monocytes are recruited to the peripheral tissues and
differentiate into cells that functionally and phenotypically resemble macrophages or
dendritic cells. When these cells acquire features characteristic of cDCs, such as high
expression of CD11c, MHCII, CD11b and CD172a, they are called monocyte-derived
dendritic cells (moDCs). These cells have been shown to migrate to lymph nodes and
can display antigen-presenting functions, thereby complicating the interpretation of
studies that were performed before their recognition. Monocyte-derived DCs were
discovered in a model of Listeria monocytogenes infection, where they secreted high
amounts of TNF and intracellular nitric oxide synthase (iNOS), hence their original
name TNF-and-iNOS-producing DC (TipDC). However, the development of monocyte-
derived DCs does not require Flt3 signals, but rather relies on macrophage-colony
stimulating factor (M-CSF) signals. In contrast to cDCs, moDCs express MERTK and
CD64, but are negative for CD26 (DPP4) and CD24. While the efficiency by which they
prime T cells may be inferior to that of cDCs, moDCs often outnumber the latter in an
infectious setting, thus contributing substantially to host defense.

1.4.4 Transcriptional regulation of conventional DCs

In the bone marrow, sequential differentiation steps from hematopoietic stem cells to
more lineage-restricted progenitors occur and eventually lead to the formation of
common dendritic precursor cells (CDP), cells that give rise to pDCs (via a pre-pDC
intermediate) and both subsets of cDCs, but have lost the potential to generate other
myeloid cells such as neutrophils, monocytes, macrophages. Most pre-cDCs in the bone
marrow are still bipotent, but gradually become committed to the cDC1 or cDC2 fate.
Committed precursors are called pre-cDC1s and pre-cDC2s, respectively136. The factors
that regulate the cDC1-cDC2 fate decision have gained much attention, but are still not
completely understood (Fig. 3). Interferon-regulatory factor 8 (IRF8), basic leucine
zipper transcriptional factor ATF-like 3 (BATF), DNA-binding protein inhibitor ID2 and
nuclear factor interleukin-3 (NFIL3) are all critically required for the development of
cDC1s135-141. The requirement for IRF8 appears to be the most stringent and is
continuously needed to maintain the cDC1 identity142. Accordingly, some cDC1s could
still develop during infection in mice that lacked ID2, NFIL3 or BATF3, while IRF8-
knockout mice remained completely devoid of all cDC1s143.

By contrast, no single transcriptional regulator has been identified yet that is required
for the development of all cDC2s. Type 2 cDCs are functionally and phenotypically more
heterogeneous than cDC1s, which is reflected in the differential dependency of cDC2s in
different tissues on distinct transcription factors. For example, the interferon regulatory

21
factor 4 (IRF4) was shown to be essential for the generation of CD4+ cDC2s in spleen,
CD24+ cDC2s in lung, and CD103+ cDC2s in the gut, while other cDC2s developed
normally144-146. Functionally, IRF4-dependent cDC2s mainly appear to be required for
the differentiation of Th17 cells, although they promoted Th2 polarization in a mouse
model of allergic asthma145-147. In addition, IRF4 appears to control the migration of
cDC2s towards draining lymph nodes, and regulates a transcriptional program that
governs MHCII-dependent antigen presentation148, 149. Interestingly, loss of Krüppel-like
factor 4 (KLF4) affected similar cDC2 subsets (CD24+ Mgl2+ lung DCs, dermal CD11b-
cDC2s and intestinal CD103+ CD11b+ cDC2s) and disrupted IRF4 expression in pre-
cDCs. KLF4-dependent dendritic cells mainly promote Th2 responses, as demonstrated
in a Schistosoma mansoni infection model150, 151. ZEB2, a transcription factor mainly
associated with epithelial-mesenchymal transition (EMT), promotes pDC and cDC2
development by repressing the cDC1 fate, although it might be required only by a subset
of cDC2s152, 153. Notch instruction, LTβR signaling and the transcription factor RUNX3
control the homeostasis of a subset of cDC2s in the spleen, as will be discussed below126,
154-157. The absence of cDC2s in mice lacking the NFκB member RELB, was mainly due to

the role of RELB in stromal cells, except for a cell-intrinsic role in CD4+ cDC2s in the
spleen158, 159. Finally, IRF2-knockout mice also display a cell-intrinsic deficit in splenic
cDC2s, although the mechanism is unclear160.

1.4.5 Conventional dendritic cells in spleen

Historically, splenic cDCs have been subdivided in CD8α+, CD4+ and double negative
cDCs. At present, it is clear that additional heterogeneity exists within both cDC1s
(largely equivalent to CD8α+ cDCs) and cDC2s (comprising CD4+ and double negative
cDCs) in the spleen162. Type 1 dendritic cells in the spleen were originally thought to
reside only in the T cell zone of the white pulp, based on staining for the C-type lectin
DEC205 (CD205)163, 164. However, several groups have now demonstrated that XCR1+
cDCs can also be found in the marginal zone and the red pulp165. Interestingly, cDC1s
that locate in the marginal zone often express CD103 and langerin (CD207) and appear
to be functionally more mature, with an increased ability to phagocytose apoptotic cells,
secrete IL-12 and prime CD8+ T cells. This marginal zone cDC1 subset has been
associated with antiviral responses and the induction of tolerance against self-
antigens162, 166. Similarly, CD205+ also stained DCs in the T cell zone and perifollicular
regions of human spleen167.

The majority of cDC2s in the spleen co-express endothelial cell selective adhesion
molecule (ESAM) and CD4 on the cell surface. In several reports, a crucial role for Notch
signaling in the terminal differentiation and function of this cDC2 subset was
demonstrated. Deletion of dendritic cell-derived Notch2 or its ligand Delta-like 1 (DLL1)
on fibroblasts abrogated the development of ESAM+ CD4+ cDC2s155-157. Notch2-deficient
dendritic cells failed to prime CD4+ T cells and generate germinal center responses to
blood-borne antigens168. Mice that lack the transcription factor RUNX3 displayed a
similar loss of ESAM+ CD4+ cDC2s, potentially through direct interactions with RBPjκ154.

22
Figure 3 | Development of conventional DCs. Most pre-cDCs in the BM are still uncommitted, but
gradually split up into IRF8-/BATF3-dependent cDC1-committed pre-cDCs and cDC2-committed pre-
cDCs. These pre-cDCs then leave the BM into the bloodstream and seed the different tissues where they
differentiate into cDC1s or cDC2s. Whereas cDC1 development requires IRF8, BATF3, NFIL3 and ID2,
cDC2s depend on the activity of IRF4, ZEB2, KLF4, RELB, Notch2, RUNX3 and Ikaros to varying degrees.
Adapted from Sichien et al.161, with permission.

23
Retinoic acid is equally required to maintain this subset of cDC2s, although it is not
known whether this dependency is cell-intrinsic169, 170. It has long been recognized that
activation of the lymphotoxin β receptor (LTβR) by the heterotrimeric ligand
lymphotoxin α1β2 (LT) also regulates the homeostasis of cDC2s in the spleen126, 171-173.
More recently, it was revealed that this pathway specifically maintained Notch2-
dependent cDC2s155, 157. However, how retinoic acid, Notch and LTβR signals intersect
in dendritic cells, remains to be elucidated.

As early as 1990, the presence of CD11c+ cell clusters in marginal zone bridging
channels was identified by Steinman and colleagues174. Later, these clusters were
shown to consist of dendritic cells positive for the dendritic cell inhibitory receptor 2
(DCIR2), a C-type lectin that is preferentially expressed by cDC2s165. The positioning of
cDC2s in bridging channels is guided by the chemokine receptor Epstein-Barr-virus
induced gene 2 (EBI2, also known as GPR183) in response to oxysterol ligands
produced by stromal cells164, 175. This location uniquely allows these cDC2s to initiate
CD4+ T cell responses to blood-borne antigens such as circulating bacteria or allogeneic
red blood cells.

1.5 Other cell types

It should be mentioned that the cells composing the marginal zone are not limited to
dendritic cells, macrophages and B cells. For example, invariant natural killer T (iNKT)
cells have been shown to patrol the marginal zone and interact with MZB cells and
dendritic cells to respond to circulating lipid antigens176. αGalCer-loaded MZB cells did
not directly activate iNKT cells ex vivo, but enhanced the NKT response induced by
dendritic cells in vivo177. As mentioned previously, neutrophils have also been identified
in the marginal zone under inflammatory conditions. Upon Streptococcus pneumoniae
infection, neutrophils were retrieved in the MZ interacting with MZM and MZB cells178.
Neutrophils have been detected in the perifollicular zone in humans as well. These so-
called ‘B-helper’ neutrophils enhanced the production of antibodies against T-
independent antigens by secreting adjuvant molecules such as pentraxin-3102, 104.
Though derived from circulating neutrophils, B-helper neutrophils displayed a distinct
transcriptional identity102. However, this costimulatory function of human splenic NFs
could not be confirmed by others179. Finally, the CD4+ CD3- cells that had been observed
in the marginal zone of mice, appeared to be innate lymphoid cells (see chapter 4) that
interacted with marginal zone B cells and marginal reticular cells. Similarly, type 3
innate lymphoid cells (ILC3s) were also identified in the perifollicular regions of human
spleen103, 180.

1.6 The marginal zone in humans

While the marginal zone of mice is relatively well characterized, the exact structure and
function of its counterpart in humans remains enigmatic. Although most information
about the human splenic micro-anatomy is derived from simple hematoxylin-eosin
stainings or immunohistochemical detection of a limited number of antigens, clear

24
differences exist between species (Fig. 4)1. In humans, the term ‘marginal zone’ refers
to a region in the B cell area that borders the germinal center and mantle zone, but does
not form the interface between white and red pulp. The term was coined because this
region contains B cells that phenotypically resemble murine marginal zone B cells
(IgMhi IgDlo CD21hi CD23- CD1c+) and contribute to the defense against encapsulated
bacteria. However, human MZB cells differ in many aspects from MZB cells in rodents.
Human MZB cells correspond to memory CD27+ IgM+ B cells that recirculate, have often
undergone somatic hypermutation, and can also be identified in other secondary
lymphoid organs181-183. Indeed, memory IgM+ B cells in peripheral blood demonstrated
a similar gene expression profile and BCR repertoire and were lost in asplenic
individuals87, 181. It is unclear whether the development of these memory B cells
depends on germinal center reactions and antigen. Interestingly, patients with hyper-
IgM syndrome that have mutations in CD40L also harbor somatic mutations in MZB
cells, indicating an extrafollicular origin184. In contrast, others have identified MZB cells
in gut-associated lymphoid tissue (GALT) that showed signatures of GC experience185.

Unlike the marginal zone in mice and rats, human marginal zone B cells only form an
outer shell around the B cell follicle, but do not surround the T cell zone. Furthermore,
the existence of a true marginal sinus around the white pulp in humans is
controversial1, 186-190. Nevertheless, the differences between human and mouse spleen
might be smaller than perceived. Between the white and red pulp lies a poorly
characterized but distinctive zone, called the perifollicular zone187, 191. This
compartment consists of multiple cavernous sinuses, that differ from typical red pulp
sinusoids, and receives afferent blood1, 192, 193. Here, macrophages were identified by
antibodies recognizing SIGLEC1 (CD169) and SIGN (CD209, closely related to SIGNR1).
Moreover, a population of CD205+ dendritic cells inhabits this region in close
association with MadCAM1+ fibroblasts, that surround human splenic follicles and bear
resemblance to the MRC network in rodents167. B cells also occur in the perifollicular
zone, as do innate lymphoid cells and B-helper neutrophils103, 104. The perifollicular zone
and red pulp in humans also contain sheathed capillaries192, 194, 195. These structures are
composed of capillaries with an endothelial lining, surrounded by cuboidal stromal
cells, and CD169+ macrophages and B cells enveloping these capillaries in different
layers196-198. The function of these structures remains to be determined. Sheathed
capillaries are absent in mice and rats, which seem to form an evolutionary
exception190.

25
Figure 4 | The marginal zone in humans and mice.
Structural differences exist between the organization of the spleen in mice (left) and humans (right). The
T cell zone (TCZ, also known as PALS) and B cell zone (BCZ) follicles (gray and shades of blue, with light
zone, LZ, and dark zone, DZ) constitute the white pulp (WP). The marginal zone (MZ) in mice and the
perifollicular zone (PFZ, dark blue) in human spleen lie at the border between the WP and red pulp (RP).
The PFZ remains poorly characterized and it is unknown to what extent its structure and function are
analogous to the mouse MZ. In mice, the marginal zone consists of a marginal sinus that is lined by a
concentric ring of CD169+ marginal metallophilic macrophages (MMM, dark blue) around the WP, while
marginal zone B cells (MZB, blue) and marginal zone macrophages (MZM, light blue) lie between the
marginal sinus and the red pulp. Blood flows from the central arteriole (CA) through the marginal zone
towards the red pulp. In humans, MZB cells surround activated B cells, containing a germinal center (GC)
and Corona (gray, “Cor”). Under homeostatic conditions, cDC1s localize in the TCZ, MZ and RP, whereas
cDC2s reside at bridging channels. From Lewis SM, Williams A, Eisenbarth SC. Structure and function of
the immune system in the spleen. Sci Immunol. 2019;4(33)1. Reprinted with permission from AAAS.

26
2. NOTCH SIGNALING AND FUNCTION OF ADAM PROTEINASES

Adapted from
Lambrecht BN, Vanderkerken M, Hammad H. The emerging role of ADAM
metalloproteinases in immunity. Nat Rev Immunol. 2018;18(12):745-58.

Author contribution
Matthias Vanderkerken co-wrote the manuscript and created the figures.

27
2.1 The Notch signaling pathway

Notch signaling constitutes a highly conserved pathway throughout evolution. Typically,

Notch signals require close intercellular contact and often determine fate choices
between pre-existing developmental programs199. Notch has widespread effects in
tissue development and renewal, and its precise effects highly depend on contextual
factors. Its mechanism of signal transduction is unique compared to other conserved
signaling pathways. Because of its pleiotropic and highly impacting effects, the Notch
pathway is highly regulated at different levels, which I will briefly highlight here.

Activation of Notch receptors relies on two subsequent proteolytic cleavage events

induced by ligand binding, ultimately leading to nuclear translocation of the Notch
intracellular domain (NICD) that will assemble a transcription complex (Fig. 5). In
vertebrates, five ligands (Delta-like (DLL) 1, 3 and 4 and Jagged1 and 2) are expressed
that can engage 4 distinct Notch receptor paralogs (Notch1-4). The ligands are
membrane-bound and thus, receptor ligation requires physical cell-cell contact.
Glycosylation of epidermal growth facter (EGF)-like repeats in the Notch extracellular
domain by the O-fucosyltransferase Pofut1 and Fringe glycosyltransferases was shown
to differentially modulate the affinity of Notch receptors for different ligands. Notch
receptors are noncovalently linked heterodimers consisting of an intracellular and
extracellular receptor domain. To liberate the Notch intracellular domain, Notch
receptors must first be proteolytically cleaved by A disintegrin and metalloproteinases
(ADAMs). However, this cleavage only occurs after the receptor undergoes a
conformational change, induced by a pulling force on the receptor. This pulling force is
generated by ligand endocytosis in the ligand-bearing cell, which depends on E3-
ubiquitin ligases such as Mindbomb and Neuralized. When ADAMs have cleaved off
most of the extracellular part of the receptor, a second proteolytic event can occur,
controlled by intramembrane γ-secretases. When NICD translocates to the nucleus, it
interacts with the DNA-binding protein RBPjκ (also known as CSL) and the coactivator
Mastermind-like 1 (MAML1). This core pathway, leading to transcriptional changes
through RBPjκ, is known as canonical Notch signaling199-202. Notch effects independent
of RBPjκ have been described, but the mechanisms remain enigmatic203, 204. Members of
the Hairy enhancer of split (Hes) or Hairy related (Hey or Hrt) genes were identified as
Notch target genes, but many genes seem to be regulated in a very cell-type and
context-specific manner205, 206. Moreover, many Notch-responsive elements are
enhancers that may only affect transcription when NICD synergizes with other
transcriptional cofactors such as NF-κB and hypoxia inducible factor-1 (HIF-1)199, 207.

The Notch pathway plays pivotal roles in the haematopoietic and immune system, and
governs many lineage decisions in developing lymphoid and myeloid cells201, 208. Below,
the roles of Notch signaling in the immune system will be highlighted as we discuss the
role of ADAMs that control their activation in different cell types.

28
Figure 5 | Notch activation by ADAM proteins. a | The classical Notch signalling pathway. Epsin-
mediated transendocytosis of Notch-engaged delta-like ligand 1 (DLL1) likely provides the pulling force
to extend the negative regulatory region (NRR) of Notch 1 into an open configuration, which renders
Notch 1 at the cell surface sensitive to cleavage by ADAM10 (a disintegrin and metalloproteinase 10) at
protease site 2 (S2). ADAM10 activity at the cell surface is triggered by a Notch-induced conformational
change. b | An alternative model of endocytic processing of Notch. According to one study249 —the
relevance of which has yet to be shown in vivo — Notch cleavage at S2 might occur only after the fusion of
Golgi-derived vesicles containing ADAM10 with endosomes containing endocytosed dissociated Notch. In
this study, endocytosis of Notch prior to ADAM10-mediated cleavage is promoted by the E3 ubiquitin
ligase deltex-like 4 (DTX4). In both scenarios, regulated intramembrane proteolysis by the γ-secretase
complex at protease site 3 (S3) subsequently liberates the Notch intracellular domain, which migrates to
the nucleus. Here, the Notch intracellular domain binds to RBPjκ, mastermind-like-1 (MAML) and p300 to
induce the transcription of Notch target genes.

29
2.2 The role of ADAM metalloproteinases and Notch signaling in immunity

The ADAM (a disintegrin and metalloproteinase) proteins are membrane-anchored

enzymes that convert nearby membrane-anchored cytokine precursors into soluble
bioactive mediators209, 210. The best-known example of this is the cleavage of
membrane-bound tumour necrosis factor (TNF) into soluble TNF (sTNF) by ADAM17,
which was originally named TNFα-converting enzyme (TACE) for this reason211. The
metalloproteinase activity of ADAM proteins can also be directed at other
transmembrane proteins, such as cytokine receptors, Notch receptors, phagocytic
receptors and cell adhesion molecules, all of which are known to have important roles
in immune system function. Cleavage of the ectodomains of these transmembrane
proteins can lead to release of the soluble ectodomain (for example, soluble TNF
receptor (sTNFR) and sCD44), which is known as ectodomain shedding or sheddase
activity. The shed ectodomains can function as activators or inhibitors of protein
function. Furthermore, in many cases, the stub that remains on the cell surface after
proteolytic cleavage by ADAM proteins can be processed further through regulated
intramembrane proteolysis by γ-secretases; the intracellular domain of the cleaved
molecule can then function as a signalling molecule and transcription factor in the
nucleus.

With the current development of and interest in RNA sequencing at population and
single-cell levels to study immune system diversity, it would be easy to ignore the
ADAM-mediated posttranslational regulation of the function of membrane proteins in
the immune system. Indeed, other posttranslational modifications such as
phosphorylation and polyubiquitylation are indisputably important regulators of the
stability, activity and localization of key proteins involved in immune responses212.
Here, recent insights into the molecular cell biology of ADAM proteins in the context of
the normal and abnormal immune system are discussed.

2.2.1 Regulation of ADAM proteins

The ADAM family contains 21 members in humans, thirteen of which encode a
proteolytically active domain and share a common structure (Fig. 6)213. Unlike the
substrates for caspases, there is no known ADAM-specific motif that can be used to
predict which molecules will be sensitive to cleavage by ADAM proteins, which has led
to a very long and complex list of potential substrates. The identification of ADAM
substrates is based mainly on in vitro cleavage assays, as well as proteomic approaches
and some in vivo models.

From immune system-wide expression scans, such as the Immunological Genome

Project (Immgen) data set214, it is clear that mRNAs encoding ADAM8, -9, -10, -12, -15, -
17, -19, -28 and -33 are found in immune cells. Among these, ADAM10 and ADAM17 are
the family members that have been studied in most detail. They have a close

30
phylogenetic relationship and are widely expressed by most immune cells. As these two
ADAM proteins have a large number of potential substrates, and the proteolysis they
mediate is irreversible, their enzymatic activity must be tightly regulated through post-
translational mechanisms. Regulation often occurs in discrete cellular compartments,
through conformational changes, and by binding to activators, inhibitors and stabilizers.

Figure 6 | Structure of human proteolytically active ADAM and ADAMTS proteins. ADAM (a
disintegrin and metalloproteinase) proteins have a common multidomain structure, with an amino-
terminal extracellular domain that includes a chaperone-like prodomain with two convertase recognition
sites, a Zn2+-dependent metalloproteinase catalytic domain, a disintegrin domain, a cysteine-rich domain
containing a hypervariable region, an epidermal growth factor domain-like (EGF-like) repeat domain, a
transmembrane domain and a carboxy-terminal SH3-binding cytoplasmic domain. In ADAM10 and
ADAM17, the EGF-like domain is absent and the membrane-proximal region is often referred to as the
stalk region. Depicted are the 13 proteolytically active ADAM proteins in humans; ADAM proteins that are
expressed only in testis are denoted with an asterisk215. ADAMDEC1 is an unusual metalloproteinase that
unlike the canonical ADAM proteins is secreted rather than membrane bound. It consists of a prodomain,
a metalloproteinase domain with catalytic activity and a shortened disintegrin domain. ADAMTS (a
disintegrin and metalloproteinase with trombospondin motifs) proteins are a closely related family of
secreted proteinases that contain a variable number of trombospondin repeats.

Regulation by intracellular retention. One way in which the proteolytic activity of ADAM
proteins at the cell membrane can be regulated is the retention of properly folded
ADAM proteins in the endoplasmic reticulum (ER) or Golgi network. In the case of
ADAM10, an arginine-rich region of the cytoplasmic tail controls ER retention by an

31
unidentified mechanism, whereas association of ADAM10 with tetraspanins and
synapse-associated protein 97 (SAP97) promote its exit from the ER and Golgi network,
respectively.216-218. Interactome studies have shown that the cytoplasmic tail of
ADAM10 binds to many adaptor proteins and non-receptor tyrosine kinases containing
SH3 domains, such as NCK, GRB2, LCK, FYN and ITK, all of which have important
functions in the immune system219. However, the functional importance of these
interactions in immune cells in vivo remains to be determined.

Regulation at the cell membrane. The enzymatic activity of ADAM proteins is also
controlled at the cell membrane220-222. The sheddase activity of ADAM17 is triggered by
exposure of its positively charged, extracellular, membrane-proximal domains to
membrane-expressed, negatively charged phosphatidylserine. The ensuing
conformational change in ADAM17 pulls its metalloproteinase domain closer to the cell
surface to cleave substrates223, 224. Although the exposure of phosphatidylserine in the
outer leaflet of the cell membrane is best known as an ‘eat-me’ signal in cells
undergoing apoptosis, regulated exposure of phosphatidylserine can also occur in
viable immune cells in response to signalling events through protein kinase C (PKC)
isoforms and a rise in intracellular Ca2+ concentration. Not surprisingly, signalling
through receptors that trigger the activation of PKC isoforms, including the B cell
receptor (BCR) and T cell receptor (TCR), as well as artificial T cell stimulation with
phorbol 12-myristate 13-acetate (PMA) and ionomycin, can stimulate sheddase activity
of ADAM17 by regulated phosphatidylserine exposure224, 225.

Regulation by phosphorylation and dimerization. Several kinases have been shown to

increase the sheddase activity of ADAM10 and ADAM17, with most studies indicating a
role for the mitogen-activated protein (MAP) kinases ERK and p38 (REFS226-228). How
exactly the phosphorylation of ADAM proteins regulates their activity is a matter of
intense study. Tissue inhibitor of metalloproteinase 3 (TIMP3) is a potent inhibitor of
ADAM17 activity through binding of the catalytic domain229. One group has reported
that ADAM10 and ADAM17 dimerize at the cell surface, which inhibits their catalytic
activity as TIMP3 preferentially binds to the dimeric conformation of the proteins230
(Fig. 7). In this model, dimerization was reduced by the phosphorylation of a key
conserved threonine in the cytoplasmic tail of ADAM17 (Thr735)227, 230, 231. Although
ERK and p38 are widely implicated in ADAM10 and ADAM17 activation, these can be
activated by many upstream kinases functioning downstream of a large number of
receptors. The fine molecular detail of the upstream kinase signalling cascade that
promotes ADAM-mediated shedding in cells of the immune system therefore awaits
further elucidation232.

32
Figure 7 | Regulation of ADAM proteins. The prodomain of ADAM (a disintegrin and metalloproteinase)
proteins has a chaperone function that prevents proteolytic activity in the lumen of the secretory system (Golgi
and endoplasmic reticulum (ER))233. As the ADAM protein moves from the Golgi and trans-Golgi network to the
cell surface, the prodomain is progressively cleaved off, by the action of prodomain convertase proteins PC7
and furin on the proprotein convertase recognition sites234. Thus, the ADAM protein gradually acquires
proteolytic activity as it reaches the cell surface. ADAM proteins are expressed at the cell surface as dimers, in
which the ADAM protein ectodomain is tightly packed and access to the catalytic site is blocked by close
proximity of the disintegrin and cysteine-rich domains, which exert an autoinhibitory function. Dimerization of
ADAM17 was shown to inhibit catalytic activity through the preferential binding of tissue inhibitor of
metalloproteinases 1 (TIMP1) and TIMP3 to the catalytic domain of dimeric ADAM proteins. Dimerization of
ADAM17 is controlled by a conserved threonine in position 735 (Thr735) of the cytoplasmic tail.
Phosphorylation of this threonine, by mitogen-activated protein (MAP) kinases, inhibits ADAM17 dimerization
and, consequently, TIMP3 is no longer able to inhibit the enzymatic activity of ADAM17. Interestingly, ADAM10
has a conserved threonine in position 719 (Thr719) of the cytoplasmic domain, although it remains to be
shown whether phosphorylation at this site controls ADAM10 dimerization and activity. At the cell surface, the
substrate specificity of ADAM10 may be controlled by six members of the tetraspanin C8 (TSPANC8) family,
which interact with the membrane-proximal regions of ADAM10 via a large extracellular loop. The activity of
ADAM17 is controlled by inactive rhomboid proteins (RHBDF1 and RHBDF2). The sheddase activity of
ADAM17 is triggered by a conformational change that occurs upon binding of membrane-expressed, negatively
charged phosphatidylserine by the positively charged, extracellular domains of ADAM17. Phosphatidylserine is
exposed in the outer leaflet of the cell membrane in cells undergoing apoptosis or in response to signalling
events through protein kinase C (PKC) isoforms and a rise in intracellular Ca2+ concentration.

33
Regulation by tetraspanins. The expression and function of ADAM10 are strongly
regulated by binding to tetraspanins, a family of 33 proteins that each have 4
transmembrane domains (Fig. 7). ADAM10 binds to 6 different tetraspanins of the C8
family (TSPAN5, TSPAN10, TSPAN14, TSPAN15, TSPAN17 and TSPAN33). Members of
the C8 family have eight conserved cysteines in the large extracellular loop, which bind
the stalk region and the cysteine-rich and disintegrin domains of ADAM10. The model
now emerging is that different tetraspanin C8 proteins may differentially regulate
ADAM10 localization, surface export, activity and, potentially even, substrate
preference. It was proposed therefore that ADAM10 could be thought of as six different
sheddases, depending on its association with the various tetraspanin molecules235. Very
little is currently known about the precise members of the tetraspanin C8 family that
control ADAM10 activity in cells of the immune system. However, the idea that the
conformation and substrate specificity of ADAM10 are regulated by a conformation-
dependent interaction between ADAM10 and a tetraspanin opens up the possibility that
small molecule inhibitors could be found that target cell-specific and substrate-specific
functions of ADAM proteins222, 236.

Regulation by rhomboid proteins. Despite the fact that ADAM10 and ADAM17 share
many substrates, the tetraspanin C8 family does not interact directly with ADAM17.
Rather, ADAM17 is regulated by rhomboid proteins, which have seven transmembrane
domains and usually carry out intramembrane proteolysis. The inactive rhomboid
proteins (RHBDF1 and RHBDF2) lack catalytic activity, but regulate the shedding of
sTNF by ADAM17. Current data suggest that RHBDF1 and RHBDF2 interact directly
with the cytoplasmic tail of ADAM17 and control its intracellular trafficking and
maturation237, 238. Deficiency of RHBDF2 in mice reduced the secretion of TNF in
response to lipopolysaccharide and bacterial infection, and these mice were more
susceptible to Listeria monocytogenes infection. The importance of rhomboid proteins in
the regulation of ADAM17 is highlighted by the fact that RHBDF1 and RHBDF2 double-
knockout mice die perinatally with developmental defects that are identical to those of
ADAM17-deficient mice.239 Recent studies have also suggested that RHBDF1 and
RHBDF2 may influence the substrate specificity of ADAM17, but the molecular
mechanism accounting for the effect is poorly understood240, 241. Little is known about
the cell-specific regulation of inactive rhomboid proteins in cells of the immune system.

In summary, while the list of potential ADAM substrates is continuously expanding, it is

now clear that the proteolytic activity of ADAM proteins in vivo is tightly controlled by
multiple regulatory mechanisms. From this section, immunologists should foremost
remember that a simple flow cytometric assessment of protein expression at the cell
membrane is insufficient to address whether or not ADAM10 or ADAM17 will be active,
or whether a particular substrate will be preferentially cleaved by the ADAM proteins.
However, in the past few years, several biologically relevant genetic models have
confirmed the role of ADAM proteases in different types of immune cell at well-defined
time points in the lifecycle of a cell. The importance of ADAM-mediated cleavage of

34
Notch receptors as a crucial cell-fate decision was soon recognized and is highlighted
here first. Next, the precise functions of ADAM proteins in T and B cells, innate
lymphocytes and myeloid cells are discussed separately.

2.2.2 Activation of Notch by ADAM proteins

Notch-mediated signalling controls many aspects of the development and function of
innate and adaptive immune cells. During transit to the cell surface, Notch proteins are
cleaved by a furin-like convertase at site 1 (S1), generating two non-covalently
associated subunits. The mature heterodimeric Notch receptor is normally held in an
‘off’ state that is resistant to further proteolysis until ligand binds. Studies in fruitflies
and mammals have led to the paradigm that Notch-mediated signalling is initiated by a
conformational change of the Notch heterodimer that is induced by ligand binding; this
conformational change renders extracellular Notch accessible to ectodomain shedding
by membrane-expressed ADAM10 or ADAM17 (REFS242-244) (Fig. 5). Although both
ADAM10 and ADAM17 have been shown to cleave Notch in vitro, accumulating evidence
from knockout mouse models indicates that ADAM10 is the relevant Notch sheddase in
vivo. Epsin-mediated endocytosis of the receptor-engaged Notch ligand by the signal-
sending cell (the cell that provides the Notch ligand) exerts a pulling force that exposes
the negative regulatory regions (NRRs) of Notch to ADAM10, which processes Notch at
S2 immediately external to the cell membrane245-247. The truncated receptor, known as
Notch external truncation (NEXT) is then a constitutive substrate for regulated
intramembrane proteolysis248 by the g-secretase complex. Cleavage of NEXT by g-
secretase at S3 in the transmembrane domain leads to the release of Notch intracellular
domain (NICD) from the cell membrane and its localization to the nucleus. In the
nucleus, NICD binds to RBPjκ and other transcriptional regulators to activate gene
expression.

This widely accepted model of Notch activation, in which ADAM-mediated cleavage

occurs at the cell surface, is known as canonical Notch signalling. Recently, this view of
ligand-induced extracellular proteolysis of Notch by ADAM proteins has been
challenged by one study that addressed the temporal localization of ADAM10 and Notch
1 after ligation by delta-like ligand 1 (DLL1) on transfected cell lines249. Ligand binding
led to the transendocytosis of Notch-engaged DLL1 in transfected cells, but also to the
endocytosis of dissociated Notch 1 in endosomes of Notch-expressing cell lines. The
endocytosis of Notch 1 was promoted by the E3 ubiquitin ligase deltex-like 4 (DTX4),
the human homologue of deltex, which is necessary for efficient Notch signaling in
Drosophila. ADAM10-containing vesicles were subsequently recruited to these Notch-
containing endocytic vesicles, and ADAM10-mediated cleavage of Notch 1 and
generation of NICD occurred in the fused vesicles rather than at the cell surface. In
support of the endocytic processing of Notch 1, ADAM10 surface staining in non-
permeabilized cells was not detected in this system249. Although the validity of this
model in vivo remains to be shown, it could potentially explain why some immune cells
that clearly depend on Notch signaling for activation do not express much ADAM10 or

35
ADAM17 at the cell surface. Regardless of the precise mechanism, ectodomain cleavage
by an ADAM protease is indispensable for Notch activation, which has a pivotal role in
the development and function of B, T and myeloid cells.

2.2.3 ADAM proteins in T cells

Through cleavage of Notch receptors, ADAM proteins can regulate cell-fate commitment
of T cells. In addition, ectodomain shedding of numerous other ADAM substrates occurs
at the surface of T cells, which generates soluble decoy receptors, growth factors and
cytokines, but also initiates intracellular signal transduction. ADAM proteins are
emerging as major regulators of T cell development and function.

Development. Notch 1 activation on thymocytes by delta-like ligand 4 (DLL4) on thymic

epithelial cells (TECs) has a crucial role in thymocyte development and T cell
commitment250. The strength of Notch-mediated signalling in developing thymocytes
determines the CD4 versus CD8 lineage choice, as well as the αβ versus γδ T cell fate
choice251-253. Few studies have reported the thymocyte stage-specific expression of
ADAM10 or ADAM17 by surface flow cytometry, which could be as a result of the
localization and function of ADAM proteins in endosomes. Expression of a dominant-
negative form of ADAM10 could suppress T cell development, and selective deficiency
of ADAM10 on thymocytes suppressed T cell commitment at the same stage of
development as has been observed for conditional Notch 1-deficient thymocytes254, 255.
These results suggest that ADAM10 is the major Notch-activating metalloproteinase in
thymocytes. However, it is not currently known how ADAM10 activity is regulated by
members of the tetraspanin C8 family in thymocytes, or how it intersects with pre-TCR
signalling. The recently published crystal structure of the ADAM10 ectodomain suggests
that binding by ligand-engaged Notch induces a conformational change in ADAM10 that
relieves autoinhibition by the membrane-proximal domains222.

ADAM17-deficient mice have a non-cell-autonomous defect in thymocyte development,

which led to the suggestion that ADAM17 on thymic stromal cells might be responsible
for the phenotype. However, specific deletion of ADAM17 in FOXN1-expressing TECs
did not affect normal T cell development, although the mRNA levels of expression of
autoimmune regulator (AIRE) by ADAM17-deficient TECs were reduced. Decreased
AIRE expression should, in theory, interfere with the negative selection of self-reactive
T cells, although this could not be shown in this model. How exactly ADAM17 controls
the expression level of AIRE remains to be elucidated 256. Deficiency of ADAM8 was
shown to lead to thymic hypercellularity, also in a non-cell-autonomous manner, which
was explained by the decreased responsiveness of ADAM8-deficient thymocytes to CC-
chemokine ligand 21 (CCL21) expressed in the thymic medulla257.

Proliferation. In peripheral T cells, Notch 1 ligation by delta-like or jagged ligands

expressed on dendritic cells (DCs) is necessary for T cell proliferation and the induction
of a T helper (TH) cell differentiation programme258-262. Several papers have shown that

36
Notch 1 and ADAM10 are recruited to the immunological synapse that forms between a
DC and an antigen-reactive peripheral T cell260, 263, 264 (Fig. 8). The T cell-specific
deletion of ADAM10 led to a severe reduction in antigen-induced T cell proliferation,
whereas T cell cytokine production was less affected. Therefore, the phenotype of these
mice closely resembled that of mice with T cell-selective Notch 1 deficiency260. In
combination with TCR signalling, early Notch-mediated signalling induces expression of
the cell cycle regulator MYC, which is important for inducing a metabolic switch in T
cells to allow full proliferation. Furthermore, tonic Notch signalling is essential to
maintain the glycolytic metabolism that is necessary for the survival of memory CD4+ T
cells. It has also been shown that ligand-independent activation of Notch can occur
during T cell activation in the absence of APCs260. In this case, Notch is activated through
diverse mechanisms during endosomal trafficking or ubiquitylation265. However, the
precise molecular details of how signalling through the TCR and costimulatory or
checkpoint molecules intersects with metabolism and ADAM-mediated Notch activation
in peripheral T cells remain largely unknown (Fig. 8). As for many aspects of the
formation of a mature immunological synapse, TCR- and CD28-mediated reorganization
of the actin cytoskeleton, through NCK, VAV1 and PKCθ, is involved in bringing ADAM10
to the center of the synapse260, 266. Regulated exposure of phosphatidylserine on the
outer leaflet of the plasma membrane at the synapse could also help in activating ADAM
sheddase activity224, 225.

Activation and survival. Lymphocyte activating gene 3 (LAG3; also known as CD223) is
an inhibitory receptor expressed by activated T cells that is cleaved by ADAM10 upon
TCR ligation. In the later stages of T cell activation, PKCθ also induces ADAM17
enzymatic activity and further LAG3 cleavage267. A subset of regulatory T (Treg) cells
exert their suppressive effects through LAG3, but it is currently unknown if ADAM-
mediated cleavage of LAG3 is differentially regulated in Treg cells compared with other
types of T cell. In tumour-infiltrating CD8+ T cells, sustained antigen stimulation leads to
persistent LAG3 expression and a state of T cell exhaustion 268. It will be interesting to
determine whether the ADAM-mediated cleavage of LAG3 during T cell activation
depends on a specific conformational state of the ADAM proteins or interaction with
specific members of the tetraspanin C8 family222, as this could lead to the development
of selective agonists of LAG3 cleavage to function as checkpoint inhibitors for cancer
therapy269. Along similar lines, T cell immunoglobulin and mucin domain-containing 3
(TIM3; also known as HAVCR2) is a marker for exhausted T cells that suppresses T cell
activation. Cleavage of TIM3 on CD8+ T cells is regulated by ADAM10, and stimulating
ADAM10 might enhance T cell effector functions. It is unclear at present if soluble LAG3
and soluble TIM3 released by ADAM10-mediated cleavage also function as antagonists
that reverse T cell exhaustion, or just as markers for the presence of exhausted T
cells270.

Many TNF family members are expressed on T cells and shed by ADAM proteins.
Activation-induced cell death in T cells is controlled by FAS ligand (FASL; also known as

37
Figure 8 | Role of ADAM10 in T cells. During early interactions between an antigen-presenting cell (APC) and
a T cell in a nascent immunological synapse, NCK1- and VAV1-induced alterations of the actin cytoskeleton
cause Notch 1 and delta-like ligand 1 (DLL1) to move to the APC–T cell contact zones and the central part of the
supramolecular activation cluster (cSMAC). The engaged T cell receptor (TCR)–CD3–CD28 complex induces
membrane expression of diacylglycerol (DAG) and activation of the serine/threonine kinase protein kinase Cθ
(PKCθ), which is necessary for the localization of ADAM10 (a disintegrin and metalloproteinase 10) and Notch
1 to the cSMAC. PKCθ phosphorylates and inactivates the negative regulator of F-actin coronin 1A (COR1A;
which is an antagonist of the ARP2/3 complex) in the cSMAC area, leading to actin polymerization and
cytoskeletal rearrangements. Cleavage of Notch 1 by ADAM10 and γ-secretase in the cSMAC of a mature
immunological synapse leads to the transcription of Notch target genes. In the thymus, this stimulates T cell
fate commitment, whereas in mature T cells, Notch signals are crucial for antigen-induced proliferation and
metabolic reprogramming.

CD95L), which interacts with FAS (also known as CD95) to induce apoptosis. FASL is
cleaved by ADAM10 and ADAM17 in mouse and human T cells upon TCR ligation, and
this could control T cell survival and effector functions, with soluble FASL trimers often
functioning as apoptosis inhibitors271,272. Another TNF family member CD154 (also
known as CD40L) is expressed on the cell surface of activated T cells, DCs, mast cells,
platelets and stromal cells, and is shed by ADAM10 and ADAM17 (REF.273). Soluble

38
CD154 prolongs B cell activation, induces the production of proinflammatory mediators,
increases the number of Treg cells and induces the immunosuppressive indoleamine 2-3-
dioxygenase (IDO) pathway of tryptophan catabolism, which has a potential negative
impact in diseases such as inflammatory bowel disease, cancer and HIV infection274. The
costimulatory TNF family member CD30 is also shed from activated T cells by ADAM10
and ADAM17 in an interferon-g (IFNγ)- and IL-2-dependent manner, although the
resulting effect on T cell function is not known. Soluble CD30 can also be used as a
biomarker in transplant rejection and Hodgkin lymphoma275, 276.

Migration. Effector and memory T cells express the adhesion molecule CD44, a well-
known substrate of ADAM10 that in its uncleaved form mediates T cell binding to
inflamed vascular endothelium and to the extracellular matrix component hyaluronic
acid. The membrane stub of CD44 that is left behind after ADAM10-mediated cleavage is
subsequently cleaved by γ-secretase, leading to release of the CD44 intracellular domain
(CD44-ICD). Although nuclear CD44-ICD has been shown to have transforming potential
in cancer, the precise role of CD44-ICD in memory lymphocytes is not known277.
However, a CD44-ICD-binding DNA consensus sequence has been identified and early
responsive genes to CD44-ICD encode crucial enzymes in the glycolytic pathway278. It is
therefore tempting to speculate that CD44-ICD contributes to a metabolic switch in
effector T cells that is required for T cell division, which might help to explain the
proliferation defect of T cells lacking ADAM10 (REF.260). Whereas CD44 cleavage in T
cells affects migration in a cell-intrinsic manner, lymphocyte migration is also affected
by loss of ADAM protein function on stromal cells. Endothelial expression of ADAM10,
TSPAN5 and TSPAN17 has been proven crucial for the extravasation of human
lymphocytes across endothelial cells. Upon lymphocyte arrest, endothelial ADAM10
cleaves vascular endothelial cell cadherin (VE-cadherin) and the transmembrane
chemokines CX3C-chemokine ligand 1 (CX3CL1) and CXC-chemokine ligand 16
(CXCL16), which promotes transendothelial migration235, 279, 280.

Differentiation. The activity of ADAM proteins in T cells also contributes to TH cell

differentiation. This could occur in a cell-intrinsic manner through the effects of ADAM
protein activity on Notch activation, which is known to directly affect TH cell skewing281,
282. Furthermore, in mice and humans, IL-23R — a receptor that controls T 17 cell
H
differentiation and proliferation — is cleaved in vitro by ADAM17 after PMA
stimulation, and to a lesser extent constitutively by ADAM10 (REF.283). It has also been
shown that ADAM17 can cleave IFNγ, the prototypical polarizing and effector cytokine
of type 1 immune responses, which could lead to non-cell-intrinsic effects of ADAM
proteins on TH cell polarization284.

2.2.4 ADAM10 in B cells

Immature B cells develop continuously in the bone marrow and further mature into
follicular B cells and marginal zone B (MZB) cells in the spleen. Follicular B cells have a
broad range of specificities, recirculate and give rise to germinal centre (GC) B cells that

39
undergo somatic hypermutation in a T cell-dependent manner. MZB cells shuttle
continuously between the marginal zone and follicles of the spleen and produce
antibodies specific for encapsulated bacteria and polysaccharide antigens in a T cell-
independent manner35, 36, 58.

MZB cell commitment. It is widely accepted that the MZB cell fate requires triggering of
Notch 2 on developing B cells by DLL1 expressed by marginal zone reticular cells
(MRCs) or red pulp sinusoidal endothelial cells49, 50, 54, 56. Notch 2 cleavage in B cells
requires ADAM10, as mice lacking ADAM10 selectively in B cells completely lack MZB
cells but have normal and slightly elevated numbers of B1 B cells and follicular B cells,
respectively53. However, it is not clear how and at what developmental stage the
expression of ADAM10 is induced. It has been suggested that signals through the BCR
might render MZB precursor cells sensitive to Notch2 signals, but whether this involves
ADAM10 remains to be determined35.

Function of follicular B cells. In addition to the clear effects of ADAM10 in MZB cell
development, it can also affect follicular B cells, although to a lesser extent. ADAM10 is
required to cleave CD23 from the surface of follicular B cells, leading to the generation
of soluble CD23 (sCD23)53, 285. The cleavage of CD23 seems to occur in an endosomal
compartment, after internalization of CD23 and routing of CD23 to exosomes286.
Understanding this pathway has important implications for allergy as CD23 is a low-
affinity IgE receptor and cleavage of CD23 increases the synthesis of IgE.

ADAM10 levels are significantly increased on GC B cells compared with naive B cells,
and mice lacking ADAM10 selectively in B cells had reduced primary and secondary
antibody responses after T cell-dependent immunization. These animals also had
impaired GC formation, fewer T follicular helper (TFH) cells, decreased follicular DC
(FDC) networks, and altered chemokine expression and lymphoid architecture in
draining lymph nodes287, 288. Although some of these effects might have been caused by
a compensatory upregulation of ADAM17 expression and, therefore, an increased
concentration of sTNF, they are more likely to result from defective TFH cell help. In
support of this, the same group reported that ADAM10 deficiency in B cells led to their
persistent upregulation of expression of the ADAM10 substrate ICOS ligand (ICOSL),
which resulted in ICOS internalization by developing TFH cells and a loss of these
cells289.

Plasma and memory B cell responses. ADAM10 also controls plasma cell differentiation
beyond the GC phase of the response, by affecting expression of the transcription
factors B lymphocyte-induced maturation protein 1 (BLIMP1; also known as PRDM1),
X-box binding protein 1 (XBP1) and IFN regulatory factor 4 (IRF4)290, but it is unclear at
present how this occurs. The cytokine B cell activating factor (BAFF; also known as
TNFSF13B) controls survival of memory B cells by binding three receptors
(TNFRSF13B, BAFFR (also known as TNFRSF13C) and TNFRSF17). Inhibition of

40
ADAM10 augments BAFF-dependent survival of primary human B cells, whereas
inhibition of ADAM17 increases BAFFR expression levels on GC B cells but does not
alter their survival291.

In summary, ADAM10-mediated cleavage of Notch 2 is indispensable for MZB cell

development. In addition, the identification of CD23, ICOSL and BAFFR as ADAM10
substrates has shown that ADAM proteins also have important roles in the homeostasis
and function of other B cell populations.

2.2.5 ADAM proteins in innate lymphoid cells

The precise role of ADAM proteins in innate lymphoid cells (ILCs, see chapter 4) is less
well studied, although there are clear roles for Notch in the development of different
subsets of ILCs, which potentially implies a role for ADAM proteins292. In natural killer
(NK) cells, ADAM10 and ADAM17 can cleave the high-affinity membrane IgG receptor
FcRγIII (also known as CD16), which is involved in antibody-dependent cytotoxicity.
Treating NK cells with an ADAM17 inhibitor increased their ability to kill breast cancer
cells treated with the monoclonal antibody trastuzumab293. NK cells are activated by a
series of receptors that are triggered by ligands on virus-infected or transformed cells
or cells undergoing genotoxic stress. Engagement of the NKG2D receptor on NK cells
and γδ T cells by its ligands MHC class I polypeptide-related sequence A (MICA), MICB
and the UL16 binding proteins (ULBPs) activates NK cell effector function. All known
NKG2D ligands are substrates of ADAM10 and ADAM17 (REFS294, 295). Not only does
ectodomain shedding due to ADAM-mediated cleavage result in decreased surface
expression of these ligands, but also the soluble NKG2D ligand that is released can
mediate NKG2D receptor downregulation on effector lymphocytes296. Consequently,
increased NKG2D ligand shedding is a mechanism by which tumour cells can
circumvent immunosurveillance. It is increasingly evident that NKG2D-mediated
cytotoxicity is hampered in a diverse array of lymphoid malignancies, and a high
concentration of soluble ADAM substrates in patient serum often correlates with a
worse prognosis297. Highly selective inhibition of ADAM10 sensitized malignant
Hodgkin lymphoma cell lines to NK cell-mediated killing in vitro, which suggests that
ADAM10 inhibitors might have therapeutic potential298, 299.

2.2.6 ADAM proteins in myeloid cells

A key feature of acute inflammation is the recruitment and tissue infiltration of innate
myeloid cells such as macrophages, DCs and neutrophils. Many soluble and membrane-
bound molecules that regulate the activation and/or migration of these cells have been
identified as ADAM substrates. In particular, ADAM10 and ADAM17 are essential
shedding enzymes that control transendothelial migration, cytokine production and
intracellular signalling cascades in myeloid cells.

Myelopoiesis. During an innate immune response to pathogens, large numbers of

neutrophils and monocytes are recruited to inflamed tissues, and this increased

41
demand needs to be supported by increased output from the bone marrow through
emergency myelopoiesis. Loss of ADAM function or loss of key components of the Notch
signalling pathway alters the development of myeloid cells and promotes uncontrolled
myeloproliferation. Conditional knockout of Adam10 in mice resulted in
myeloproliferative disease characterized by extramedullary haematopoiesis and the
accumulation of immature granulocytes300, 301. These effects are both intrinsic and
extrinsic to the haematopoietic system, and are driven by increased circulating levels of
granulocyte colony-stimulating factor (GCSF) and thymic stromal lymphopoietin
(TSLP). It is currently unknown if the long-term blockade of ADAM protein function for
other disease indications such as cancer or inflammatory disease could lead to
unwanted side effects on haematopoiesis.

Inflammatory cytokine production. In neutrophils and macrophages, ADAM17 controls

the cleavage of membrane-bound TNF into pro-inflammatory sTNF, as well as cleavage
of TNFR into sTNFR, a process that is controlled by the inactive rhomboid proteins
(RHBDF1 and RHBDF2) and Polo-like kinases232, 302. Inhibiting ADAM17
metalloproteinase activity using a peptide variant of the autoinhibitory prodomain was
able to reduce inflammation in models of inflammatory bowel disease and rheumatoid
arthritis233. ADAM17 activity was also associated with metabolic inflammation, and
haploinsufficiency of ADAM17 protected against diet-induced obesity and insulin
resistance303. Another prominent substrate of ADAM proteins on myeloid cells is the IL-
6 receptor (IL-6R), which is shed mainly by ADAM17; soluble IL-6R binds to GP130 to
mediate alternative IL-6R trans-signalling and thus render other cells responsive to IL-6
304.

Migration. Neutrophils lacking ADAM17 showed enhanced recruitment to the

peritoneal cavity in a sepsis model owing to higher levels of expression of L-selectin —a
substrate for ADAM proteins — which mediates the rolling of neutrophils on inflamed
blood vessels305. In a model of recruitment of neutrophils to the alveolar space,
ADAM10 but not ADAM17 was required for neutrophil influx306. During bacterial
infections, it has emerged that ADAM proteins on myeloid cells can function as direct
receptors for bacterial toxins and thus contribute importantly to the early phases of
infection.

Phagocytosis. Macrophages are versatile cells that reside in tissues and carry out
essential tasks in tissue homeostasis and host defence. Macrophage activation is tightly
regulated, and ADAM proteins contribute to this process. Tissue-resident macrophages
ingest apoptotic cells in a process known as efferocytosis that depends on the
recognition of ‘eat me’ signals by receptors such as CD36 and TIM3; this often leads to
the induction of an anti-inflammatory phenotype in the macrophages. Macrophages
lacking ADAM17 had increased surface expression of CD36 and were better able to
ingest phosphatidylserine-containing liposomes and apoptotic cell corpses, which led to
accelerated clearance of inflammatory cells in a model of peritonitis307. TIM3, which

42
binds to galectin 9 on apoptotic cells, is shed by ADAM10 and ADAM17 (REF.308).
Soluble TIM3 in the serum of patients with systemic lupus erythematosus was shown to
inhibit the uptake of apoptotic cells by macrophages, contributing to the induction of
autoimmunity309. Signal regulatory protein-α (SIRPα) is expressed abundantly by cells
of the monocyte lineage. Ligation of SIRPα by the broadly expressed transmembrane
protein CD47 represents a ‘don’t eat me’ signal to the immune system that results in the
inhibition of nuclear factor-kB signalling in macrophages. During inflammation, the
cleavage of SIRPα by ADAM10 could lead to loss of inhibition of macrophages and the
induction of autoimmunity310. In an atherosclerosis model, ADAM10 deficiency in
myeloid cells resulted in their impaired migration and the acquisition of an anti-
inflammatory phenotype by macrophages, which increased plaque stability311.
Triggering receptors expressed by myeloid cells (TREMs) include activating and
inhibitory isoforms. TREM1 and TREM2 activate myeloid cells by signalling through the
adaptor protein DAP12. Remarkably, in mice, TREM2 deficiency leads to a severe
disease characterized by bone cysts and dementia. In humans, genetic variants of
TREM2 confer an altered risk of Alzheimer disease; these variants either increase or
decrease shedding of TREM2 by ADAM10 or ADAM17 on brain microglia312, 313.

Antigen-presenting functions. A key function of DCs and macrophages is to present

antigen to T cells, after uptake by phagocytosis, and subsequently to induce a polarized
T cell-mediated immune response. The specific deficiency of ADAM10 in CD11c-
expressing cells was shown to reduce the potential of lung DCs to prime for allergic
airway inflammation, which was restored by overexpressing a constitutively active
form of Notch 1 (REF.314). However, it is also possible that lack of ADAM10 affects the
terminal differentiation of DCs, which depends on signalling through Notch 2 (REF.157).
It was shown that Notch 2 blockade or lack of Notch 2 ligation affects the expression of
CC-chemokine receptor 7 (CCR7), which is the main chemokine receptor controlling the
migration of DCs towards lymph nodes via the afferent lymph315. After antigen uptake,
DCs reach the lymph nodes via afferent lymphatics that express the CD44-like molecule
lymphatic vessel endothelial hyaluronic acid receptor 1 (LYVE1), which is yet another
substrate of ADAM10 and ADAM17 (REF.316). Therefore, multiple pathways might
explain why ADAM10-deficient DCs are less well able to prime for T cell responses.

2.2.7 Conclusions
The list of substrates of ADAM sheddase activity in the immune system is long and, not
surprisingly, many mouse models and inhibitor studies have provided genetic and
biochemical evidence that ADAM10 and ADAM17 have crucial roles in the functioning of
almost every type of immune cell. From these studies, it has emerged that the role of
ADAM proteins is highly cell-type and context specific; sometimes they function mainly
by activating Notch signalling pathways, whereas at other times they function by
shedding key adhesion molecules, cytokines or receptors from the surface of cells. Many
questions remain about how ADAM proteins influence specific cell types in a context-
dependent manner.

43
3. THOUSAND-AND-ONE KINASE 3

3.1 Ste20 kinases

Thousand-and-one kinase 3 is a member of the family of Ste20-like kinases. Sterile 20
was discovered in Saccharomyces cerevisiae as a factor that controlled mating responses
and osmoregulation in hypertonic stress conditions317. Mechanistically, the Ste20
protein constituted a crucial serine/threonine kinase that linked signals from G-protein
coupled receptors to the mitogen-activated protein kinase (MAPK) cascade318, 319. In
mammals, 28 Ste20-like kinases have been identified that display similarity with their
yeast homolog. Many of these proteins were confirmed to act as genuine MAP kinase
kinase kinase kinases (MAP4K), initiating a phosphorylation cascade that eventually
activates MAP kinases like p38, c-Jun N-terminal kinase (JNK) and extracellular signal
regulated kinase (ERK). Based on sequence alignment, Ste20-like proteins can be
divided into 2 groups with different subfamilies: p21-activated kinases (PAK-I and PAK-
II) and germinal center kinases (GCK I-VIII, see Fig. 9)320. In contrast to their conserved
kinase domain, the structure of the noncatalytic region is highly diverse321. In PAKs, the
kinase domain is located at the C-terminus, while their aminoterminal domain mediates
binding to small GTPases like Cdc42 and Rac. In vitro and in vivo studies have
established that PAKs are crucial mediators of the functions of these Rho-type GTPases
in controlling actin reorganization and motility322. GCKs, in contrast, have an N-terminal
kinase domain and a C-terminal regulatory region. Whereas some GCKs have been
associated with the induction of the c-Jun N-terminal kinase/stress-activated protein
kinase (JNK/SAPK) cascade, their specific functions remain largely unknown323.

44
Figure 9 | Phylogenetic tree and domain structure of the GCK and PAK subfamilies of STE20-like
kinases from humans and C. elegans. STE20-related kinases were grouped based on protein alignment.
p21-activated kinases (PAKs) contain two structurally similar subfamilies, PAK-I and PAK-II. Within the
germinal center kinase subfamily (GCK), eight subfamilies can be discerned. Conserved domains are
indicated by colored boxes, with sequence alignment results indicated below. JIK (Jnk-inhibitory kinase)
equals TAOK3. For alternative names of these proteins, we refer to the HUGO website
(https://www.genenames.org). Reprinted from Strange et al.320, with permission.

3.2 The TAO-family

GCK-VIII subfamily members are phylogenetically more distant from other GCKs. This
family is also known as the TAO-family, and comprises 3 paralogs in mammals: TAOK1
(also known as MAP3K16, PSK2 or MARKK), TAOK2 (MAP3K17 or PSK1) and TAOK3
(MAP3K18, JIK or DPK)324-326. All three kinases display many similarities including an
aminoterminal kinase domain, a serine-rich region, and a coiled-coil configuration
within the C-terminus. However, they differ in tissue distribution, subcellular
localization and have demonstrated structural and functional differences324.

TAOK1 was first cloned as a rat homolog of Ste20 that contained one thousand and one
amino acids, explaining the name327. While it is expressed in many tissues, the highest
levels were found in brain, heart and lung324. Similarly, TAOK2 was found to be widely
expressed in rats, with predominance in brain and testes324, 328. Rat TAOK1 and TAOK2
were both able to activate p38 in vitro by interacting with MEKK3 or MEKK3 and
MEKK6, respectively327. Thus, TAOK1 and TAOK2 function as MAP3Ks, as opposed to
most other members from the Ste20 family of kinases. TAOK2 was also found to
stabilize microtubules, independent of its catalytic domain329. In 1999, Tassi et al.
cloned a human Ste20 homolog that did not activate MAP kinases, but instead inhibited
the basal activity of JNK/SAPK in response to human epidermal growth factor (EGF) in
vitro. The kinase was named JNK/SAPK-inhibitory kinase (JIK), and was later found to
be identical to human dendritic cell protein kinase (DPK), a kinase identified in human
dendritic cells323, 325, 330. Phylogenetic analysis later revealed that these kinases form a
third TAOK paralog. TAOK3 transcripts were found predominantly in hematopoietic
organs (blood leukocytes, spleen, thymus), but also in pancreas, prostate and testis324.

It should be mentioned that the effects of TAOK3 activation on JNK/SAPK and p38
remain controversial, with opposing effects reported by independent investigators324-
326, 330-332. Even less is known about the upstream activators of TAO kinases. Apart from

the aforementioned activation by EGF, TAOK3 was activated in vitro by the DNA
damage-sensing kinase Ataxia telangiectasia mutated (ATM) upon ultraviolet
radiation326. Recently, one group has reported a role for TAOK3 downstream of T-cell
receptor signaling. TAOK3 positively regulated TCR signals by inhibiting the negative
feedback loop mediated by the phosphatase SHP-1 in vitro333. Preliminary data from our
lab confirmed a cell-intrinsic role of TAOK3 in T cells in vivo (Vandersarren et al.,
unpublished results), although the mechanism remains unclear.

45
4. INNATE LYMPHOID CELLS

Innate lymphoid cells (ILCs) constitute a heterogeneous group of immune cells with
essential roles in tissue homeostasis and early defense against pathogens. Like other
cellular components of innate immunity (myeloid cells, MAIT cells, NKT and γδ T cells),
their function is most crucial in the early phases of infection, before an adaptive
immune response has developed. However, the function of ILCs appears to extend
beyond classical immunology, as roles have been demonstrated in metabolic
homeostasis and tissue remodeling. Ontogenically, ILCs arise from the common
lymphoid precursor in bone marrow. However, in contrast to adaptive lymphocytes,
ILCs lack somatic rearrangements of surface receptor genes and use RAG-independent,
germline-encoded receptors to identify environmental threats. Nevertheless, many
similarities exist in the transcriptional programs and effector functions of ILCs and T
cells, leading to the view of ILCs as innate counterparts of T lymphocytes. Classically,
natural killer (NK) cells have been denoted as ‘killer’ ILCs, while ‘helper’ ILCs are further
subdivided into group 1, 2 and 3 ILCs based on their transcriptional requirements and
cytokine expression patterns. Whereas ILCs have been identified across almost all
lymphoid and non-lymphoid tissues, they are especially well represented in barrier
tissues, including the skin, gastro-intestinal tract, lung, and urogenital tract. Helper ILCs
are regarded as tissue-resident cells that renew locally, although some exceptions have
been noted under inflammatory conditions334, 335. Adult bone marrow contains
precursor cells that can give rise to ILCs, but it remains unclear whether and under
what circumstances renewal by circulating BM-derived precursors is needed336-340.

4.1 ILC1s
Type 1 ILCs (ILC1s) are hard to distinguish from NK cells phenotypically, but arise
through a clearly distinct developmental trajectory. Both cell types express activating
NK cells receptors such as NK1.1 and NKp46, as well as the transcription factor T-bet
(TBX21). NK cells and ILC1s readily secrete the cytokines IFNγ and TNF in response to
IL-12, IL-15 and IL-18, and are important in the defense against intracellular pathogens.
In contrast to NK cells, ILC1s express CD127 (IL-7Rα) but lack activating or inhibitory
receptors of the Ly49 family. NK cells can further be distinguished by their requirement
for the transcription factor Eomesodermin (Eomes), and the production of perforin and
granzyme B as hallmarks of their cytotoxic potential. Unlike other ILCs, NK cells
circulate and are constantly renewed from the bone marrow. Cytokine production by
ILC1s was shown to be crucial in the early defense against infections with viruses,
Toxoplasma gondii and Clostridium difficile336-338, 341-343.

4.2 ILC2s
Group 2 ILCs (ILC2s) are tissue-resident cells that are capable of producing IL-4, IL-5
and IL-13, eliciting a Th2 response. Predominantly located at mucosal surfaces, their
function has been implicated in the defense against helminthic infection and in allergic
airway disease. ILC2s depend on the transcription factor GATA-3 and respond to the

46
cytokines IL-25, IL-33 and thymic stromal lymphopoietin (TSLP). ILC2s from different
tissues are phenotypically distinct, and display preferential responses to different
cytokines338, 339. For example, while lung ILC2s express the IL-33 receptor ST2 but only
low levels of the IL-25 receptor IL17-RB, ILC2s in the small intestine and skin do not
express ST2338. Apart from its responses to parasitic infections and allergens, ILC2s also
play crucial roles in tissue repair after mucosal damage344. Additionally, ILC2s influence
metabolic homeostasis through their regulation of eosinophil activation and of
alternatively activated macrophages that influence insulin sensitivity336, 337. Further
supporting their metabolic role, ILC2s also play a crucial role in thermogenesis through
a process termed “beiging” of adipose tissue345, 346.

4.3 ILC3s
The requirement for the transcription factor RORγt sets type 3 ILCs (ILC3s) apart from
other innate lymphoid cells347. RORγt controls the development of ILC3s and is
continuously required to maintain the ILC3 identity. During fetal life, a population of
RORγt-dependent CD4+ CD3- lymphocytes instructs the formation of lymphoid organs
by providing lymphotoxin α1β2 to stromal cells348. Accordingly, these cells were named
lymphoid-tissue inducer (LTi) cells. In the mouse embryo, LTis can be identified as early
as embryonic day (E) 12.5-13.5. Through their actions on the stromal lymphotoxin β
receptor (LTβR), LTis induce the production of the chemokines CXCL13 and CCL19, and
initiate the structural organisation of secondary lymphoid tissues. In absence of LTis or
lymphotoxin, lymph nodes are absent, as are Peyer’s patches. The spleen still develops,
but displays a disturbed microanatomy with lack of segregating B and T cell zones348.
Similarly, ILC3s that correspond to their fetal LTi counterparts are also involved in the
development and maintenance of lymphoid structures after birth, such as intestinal
cryptopatches and isolated lymphoid follicles (ILFs)349.

In adult mice, RORγt+ ILCs are classically subdivided in NKp46- CD4+/- CCR6+ LTi-like
ILC3s and NKp46+ CD4- CCR6- ILC3s. However, reports of additional heterogeneity
within ILC3s and plasticity between ILC1s and ILC3s have highlighted a more complex
reality, where environmental cues might direct dynamic changes between several
transcriptional states350. Nevertheless, a conserved ILC3 signature can be recognized in
mice and humans351, 352. Similar to ILC2s, type 3 ILCs (ILC3s) mainly populate mucosal
tissues. In the gut, they reside in the lamina propria, but also ILFs and cryptopatches. In
the immune response against intestinal commensals and attaching and effacing bacteria
like Citrobacter rodentium, ILC3s maintain the integrity and function of the intestinal
epithelium353, 354. In response to IL-23, ILC3s are activated in a STAT3-dependent
manner and secrete IL-22, a cytokine that triggers goblet cell hyperplasia and the
production of antimicrobial peptides such as RegIII proteins, S100a8 and S100a9 by
epithelial cells339, 350, 353, 354.

In addition to RORγt, the transcriptional program of ILC3s also depends on

environmental factors such as dietary metabolites. Maternal retinoic acid is required for

47
the development of LTi cells in embryos, and vitamin A deprivation depleted intestinal
ILC3s in adult mice355-357. Similarly, ligation of the aryl hydrocarbon receptor in ILC3s
by aromatic hydrocarbons, derived from food or microflora, was demonstrated to be
crucially required for ILC3 development throughout life358-361. Notably, it has been
described that some ILC3s may downregulate RORγt, upregulate the transcription factor
Tbet and acquire ILC1 functions362. In conclusion, ILC3s constitute a heterogeneous
group of cells that controls lymphoid organogenesis, but also exerts effector functions
against various pathogens. Furthermore, ILC3 development is closely connected with
environmental stimuli, and accumulating evidence supports bidirectional interactions
between ILC3s and the microflora.

5. LYMPHOTOXIN

The name lymphotoxin was originally attributed to a lymphocyte-derived factor that

mediated killing of neoplastic cell lines in vitro, resembling the activity of tumor
necrosis factor (TNF)363, 364. This led to the cloning of lymphotoxin (LT) and TNF, which
appeared to be structurally very similar, in two seminal 1984 papers365, 366. It was later
discovered that LT existed in two structural forms: a soluble form and a membrane-
bound form that complexed with another protein. That protein appeared to be yet
another TNF family member and was named LTβ, while the original lymphotoxin was
renamed LTα367, 368. The soluble form was identified as an LTα3 homotrimer that binds
the same receptors (TNFR1 and TNFR2) as TNF. Due to its many structural and
functional similarities, LTα3 was even named TNFβ for a short time. In contrast, the
membrane bound form of LT is an α1β2 heterotrimer that binds the lymphotoxin β
receptor (LTβR), another member of the TNF-receptor superfamily369.

5.1 Lymphotoxin signaling in lymphoid tissue development

The first evidence for a role of LT in the immune system came from mice that were
deficient in LTα. These mice completely lacked all lymph nodes and Peyer’s patches and
displayed a disturbed microarchitecture of the splenic white pulp370. It is now
commonly accepted that the LTα1β2 - LTβR axis governs the fetal development of
secondary lymphoid organs. This is based on the fact that the phenotype of LTβ- or
LTβR-deficient mice highly resembled that of LTα knockout mice, whereas deficiency in
TNFR1 and TNFR2 yielded much more subtle effects371, 372. In further support,
treatment of pregnant mice with an inhibitory LTβR-Ig fusion protein abrogated lymph
node development in embryos373. However, careful comparison of multiple single and
double knockout mice has demonstrated that signaling through TNFR1 (by LTα3 or
TNF) might contribute to some features, such as Peyer’s patch development or the
formation of FDC networks and germinal centers374-376.

Though LTα1β2 is highly expressed by B and T lymphocytes, it was discovered that

secondary lymphoid organ development was instructed by LTα1β2 expressed on LTi

48
cells. Indeed, ILC3-deficient mice fail to develop lymph nodes, Peyer’s patches or a
normal B-T segregation in the white pulp348, 377, 378. LTis were found to engage the LTβR
expressed on mesenchymal lymphoid tissue organizer (LTo) cells. This induced the
expression of adhesion molecules like VCAM1 and MadCAM1, but also of CXCL13, CCL19
and CCL21 by LTo cells, thus organizing B and T cell zones379. Downstream, stromal
LTβR signals activate NFκB signaling, predominantly via the alternative (NFκB2)
pathway. In line, mice that carry a mutation in the gene encoding NFκB inducing kinase
(NIK) fail to transmit LTβR signals and lack SLOs380, 381. The same LTα1β2 - LTβR
pathway also commands the development of tertiary lymphoid organs, and therefore
forms an attractive therapeutic target for autoimmune diseases and cancer382.

It should be noted that the LTβR can also bind another member of the TNF family, LT-
related inducible ligand that competes for glycoprotein D binding to herpesvirus entry
mediator on T cells (LIGHT). LIGHT also engages the herpesvirus entry mediator
(HVEM), a pathway that confers co-stimulatory signals to T cells (Fig. 10) 383, 384. In
addition, the soluble decoy receptor 3 (DCR3) also binds LIGHT. DCR3 expression is
increased in tumors, indicating that this protein might function as a decoy receptor to
escape immune surveillance385. Despite the clear role of the LIGHT-HVEM axis in T cell
biology, LIGHT does not seem to play an important role in lymphoid tissue
development386, 387.

49
FIGURE 10 | Tumor necrosis factor (TNF) and lymphotoxin α3 (LTα3) can induce signaling through
the two TNF receptors, TNFR55 (= TNFR1) and TNFR75 (= TNFR2). The LTβ receptor (LTβR) binds
two ligands: the LTα1β2 heterotrimer and homotrimeric LIGHT. LIGHT also binds to two additional
TNF family receptors: herpesvirus entry mediator (HVEM) and the soluble decoy receptor 3 (DCR3).
Reprinted from Gommerman and Browning385, with permission.

5.2 Roles beyond lymphoid organogenesis

The advent of Cre-Lox technology, pharmacological LTβR-inhibitors and the use of bone
marrow chimeric mice has allowed the identification of LT-dependent processes that
were not due to impaired SLO development. Indeed, LTα1β2 exerts many functions in
the adult immune system. The development of isolated lymphoid follicles and colonic
cryptopatches in adult mice requires LT, as does tertiary lymphoid organ formation382,
388. Formation of germinal centers is abrogated in mice lacking components of the LT

pathway374. Furthermore, many stromal cells in SLOs display a continued requirement

for homeostatic LT signals. B-cell derived LT crucially maintains FDCs and induces
CXCL13, establishing a positive feedback circuit. For this reason, postnatal blockade of
LT signaling abrogates normal follicular architecture and germinal center formation374,
389-392. Similarly, LT-LTβR signals instructed both the development and adult

maintenance of MadCAM+ endothelial cells of the marginal sinus393.

B cell-derived LT was also required for the homeostasis and antiviral responses of
subcapsular macrophages in the lymph node394. Lymphotoxin on activated antigen-
specific T cells was reported to be essential to license CD8+ T cell priming by dendritic
cells395. As mentioned above, lymphotoxin β receptor signals are crucial for the
homeostasis of CD4+ dendritic cells in the spleen126. On the other hand, dendritic cells
also express LT themselves, with crucial roles in controlling the expansion and survival
of reticular stromal cells in inflamed lymph nodes396.

Even in the adult, ILC3 derived LTα1β2 remains crucial for immune homeostasis. In the
intestine, ILC3-derived LT is required for IgA responses through actions on dendritic
cells in the lamina propria397, 398. Lymphotoxin signaling proved crucial for protective
immune responses against Citrobacter rodentium, whereby both ILC-DC interactions
that lead to IL-22 production, and epithelial-expressed LTβR seem to contribute399-401.
Another study reported the onset of splenomegaly and systemic inflammation in
absence of ILC3-derived LTα1β2402. Additionally, ILC3s controlled normal NK cell
development in the bone marrow in a LT-dependent manner403. Different subsets of
unconventional T cells, such as γδ T cells and invariant NKT cells, also require LTβR
signals for their development, probably through effects on stromal cells404, 405. Finally,
regulatory T cells (Tregs) were shown to employ the LT-LTβR interaction to migrate
from allografts to the draining lymph node406.

50
REFERENCES

1. Lewis SM, Williams A, Eisenbarth SC. Structure and function of the immune system in the spleen. Sci
Immunol. 2019;4(33).
2. Groom AC, Schmidt EE, MacDonald IC. Microcirculatory pathways and blood flow in spleen: new insights
from washout kinetics, corrosion casts, and quantitative intravital videomicroscopy. Scanning Microsc.
1991;5(1):159-73; discussion 73-4.
3. Drenckhahn D, Wagner J. Stress fibers in the splenic sinus endothelium in situ: molecular structure,
relationship to the extracellular matrix, and contractility. J Cell Biol. 1986;102(5):1738-47.
4. MacDonald IC, Ragan DM, Schmidt EE, Groom AC. Kinetics of red blood cell passage through interendothelial
slits into venous sinuses in rat spleen, analyzed by in vivo microscopy. Microvasc Res. 1987;33(1):118-34.
5. Kurotaki D, Uede T, Tamura T. Functions and development of red pulp macrophages. Microbiol Immunol.
2015;59(2):55-62.
6. Nolte MA, Hoen EN, van Stijn A, Kraal G, Mebius RE. Isolation of the intact white pulp. Quantitative and
qualitative analysis of the cellular composition of the splenic compartments. Eur J Immunol. 2000;30(2):626-34.
7. Bronte V, Pittet MJ. The spleen in local and systemic regulation of immunity. Immunity. 2013;39(5):806-18.
8. Mebius RE, Kraal G. Structure and function of the spleen. Nat Rev Immunol. 2005;5(8):606-16.
9. Sasou S, Satodate R, Katsura S. The marginal sinus in the perifollicular region of the rat spleen. Cell Tissue
Res. 1976;172(2):195-203.
10. Rodda LB, Lu E, Bennett ML, Sokol CL, Wang X, Luther SA, et al. Single-Cell RNA Sequencing of Lymph Node
Stromal Cells Reveals Niche-Associated Heterogeneity. Immunity. 2018;48(5):1014-28 e6.
11. Katakai T, Suto H, Sugai M, Gonda H, Togawa A, Suematsu S, et al. Organizer-like reticular stromal cell layer
common to adult secondary lymphoid organs. J Immunol. 2008;181(9):6189-200.
12. Munday J, Floyd H, Crocker PR. Sialic acid binding receptors (siglecs) expressed by macrophages. J Leukoc
Biol. 1999;66(5):705-11.
13. Martinez-Pomares L, Gordon S. CD169+ macrophages at the crossroads of antigen presentation. Trends
Immunol. 2012;33(2):66-70.
14. Habbeddine M, Verthuy C, Rastoin O, Chasson L, Bebien M, Bajenoff M, et al. Receptor Activator of NF-
kappaB Orchestrates Activation of Antiviral Memory CD8 T Cells in the Spleen Marginal Zone. Cell Rep.
2017;21(9):2515-27.
15. Backer R, Schwandt T, Greuter M, Oosting M, Jungerkes F, Tuting T, et al. Effective collaboration between
marginal metallophilic macrophages and CD8+ dendritic cells in the generation of cytotoxic T cells. Proc Natl Acad Sci
U S A. 2010;107(1):216-21.
16. Bernhard CA, Ried C, Kochanek S, Brocker T. CD169+ macrophages are sufficient for priming of CTLs with
specificities left out by cross-priming dendritic cells. Proc Natl Acad Sci U S A. 2015;112(17):5461-6.
17. Asano K, Nabeyama A, Miyake Y, Qiu CH, Kurita A, Tomura M, et al. CD169-positive macrophages dominate
antitumor immunity by crosspresenting dead cell-associated antigens. Immunity. 2011;34(1):85-95.
18. Yu P, Wang Y, Chin RK, Martinez-Pomares L, Gordon S, Kosco-Vibois MH, et al. B cells control the migration
of a subset of dendritic cells into B cell follicles via CXC chemokine ligand 13 in a lymphotoxin-dependent fashion. J
Immunol. 2002;168(10):5117-23.
19. Kang YS, Kim JY, Bruening SA, Pack M, Charalambous A, Pritsker A, et al. The C-type lectin SIGN-R1 mediates
uptake of the capsular polysaccharide of Streptococcus pneumoniae in the marginal zone of mouse spleen. Proc Natl
Acad Sci U S A. 2004;101(1):215-20.
20. Kang YS, Yamazaki S, Iyoda T, Pack M, Bruening SA, Kim JY, et al. SIGN-R1, a novel C-type lectin expressed by
marginal zone macrophages in spleen, mediates uptake of the polysaccharide dextran. Int Immunol. 2003;15(2):177-
86.
21. Kraal G, Mebius R. New insights into the cell biology of the marginal zone of the spleen. Int Rev Cytol.
2006;250:175-215.
22. You Y, Myers RC, Freeberg L, Foote J, Kearney JF, Justement LB, et al. Marginal zone B cells regulate antigen
capture by marginal zone macrophages. J Immunol. 2011;186(4):2172-81.
23. Aichele P, Zinke J, Grode L, Schwendener RA, Kaufmann SH, Seiler P. Macrophages of the splenic marginal
zone are essential for trapping of blood-borne particulate antigen but dispensable for induction of specific T cell
responses. J Immunol. 2003;171(3):1148-55.
24. Seiler P, Aichele P, Odermatt B, Hengartner H, Zinkernagel RM, Schwendener RA. Crucial role of marginal
zone macrophages and marginal zone metallophils in the clearance of lymphocytic choriomeningitis virus infection.
Eur J Immunol. 1997;27(10):2626-33.

51
25. Phillips R, Svensson M, Aziz N, Maroof A, Brown N, Beattie L, et al. Innate killing of Leishmania donovani by
macrophages of the splenic marginal zone requires IRF-7. PLoS Pathog. 2010;6(3):e1000813.
26. Prokopec KE, Georgoudaki AM, Sohn S, Wermeling F, Gronlund H, Lindh E, et al. Cutting Edge: Marginal Zone
Macrophages Regulate Antigen Transport by B Cells to the Follicle in the Spleen via CD21. J Immunol.
2016;197(6):2063-8.
27. Karlsson MC, Guinamard R, Bolland S, Sankala M, Steinman RM, Ravetch JV. Macrophages control the
retention and trafficking of B lymphocytes in the splenic marginal zone. J Exp Med. 2003;198(2):333-40.
28. McGaha TL, Karlsson MC. Apoptotic cell responses in the splenic marginal zone: a paradigm for immunologic
reactions to apoptotic antigens with implications for autoimmunity. Immunol Rev. 2016;269(1):26-43.
29. McGaha TL, Chen Y, Ravishankar B, van Rooijen N, Karlsson MC. Marginal zone macrophages suppress
innate and adaptive immunity to apoptotic cells in the spleen. Blood. 2011;117(20):5403-12.
30. Gordon S, Pluddemann A. Tissue macrophages: heterogeneity and functions. BMC Biol. 2017;15(1):53.
31. Bajenoff M, Glaichenhaus N, Germain RN. Fibroblastic reticular cells guide T lymphocyte entry into and
migration within the splenic T cell zone. J Immunol. 2008;181(6):3947-54.
32. Arnon TI, Horton RM, Grigorova IL, Cyster JG. Visualization of splenic marginal zone B-cell shuttling and
follicular B-cell egress. Nature. 2013;493(7434):684-8.
33. Nolte MA, Belien JA, Schadee-Eestermans I, Jansen W, Unger WW, van Rooijen N, et al. A conduit system
distributes chemokines and small blood-borne molecules through the splenic white pulp. J Exp Med.
2003;198(3):505-12.
34. Hao Z, Rajewsky K. Homeostasis of peripheral B cells in the absence of B cell influx from the bone marrow. J
Exp Med. 2001;194(8):1151-64.
35. Pillai S, Cariappa A. The follicular versus marginal zone B lymphocyte cell fate decision. Nat Rev Immunol.
2009;9(11):767-77.
36. Cerutti A, Cols M, Puga I. Marginal zone B cells: virtues of innate-like antibody-producing lymphocytes. Nat
Rev Immunol. 2013;13(2):118-32.
37. Pillai S, Cariappa A, Moran ST. Marginal zone B cells. Annu Rev Immunol. 2005;23:161-96.
38. Martin F, Kearney JF. Marginal-zone B cells. Nat Rev Immunol. 2002;2(5):323-35.
39. Cariappa A, Boboila C, Moran ST, Liu H, Shi HN, Pillai S. The recirculating B cell pool contains two
functionally distinct, long-lived, posttransitional, follicular B cell populations. J Immunol. 2007;179(4):2270-81.
40. Samardzic T, Marinkovic D, Danzer CP, Gerlach J, Nitschke L, Wirth T. Reduction of marginal zone B cells in
CD22-deficient mice. Eur J Immunol. 2002;32(2):561-7.
41. Cariappa A, Takematsu H, Liu H, Diaz S, Haider K, Boboila C, et al. B cell antigen receptor signal strength and
peripheral B cell development are regulated by a 9-O-acetyl sialic acid esterase. J Exp Med. 2009;206(1):125-38.
42. Cariappa A, Tang M, Parng C, Nebelitskiy E, Carroll M, Georgopoulos K, et al. The follicular versus marginal
zone B lymphocyte cell fate decision is regulated by Aiolos, Btk, and CD21. Immunity. 2001;14(5):603-15.
43. Wang JH, Avitahl N, Cariappa A, Friedrich C, Ikeda T, Renold A, et al. Aiolos regulates B cell activation and
maturation to effector state. Immunity. 1998;9(4):543-53.
44. Hikida M, Johmura S, Hashimoto A, Takezaki M, Kurosaki T. Coupling between B cell receptor and
phospholipase C-gamma2 is essential for mature B cell development. J Exp Med. 2003;198(4):581-9.
45. Pani G, Siminovitch KA, Paige CJ. The motheaten mutation rescues B cell signaling and development in
CD45-deficient mice. J Exp Med. 1997;186(4):581-8.
46. Martin F, Kearney JF. Positive selection from newly formed to marginal zone B cells depends on the rate of
clonal production, CD19, and btk. Immunity. 2000;12(1):39-49.
47. Cariappa A, Chase C, Liu H, Russell P, Pillai S. Naive recirculating B cells mature simultaneously in the spleen
and bone marrow. Blood. 2007;109(6):2339-45.
48. Wen L, Brill-Dashoff J, Shinton SA, Asano M, Hardy RR, Hayakawa K. Evidence of marginal-zone B cell-
positive selection in spleen. Immunity. 2005;23(3):297-308.
49. Saito T, Chiba S, Ichikawa M, Kunisato A, Asai T, Shimizu K, et al. Notch2 is preferentially expressed in
mature B cells and indispensable for marginal zone B lineage development. Immunity. 2003;18(5):675-85.
50. Tanigaki K, Han H, Yamamoto N, Tashiro K, Ikegawa M, Kuroda K, et al. Notch-RBP-J signaling is involved in
cell fate determination of marginal zone B cells. Nat Immunol. 2002;3(5):443-50.
51. Oyama T, Harigaya K, Muradil A, Hozumi K, Habu S, Oguro H, et al. Mastermind-1 is required for Notch
signal-dependent steps in lymphocyte development in vivo. Proc Natl Acad Sci U S A. 2007;104(23):9764-9.
52. Kuroda K, Han H, Tani S, Tanigaki K, Tun T, Furukawa T, et al. Regulation of marginal zone B cell
development by MINT, a suppressor of Notch/RBP-J signaling pathway. Immunity. 2003;18(2):301-12.
53. Gibb DR, El Shikh M, Kang DJ, Rowe WJ, El Sayed R, Cichy J, et al. ADAM10 is essential for Notch2-dependent
marginal zone B cell development and CD23 cleavage in vivo. J Exp Med. 2010;207(3):623-35.

52
54. Hozumi K, Negishi N, Suzuki D, Abe N, Sotomaru Y, Tamaoki N, et al. Delta-like 1 is necessary for the
generation of marginal zone B cells but not T cells in vivo. Nat Immunol. 2004;5(6):638-44.
55. Sekine C, Moriyama Y, Koyanagi A, Koyama N, Ogata H, Okumura K, et al. Differential regulation of splenic
CD8- dendritic cells and marginal zone B cells by Notch ligands. Int Immunol. 2009;21(3):295-301.
56. Fasnacht N, Huang HY, Koch U, Favre S, Auderset F, Chai Q, et al. Specific fibroblastic niches in secondary
lymphoid organs orchestrate distinct Notch-regulated immune responses. J Exp Med. 2014;211(11):2265-79.
57. Quong MW, Martensson A, Langerak AW, Rivera RR, Nemazee D, Murre C. Receptor editing and marginal
zone B cell development are regulated by the helix-loop-helix protein, E2A. J Exp Med. 2004;199(8):1101-12.
58. Guinamard R, Okigaki M, Schlessinger J, Ravetch JV. Absence of marginal zone B cells in Pyk-2-deficient mice
defines their role in the humoral response. Nat Immunol. 2000;1(1):31-6.
59. Batten M, Groom J, Cachero TG, Qian F, Schneider P, Tschopp J, et al. BAFF mediates survival of peripheral
immature B lymphocytes. J Exp Med. 2000;192(10):1453-66.
60. Schiemann B, Gommerman JL, Vora K, Cachero TG, Shulga-Morskaya S, Dobles M, et al. An essential role for
BAFF in the normal development of B cells through a BCMA-independent pathway. Science. 2001;293(5537):2111-4.
61. Otipoby KL, Sasaki Y, Schmidt-Supprian M, Patke A, Gareus R, Pasparakis M, et al. BAFF activates Akt and
Erk through BAFF-R in an IKK1-dependent manner in primary mouse B cells. Proc Natl Acad Sci U S A.
2008;105(34):12435-8.
62. Cariappa A, Liou HC, Horwitz BH, Pillai S. Nuclear factor kappa B is required for the development of
marginal zone B lymphocytes. J Exp Med. 2000;192(8):1175-82.
63. Sasaki Y, Derudder E, Hobeika E, Pelanda R, Reth M, Rajewsky K, et al. Canonical NF-kappaB activity,
dispensable for B cell development, replaces BAFF-receptor signals and promotes B cell proliferation upon activation.
Immunity. 2006;24(6):729-39.
64. Ruefli-Brasse AA, French DM, Dixit VM. Regulation of NF-kappaB-dependent lymphocyte activation and
development by paracaspase. Science. 2003;302(5650):1581-4.
65. Xue L, Morris SW, Orihuela C, Tuomanen E, Cui X, Wen R, et al. Defective development and function of Bcl10-
deficient follicular, marginal zone and B1 B cells. Nat Immunol. 2003;4(9):857-65.
66. Pappu BP, Lin X. Potential role of CARMA1 in CD40-induced splenic B cell proliferation and marginal zone B
cell maturation. Eur J Immunol. 2006;36(11):3033-43.
67. Siebenlist U, Brown K, Claudio E. Control of lymphocyte development by nuclear factor-kappaB. Nat Rev
Immunol. 2005;5(6):435-45.
68. Moran ST, Cariappa A, Liu H, Muir B, Sgroi D, Boboila C, et al. Synergism between NF-kappa B1/p50 and
Notch2 during the development of marginal zone B lymphocytes. J Immunol. 2007;179(1):195-200.
69. Cinamon G, Zachariah MA, Lam OM, Foss FW, Jr., Cyster JG. Follicular shuttling of marginal zone B cells
facilitates antigen transport. Nat Immunol. 2008;9(1):54-62.
70. Ramos-Perez WD, Fang V, Escalante-Alcalde D, Cammer M, Schwab SR. A map of the distribution of
sphingosine 1-phosphate in the spleen. Nat Immunol. 2015;16(12):1245-52.
71. Basu S, Ray A, Dittel BN. Cannabinoid receptor 2 is critical for the homing and retention of marginal zone B
lineage cells and for efficient T-independent immune responses. J Immunol. 2011;187(11):5720-32.
72. Muppidi JR, Arnon TI, Bronevetsky Y, Veerapen N, Tanaka M, Besra GS, et al. Cannabinoid receptor 2
positions and retains marginal zone B cells within the splenic marginal zone. J Exp Med. 2011;208(10):1941-8.
73. Lu TT, Cyster JG. Integrin-mediated long-term B cell retention in the splenic marginal zone. Science.
2002;297(5580):409-12.
74. Cinamon G, Matloubian M, Lesneski MJ, Xu Y, Low C, Lu T, et al. Sphingosine 1-phosphate receptor 1
promotes B cell localization in the splenic marginal zone. Nat Immunol. 2004;5(7):713-20.
75. Westerberg LS, de la Fuente MA, Wermeling F, Ochs HD, Karlsson MC, Snapper SB, et al. WASP confers
selective advantage for specific hematopoietic cell populations and serves a unique role in marginal zone B-cell
homeostasis and function. Blood. 2008;112(10):4139-47.
76. Croker BA, Tarlinton DM, Cluse LA, Tuxen AJ, Light A, Yang FC, et al. The Rac2 guanosine triphosphatase
regulates B lymphocyte antigen receptor responses and chemotaxis and is required for establishment of B-1a and
marginal zone B lymphocytes. J Immunol. 2002;168(7):3376-86.
77. Walmsley MJ, Ooi SK, Reynolds LF, Smith SH, Ruf S, Mathiot A, et al. Critical roles for Rac1 and Rac2 GTPases
in B cell development and signaling. Science. 2003;302(5644):459-62.
78. Girkontaite I, Missy K, Sakk V, Harenberg A, Tedford K, Potzel T, et al. Lsc is required for marginal zone B
cells, regulation of lymphocyte motility and immune responses. Nat Immunol. 2001;2(9):855-62.
79. Rubtsov A, Strauch P, Digiacomo A, Hu J, Pelanda R, Torres RM. Lsc regulates marginal-zone B cell migration
and adhesion and is required for the IgM T-dependent antibody response. Immunity. 2005;23(5):527-38.

53
80. Rieken S, Sassmann A, Herroeder S, Wallenwein B, Moers A, Offermanns S, et al. G12/G13 family G proteins
regulate marginal zone B cell maturation, migration, and polarization. J Immunol. 2006;177(5):2985-93.
81. Hwang IY, Boularan C, Harrison K, Kehrl JH. Galphai Signaling Promotes Marginal Zone B Cell Development
by Enabling Transitional B Cell ADAM10 Expression. Front Immunol. 2018;9:687.
82. Tedford K, Steiner M, Koshutin S, Richter K, Tech L, Eggers Y, et al. The opposing forces of shear flow and
sphingosine-1-phosphate control marginal zone B cell shuttling. Nat Commun. 2017;8(1):2261.
83. Carsetti R, Rosado MM, Donnanno S, Guazzi V, Soresina A, Meini A, et al. The loss of IgM memory B cells
correlates with clinical disease in common variable immunodeficiency. J Allergy Clin Immunol. 2005;115(2):412-7.
84. Hart M, Steel A, Clark SA, Moyle G, Nelson M, Henderson DC, et al. Loss of discrete memory B cell subsets is
associated with impaired immunization responses in HIV-1 infection and may be a risk factor for invasive
pneumococcal disease. J Immunol. 2007;178(12):8212-20.
85. Weller S, Bonnet M, Delagreverie H, Israel L, Chrabieh M, Marodi L, et al. IgM+IgD+CD27+ B cells are
markedly reduced in IRAK-4-, MyD88-, and TIRAP- but not UNC-93B-deficient patients. Blood. 2012;120(25):4992-
5001.
86. Amlot PL, Hayes AE. Impaired human antibody response to the thymus-independent antigen, DNP-Ficoll,
after splenectomy. Implications for post-splenectomy infections. Lancet. 1985;1(8436):1008-11.
87. Kruetzmann S, Rosado MM, Weber H, Germing U, Tournilhac O, Peter HH, et al. Human immunoglobulin M
memory B cells controlling Streptococcus pneumoniae infections are generated in the spleen. J Exp Med.
2003;197(7):939-45.
88. Wasserstrom H, Bussel J, Lim LC, Cunningham-Rundles C. Memory B cells and pneumococcal antibody after
splenectomy. J Immunol. 2008;181(5):3684-9.
89. Di Sabatino A, Rosado MM, Ciccocioppo R, Cazzola P, Morera R, Corazza GR, et al. Depletion of
immunoglobulin M memory B cells is associated with splenic hypofunction in inflammatory bowel disease. Am J
Gastroenterol. 2005;100(8):1788-95.
90. Timens W, Boes A, Rozeboom-Uiterwijk T, Poppema S. Immaturity of the human splenic marginal zone in
infancy. Possible contribution to the deficient infant immune response. J Immunol. 1989;143(10):3200-6.
91. Kruschinski C, Zidan M, Debertin AS, von Horsten S, Pabst R. Age-dependent development of the splenic
marginal zone in human infants is associated with different causes of death. Hum Pathol. 2004;35(1):113-21.
92. Lortan J, Gray D, Kumararatne DS, Platteau B, Bazin H, MacLennan IC. Regulation of the size of the
recirculating B cell pool of adult rats. Adv Exp Med Biol. 1985;186:593-601.
93. Cowan MJ, Ammann AJ, Wara DW, Howie VM, Schultz L, Doyle N, et al. Pneumococcal polysaccharide
immunization in infants and children. Pediatrics. 1978;62(5):721-7.
94. Totapally BR, Walsh WT. Pneumococcal bacteremia in childhood: a 6-year experience in a community
hospital. Chest. 1998;113(5):1207-14.
95. Sha WC, Liou HC, Tuomanen EI, Baltimore D. Targeted disruption of the p50 subunit of NF-kappa B leads to
multifocal defects in immune responses. Cell. 1995;80(2):321-30.
96. Gatto D, Ruedl C, Odermatt B, Bachmann MF. Rapid response of marginal zone B cells to viral particles. J
Immunol. 2004;173(7):4308-16.
97. Patel SR, Gibb DR, Girard-Pierce K, Zhou X, Rodrigues LC, Arthur CM, et al. Marginal Zone B Cells Induce
Alloantibody Formation Following RBC Transfusion. Front Immunol. 2018;9:2516.
98. Balazs M, Martin F, Zhou T, Kearney J. Blood dendritic cells interact with splenic marginal zone B cells to
initiate T-independent immune responses. Immunity. 2002;17(3):341-52.
99. Martin F, Oliver AM, Kearney JF. Marginal zone and B1 B cells unite in the early response against T-
independent blood-borne particulate antigens. Immunity. 2001;14(5):617-29.
100. Leadbetter EA, Brigl M, Illarionov P, Cohen N, Luteran MC, Pillai S, et al. NK T cells provide lipid antigen-
specific cognate help for B cells. Proc Natl Acad Sci U S A. 2008;105(24):8339-44.
101. Barral P, Eckl-Dorna J, Harwood NE, De Santo C, Salio M, Illarionov P, et al. B cell receptor-mediated uptake
of CD1d-restricted antigen augments antibody responses by recruiting invariant NKT cell help in vivo. Proc Natl Acad
Sci U S A. 2008;105(24):8345-50.
102. Chorny A, Casas-Recasens S, Sintes J, Shan M, Polentarutti N, Garcia-Escudero R, et al. The soluble pattern
recognition receptor PTX3 links humoral innate and adaptive immune responses by helping marginal zone B cells. J
Exp Med. 2016;213(10):2167-85.
103. Magri G, Miyajima M, Bascones S, Mortha A, Puga I, Cassis L, et al. Innate lymphoid cells integrate stromal
and immunological signals to enhance antibody production by splenic marginal zone B cells. Nat Immunol.
2014;15(4):354-64.

54
104. Puga I, Cols M, Barra CM, He B, Cassis L, Gentile M, et al. B cell-helper neutrophils stimulate the
diversification and production of immunoglobulin in the marginal zone of the spleen. Nat Immunol. 2011;13(2):170-
80.
105. Attanavanich K, Kearney JF. Marginal zone, but not follicular B cells, are potent activators of naive CD4 T
cells. J Immunol. 2004;172(2):803-11.
106. Song H, Cerny J. Functional heterogeneity of marginal zone B cells revealed by their ability to generate both
early antibody-forming cells and germinal centers with hypermutation and memory in response to a T-dependent
antigen. J Exp Med. 2003;198(12):1923-35.
107. William J, Euler C, Christensen S, Shlomchik MJ. Evolution of autoantibody responses via somatic
hypermutation outside of germinal centers. Science. 2002;297(5589):2066-70.
108. MacLennan IC, Toellner KM, Cunningham AF, Serre K, Sze DM, Zuniga E, et al. Extrafollicular antibody
responses. Immunol Rev. 2003;194:8-18.
109. Steinman RM, Cohn ZA. Identification of a novel cell type in peripheral lymphoid organs of mice. I.
Morphology, quantitation, tissue distribution. J Exp Med. 1973;137(5):1142-62.
110. Mildner A, Jung S. Development and function of dendritic cell subsets. Immunity. 2014;40(5):642-56.
111. Merad M, Sathe P, Helft J, Miller J, Mortha A. The dendritic cell lineage: ontogeny and function of dendritic
cells and their subsets in the steady state and the inflamed setting. Annu Rev Immunol. 2013;31:563-604.
112. Durai V, Murphy KM. Functions of Murine Dendritic Cells. Immunity. 2016;45(4):719-36.
113. Maldonado-Lopez R, De Smedt T, Michel P, Godfroid J, Pajak B, Heirman C, et al. CD8alpha+ and CD8alpha-
subclasses of dendritic cells direct the development of distinct T helper cells in vivo. J Exp Med. 1999;189(3):587-92.
114. Vremec D, Zorbas M, Scollay R, Saunders DJ, Ardavin CF, Wu L, et al. The surface phenotype of dendritic cells
purified from mouse thymus and spleen: investigation of the CD8 expression by a subpopulation of dendritic cells. J
Exp Med. 1992;176(1):47-58.
115. Guilliams M, Ginhoux F, Jakubzick C, Naik SH, Onai N, Schraml BU, et al. Dendritic cells, monocytes and
macrophages: a unified nomenclature based on ontogeny. Nat Rev Immunol. 2014;14(8):571-8.
116. Naik SH, Corcoran LM, Wu L. Development of murine plasmacytoid dendritic cell subsets. Immunol Cell Biol.
2005;83(5):563-70.
117. Osorio F, Fuentes C, Lopez MN, Salazar-Onfray F, Gonzalez FE. Role of Dendritic Cells in the Induction of
Lymphocyte Tolerance. Front Immunol. 2015;6:535.
118. Naik SH, Sathe P, Park HY, Metcalf D, Proietto AI, Dakic A, et al. Development of plasmacytoid and
conventional dendritic cell subtypes from single precursor cells derived in vitro and in vivo. Nat Immunol.
2007;8(11):1217-26.
119. Onai N, Kurabayashi K, Hosoi-Amaike M, Toyama-Sorimachi N, Matsushima K, Inaba K, et al. A clonogenic
progenitor with prominent plasmacytoid dendritic cell developmental potential. Immunity. 2013;38(5):943-57.
120. Onai N, Obata-Onai A, Schmid MA, Ohteki T, Jarrossay D, Manz MG. Identification of clonogenic common
Flt3+M-CSFR+ plasmacytoid and conventional dendritic cell progenitors in mouse bone marrow. Nat Immunol.
2007;8(11):1207-16.
121. Corcoran L, Ferrero I, Vremec D, Lucas K, Waithman J, O'Keeffe M, et al. The lymphoid past of mouse
plasmacytoid cells and thymic dendritic cells. J Immunol. 2003;170(10):4926-32.
122. Pelayo R, Hirose J, Huang J, Garrett KP, Delogu A, Busslinger M, et al. Derivation of 2 categories of
plasmacytoid dendritic cells in murine bone marrow. Blood. 2005;105(11):4407-15.
123. Rodrigues PF, Alberti-Servera L, Eremin A, Grajales-Reyes GE, Ivanek R, Tussiwand R. Distinct progenitor
lineages contribute to the heterogeneity of plasmacytoid dendritic cells. Nat Immunol. 2018;19(7):711-22.
124. Cabeza-Cabrerizo M, van Blijswijk J, Wienert S, Heim D, Jenkins RP, Chakravarty P, et al. Tissue clonality of
dendritic cell subsets and emergency DCpoiesis revealed by multicolor fate mapping of DC progenitors. Sci Immunol.
2019;4(33).
125. Kamath AT, Pooley J, O'Keeffe MA, Vremec D, Zhan Y, Lew AM, et al. The development, maturation, and
turnover rate of mouse spleen dendritic cell populations. J Immunol. 2000;165(12):6762-70.
126. Kabashima K, Banks TA, Ansel KM, Lu TT, Ware CF, Cyster JG. Intrinsic lymphotoxin-beta receptor
requirement for homeostasis of lymphoid tissue dendritic cells. Immunity. 2005;22(4):439-50.
127. Guilliams M, Dutertre CA, Scott CL, McGovern N, Sichien D, Chakarov S, et al. Unsupervised High-
Dimensional Analysis Aligns Dendritic Cells across Tissues and Species. Immunity. 2016;45(3):669-84.
128. Desch AN, Gibbings SL, Clambey ET, Janssen WJ, Slansky JE, Kedl RM, et al. Dendritic cell subsets require cis-
activation for cytotoxic CD8 T-cell induction. Nat Commun. 2014;5:4674.
129. Idoyaga J, Fiorese C, Zbytnuik L, Lubkin A, Miller J, Malissen B, et al. Specialized role of migratory dendritic
cells in peripheral tolerance induction. J Clin Invest. 2013;123(2):844-54.

55
130. Chung Y, Chang JH, Kweon MN, Rennert PD, Kang CY. CD8alpha-11b+ dendritic cells but not CD8alpha+
dendritic cells mediate cross-tolerance toward intestinal antigens. Blood. 2005;106(1):201-6.
131. Nutsch K, Chai JN, Ai TL, Russler-Germain E, Feehley T, Nagler CR, et al. Rapid and Efficient Generation of
Regulatory T Cells to Commensal Antigens in the Periphery. Cell Rep. 2016;17(1):206-20.
132. Gottschalk C, Damuzzo V, Gotot J, Kroczek RA, Yagita H, Murphy KM, et al. Batf3-dependent dendritic cells in
the renal lymph node induce tolerance against circulating antigens. J Am Soc Nephrol. 2013;24(4):543-9.
133. Leventhal DS, Gilmore DC, Berger JM, Nishi S, Lee V, Malchow S, et al. Dendritic Cells Coordinate the
Development and Homeostasis of Organ-Specific Regulatory T Cells. Immunity. 2016;44(4):847-59.
134. Esterhazy D, Loschko J, London M, Jove V, Oliveira TY, Mucida D. Classical dendritic cells are required for
dietary antigen-mediated induction of peripheral T(reg) cells and tolerance. Nat Immunol. 2016;17(5):545-55.
135. Hildner K, Edelson BT, Purtha WE, Diamond M, Matsushita H, Kohyama M, et al. Batf3 deficiency reveals a
critical role for CD8alpha+ dendritic cells in cytotoxic T cell immunity. Science. 2008;322(5904):1097-100.
136. Grajales-Reyes GE, Iwata A, Albring J, Wu X, Tussiwand R, Kc W, et al. Batf3 maintains autoactivation of Irf8
for commitment of a CD8alpha(+) conventional DC clonogenic progenitor. Nat Immunol. 2015;16(7):708-17.
137. Aliberti J, Schulz O, Pennington DJ, Tsujimura H, Reis e Sousa C, Ozato K, et al. Essential role for ICSBP in the
in vivo development of murine CD8alpha + dendritic cells. Blood. 2003;101(1):305-10.
138. Luda KM, Joeris T, Persson EK, Rivollier A, Demiri M, Sitnik KM, et al. IRF8 Transcription-Factor-Dependent
Classical Dendritic Cells Are Essential for Intestinal T Cell Homeostasis. Immunity. 2016;44(4):860-74.
139. Schiavoni G, Mattei F, Sestili P, Borghi P, Venditti M, Morse HC, 3rd, et al. ICSBP is essential for the
development of mouse type I interferon-producing cells and for the generation and activation of CD8alpha(+)
dendritic cells. J Exp Med. 2002;196(11):1415-25.
140. Jackson JT, Hu Y, Liu R, Masson F, D'Amico A, Carotta S, et al. Id2 expression delineates differential
checkpoints in the genetic program of CD8alpha+ and CD103+ dendritic cell lineages. EMBO J. 2011;30(13):2690-
704.
141. Kashiwada M, Pham NL, Pewe LL, Harty JT, Rothman PB. NFIL3/E4BP4 is a key transcription factor for
CD8alpha(+) dendritic cell development. Blood. 2011;117(23):6193-7.
142. Sichien D, Scott CL, Martens L, Vanderkerken M, Van Gassen S, Plantinga M, et al. IRF8 Transcription Factor
Controls Survival and Function of Terminally Differentiated Conventional and Plasmacytoid Dendritic Cells,
Respectively. Immunity. 2016;45(3):626-40.
143. Seillet C, Jackson JT, Markey KA, Brady HJ, Hill GR, Macdonald KP, et al. CD8alpha+ DCs can be induced in the
absence of transcription factors Id2, Nfil3, and Batf3. Blood. 2013;121(9):1574-83.
144. Bajana S, Turner S, Paul J, Ainsua-Enrich E, Kovats S. IRF4 and IRF8 Act in CD11c+ Cells To Regulate
Terminal Differentiation of Lung Tissue Dendritic Cells. J Immunol. 2016;196(4):1666-77.
145. Schlitzer A, McGovern N, Teo P, Zelante T, Atarashi K, Low D, et al. IRF4 transcription factor-dependent
CD11b+ dendritic cells in human and mouse control mucosal IL-17 cytokine responses. Immunity. 2013;38(5):970-
83.
146. Persson EK, Uronen-Hansson H, Semmrich M, Rivollier A, Hagerbrand K, Marsal J, et al. IRF4 transcription-
factor-dependent CD103(+)CD11b(+) dendritic cells drive mucosal T helper 17 cell differentiation. Immunity.
2013;38(5):958-69.
147. Williams JW, Tjota MY, Clay BS, Vander Lugt B, Bandukwala HS, Hrusch CL, et al. Transcription factor IRF4
drives dendritic cells to promote Th2 differentiation. Nat Commun. 2013;4:2990.
148. Bajana S, Roach K, Turner S, Paul J, Kovats S. IRF4 promotes cutaneous dendritic cell migration to lymph
nodes during homeostasis and inflammation. J Immunol. 2012;189(7):3368-77.
149. Vander Lugt B, Khan AA, Hackney JA, Agrawal S, Lesch J, Zhou M, et al. Transcriptional programming of
dendritic cells for enhanced MHC class II antigen presentation. Nat Immunol. 2014;15(2):161-7.
150. Park CS, Lee PH, Yamada T, Burns A, Shen Y, Puppi M, et al. Kruppel-like factor 4 (KLF4) promotes the
survival of natural killer cells and maintains the number of conventional dendritic cells in the spleen. J Leukoc Biol.
2012;91(5):739-50.
151. Tussiwand R, Everts B, Grajales-Reyes GE, Kretzer NM, Iwata A, Bagaitkar J, et al. Klf4 expression in
conventional dendritic cells is required for T helper 2 cell responses. Immunity. 2015;42(5):916-28.
152. Scott CL, Soen B, Martens L, Skrypek N, Saelens W, Taminau J, et al. The transcription factor Zeb2 regulates
development of conventional and plasmacytoid DCs by repressing Id2. J Exp Med. 2016;213(6):897-911.
153. Wu X, Briseno CG, Grajales-Reyes GE, Haldar M, Iwata A, Kretzer NM, et al. Transcription factor Zeb2
regulates commitment to plasmacytoid dendritic cell and monocyte fate. Proc Natl Acad Sci U S A.
2016;113(51):14775-80.
154. Dicken J, Mildner A, Leshkowitz D, Touw IP, Hantisteanu S, Jung S, et al. Transcriptional reprogramming of
CD11b+Esam(hi) dendritic cell identity and function by loss of Runx3. PLoS One. 2013;8(10):e77490.

56
155. Satpathy AT, Briseno CG, Lee JS, Ng D, Manieri NA, Kc W, et al. Notch2-dependent classical dendritic cells
orchestrate intestinal immunity to attaching-and-effacing bacterial pathogens. Nat Immunol. 2013;14(9):937-48.
156. Caton ML, Smith-Raska MR, Reizis B. Notch-RBP-J signaling controls the homeostasis of CD8- dendritic cells
in the spleen. J Exp Med. 2007;204(7):1653-64.
157. Lewis KL, Caton ML, Bogunovic M, Greter M, Grajkowska LT, Ng D, et al. Notch2 receptor signaling controls
functional differentiation of dendritic cells in the spleen and intestine. Immunity. 2011;35(5):780-91.
158. Briseno CG, Gargaro M, Durai V, Davidson JT, Theisen DJ, Anderson DA, 3rd, et al. Deficiency of transcription
factor RelB perturbs myeloid and DC development by hematopoietic-extrinsic mechanisms. Proc Natl Acad Sci U S A.
2017;114(15):3957-62.
159. Wu L, D'Amico A, Winkel KD, Suter M, Lo D, Shortman K. RelB is essential for the development of myeloid-
related CD8alpha- dendritic cells but not of lymphoid-related CD8alpha+ dendritic cells. Immunity. 1998;9(6):839-
47.
160. Ichikawa E, Hida S, Omatsu Y, Shimoyama S, Takahara K, Miyagawa S, et al. Defective development of splenic
and epidermal CD4+ dendritic cells in mice deficient for IFN regulatory factor-2. Proc Natl Acad Sci U S A.
2004;101(11):3909-14.
161. Sichien D, Lambrecht BN, Guilliams M, Scott CL. Development of conventional dendritic cells: from common
bone marrow progenitors to multiple subsets in peripheral tissues. Mucosal Immunol. 2017;10(4):831-44.
162. Backer RA, Diener N, Clausen BE. Langerin(+)CD8(+) Dendritic Cells in the Splenic Marginal Zone: Not So
Marginal After All. Front Immunol. 2019;10:741.
163. Miyake Y, Asano K, Kaise H, Uemura M, Nakayama M, Tanaka M. Critical role of macrophages in the marginal
zone in the suppression of immune responses to apoptotic cell-associated antigens. J Clin Invest. 2007;117(8):2268-
78.
164. Yi T, Cyster JG. EBI2-mediated bridging channel positioning supports splenic dendritic cell homeostasis and
particulate antigen capture. Elife. 2013;2:e00757.
165. Calabro S, Liu D, Gallman A, Nascimento MS, Yu Z, Zhang TT, et al. Differential Intrasplenic Migration of
Dendritic Cell Subsets Tailors Adaptive Immunity. Cell Rep. 2016;16(9):2472-85.
166. Idoyaga J, Suda N, Suda K, Park CG, Steinman RM. Antibody to Langerin/CD207 localizes large numbers of
CD8alpha+ dendritic cells to the marginal zone of mouse spleen. Proc Natl Acad Sci U S A. 2009;106(5):1524-9.
167. Pack M, Trumpfheller C, Thomas D, Park CG, Granelli-Piperno A, Munz C, et al. DEC-205/CD205+ dendritic
cells are abundant in the white pulp of the human spleen, including the border region between the red and white
pulp. Immunology. 2008;123(3):438-46.
168. Briseno CG, Satpathy AT, Davidson JTt, Ferris ST, Durai V, Bagadia P, et al. Notch2-dependent DC2s mediate
splenic germinal center responses. Proc Natl Acad Sci U S A. 2018;115(42):10726-31.
169. Beijer MR, Molenaar R, Goverse G, Mebius RE, Kraal G, den Haan JM. A crucial role for retinoic acid in the
development of Notch-dependent murine splenic CD8- CD4- and CD4+ dendritic cells. Eur J Immunol.
2013;43(6):1608-16.
170. Klebanoff CA, Spencer SP, Torabi-Parizi P, Grainger JR, Roychoudhuri R, Ji Y, et al. Retinoic acid controls the
homeostasis of pre-cDC-derived splenic and intestinal dendritic cells. J Exp Med. 2013;210(10):1961-76.
171. Wu Q, Wang Y, Wang J, Hedgeman EO, Browning JL, Fu YX. The requirement of membrane lymphotoxin for
the presence of dendritic cells in lymphoid tissues. J Exp Med. 1999;190(5):629-38.
172. Wang YG, Kim KD, Wang J, Yu P, Fu YX. Stimulating lymphotoxin beta receptor on the dendritic cells is
critical for their homeostasis and expansion. J Immunol. 2005;175(10):6997-7002.
173. Abe K, Yarovinsky FO, Murakami T, Shakhov AN, Tumanov AV, Ito D, et al. Distinct contributions of TNF and
LT cytokines to the development of dendritic cells in vitro and their recruitment in vivo. Blood. 2003;101(4):1477-83.
174. Metlay JP, Witmer-Pack MD, Agger R, Crowley MT, Lawless D, Steinman RM. The distinct leukocyte integrins
of mouse spleen dendritic cells as identified with new hamster monoclonal antibodies. J Exp Med. 1990;171(5):1753-
71.
175. Lu E, Dang EV, McDonald JG, Cyster JG. Distinct oxysterol requirements for positioning naive and activated
dendritic cells in the spleen. Sci Immunol. 2017;2(10).
176. Barral P, Sanchez-Nino MD, van Rooijen N, Cerundolo V, Batista FD. The location of splenic NKT cells favours
their rapid activation by blood-borne antigen. EMBO J. 2012;31(10):2378-90.
177. Bialecki E, Paget C, Fontaine J, Capron M, Trottein F, Faveeuw C. Role of marginal zone B lymphocytes in
invariant NKT cell activation. J Immunol. 2009;182(10):6105-13.
178. Deniset JF, Surewaard BG, Lee WY, Kubes P. Splenic Ly6G(high) mature and Ly6G(int) immature neutrophils
contribute to eradication of S. pneumoniae. J Exp Med. 2017;214(5):1333-50.
179. Nagelkerke SQ, aan de Kerk DJ, Jansen MH, van den Berg TK, Kuijpers TW. Failure to detect functional
neutrophil B helper cells in the human spleen. PLoS One. 2014;9(2):e88377.

57
180. Kim MY, McConnell FM, Gaspal FM, White A, Glanville SH, Bekiaris V, et al. Function of CD4+CD3- cells in
relation to B- and T-zone stroma in spleen. Blood. 2007;109(4):1602-10.
181. Weller S, Braun MC, Tan BK, Rosenwald A, Cordier C, Conley ME, et al. Human blood IgM "memory" B cells
are circulating splenic marginal zone B cells harboring a prediversified immunoglobulin repertoire. Blood.
2004;104(12):3647-54.
182. Klein U, Rajewsky K, Kuppers R. Human immunoglobulin (Ig)M+IgD+ peripheral blood B cells expressing
the CD27 cell surface antigen carry somatically mutated variable region genes: CD27 as a general marker for
somatically mutated (memory) B cells. J Exp Med. 1998;188(9):1679-89.
183. Tangye SG, Liu YJ, Aversa G, Phillips JH, de Vries JE. Identification of functional human splenic memory B
cells by expression of CD148 and CD27. J Exp Med. 1998;188(9):1691-703.
184. Weller S, Faili A, Garcia C, Braun MC, Le Deist FF, de Saint Basile GG, et al. CD40-CD40L independent Ig gene
hypermutation suggests a second B cell diversification pathway in humans. Proc Natl Acad Sci U S A.
2001;98(3):1166-70.
185. Zhao Y, Uduman M, Siu JHY, Tull TJ, Sanderson JD, Wu YB, et al. Spatiotemporal segregation of human
marginal zone and memory B cell populations in lymphoid tissue. Nat Commun. 2018;9(1):3857.
186. Schmidt EE, MacDonald IC, Groom AC. Changes in splenic microcirculatory pathways in chronic idiopathic
thrombocytopenic purpura. Blood. 1991;78(6):1485-9.
187. Steiniger B, Barth P, Herbst B, Hartnell A, Crocker PR. The species-specific structure of microanatomical
compartments in the human spleen: strongly sialoadhesin-positive macrophages occur in the perifollicular zone, but
not in the marginal zone. Immunology. 1997;92(2):307-16.
188. van Krieken JH, te Velde J. Immunohistology of the human spleen: an inventory of the localization of
lymphocyte subpopulations. Histopathology. 1986;10(3):285-94.
189. van Krieken JH, Te Velde J, Hermans J, Welvaart K. The splenic red pulp; a histomorphometrical study in
splenectomy specimens embedded in methylmethacrylate. Histopathology. 1985;9(4):401-16.
190. Steiniger BS. Human spleen microanatomy: why mice do not suffice. Immunology. 2015;145(3):334-46.
191. Steiniger B, Barth P, Hellinger A. The perifollicular and marginal zones of the human splenic white pulp : do
fibroblasts guide lymphocyte immigration? Am J Pathol. 2001;159(2):501-12.
192. Schmidt EE, MacDonald IC, Groom AC. Comparative aspects of splenic microcirculatory pathways in
mammals: the region bordering the white pulp. Scanning Microsc. 1993;7(2):613-28.
193. Yamamoto K, Kobayashi T, Murakami T. Arterial terminals in the rat spleen as demonstrated by scanning
electron microscopy of vascular casts. Scan Electron Microsc. 1982(Pt 1):455-8.
194. Snook T. Studies on the Perifollicular Region of the Rat's Spleen. Anat Rec. 1964;148:149-59.
195. Steiniger BS, Wilhelmi V, Berthold M, Guthe M, Lobachev O. Locating human splenic capillary sheaths in
virtual reality. Sci Rep. 2018;8(1):15720.
196. Steiniger B, Bette M, Schwarzbach H. The open microcirculation in human spleens: a three-dimensional
approach. J Histochem Cytochem. 2011;59(6):639-48.
197. Steiniger BS, Seiler A, Lampp K, Wilhelmi V, Stachniss V. B lymphocyte compartments in the human splenic
red pulp: capillary sheaths and periarteriolar regions. Histochem Cell Biol. 2014;141(5):507-18.
198. Buyssens N, Paulus G, Bourgeois N. Ellipsoids in the human spleen. Virchows Arch A Pathol Anat
Histopathol. 1984;403(1):27-40.
199. Kopan R, Ilagan MX. The canonical Notch signaling pathway: unfolding the activation mechanism. Cell.
2009;137(2):216-33.
200. D'Souza B, Meloty-Kapella L, Weinmaster G. Canonical and non-canonical Notch ligands. Curr Top Dev Biol.
2010;92:73-129.
201. Radtke F, Fasnacht N, Macdonald HR. Notch signaling in the immune system. Immunity. 2010;32(1):14-27.
202. Maillard I, Adler SH, Pear WS. Notch and the immune system. Immunity. 2003;19(6):781-91.
203. Andersen P, Uosaki H, Shenje LT, Kwon C. Non-canonical Notch signaling: emerging role and mechanism.
Trends Cell Biol. 2012;22(5):257-65.
204. Ayaz F, Osborne BA. Non-canonical notch signaling in cancer and immunity. Front Oncol. 2014;4:345.
205. Weng AP, Millholland JM, Yashiro-Ohtani Y, Arcangeli ML, Lau A, Wai C, et al. c-Myc is an important direct
target of Notch1 in T-cell acute lymphoblastic leukemia/lymphoma. Genes Dev. 2006;20(15):2096-109.
206. Palomero T, Lim WK, Odom DT, Sulis ML, Real PJ, Margolin A, et al. NOTCH1 directly regulates c-MYC and
activates a feed-forward-loop transcriptional network promoting leukemic cell growth. Proc Natl Acad Sci U S A.
2006;103(48):18261-6.
207. Poellinger L, Lendahl U. Modulating Notch signaling by pathway-intrinsic and pathway-extrinsic
mechanisms. Curr Opin Genet Dev. 2008;18(5):449-54.

58
208. Yuan JS, Kousis PC, Suliman S, Visan I, Guidos CJ. Functions of notch signaling in the immune system:
consensus and controversies. Annu Rev Immunol. 2010;28:343-65.
209. Wolfsberg TG, Bazan JF, Blobel CP, Myles DG, Primakoff P, White JM. The precursor region of a protein active
in sperm-egg fusion contains a metalloprotease and a disintegrin domain: structural, functional, and evolutionary
implications. Proc Natl Acad Sci U S A. 1993;90(22):10783-7.
210. Wolfsberg TG, Primakoff P, Myles DG, White JM. ADAM, a novel family of membrane proteins containing A
Disintegrin And Metalloprotease domain: multipotential functions in cell-cell and cell-matrix interactions. J Cell Biol.
1995;131(2):275-8.
211. Black RA, Rauch CT, Kozlosky CJ, Peschon JJ, Slack JL, Wolfson MF, et al. A metalloproteinase disintegrin that
releases tumour-necrosis factor-alpha from cells. Nature. 1997;385(6618):729-33.
212. Liu J, Qian C, Cao X. Post-Translational Modification Control of Innate Immunity. Immunity. 2016;45(1):15-
30.
213. Edwards DR, Handsley MM, Pennington CJ. The ADAM metalloproteinases. Mol Aspects Med.
2008;29(5):258-89.
214. Heng TS, Painter MW, Immunological Genome Project C. The Immunological Genome Project: networks of
gene expression in immune cells. Nat Immunol. 2008;9(10):1091-4.
215. Takeda S. ADAM and ADAMTS Family Proteins and Snake Venom Metalloproteinases: A Structural
Overview. Toxins (Basel). 2016;8(5).
216. Prox J, Willenbrock M, Weber S, Lehmann T, Schmidt-Arras D, Schwanbeck R, et al. Tetraspanin15 regulates
cellular trafficking and activity of the ectodomain sheddase ADAM10. Cell Mol Life Sci. 2012;69(17):2919-32.
217. Saraceno C, Marcello E, Di Marino D, Borroni B, Claeysen S, Perroy J, et al. SAP97-mediated ADAM10
trafficking from Golgi outposts depends on PKC phosphorylation. Cell Death Dis. 2014;5:e1547.
218. Marcello E, Gardoni F, Di Luca M, Perez-Otano I. An arginine stretch limits ADAM10 exit from the
endoplasmic reticulum. J Biol Chem. 2010;285(14):10376-84.
219. Ebsen H, Lettau M, Kabelitz D, Janssen O. Identification of SH3 domain proteins interacting with the
cytoplasmic tail of the a disintegrin and metalloprotease 10 (ADAM10). PLoS One. 2014;9(7):e102899.
220. Lorenzen I, Lokau J, Korpys Y, Oldefest M, Flynn CM, Kunzel U, et al. Control of ADAM17 activity by
regulation of its cellular localisation. Sci Rep. 2016;6:35067.
221. Dusterhoft S, Michalek M, Kordowski F, Oldefest M, Sommer A, Roseler J, et al. Extracellular Juxtamembrane
Segment of ADAM17 Interacts with Membranes and Is Essential for Its Shedding Activity. Biochemistry.
2015;54(38):5791-801.
222. Seegar TCM, Killingsworth LB, Saha N, Meyer PA, Patra D, Zimmerman B, et al. Structural Basis for Regulated
Proteolysis by the alpha-Secretase ADAM10. Cell. 2017;171(7):1638-48 e7.
223. Grotzinger J, Lorenzen I, Dusterhoft S. Molecular insights into the multilayered regulation of ADAM17: The
role of the extracellular region. Biochim Biophys Acta. 2017;1864(11 Pt B):2088-95.
224. Sommer A, Kordowski F, Buch J, Maretzky T, Evers A, Andra J, et al. Phosphatidylserine exposure is required
for ADAM17 sheddase function. Nat Commun. 2016;7:11523.
225. Fischer K, Voelkl S, Berger J, Andreesen R, Pomorski T, Mackensen A. Antigen recognition induces
phosphatidylserine exposure on the cell surface of human CD8+ T cells. Blood. 2006;108(13):4094-101.
226. Cisse M, Duplan E, Guillot-Sestier MV, Rumigny J, Bauer C, Pages G, et al. The extracellular regulated kinase-1
(ERK1) controls regulated alpha-secretase-mediated processing, promoter transactivation, and mRNA levels of the
cellular prion protein. J Biol Chem. 2011;286(33):29192-206.
227. Diaz-Rodriguez E, Montero JC, Esparis-Ogando A, Yuste L, Pandiella A. Extracellular signal-regulated kinase
phosphorylates tumor necrosis factor alpha-converting enzyme at threonine 735: a potential role in regulated
shedding. Mol Biol Cell. 2002;13(6):2031-44.
228. Soond SM, Everson B, Riches DW, Murphy G. ERK-mediated phosphorylation of Thr735 in TNFalpha-
converting enzyme and its potential role in TACE protein trafficking. J Cell Sci. 2005;118(Pt 11):2371-80.
229. Wisniewska M, Goettig P, Maskos K, Belouski E, Winters D, Hecht R, et al. Structural determinants of the
ADAM inhibition by TIMP-3: crystal structure of the TACE-N-TIMP-3 complex. J Mol Biol. 2008;381(5):1307-19.
230. Xu P, Liu J, Sakaki-Yumoto M, Derynck R. TACE activation by MAPK-mediated regulation of cell surface
dimerization and TIMP3 association. Sci Signal. 2012;5(222):ra34.
231. Deng W, Cho S, Su PC, Berger BW, Li R. Membrane-enabled dimerization of the intrinsically disordered
cytoplasmic domain of ADAM10. Proc Natl Acad Sci U S A. 2014;111(45):15987-92.
232. Schwarz J, Schmidt S, Will O, Koudelka T, Kohler K, Boss M, et al. Polo-like kinase 2, a novel ADAM17
signaling component, regulates tumor necrosis factor alpha ectodomain shedding. J Biol Chem. 2014;289(5):3080-93.
233. Wong E, Cohen T, Romi E, Levin M, Peleg Y, Arad U, et al. Harnessing the natural inhibitory domain to
control TNFalpha Converting Enzyme (TACE) activity in vivo. Sci Rep. 2016;6:35598.

59
234. Wong E, Maretzky T, Peleg Y, Blobel CP, Sagi I. The Functional Maturation of A Disintegrin and
Metalloproteinase (ADAM) 9, 10, and 17 Requires Processing at a Newly Identified Proprotein Convertase (PC)
Cleavage Site. J Biol Chem. 2015;290(19):12135-46.
235. Reyat JS, Chimen M, Noy PJ, Szyroka J, Rainger GE, Tomlinson MG. ADAM10-Interacting Tetraspanins Tspan5
and Tspan17 Regulate VE-Cadherin Expression and Promote T Lymphocyte Transmigration. J Immunol.
2017;99(2):666-76.
236. Matthews AL, Szyroka J, Collier R, Noy PJ, Tomlinson MG. Scissor sisters: regulation of ADAM10 by the
TspanC8 tetraspanins. Biochem Soc Trans. 2017;45(3):719-30.
237. Adrain C, Zettl M, Christova Y, Taylor N, Freeman M. Tumor necrosis factor signaling requires iRhom2 to
promote trafficking and activation of TACE. Science. 2012;335(6065):225-8.
238. McIlwain DR, Lang PA, Maretzky T, Hamada K, Ohishi K, Maney SK, et al. iRhom2 regulation of TACE controls
TNF-mediated protection against Listeria and responses to LPS. Science. 2012;335(6065):229-32.
239. Li X, Maretzky T, Weskamp G, Monette S, Qing X, Issuree PD, et al. iRhoms 1 and 2 are essential upstream
regulators of ADAM17-dependent EGFR signaling. Proc Natl Acad Sci U S A. 2015;112(19):6080-5.
240. Maretzky T, McIlwain DR, Issuree PD, Li X, Malapeira J, Amin S, et al. iRhom2 controls the substrate
selectivity of stimulated ADAM17-dependent ectodomain shedding. Proc Natl Acad Sci U S A. 2013;110(28):11433-8.
241. Matthews AL, Noy PJ, Reyat JS, Tomlinson MG. Regulation of A disintegrin and metalloproteinase (ADAM)
family sheddases ADAM10 and ADAM17: The emerging role of tetraspanins and rhomboids. Platelets.
2017;28(4):333-41.
242. Hartmann D, de Strooper B, Serneels L, Craessaerts K, Herreman A, Annaert W, et al. The
disintegrin/metalloprotease ADAM 10 is essential for Notch signalling but not for alpha-secretase activity in
fibroblasts. Hum Mol Genet. 2002;11(21):2615-24.
243. Brou C, Logeat F, Gupta N, Bessia C, LeBail O, Doedens JR, et al. A novel proteolytic cleavage involved in
Notch signaling: the role of the disintegrin-metalloprotease TACE. Mol Cell. 2000;5(2):207-16.
244. Mumm JS, Schroeter EH, Saxena MT, Griesemer A, Tian X, Pan DJ, et al. A ligand-induced extracellular
cleavage regulates gamma-secretase-like proteolytic activation of Notch1. Mol Cell. 2000;5(2):197-206.
245. Gordon WR, Zimmerman B, He L, Miles LJ, Huang J, Tiyanont K, et al. Mechanical Allostery: Evidence for a
Force Requirement in the Proteolytic Activation of Notch. Dev Cell. 2015;33(6):729-36.
246. Musse AA, Meloty-Kapella L, Weinmaster G. Notch ligand endocytosis: mechanistic basis of signaling
activity. Semin Cell Dev Biol. 2012;23(4):429-36.
247. Langridge PD, Struhl G. Epsin-Dependent Ligand Endocytosis Activates Notch by Force. Cell.
2017;171(6):1383-96 e12.
248. De Strooper B, Annaert W, Cupers P, Saftig P, Craessaerts K, Mumm JS, et al. A presenilin-1-dependent
gamma-secretase-like protease mediates release of Notch intracellular domain. Nature. 1999;398(6727):518-22.
249. Chastagner P, Rubinstein E, Brou C. Ligand-activated Notch undergoes DTX4-mediated ubiquitylation and
bilateral endocytosis before ADAM10 processing. Sci Signal. 2017;10(483).
250. Kueh HY, Yui MA, Ng KK, Pease SS, Zhang JA, Damle SS, et al. Asynchronous combinatorial action of four
regulatory factors activates Bcl11b for T cell commitment. Nat Immunol. 2016;17(8):956-65.
251. Robey E, Chang D, Itano A, Cado D, Alexander H, Lans D, et al. An activated form of Notch influences the
choice between CD4 and CD8 T cell lineages. Cell. 1996;87(3):483-92.
252. Washburn T, Schweighoffer E, Gridley T, Chang D, Fowlkes BJ, Cado D, et al. Notch activity influences the
alphabeta versus gammadelta T cell lineage decision. Cell. 1997;88(6):833-43.
253. Yasutomo K, Doyle C, Miele L, Fuchs C, Germain RN. The duration of antigen receptor signalling determines
CD4+ versus CD8+ T-cell lineage fate. Nature. 2000;404(6777):506-10.
254. Tian L, Wu X, Chi C, Han M, Xu T, Zhuang Y. ADAM10 is essential for proteolytic activation of Notch during
thymocyte development. Int Immunol. 2008;20(9):1181-7.
255. Manilay JO, Anderson AC, Kang C, Robey EA. Impairment of thymocyte development by dominant-negative
Kuzbanian (ADAM-10) is rescued by the Notch ligand, delta-1. J Immunol. 2005;174(11):6732-41.
256. Gravano DM, McLelland BT, Horiuchi K, Manilay JO. ADAM17 deletion in thymic epithelial cells alters aire
expression without affecting T cell developmental progression. PLoS One. 2010;5(10):e13528.
257. Gossens K, Naus S, Hollander GA, Ziltener HJ. Deficiency of the metalloproteinase-disintegrin ADAM8 is
associated with thymic hyper-cellularity. PLoS One. 2010;5(9):e12766.
258. Amsen D, Blander JM, Lee GR, Tanigaki K, Honjo T, Flavell RA. Instruction of distinct CD4 T helper cell fates
by different notch ligands on antigen-presenting cells. Cell. 2004;117(4):515-26.
259. Helbig C, Gentek R, Backer RA, de Souza Y, Derks IA, Eldering E, et al. Notch controls the magnitude of T
helper cell responses by promoting cellular longevity. Proc Natl Acad Sci U S A. 2012;109(23):9041-6.

60
260. Guy CS, Vignali KM, Temirov J, Bettini ML, Overacre AE, Smeltzer M, et al. Distinct TCR signaling pathways
drive proliferation and cytokine production in T cells. Nat Immunol. 2013;14(3):262-70.
261. Maekawa Y, Ishifune C, Tsukumo S, Hozumi K, Yagita H, Yasutomo K. Notch controls the survival of memory
CD4+ T cells by regulating glucose uptake. Nat Med. 2015;21(1):55-61.
262. Laky K, Evans S, Perez-Diez A, Fowlkes BJ. Notch signaling regulates antigen sensitivity of naive CD4+ T cells
by tuning co-stimulation. Immunity. 2015;42(1):80-94.
263. Luty WH, Rodeberg D, Parness J, Vyas YM. Antiparallel segregation of notch components in the
immunological synapse directs reciprocal signaling in allogeneic Th:DC conjugates. J Immunol. 2007;179(2):819-29.
264. Anderson AC, Kitchens EA, Chan SW, St Hill C, Jan YN, Zhong W, et al. The Notch regulator Numb links the
Notch and TCR signaling pathways. J Immunol. 2005;174(2):890-7.
265. Palmer WH, Deng WM. Ligand-Independent Mechanisms of Notch Activity. Trends Cell Biol.
2015;25(11):697-707.
266. Britton GJ, Ambler R, Clark DJ, Hill EV, Tunbridge HM, McNally KE, et al. PKCtheta links proximal T cell and
Notch signaling through localized regulation of the actin cytoskeleton. Elife. 2017;6:e20003.
267. Li N, Wang Y, Forbes K, Vignali KM, Heale BS, Saftig P, et al. Metalloproteases regulate T-cell proliferation
and effector function via LAG-3. EMBO J. 2007;26(2):494-504.
268. Woo SR, Turnis ME, Goldberg MV, Bankoti J, Selby M, Nirschl CJ, et al. Immune inhibitory molecules LAG-3
and PD-1 synergistically regulate T-cell function to promote tumoral immune escape. Cancer Res. 2012;72(4):917-27.
269. Andrews LP, Marciscano AE, Drake CG, Vignali DA. LAG3 (CD223) as a cancer immunotherapy target.
Immunol Rev. 2017;276(1):80-96.
270. Clayton KL, Douglas-Vail MB, Nur-ur Rahman AK, Medcalf KE, Xie IY, Chew GM, et al. Soluble T cell
immunoglobulin mucin domain 3 is shed from CD8+ T cells by the sheddase ADAM10, is increased in plasma during
untreated HIV infection, and correlates with HIV disease progression. J Virol. 2015;89(7):3723-36.
271. Schulte M, Reiss K, Lettau M, Maretzky T, Ludwig A, Hartmann D, et al. ADAM10 regulates FasL cell surface
expression and modulates FasL-induced cytotoxicity and activation-induced cell death. Cell Death Differ.
2007;14(5):1040-9.
272. Ebsen H, Lettau M, Kabelitz D, Janssen O. Subcellular localization and activation of ADAM proteases in the
context of FasL shedding in T lymphocytes. Mol Immunol. 2015;65(2):416-28.
273. Yacoub D, Benslimane N, Al-Zoobi L, Hassan G, Nadiri A, Mourad W. CD154 is released from T-cells by a
disintegrin and metalloproteinase domain-containing protein 10 (ADAM10) and ADAM17 in a CD40 protein-
dependent manner. J Biol Chem. 2013;288(50):36083-93.
274. Jenabian MA, Patel M, Kema I, Vyboh K, Kanagaratham C, Radzioch D, et al. Soluble CD40-ligand (sCD40L,
sCD154) plays an immunosuppressive role via regulatory T cell expansion in HIV infection. Clin Exp Immunol.
2014;178(1):102-11.
275. Eichenauer DA, Simhadri VL, von Strandmann EP, Ludwig A, Matthews V, Reiners KS, et al. ADAM10
inhibition of human CD30 shedding increases specificity of targeted immunotherapy in vitro. Cancer Res.
2007;67(1):332-8.
276. Velasquez SY, Garcia LF, Opelz G, Alvarez CM, Susal C. Release of soluble CD30 after allogeneic stimulation is
mediated by memory T cells and regulated by IFN-gamma and IL-2. Transplantation. 2013;96(2):154-61.
277. Nagano O, Saya H. Mechanism and biological significance of CD44 cleavage. Cancer Sci. 2004;95(12):930-5.
278. Miletti-Gonzalez KE, Murphy K, Kumaran MN, Ravindranath AK, Wernyj RP, Kaur S, et al. Identification of
function for CD44 intracytoplasmic domain (CD44-ICD): modulation of matrix metalloproteinase 9 (MMP-9)
transcription via novel promoter response element. J Biol Chem. 2012;287(23):18995-9007.
279. Schulz B, Pruessmeyer J, Maretzky T, Ludwig A, Blobel CP, Saftig P, et al. ADAM10 regulates endothelial
permeability and T-Cell transmigration by proteolysis of vascular endothelial cadherin. Circ Res. 2008;102(10):1192-
201.
280. Hundhausen C, Schulte A, Schulz B, Andrzejewski MG, Schwarz N, von Hundelshausen P, et al. Regulated
shedding of transmembrane chemokines by the disintegrin and metalloproteinase 10 facilitates detachment of
adherent leukocytes. J Immunol. 2007;178(12):8064-72.
281. Amsen D, Antov A, Flavell RA. The different faces of Notch in T-helper-cell differentiation. Nat Rev Immunol.
2009;9(2):116-24.
282. Tanigaki K, Tsuji M, Yamamoto N, Han H, Tsukada J, Inoue H, et al. Regulation of alphabeta/gammadelta T
cell lineage commitment and peripheral T cell responses by Notch/RBP-J signaling. Immunity. 2004;20(5):611-22.
283. Franke M, Schroder J, Monhasery N, Ackfeld T, Hummel TM, Rabe B, et al. Human and Murine Interleukin 23
Receptors Are Novel Substrates for A Disintegrin and Metalloproteases ADAM10 and ADAM17. J Biol Chem.
2016;291(20):10551-61.

61
284. Kanzaki H, Shinohara F, Suzuki M, Wada S, Miyamoto Y, Yamaguchi Y, et al. A-Disintegrin and
Metalloproteinase (ADAM) 17 Enzymatically Degrades Interferon-gamma. Sci Rep. 2016;6:32259.
285. Weskamp G, Ford JW, Sturgill J, Martin S, Docherty AJ, Swendeman S, et al. ADAM10 is a principal 'sheddase'
of the low-affinity immunoglobulin E receptor CD23. Nat Immunol. 2006;7(12):1293-8.
286. Mathews JA, Gibb DR, Chen BH, Scherle P, Conrad DH. CD23 Sheddase A disintegrin and metalloproteinase
10 (ADAM10) is also required for CD23 sorting into B cell-derived exosomes. J Biol Chem. 2010;285(48):37531-41.
287. Chaimowitz NS, Martin RK, Cichy J, Gibb DR, Patil P, Kang DJ, et al. A disintegrin and metalloproteinase 10
regulates antibody production and maintenance of lymphoid architecture. J Immunol. 2011;187(10):5114-22.
288. Folgosa L, Zellner HB, El Shikh ME, Conrad DH. Disturbed follicular architecture in B cell A disintegrin and
metalloproteinase (ADAM)10 knockouts is mediated by compensatory increases in ADAM17 and TNF-alpha
shedding. J Immunol. 2013;191(12):5951-8.
289. Lownik JC, Luker AJ, Damle SR, Cooley LF, El Sayed R, Hutloff A, et al. ADAM10-Mediated ICOS Ligand
Shedding on B Cells Is Necessary for Proper T Cell ICOS Regulation and T Follicular Helper Responses. J Immunol.
2017;199(7):2305-15.
290. Chaimowitz NS, Kang DJ, Dean LM, Conrad DH. ADAM10 regulates transcription factor expression required
for plasma cell function. PLoS One. 2012;7(8):e42694.
291. Smulski CR, Kury P, Seidel LM, Staiger HS, Edinger AK, Willen L, et al. BAFF- and TACI-Dependent Processing
of BAFFR by ADAM Proteases Regulates the Survival of B Cells. Cell Rep. 2017;18(9):2189-202.
292. Possot C, Schmutz S, Chea S, Boucontet L, Louise A, Cumano A, et al. Notch signaling is necessary for adult,
but not fetal, development of RORgammat(+) innate lymphoid cells. Nat Immunol. 2011;12(10):949-58.
293. Pham DH, Kim JS, Kim SK, Shin DJ, Uong NT, Hyun H, et al. Effects of ADAM10 and ADAM17 Inhibitors on
Natural Killer Cell Expansion and Antibody-dependent Cellular Cytotoxicity Against Breast Cancer Cells In Vitro.
Anticancer Res. 2017;37(10):5507-13.
294. Chitadze G, Lettau M, Bhat J, Wesch D, Steinle A, Furst D, et al. Shedding of endogenous MHC class I-related
chain molecules A and B from different human tumor entities: heterogeneous involvement of the "a disintegrin and
metalloproteases" 10 and 17. Int J Cancer. 2013;133(7):1557-66.
295. Kohga K, Takehara T, Tatsumi T, Miyagi T, Ishida H, Ohkawa K, et al. Anticancer chemotherapy inhibits MHC
class I-related chain a ectodomain shedding by downregulating ADAM10 expression in hepatocellular carcinoma.
Cancer Res. 2009;69(20):8050-7.
296. Matusali G, Tchidjou HK, Pontrelli G, Bernardi S, D'Ettorre G, Vullo V, et al. Soluble ligands for the NKG2D
receptor are released during HIV-1 infection and impair NKG2D expression and cytotoxicity of NK cells. FASEB J.
2013;27(6):2440-50.
297. Nuckel H, Switala M, Sellmann L, Horn PA, Durig J, Duhrsen U, et al. The prognostic significance of soluble
NKG2D ligands in B-cell chronic lymphocytic leukemia. Leukemia. 2010;24(6):1152-9.
298. Zocchi MR, Camodeca C, Nuti E, Rossello A, Vene R, Tosetti F, et al. ADAM10 new selective inhibitors reduce
NKG2D ligand release sensitizing Hodgkin lymphoma cells to NKG2D-mediated killing. Oncoimmunology.
2016;5(5):e1123367.
299. Camodeca C, Nuti E, Tepshi L, Boero S, Tuccinardi T, Stura EA, et al. Discovery of a new selective inhibitor of
A Disintegrin And Metalloprotease 10 (ADAM-10) able to reduce the shedding of NKG2D ligands in Hodgkin's
lymphoma cell models. Eur J Med Chem. 2016;111:193-201.
300. Weber S, Wetzel S, Prox J, Lehmann T, Schneppenheim J, Donners M, et al. Regulation of adult hematopoiesis
by the a disintegrin and metalloproteinase 10 (ADAM10). Biochem Biophys Res Commun. 2013;442(3-4):234-41.
301. Yoda M, Kimura T, Tohmonda T, Uchikawa S, Koba T, Takito J, et al. Dual functions of cell-autonomous and
non-cell-autonomous ADAM10 activity in granulopoiesis. Blood. 2011;118(26):6939-42.
302. Maney SK, McIlwain DR, Polz R, Pandyra AA, Sundaram B, Wolff D, et al. Deletions in the cytoplasmic domain
of iRhom1 and iRhom2 promote shedding of the TNF receptor by the protease ADAM17. Sci Signal.
2015;8(401):ra109.
303. Menghini R, Fiorentino L, Casagrande V, Lauro R, Federici M. The role of ADAM17 in metabolic
inflammation. Atherosclerosis. 2013;228(1):12-7.
304. Schumacher N, Meyer D, Mauermann A, von der Heyde J, Wolf J, Schwarz J, et al. Shedding of Endogenous
Interleukin-6 Receptor (IL-6R) Is Governed by A Disintegrin and Metalloproteinase (ADAM) Proteases while a Full-
length IL-6R Isoform Localizes to Circulating Microvesicles. J Biol Chem. 2015;290(43):26059-71.
305. Tang J, Zarbock A, Gomez I, Wilson CL, Lefort CT, Stadtmann A, et al. Adam17-dependent shedding limits
early neutrophil influx but does not alter early monocyte recruitment to inflammatory sites. Blood. 2011;118(3):786-
94.
306. Pruessmeyer J, Hess FM, Alert H, Groth E, Pasqualon T, Schwarz N, et al. Leukocytes require ADAM10 but not
ADAM17 for their migration and inflammatory recruitment into the alveolar space. Blood. 2014;123(26):4077-88.

62
307. Driscoll WS, Vaisar T, Tang J, Wilson CL, Raines EW. Macrophage ADAM17 deficiency augments CD36-
dependent apoptotic cell uptake and the linked anti-inflammatory phenotype. Circ Res. 2013;113(1):52-61.
308. Moller-Hackbarth K, Dewitz C, Schweigert O, Trad A, Garbers C, Rose-John S, et al. A disintegrin and
metalloprotease (ADAM) 10 and ADAM17 are major sheddases of T cell immunoglobulin and mucin domain 3 (Tim-
3). J Biol Chem. 2013;288(48):34529-44.
309. Zhao D, Guo M, Liu B, Lin Q, Xie T, Zhang Q, et al. Frontline Science: Tim-3-mediated dysfunctional
engulfment of apoptotic cells in SLE. J Leukoc Biol. 2017;102(6):1313-22.
310. Londino JD, Gulick D, Isenberg JS, Mallampalli RK. Cleavage of Signal Regulatory Protein alpha (SIRPalpha)
Enhances Inflammatory Signaling. J Biol Chem. 2015;290(52):31113-25.
311. van der Vorst EP, Jeurissen M, Wolfs IM, Keijbeck A, Theodorou K, Wijnands E, et al. Myeloid A disintegrin
and metalloproteinase domain 10 deficiency modulates atherosclerotic plaque composition by shifting the balance
from inflammation toward fibrosis. Am J Pathol. 2015;185(4):1145-55.
312. Thornton P, Sevalle J, Deery MJ, Fraser G, Zhou Y, Stahl S, et al. TREM2 shedding by cleavage at the H157-
S158 bond is accelerated for the Alzheimer's disease-associated H157Y variant. EMBO Mol Med. 2017;9(10):1366-78.
313. Kleinberger G, Yamanishi Y, Suarez-Calvet M, Czirr E, Lohmann E, Cuyvers E, et al. TREM2 mutations
implicated in neurodegeneration impair cell surface transport and phagocytosis. Sci Transl Med.
2014;6(243):243ra86.
314. Damle SR, Martin RK, Cockburn CL, Lownik JC, Carlyon JA, Smith AD, et al. ADAM10 and Notch1 on murine
dendritic cells control the development of type 2 immunity and IgE production. Allergy. 2018;73(1):125-36.
315. Kirkling ME, Cytlak U, Lau CM, Lewis KL, Resteu A, Khodadadi-Jamayran A, et al. Notch Signaling Facilitates
In Vitro Generation of Cross-Presenting Classical Dendritic Cells. Cell Rep. 2018;23(12):3658-72 e6.
316. Nishida-Fukuda H, Araki R, Shudou M, Okazaki H, Tomono Y, Nakayama H, et al. Ectodomain Shedding of
Lymphatic Vessel Endothelial Hyaluronan Receptor 1 (LYVE-1) Is Induced by Vascular Endothelial Growth Factor A
(VEGF-A). J Biol Chem. 2016;291(20):10490-500.
317. Wu C, Whiteway M, Thomas DY, Leberer E. Molecular characterization of Ste20p, a potential mitogen-
activated protein or extracellular signal-regulated kinase kinase (MEK) kinase kinase from Saccharomyces cerevisiae.
J Biol Chem. 1995;270(27):15984-92.
318. Ramer SW, Davis RW. A dominant truncation allele identifies a gene, STE20, that encodes a putative protein
kinase necessary for mating in Saccharomyces cerevisiae. Proc Natl Acad Sci U S A. 1993;90(2):452-6.
319. Leberer E, Dignard D, Harcus D, Thomas DY, Whiteway M. The protein kinase homologue Ste20p is required
to link the yeast pheromone response G-protein beta gamma subunits to downstream signalling components. EMBO J.
1992;11(13):4815-24.
320. Strange K, Denton J, Nehrke K. Ste20-type kinases: evolutionarily conserved regulators of ion transport and
cell volume. Physiology (Bethesda). 2006;21:61-8.
321. Dan I, Watanabe NM, Kusumi A. The Ste20 group kinases as regulators of MAP kinase cascades. Trends Cell
Biol. 2001;11(5):220-30.
322. Rane CK, Minden A. P21 activated kinases: structure, regulation, and functions. Small GTPases. 2014;5.
323. Delpire E. The mammalian family of sterile 20p-like protein kinases. Pflugers Arch. 2009;458(5):953-67.
324. Yustein JT, Xia L, Kahlenburg JM, Robinson D, Templeton D, Kung HJ. Comparative studies of a new
subfamily of human Ste20-like kinases: homodimerization, subcellular localization, and selective activation of MKK3
and p38. Oncogene. 2003;22(40):6129-41.
325. Zhang W, Chen T, Wan T, He L, Li N, Yuan Z, et al. Cloning of DPK, a novel dendritic cell-derived protein
kinase activating the ERK1/ERK2 and JNK/SAPK pathways. Biochem Biophys Res Commun. 2000;274(3):872-9.
326. Raman M, Earnest S, Zhang K, Zhao Y, Cobb MH. TAO kinases mediate activation of p38 in response to DNA
damage. EMBO J. 2007;26(8):2005-14.
327. Hutchison M, Berman KS, Cobb MH. Isolation of TAO1, a protein kinase that activates MEKs in stress-
activated protein kinase cascades. J Biol Chem. 1998;273(44):28625-32.
328. Chen Z, Hutchison M, Cobb MH. Isolation of the protein kinase TAO2 and identification of its mitogen-
activated protein kinase/extracellular signal-regulated kinase kinase binding domain. J Biol Chem.
1999;274(40):28803-7.
329. Mitsopoulos C, Zihni C, Garg R, Ridley AJ, Morris JD. The prostate-derived sterile 20-like kinase (PSK)
regulates microtubule organization and stability. J Biol Chem. 2003;278(20):18085-91.
330. Tassi E, Biesova Z, Di Fiore PP, Gutkind JS, Wong WT. Human JIK, a novel member of the STE20 kinase family
that inhibits JNK and is negatively regulated by epidermal growth factor. J Biol Chem. 1999;274(47):33287-95.
331. Kapfhamer D, King I, Zou ME, Lim JP, Heberlein U, Wolf FW. JNK pathway activation is controlled by
Tao/TAOK3 to modulate ethanol sensitivity. PLoS One. 2012;7(12):e50594.

63
332. Yoneda T, Imaizumi K, Oono K, Yui D, Gomi F, Katayama T, et al. Activation of caspase-12, an endoplastic
reticulum (ER) resident caspase, through tumor necrosis factor receptor-associated factor 2-dependent mechanism
in response to the ER stress. J Biol Chem. 2001;276(17):13935-40.
333. Ormonde JVS, Li Z, Stegen C, Madrenas J. TAOK3 Regulates Canonical TCR Signaling by Preventing Early
SHP-1-Mediated Inactivation of LCK. J Immunol. 2018;201(11):3431-42.
334. Huang Y, Mao K, Chen X, Sun MA, Kawabe T, Li W, et al. S1P-dependent interorgan trafficking of group 2
innate lymphoid cells supports host defense. Science. 2018;359(6371):114-9.
335. Gasteiger G, Fan X, Dikiy S, Lee SY, Rudensky AY. Tissue residency of innate lymphoid cells in lymphoid and
nonlymphoid organs. Science. 2015;350(6263):981-5.
336. Zook EC, Kee BL. Development of innate lymphoid cells. Nat Immunol. 2016;17(7):775-82.
337. Klose CS, Artis D. Innate lymphoid cells as regulators of immunity, inflammation and tissue homeostasis. Nat
Immunol. 2016;17(7):765-74.
338. Vivier E, Artis D, Colonna M, Diefenbach A, Di Santo JP, Eberl G, et al. Innate Lymphoid Cells: 10 Years On.
Cell. 2018;174(5):1054-66.
339. Huang Y, Mao K, Germain RN. Thinking differently about ILCs-Not just tissue resident and not just the same
as CD4(+) T-cell effectors. Immunol Rev. 2018;286(1):160-71.
340. Gronke K, Kofoed-Nielsen M, Diefenbach A. Innate lymphoid cells, precursors and plasticity. Immunol Lett.
2016;179:9-18.
341. Klose CSN, Flach M, Mohle L, Rogell L, Hoyler T, Ebert K, et al. Differentiation of type 1 ILCs from a common
progenitor to all helper-like innate lymphoid cell lineages. Cell. 2014;157(2):340-56.
342. Abt MC, Lewis BB, Caballero S, Xiong H, Carter RA, Susac B, et al. Innate Immune Defenses Mediated by Two
ILC Subsets Are Critical for Protection against Acute Clostridium difficile Infection. Cell Host Microbe. 2015;18(1):27-
37.
343. Spits H, Bernink JH, Lanier L. NK cells and type 1 innate lymphoid cells: partners in host defense. Nat
Immunol. 2016;17(7):758-64.
344. Monticelli LA, Osborne LC, Noti M, Tran SV, Zaiss DM, Artis D. IL-33 promotes an innate immune pathway of
intestinal tissue protection dependent on amphiregulin-EGFR interactions. Proc Natl Acad Sci U S A.
2015;112(34):10762-7.
345. Lee MW, Odegaard JI, Mukundan L, Qiu Y, Molofsky AB, Nussbaum JC, et al. Activated type 2 innate lymphoid
cells regulate beige fat biogenesis. Cell. 2015;160(1-2):74-87.
346. Brestoff JR, Kim BS, Saenz SA, Stine RR, Monticelli LA, Sonnenberg GF, et al. Group 2 innate lymphoid cells
promote beiging of white adipose tissue and limit obesity. Nature. 2015;519(7542):242-6.
347. Sawa S, Cherrier M, Lochner M, Satoh-Takayama N, Fehling HJ, Langa F, et al. Lineage relationship analysis of
RORgammat+ innate lymphoid cells. Science. 2010;330(6004):665-9.
348. van de Pavert SA, Mebius RE. New insights into the development of lymphoid tissues. Nat Rev Immunol.
2010;10(9):664-74.
349. Eberl G. Inducible lymphoid tissues in the adult gut: recapitulation of a fetal developmental pathway? Nat
Rev Immunol. 2005;5(5):413-20.
350. Melo-Gonzalez F, Hepworth MR. Functional and phenotypic heterogeneity of group 3 innate lymphoid cells.
Immunology. 2017;150(3):265-75.
351. Bjorklund AK, Forkel M, Picelli S, Konya V, Theorell J, Friberg D, et al. The heterogeneity of human CD127(+)
innate lymphoid cells revealed by single-cell RNA sequencing. Nat Immunol. 2016;17(4):451-60.
352. Gury-BenAri M, Thaiss CA, Serafini N, Winter DR, Giladi A, Lara-Astiaso D, et al. The Spectrum and
Regulatory Landscape of Intestinal Innate Lymphoid Cells Are Shaped by the Microbiome. Cell. 2016;166(5):1231-46
e13.
353. Song C, Lee JS, Gilfillan S, Robinette ML, Newberry RD, Stappenbeck TS, et al. Unique and redundant
functions of NKp46+ ILC3s in models of intestinal inflammation. J Exp Med. 2015;212(11):1869-82.
354. Sonnenberg GF, Monticelli LA, Elloso MM, Fouser LA, Artis D. CD4(+) lymphoid tissue-inducer cells promote
innate immunity in the gut. Immunity. 2011;34(1):122-34.
355. Gomez de Aguero M, Ganal-Vonarburg SC, Fuhrer T, Rupp S, Uchimura Y, Li H, et al. The maternal microbiota
drives early postnatal innate immune development. Science. 2016;351(6279):1296-302.
356. van de Pavert SA, Ferreira M, Domingues RG, Ribeiro H, Molenaar R, Moreira-Santos L, et al. Maternal
retinoids control type 3 innate lymphoid cells and set the offspring immunity. Nature. 2014;508(7494):123-7.
357. Goverse G, Labao-Almeida C, Ferreira M, Molenaar R, Wahlen S, Konijn T, et al. Vitamin A Controls the
Presence of RORgamma+ Innate Lymphoid Cells and Lymphoid Tissue in the Small Intestine. J Immunol.
2016;196(12):5148-55.

64
358. Qiu J, Heller JJ, Guo X, Chen ZM, Fish K, Fu YX, et al. The aryl hydrocarbon receptor regulates gut immunity
through modulation of innate lymphoid cells. Immunity. 2012;36(1):92-104.
359. Lee JS, Cella M, Colonna M. AHR and the Transcriptional Regulation of Type-17/22 ILC. Front Immunol.
2012;3:10.
360. Qiu J, Guo X, Chen ZM, He L, Sonnenberg GF, Artis D, et al. Group 3 innate lymphoid cells inhibit T-cell-
mediated intestinal inflammation through aryl hydrocarbon receptor signaling and regulation of microflora.
Immunity. 2013;39(2):386-99.
361. Kiss EA, Vonarbourg C, Kopfmann S, Hobeika E, Finke D, Esser C, et al. Natural aryl hydrocarbon receptor
ligands control organogenesis of intestinal lymphoid follicles. Science. 2011;334(6062):1561-5.
362. Vonarbourg C, Mortha A, Bui VL, Hernandez PP, Kiss EA, Hoyler T, et al. Regulated expression of nuclear
receptor RORgammat confers distinct functional fates to NK cell receptor-expressing RORgammat(+) innate
lymphocytes. Immunity. 2010;33(5):736-51.
363. Williams TW, Granger GA. Lymphocyte in vitro cytotoxicity: lymphotoxins of several mammalian species.
Nature. 1968;219(5158):1076-7.
364. Ruddle NH, Waksman BH. Cytotoxic effect of lymphocyte-antigen interaction in delayed hypersensitivity.
Science. 1967;157(3792):1060-2.
365. Gray PW, Aggarwal BB, Benton CV, Bringman TS, Henzel WJ, Jarrett JA, et al. Cloning and expression of cDNA
for human lymphotoxin, a lymphokine with tumour necrosis activity. Nature. 1984;312(5996):721-4.
366. Pennica D, Nedwin GE, Hayflick JS, Seeburg PH, Derynck R, Palladino MA, et al. Human tumour necrosis
factor: precursor structure, expression and homology to lymphotoxin. Nature. 1984;312(5996):724-9.
367. Ware CF, VanArsdale TL, Crowe PD, Browning JL. The ligands and receptors of the lymphotoxin system. Curr
Top Microbiol Immunol. 1995;198:175-218.
368. Browning JL, Ngam-ek A, Lawton P, DeMarinis J, Tizard R, Chow EP, et al. Lymphotoxin beta, a novel
member of the TNF family that forms a heteromeric complex with lymphotoxin on the cell surface. Cell.
1993;72(6):847-56.
369. Crowe PD, VanArsdale TL, Walter BN, Ware CF, Hession C, Ehrenfels B, et al. A lymphotoxin-beta-specific
receptor. Science. 1994;264(5159):707-10.
370. De Togni P, Goellner J, Ruddle NH, Streeter PR, Fick A, Mariathasan S, et al. Abnormal development of
peripheral lymphoid organs in mice deficient in lymphotoxin. Science. 1994;264(5159):703-7.
371. Alimzhanov MB, Kuprash DV, Kosco-Vilbois MH, Luz A, Turetskaya RL, Tarakhovsky A, et al. Abnormal
development of secondary lymphoid tissues in lymphotoxin beta-deficient mice. Proc Natl Acad Sci U S A.
1997;94(17):9302-7.
372. Koni PA, Sacca R, Lawton P, Browning JL, Ruddle NH, Flavell RA. Distinct roles in lymphoid organogenesis
for lymphotoxins alpha and beta revealed in lymphotoxin beta-deficient mice. Immunity. 1997;6(4):491-500.
373. Rennert PD, Browning JL, Mebius R, Mackay F, Hochman PS. Surface lymphotoxin alpha/beta complex is
required for the development of peripheral lymphoid organs. J Exp Med. 1996;184(5):1999-2006.
374. Tumanov AV, Kuprash DV, Nedospasov SA. The role of lymphotoxin in development and maintenance of
secondary lymphoid tissues. Cytokine Growth Factor Rev. 2003;14(3-4):275-88.
375. Fu YX, Chaplin DD. Development and maturation of secondary lymphoid tissues. Annu Rev Immunol.
1999;17:399-433.
376. Rennert PD, James D, Mackay F, Browning JL, Hochman PS. Lymph node genesis is induced by signaling
through the lymphotoxin beta receptor. Immunity. 1998;9(1):71-9.
377. Adachi S, Yoshida H, Kataoka H, Nishikawa S. Three distinctive steps in Peyer's patch formation of murine
embryo. Int Immunol. 1997;9(4):507-14.
378. Mebius RE, Rennert P, Weissman IL. Developing lymph nodes collect CD4+CD3- LTbeta+ cells that can
differentiate to APC, NK cells, and follicular cells but not T or B cells. Immunity. 1997;7(4):493-504.
379. Ngo VN, Korner H, Gunn MD, Schmidt KN, Riminton DS, Cooper MD, et al. Lymphotoxin alpha/beta and
tumor necrosis factor are required for stromal cell expression of homing chemokines in B and T cell areas of the
spleen. J Exp Med. 1999;189(2):403-12.
380. Dejardin E, Droin NM, Delhase M, Haas E, Cao Y, Makris C, et al. The lymphotoxin-beta receptor induces
different patterns of gene expression via two NF-kappaB pathways. Immunity. 2002;17(4):525-35.
381. Miyawaki S, Nakamura Y, Suzuka H, Koba M, Yasumizu R, Ikehara S, et al. A new mutation, aly, that induces a
generalized lack of lymph nodes accompanied by immunodeficiency in mice. Eur J Immunol. 1994;24(2):429-34.
382. Tang H, Zhu M, Qiao J, Fu YX. Lymphotoxin signalling in tertiary lymphoid structures and immunotherapy.
Cell Mol Immunol. 2017;14(10):809-18.
383. Granger SW, Ware CF. Turning on LIGHT. J Clin Invest. 2001;108(12):1741-2.

65
384. Wang J, Lo JC, Foster A, Yu P, Chen HM, Wang Y, et al. The regulation of T cell homeostasis and autoimmunity
by T cell-derived LIGHT. J Clin Invest. 2001;108(12):1771-80.
385. Gommerman JL, Browning JL. Lymphotoxin/light, lymphoid microenvironments and autoimmune disease.
Nat Rev Immunol. 2003;3(8):642-55.
386. Wang J, Chun T, Lo JC, Wu Q, Wang Y, Foster A, et al. The critical role of LIGHT, a TNF family member, in T
cell development. J Immunol. 2001;167(9):5099-105.
387. Scheu S, Alferink J, Potzel T, Barchet W, Kalinke U, Pfeffer K. Targeted disruption of LIGHT causes defects in
costimulatory T cell activation and reveals cooperation with lymphotoxin beta in mesenteric lymph node genesis. J
Exp Med. 2002;195(12):1613-24.
388. Dohi T, Rennert PD, Fujihashi K, Kiyono H, Shirai Y, Kawamura YI, et al. Elimination of colonic patches with
lymphotoxin beta receptor-Ig prevents Th2 cell-type colitis. J Immunol. 2001;167(5):2781-90.
389. Gonzalez M, Mackay F, Browning JL, Kosco-Vilbois MH, Noelle RJ. The sequential role of lymphotoxin and B
cells in the development of splenic follicles. J Exp Med. 1998;187(7):997-1007.
390. Ngo VN, Cornall RJ, Cyster JG. Splenic T zone development is B cell dependent. J Exp Med.
2001;194(11):1649-60.
391. Crowley MT, Reilly CR, Lo D. Influence of lymphocytes on the presence and organization of dendritic cell
subsets in the spleen. J Immunol. 1999;163(9):4894-900.
392. Matsumoto M, Mariathasan S, Nahm MH, Baranyay F, Peschon JJ, Chaplin DD. Role of lymphotoxin and the
type I TNF receptor in the formation of germinal centers. Science. 1996;271(5253):1289-91.
393. Zindl CL, Kim TH, Zeng M, Archambault AS, Grayson MH, Choi K, et al. The lymphotoxin LTalpha(1)beta(2)
controls postnatal and adult spleen marginal sinus vascular structure and function. Immunity. 2009;30(3):408-20.
394. Moseman EA, Iannacone M, Bosurgi L, Tonti E, Chevrier N, Tumanov A, et al. B cell maintenance of
subcapsular sinus macrophages protects against a fatal viral infection independent of adaptive immunity. Immunity.
2012;36(3):415-26.
395. Summers-DeLuca LE, McCarthy DD, Cosovic B, Ward LA, Lo CC, Scheu S, et al. Expression of lymphotoxin-
alphabeta on antigen-specific T cells is required for DC function. J Exp Med. 2007;204(5):1071-81.
396. Kumar V, Dasoveanu DC, Chyou S, Tzeng TC, Rozo C, Liang Y, et al. A dendritic-cell-stromal axis maintains
immune responses in lymph nodes. Immunity. 2015;42(4):719-30.
397. Kruglov AA, Grivennikov SI, Kuprash DV, Winsauer C, Prepens S, Seleznik GM, et al. Nonredundant function
of soluble LTalpha3 produced by innate lymphoid cells in intestinal homeostasis. Science. 2013;342(6163):1243-6.
398. Reboldi A, Arnon TI, Rodda LB, Atakilit A, Sheppard D, Cyster JG. IgA production requires B cell interaction
with subepithelial dendritic cells in Peyer's patches. Science. 2016;352(6287):aaf4822.
399. Tumanov AV, Koroleva EP, Guo X, Wang Y, Kruglov A, Nedospasov S, et al. Lymphotoxin controls the IL-22
protection pathway in gut innate lymphoid cells during mucosal pathogen challenge. Cell Host Microbe.
2011;10(1):44-53.
400. Wang Y, Koroleva EP, Kruglov AA, Kuprash DV, Nedospasov SA, Fu YX, et al. Lymphotoxin beta receptor
signaling in intestinal epithelial cells orchestrates innate immune responses against mucosal bacterial infection.
Immunity. 2010;32(3):403-13.
401. Spahn TW, Maaser C, Eckmann L, Heidemann J, Lugering A, Newberry R, et al. The lymphotoxin-beta
receptor is critical for control of murine Citrobacter rodentium-induced colitis. Gastroenterology. 2004;127(5):1463-
73.
402. Zhang Y, Kim TJ, Wroblewska JA, Tesic V, Upadhyay V, Weichselbaum RR, et al. Type 3 innate lymphoid cell-
derived lymphotoxin prevents microbiota-dependent inflammation. Cell Mol Immunol. 2018;15(7):697-709.
403. Kim TJ, Upadhyay V, Kumar V, Lee KM, Fu YX. Innate lymphoid cells facilitate NK cell development through a
lymphotoxin-mediated stromal microenvironment. J Exp Med. 2014;211(7):1421-31.
404. Elewaut D, Ware CF. The unconventional role of LT alpha beta in T cell differentiation. Trends Immunol.
2007;28(4):169-75.
405. Elewaut D, Brossay L, Santee SM, Naidenko OV, Burdin N, De Winter H, et al. Membrane lymphotoxin is
required for the development of different subpopulations of NK T cells. J Immunol. 2000;165(2):671-9.
406. Brinkman CC, Iwami D, Hritzo MK, Xiong Y, Ahmad S, Simon T, et al. Treg engage lymphotoxin beta receptor
for afferent lymphatic transendothelial migration. Nat Commun. 2016;7:12021.

66
 II. Aims and outline

Despite differences between the splenic marginal zone in humans and mice, the crucial
contribution of this compartment to immune responses against blood-borne antigens is
beyond doubt. Nevertheless, the molecular interactions that control the development,
homeostasis and function of immune cells inhabiting the marginal zone are
incompletely understood. Marginal zone B (MZB) cells require Notch2 instruction and B
cell receptor (BCR) signals for their development and shuttle continuously between the
marginal zone and the B cell follicle. An unresolved question in the field is how Notch2
activation is controlled, and how it intersects with BCR signaling in the regulation of B
cell fate decisions. Similarly, type 2 conventional dendritic cells (cDC2s) that inhabit
marginal zone bridging channels also depend on Notch2, but the factors that regulate
activation of the Notch pathway in cDCs have remained elusive.

Serendipitously, we found that the development of MZB cells and bridging channel
cDC2s is impaired in mice that lack the Ste20 kinase thousand-and-one kinase 3
(TAOK3). In chapter III, we set out to identify the molecular mechanism by which
TAOK3 regulates MZB cell development. To discriminate between cell-intrinsic and cell-
extrinsic roles of TAOK3, we made use of competitive mixed bone marrow (BM)
chimeras and Cre-Lox recombination technology. We studied in chapter IV how loss of
TAOK3 affects the phenotype, function and transcriptional program of dendritic cells,
and addressed whether the DC defects in Taok3-/- mice are secondary to, or
independent of the absence of MZB cells. Using both in vivo and in vitro approaches, we
also investigated whether the function of Taok3 related to Notch signaling.

In our efforts to further understand the cellular interactions that are required for the
development and function of cDC2s, we also investigated the role of the lymphotoxin
(LT) pathway. While the DC-intrinsic requirement for the lymphotoxin β receptor
(LTβR) is well established, it is less clear which cells provide the necessary ligands. In
chapter V, we analyze the niche cells that control cDC2 development through the LT-
LTβR axis by studying the composition of the dendritic cell compartment in mice that
lack various lymphocyte subsets.

We observed that Taok3-/- mice, which lack a subset of cDCs, develop neutrophilia and
splenomegaly when aged. This phenotype was reminiscent of several reports describing
the onset of a myeloproliferative disorder (MPD) in various mouse models with
defective cDC development. Accordingly, it has been suggested that cDCs might control
myeloproliferation. In chapter VI, we addressed this hypothesis using mice that lack
the transcription factor IRF8. These mice harbor defects in multiple subtypes of

67
dendritic cells and develop MPD. Using conditional Irf8fl/fl mice and mixed bone marrow
chimeras, we studied how loss of specific DC subsets affects hematopoiesis.

In chapter VII, key findings of this thesis are discussed and placed in perspective.
Finally, we highlight unresolved questions that particularly warrant further
investigation.

68
 III. Results part one

Transitional B cells commit to marginal zone B

cell fate by Taok3-mediated surface expression
of ADAM10

Hamida Hammad*, Matthias Vanderkerken*, Philippe Pouliot*, Kim Deswarte, Wendy

Toussaint, Karl Vergote, Lana Vandersarren, Sophie Janssens, Ioanna Ramou, Savvas N.
Savvides, Jody J. Haigh, Rudi Hendriks, Manfred Kopf, Katleen Craessaerts, Bart de
Strooper, John F. Kearney, Daniel H. Conrad & Bart N. Lambrecht

* These authors contributed equally to this work

Nature Immunology 18, 313-320 (2017).

Author contribution:
Matthias Vanderkerken performed flow cytometry and imaging experiments in figures
1b-g, 2a-d, 3c-d, 4a-b, 5c-f, analyzed the data, and contributed to the writing of the
manuscript.

69
ABSTRACT

Notch2 and B cell antigen receptor (BCR) signaling determine if transitional B cells
become marginal zone B (MZB) or follicular B (FoB) cells in the spleen, but it is
unknown how these pathways are related. We generated Taok3–/– mice and found cell-
intrinsic defects in the development of MZB, but not FoB cells. Type 1 transitional (T1) B
cells required Taok3 to rapidly respond to ligation with the Notch ligand Delta-like 1.
BCR ligation by endogenous or exogenous ligands induced the surface expression of the
metalloproteinase ADAM10 on T1 cells in a Taok3-dependent manner. T1 B cells
expressing surface ADAM10 were committed to become MZB cells in vivo, whereas T1 B
cells lacking expression of ADAM10 were not. Thus, during positive selection in the
spleen, BCR signaling causes immature T1 cells to become receptive to Notch ligands via
Taok3-mediated surface expression of ADAM10.

INTRODUCTION
B lymphocytes are categorically divided in B1 and B2 cells. B1 cells derive from fetal
progenitors and react to a restricted set of microbial ligands in a T cell-independent (TI)
manner in serosal cavities and spleen1. B2 cells develop continuously in the bone
marrow and further mature into follicular B (FoB) cells and marginal zone B (MZB)
cells. FoB cells have a broad repertoire of specificity, recirculate between lymphoid
organs and give rise to germinal center B cells that undergo somatic hypermutation in a
T-cell dependent (TD) manner. MZB cells shuttle continuously between the marginal
zone and follicles of the spleen and produce antibodies to encapsulated bacterial and
polysaccharide TI antigens. It is still poorly understood how immature transitional B
cells are instructed to become a FoB or MZB cell, and when exactly this lineage choice is
made. It is widely accepted that MZB instruction requires triggering of Notch2 on
developing B cells by Delta-like 1 (Dll1) expressed by splenic red pulp sinus endothelial
cells or marginal zone reticular cells (MRCs). Notch2 cleavage by a metalloproteinase
and disintegrin-10 (ADAM10) and g-secretase then releases the intracellular domain of
Notch (NICD) that binds to the transcription factor RBPJκ in the nucleus, and instructs
MZB development, together with NFκB signaling emanating from the BAFF receptor 2, 3,
4, 5, 6, 7, 8, 9, 10, 11, 12, 13. The quality of B cell antigen receptor (BCR) signals during positive

selection of B cell precursors in the spleen is equally important in B cell fate decisions14,
15, 16, and it was proposed that weak or strong BCR signals might render cells receptive

or resistant to Notch instruction17, 18. Yet how BCR repertoire or signaling controls
Notch responsiveness is currently poorly understood.

The Ste20 family kinases are serine-threonine kinases that participate in a variety of
signaling pathways triggered by cellular stress19, 20. The Tao kinase subfamily has three
members in mammals (TAOK1, also known as proteins MAP3K16, PSK2 or MARKK;
TAOK2 (MAP3K17, PSK1) and TAOK3 (MAP3K18, JIK or DPK)), whose function is
largely unknown. Here, we generated Taok3–/– mice and found that these mice lacked

70
MZB cells, whereas FoB cells were intact. By carefully unraveling the molecular
mechanism of this deficiency we have discovered how BCR signaling intersects with
Notch signaling in immature transitional B cells undergoing positive selection.

RESULTS

Taok3–/– mice lack MZB cells

In wild-type mice, the expression of mRNA for Taok3 was predominantly found in bone
marrow and immune tissues like spleen, thymus and lymph nodes, but also in lung and
gut (Fig. 1a). To gain insight into the biology of Taok3, we generated Taok3–/– mice
(Supplementary Fig. 1a-e). Overall there was no difference in the cellularity of the
various lymphoid organs in 6-12 week old mice (Supplementary Fig. 1f). In the spleen,
there were no gross differences in the percentage of eosinophils, monocytes, natural
killer (NK) cells and NKT cells between Taok3–/– and wild-type mice, yet there was a
consistent reduction in the amount of CD11c+ dendritic cells in Taok3–/– mice (Fig. 1b).
There was a small yet consistent increase in the percentage of Ly6G+ granulocytes in the
spleen of Taok3–/– mice. The distribution of CD3+ T cells and CD19+ B cells was
comparable, yet there was a 25-30% reduction in the amount of CD8+ T cells (Fig. 1c).
Within the B cells of the spleen, analysis of cell surface expression of CD21/35 and CD23
discriminates between transitional CD21/35–CD23–B cells that are immature B cells
that have just arrived from the bone marrow, CD21/35intCD23hi FoB cells and
CD21/35hiCD23– MZB cells. Transitional B cells express CD93 and can be further
divided in CD93+ IgMhi CD23– T1, and CD93+IgMhiCD23hi T2 cells. In Taok3–/–
splenocytes, the percentage of T1, T2, and FoB cells were comparable to wild-type,
whereas MZB cells were almost completely absent (Fig. 1d-f).

Histological examination of the spleen revealed that the characteristic rim of IgM+CD1d+
MZB cells separated from the IgMlo B cell follicles by the marginal sinus was absent in
Taok3–/– mice (Fig. 1g). Resident marginal zone metallophillic macrophages (MMM,
expressing CD169) that line the marginal sinus were present and correctly localized in
Taok3–/– mice (Fig. 1h). MZB cells are specialized in capturing TI type 2 antigens like
Ficoll21. Two h after i.v. injection of FITC-Ficoll, we detected labeled B cells in vicinity of
CD169+ MMM in wild-type mice but not Taok3–/– mice (Fig. 1h). Collectively, Taok3–/–
mice lack MZB cells without compensatory alterations in other B cell subsets.

Humoral immune response of Taok3–/– mice

The baseline serum concentration of immunoglobulins (Ig) was comparable to wild
type, with a tendency for increased IgG3 in Taok3–/– mice (Supplementary Fig 2). MZB
cells acutely produce TI Ig to polysaccharide particulate antigens like Ficoll1,22.
However, in response to immunization with trinitrophenyl (TNP) hapten-conjugated
Ficoll, there was no reduction in the concentration of TNP-specific IgG1, IgG3 and IgM in
Taok3–/– compared with wild-type mice (Fig. 1i and Supplementary Fig 2). The intact
TNP-specific Ig response as not due to compensatory increase in recirculating MZB cells

71
Figure 1: Taok3–/– mice lack marginal zone B cells and have reduced humoral responses to T-
independent antigens (legend on next page).

72
Figure 1: Taok3–/– mice lack marginal zone B cells and have reduced humoral responses to T-
independent antigens. (a) Taok3 mRNA expression in different tissues of C57Bl/6 mice. (b) Percentage
of innate immune cells in the spleens of Taok3+/+ and Taok3–/– mice. (c) Percentage of T cell subsets in the
spleens of Taok3+/+ and Taok3–/– mice. (d) Flow cytometry staining of spleens for CD21/35hi CD23lo
marginal zone B cells (MZB) and CD21/35- CD23hi follicular B cells (FoB). (e) Percentage of FoB cells and
MZB cells within splenic CD19+ B cells in Taok3+/+ and Taok3–/– mice. (f) Percentage of CD93hi transitional
T1B and T2B cells in the spleens of Taok3+/+ and Taok3–/– mice. (g) Immunofluorescence staining of
CD1d+ (green) IgM+ (red) marginal zone B cells (double positive MZB cells are yellow) in the spleens of
Taok3+/+ and Taok3–/– mice. Scale bar, 100 mm. (h) Immunofluorescence uptake of i.v-injected FITC-
labeled Ficoll (green) in the spleen of Taok3+/+ and Taok3–/– mice. CD169+ metallophilic macrophages are
stained in red. DAPI nuclear counterstaining in blue. Scale bar, 100 mm. (i) Serum IgM titers at baseline
and 7 days following TNP-Ficoll injection in Taok3+/+ and Taok3–/– mice. (j) Percentage of B1a and B1b
cells in peritoneal lavages of Taok3+/+ and Taok3–/– mice. (k) OD values for phosphorylcholine-specific
IgM titers following injection of Taok3+/+ and Taok3–/– mice with Streptococcus pneumoniae. *p<0.05
(Mann and Whitney test (b, c, e, k)). Data are representative of three experiments (b-f, j; mean + s.e.m.; n =
4-5 mice per experiment), or one experiment (g-I, k; mean + s.e.m of n = 6 mice per group in i and k, and
at least 4 images in g-h).

outside the spleen (data not shown). B1 B cells can also respond to TNP-Ficoll antigen,
in the complete absence of MZB cells2. The numbers of B220+CD5+ B1a and CD5- B1b B
cells of the peritoneal cavity were identical in wild-type and Taok3–/– mice (Fig. 1j),
which may explain the intact humoral immune response to TNP-Ficoll. MZB cells are
indispensable for mounting low affinity IgM phosphorylcholine (PC)-specific antibody
responses to encapsulated bacteria like Streptococcus pneumoniae when these reach the
bloodstream1, 2, 23. We injected 1 × 108 heat inactivated bacteria i.v. and measured the
IgM response antigen 5 days later. Whereas wild-type mice readily mounted an anti-PC
IgM response, the titer of PC-specific IgM was severely reduced in Taok3–/– mice (Fig.
1k). These findings show that Taok3–/– mice form normal responses to the TI-2 antigen
Ficoll, but not to the TI-1 antigen S. pneumoniae.

MZB defect of Taok3–/– mice is B cell-intrinsic

The development and survival of MZB cells depends on integrin signaling and correct
positioning in the marginal zone18, 24, 25, 26. As the structure of the marginal zone was not
normal in Taok3–/– mice, and other defects were found in neutrophils, DCs and CD8+ T
cells, we next addressed if the defect in MZB cell development was cell intrinsic, or
caused by changes in the stromal structures or other hematopoietic cells. We therefore
created mixed bone marrow chimeric mice by lethally irradiating CD45.1.CD45.2
C57BL/6 mice and reconstituting them with an equal mix of CD45.1 wild-type and
CD45.2 Taok3–/– BM cells. Chimerism was complete 6-8 weeks after transfer with each
donor genotype contributing to 50% of monocytes and neutrophils in the blood and
various organs (data not shown). Whereas CD23hiCD21/35lo Fo B were completely
chimeric (Fig. 2a), CD21hiCD23loMZB cells were generated almost exclusively from the
CD45.1 wild-type hematopoietic cells. As the marginal zone was fully restored in
chimeric mice (data not shown), these findings demonstrate that the defect in MZB
development is not due to changes in environment, but cell-intrinsic. We next evaluated
at which stage of B cell development the MZB development might be compromised. In

73
the bone marrow of chimeric mice, pro-B cells, cycling pre-B cells, pre-B cells and
immature B cells were all equally distributed amongst both genotypes (Fig. 2b). In the
spleen however, there was an overrepresentation of T1 cells of the CD45.1 wild-type
genotype. This effect was less marked in T2 B cells. Collectively, loss of MZB cells in
Taok3–/– mice is cell intrinsic and B cell development is compromised from the T1 stage
onwards.

Gene dosage effect of Taok3 deficiency

To address if the amount of Taok3 affects B cell development, we analyzed
heterozygous Taok3+/– mice and took advantage of the fact that the gene trap construct
made to inactivate the Taok3 locus contained a splice acceptor that was flanked by loxP
sites, allowing the partial reversal of the gene trap in cells of interest (Supplementary
Fig. 1a). We therefore crossed wild-type, Taok3+/– and Taok3–/– mice to mb1Cre mice,
in which Cre recombinase expression is under the control of the Cd79a promoter, active
from the pre B cell stage onwards27. When we analyzed the composition of splenic B
cells in wild-type Cre+, Taok3–/– Cre– or Taok3–/– Cre+ mice, we found that MZB cells
were partially recovered by gene-trap reversal exclusively in the B cell lineage (Fig.
2c,d). The numbers of MZB cells of heterozygous Taok3+/– were intermediate between
those of wild-type and Taok3–/– mice, and gene trap reversal of Taok3+/– mice led to a
stronger recovery of MZB cells compared with reversal in Taok3–/– mice. Collectively,
the defect in MZB development in Taok3–/– mice can be partially reversed by B cell
specific gene trap removal.

Defective Notch activation in Taok3–/– B cells

The gene dose-dependent lack of MZB cells and preserved induction of immune
responses to TNP-Ficoll in Taok3–/– mice resembles the phenotype of mice lacking
Notch2 or RBP-Jk in the B cell lineage2, 6, 7. It has been shown that transitional B cells are
instructed to become MZB cells via interaction with the Notch ligand Delta-like 1 (Dll1)
expressed on stromal cells4, 5. To address if Notch–mediated development of MZB cells
was disrupted in Taok3–/– mice, we cultured CD93+ transitional splenic B cells on OP9-
GFP cells, or on OP9 cells stably transfected with Delta-like 1 (OP9-Dll1). Notch
signaling was studied by measuring the mRNA of direct Notch targets genes Dtx1, Hes1,
Hes5 and Hey111. When transitional B cells were cultured for 4h on OP9-Dll1 cells, there
was induction of Notch target genes in wild-type but much less efficiently so in Taok3–/–
mice (Fig. 3a). Strikingly, these differences were no longer apparent 18h later (data not
shown), demonstrating that Taok3 was mainly involved in controlling rapid
responsiveness to Dll1 ligation.

It is exceedingly difficult to model MZB development in vitro28, 29. However, when wild-
type CD93+ transitional B cells were co-cultured with OP9-GFP cells in the presence of
the B cell growth factor BAFF, there was induction of CD21 and IgM on 8-10% of
Taok3+/+ cells. Culture on OP9-Dll1 induced expression of IgM and CD21 on roughly
20% of cells. However, Taok3–/– transitional B cells only upregulated CD21 and IgM on

74
10% of B cells when cultured on OP9-Dll1 cells (Fig. 3b). These differences were not
due to alterations in BAFF signaling (Supplementary Fig. 3a) or BAFF induced survival
(Supplementary Fig. 3b). Thus, Taok3 is required in transitional B cells for rapid Notch
signaling and differentiatiation towards the MZB phenotype.

Figure 2: The marginal zone B cell defect in Taok3–/– mice is B cell-intrinsic and subject to Taok3
gene dosage. (a) Flow cytometry staining of spleens from chimeric mice to evaluate the ratio between
CD45.1 Taok3+/+ and CD45.2 Taok3–/– cells within the marginal zone B (MZB) cell (upper right plot) and
the follicular B (FoB) cell (lower right plot) gates. The numbers adjacent to the gates represent the
percentage of cells within the gates. (b) Quantification of the ratios between CD45.1 Taok3+/+ and CD45.2
Taok3–/– cells on the populations gated as in (a). (c) Gene-trap reversal recovery of marginal zone B cells
in Taok3–/– mice crossed to mb1Cre mice was analyzed by flow cytometry. (d) Quantification of the Gene-
trap reversal recovery shown in (d) in different mouse genotypes. Data are representative of three
experiments (b), two experiments (d-e), or a pool of 2 independent experiments (c); n = 5-7 mice per
experiment).

75
To address this point more directly, we performed a rescue experiment in which Notch2
expression was increased in Taok3–/– B cells. It was recently reported that Irf4–/– mice
have an increase in MZB cells explained through a stabilization of intracellular Notch212.
We set up breedings of Taok3–/– and Irf4–/– mice. We confirmed that Irf4–/– mice had a
strong increase in MZB cells compared with Irf4+/+ mice (Fig. 3c-d). When Taok3–/–
Irf4–/– mice were analyzed, the deficiency of MZB cells was completely reversed, and
there were even higher percentages MZB cells compared with wild-type animals. We
validated that Taok3–/– mice did not have an increased intracellular accumulation of
IRF4 protein (data not shown). These data show that Taok3-deficiency leads to
defective Notch signaling in transitional B cells and that the MZB phenotype of Taok3–/–
mice can be rescued by IRF4 deficiency.

Taok3 controls the surface expression of ADAM10

We next studied how lack of a kinase might lead to reduced Notch signaling. By
immunoblot analysis on lysates, Notch2 expression in Taok3–/– was similar to wild-type
mice (Supplementary Fig 4). Looking at immunostained splenic sections and flow
cytometry we found that the intensity of CD23 staining on B cell follicles (Fig. 4a) and
FoB cells (Fig. 4b) were consistently higher in Taok3–/– compared with Taok3+/+ mice.
This was not caused by altered Cd23 mRNA (data not shown). We have previously
shown that the metalloproteinase ADAM10 determines the intensity of CD23 on the B
cell plasma membrane, by cleaving CD23 to generate a sCD23 fragment10, 30. We found
that the baseline serum concentration of sCD23 was significantly reduced in Taok3–/–
mice compared with wild-type mice (Fig. 4c). Injection of the 19G5 antibody against the
CD23 stalk causes a conformational change in CD23 rendering it highly sensitive to
cleavage by ADAM1010. After 19G5 injection in wild-type mice, there was an almost
100-fold increase in the serum levels of sCD23. This effect was strongly reduced in
Taok3–/– mice, suggesting a defect in the enzymatic activity of ADAM10 (Fig. 4d). The
defect in ADAM10 bioactivity was addressed further by studying the cleavage of other
ADAM10 substrates. A well-known substrate of ADAM10 is the amyloid precursor
protein (APP), involved in the pathogenesis of Alzheimer’s disease31. The ADAM10
causes the cleavage of cell bound APP into a membrane bound APP a-stub and the
release of soluble APPa. As APP is produced by mouse fibroblasts as well as neuronal
cells, we generated mouse embryonic fibroblast (MEF) cell lines of wild-type and Taok3-
/– mice and found that Taok3–/– MEFs had a consistently lower expression of the APP a-

stub on their cell surface, suggestive of reduced ADAM10 activity in Taok3–/– mice.
When MEFs were transfected with full length APP, we could also measure the release of
sAPPa in the concentrated supernatant, and found that MEFs derived from Taok3–/–
mice generated less sAPPa compared with Taok3+/+ mice, suggestive indeed of reduced
ADAM10 bioactivity (Supplementary Fig. 5).

ADAM10 is known to cleave substrates when expressed on the cell surface, and we have
reported that ADAM10 can be detected by flow cytometry on lymphocytes10. We found
that 5-6% of CD93+ transitional B cells of the spleen of wild-type mice expressed

76
Figure 3: Taok3–/– transitional B cells have a defect in Notch activation (legend on next page).

77
Figure 3: Taok3–/– transitional B cells have a defect in Notch activation.
(a) Quantification of Notch target gene induction in CD93hi transitional B cells from Taok3+/+ and Taok3–/–
mice after 4 hours of culture on OP9-Dll1 in the absence of BAFF. (b) Flow cytometry analysis of the
induction of CD21/35 expression on Transitional B cells cultured on OP9-GFP or OP9-Dll1 for 5 days in
the presence of BAFF. (c) Flow cytometry analysis (upper panel) and quantification (lower panel) of
marginal zone B cells in the spleens of Taok3+/+ and Taok3–/– mice crossed to Irf4-/- mice. (d)
Immunofluorescence staining of marginal zone B cells in the spleen of Taok3+/+ and Taok3–/– mice
crossed to Irf4-/- mice. Data are representative of two experiments (b-d). d: mean + s.e.m.; n = 2-6 mice
per experiment. i, at least 3 images per group were acquired.

ADAM10 on their cell surface, yet this expression was almost completely absent in
Taok3–/– mice (Fig. 4e). However, the expression of the pool of ADAM10, composed of
pro- and of mature ADAM10, in cell lysates of transitional B cells was identical
betweenwild-type and Taok3–/– mice (Supplementary Fig. 6). In Irf4-/- T1 B cells there
was a marked increase in surface ADAM10 staining compared with Irf4+/+ mice. In
compound Taok3–/– Irf4–/– mice the defect of ADAM10 expression seen in Taok3–/– mice
was restored well above the intensity seen in Taok3 wild-type mice, suggesting that lack
of IRF4 restored MZB numbers through upregulation of ADAM10 on T1 B cells (Fig. 4f).
There was a strong correlation between the number of T1 cells expressing surface
ADAM10 and the final pool of mature MZB cells in mice of various genotype (Fig 4g).
Collectively, Taok3-/- mice lacked surface ADAM10 on T1 B cells and the return of
ADAM10 on T1 B cells was accompanied by return of MZB cells, suggesting that the
expression of ADAM10 on the surface of might be a crucial event in MZB development.

ADAM10 marks MZB commitment in T1B cells

The lineage choice of transitional B cells to become FoB or MZB cell depends on the
repertoire of the BCR and the strength of BCR signaling14, 15, 16. We therefore reasoned
that positive selection events of MZB cells acting through the BCR in transitional B cells
might affect the surface expression of ADAM10. To test this, we first purified CD93+
transitional B cells and stimulated them with soluble BCR crosslinking using anti-IgM
F(ab')2 fragments. In unstimulated wild-type cells, ADAM10 expression was mainly
found in discrete punctate area inside transitional B cells. Within 20 min following BCR
stimulation, there was a relocalization of ADAM10 to a single cap-like region on the B
cell, a lipid-raft and tetraspanin rich region concentrating the BCR, Iga, Ras, and BLNK32.
Simultaneous staining for IgM revealed colocalization with ADAM10 with the capped
BCR (data not shown). This capping of ADAM10 was not seen in BCR cross-linked
Taok3–/– transitional B cells (Fig. 5a).

We next studied if positive selection in a more physiological context would also affect
ADAM10 surface expression on T1 B cells. Previously, we have created transgenic mice
expressing the VH81x heavy chain, that pairs with a limited repertoire of endogenous
light chains through binding constraints, thus generating a BCR that recognizes an
endogenous self- or microbiome-derived ligand. These mice have been used to
understand the mechanisms of positive selection of cells in the MZB pool14. We crossed

78
Taok3–/– mice to VH81x Tg mice, and stained spleen sections for the presence of
clonotype specific B cells (Fig 5b). In VH81x Tg mice with wild-type levels of Taok3, the
clonotype+ cells were almost exclusively found in the splenic marginal zone, whereas in
VH81x Tg Taok3–/– mice, these cells were absent. Flow cytometry also showed massive
expansion of MZB cells in VH81x Tg mice, yet when Taok3 was inactivated, MZB cells
were absent (Fig. 5c). We also performed flow cytometry using antibodies recognizing
the VH81x Ig heavy chain and the pairing Vk1C light chain. This staining revealed that in
Taok3+/+ VH81x Tg mice the majority of the clone was found in the CD21hi MZB cell gate,
whereas in Taok3–/– VH81x Tg there were less clonotype specific B cells among splenic
CD19+ cells, and the majority of remaining clonotype specific cells had a FoB cell
phenotype (Fig 5d-e). The percentage of T1 cells expressing surface ADAM10 was
strongly increased in VH81x Tg mice, an effect that was abolished by crossing these
mice to Taok3–/– mice (Fig. 5f). Again there was a strong correlation between the
percentage of mature MZB cells and the percentage of surface ADAM10+ T1 cells in mice
across all genotypes (data not shown).

We finally reasoned that expression of ADAM10 on T1B cells might mark commitment
of these cells to become MZB cells. The high numbers of T1 B cells in the spleens of
VH81x Tg mice allowed us to obtain sufficient cells to perform adoptive transfer
experiments into B cell-deficient Rag2-/- mice. We therefore purified CD45.2+ CD93+
transitional B cells by magnetic pre-enrichment and subsequently sorted cells into
ADAM10 surface-positive CD23- T1B cells, ADAM10 surface-negative CD23- T1B cells,
and CD23+ T2B cells (which do not express high levels of ADAM10 anyway, see fig. 4e).
These were injected simultaneously with ten times higher numbers of CD45.1 wild type
splenic cells, to avoid homeostatic proliferation that might bias to MZB development
(Fig. 5g). Five days after transfer, the fate of transferred CD45.2 cells was studied by
flow cytometry on spleen cells and confocal analysis of the spleen. Whereas T2B cells
and ADAM10 surface-negative T1 B cells gave rise to both FoB and MZB cells, the
ADAM10 surface-positive T1 B cells exclusively gave rise to CD21hi CD23- MZB cells
(Fig. 5h). Histological analysis of the spleen also revealed that these CD45.2+ ADAM10+
T1 derived MZB cells were predominantly found in the marginal zone of the spleen (Fig.
5i). Previously, others have proposed that preMZB cells can be identified amongst a
pool of TB cells, and that these cells exhibited higher expression of CD21 and CD1d, but
it is unclear if this would be a T1B or T2B stage6, 16. We found that T1 B cells that
expressed surface levels of ADAM10 were higher in the expression of CD21 and CD1d,
compared with T1B cells lacking ADAM10 surface expression (Supplementary Fig 7).
These data confirm our hypothesis that ADAM10 surface expression marks progenitor
T1B cells committed to become MZB cells after positive selection.

79
Figure 4: Taok3 controls the surface expression of ADAM10. (a) Immunofluorescence staining of
CD23 (red) in the spleen of Taok3+/+ and Taok3–/– mice. White, CD169+ metallophilic macrophages. Blue,
DAPI nuclear counterstaining. Scale bar, 100mm. (b) Flow cytometry analysis of CD23 expression on
Taok3+/+ and Taok3–/– splenic Follicular B cells. (c) Concentrations of soluble CD23 in serum of Taok3+/+
and Taok3–/– mice. (d) Concentrations of soluble CD23 in serum of Taok3+/+ and Taok3–/– mice mice
treated or not with 19G5 antibody. (e) Flow cytometry analysis of surface ADAM10 expression on non-
permeabilized splenic CD93hi transitional B cells of Taok3+/+ and Taok3–/– mice. (f) Flow cytometry
analysis (left panels) and quantification (right panel) of ADAM10 expression on splenic CD93hi
transitional B cells of Taok3+/+ and Taok3–/– mice crossed to Irf4-/- mice. (g) Correlation between the
percentage of marginal zone B cells and the percentage of ADAM10+ B cells in the spleen. *p<0.05 (Mann
and Whitney test (c, d, f)). Data are representative of at least three experiments (b, e), two experiments (c,
d), or one experiment (f). (c, d): mean + s.e.m.; n = 4-6 mice per experiment. (f): mean + s.e.m.; n = 2 mice
per experiment.

80
DISCUSSION

Our understanding of the development of splenic MZB from immature transitional B

cells is dominated by three lines of thought18. First, integrin and chemokine signals are
crucial for retention and complex shuttling of MZB cells in and around the marginal
zone. Loss of these interactions led to loss of MZB cells in many mouse strains21, 25, 26, 33.
Secondly, Notch2 ligation by Dll1 expressed on stromal cells is required for MZB
development2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, and maintenance of MZB identity12, 34. Thirdly, the
quality and the strength of the BCR repertoire determine whether positive selection of
immature B cells by self ligands or microbiome derived ligands leads to deletion, FoB
cell or MZB development14, 15, 16, 35. It has been unclear how these three pathways are
related.

Pillai et al. proposed that strong BCR signals favour FoB, whereas weak BCR signals
promote MZB cell development17, 18, 36, 37, although other investigators refuted this
idea11, 15. It was proposed that strong BCR signals render transitional cells in the follicle
impervious to the presence of Dll1 mediated triggering of Notch2, whereas weak BCR
signaling may enhance the expression of one or more components of the Notch2
signaling pathway36. These inhibitory or enhancing signals or the precise stage of B cell
development where BCR signaling and Notch permissiveness intersect have never been
identified to date. Here, by careful analysis of Taok3–/– mice we show that BCR
mediated positive selection of B cell progenitors at the T1B cell stage is linked to MZB
development through acquisition of membrane expression of ADAM10 that cleaves and
activates Notch2. ADAM10 is expressed on a subset of transitional B cells, is
redistributed to the immunocap following BCR crosslinking, and highly expressed on
the surface of T1B cells. We believe that the 5-10% of ADAM10+ transitional B cells that
are found in the steady state represent the cells that are undergoing positive selection
to become MZB cells, supported by our observation that these cells only become MZB
cell upon adoptive transfer.

How exactly B cell positive selection and BCR signaling intermediates cause Taok3
activation and ADAM10 surface expression will require further study. The levels of
Taok3mRNA remain stable throughout B cell development from hematopoietic cells,
and are not altered during B cell activation (data not shown) suggesting that Taok3
might be mainly regulated by posttranslational modifications or protein stabilization. It
was recently proposed that ADAM10 forms a homodimer in the cell membrane as does
ADAM17, a key feature in the proposed regulation mechanism of ADAM activation38.
Although we did observe differences in the processing and oligomerization of ADAM10
in Taok3–/– mice (data not shown), more research is warranted. Much more is known
about the regulation of ADAM17, which can also cleave Notch. The phosphorylation of
ADAM17 by ERK kinase on Threonine735 regulates ADAM17 dimerization and
enzymatic activity39. A conserved threonine in position 719 might have the same effect
on dimerization of ADAM10, a hypothesis we are currently testing. Taok3 is a

81
Figure 5: ADAM10 expression on transitional B cells marks commitment to the MZB cell fate
(legend on next page).

82
Figure 5: ADAM10 expression on transitional B cells marks commitment to the MZB cell fate. (a)
Immunofluorescence staining of ADAM10 (red) on purified CD93+ transitional B cells obtained from
Taok3+/+ and Taok3–/– mice, and stimulated or not with anti-IgM F(ab')2 fragments for 15 minutes. Blue,
nuclear counterstaining. Scale bar, 5 mm. (b) Immunofluorescence staining of marginal zone B cells
(green) in the spleens of Taok3+/+ and Taok3–/– mice crossed to VH81x Tg mice using MZ-21 clonotypic
antibodies recognizing the heavy chain Ig transgene. Blue, nuclear counterstaining. Scale bar, 100 mm.
(c) Percentage of marginal zone B cells within CD19+ B cells of Taok3+/+ and Taok3–/– mice crossed to
VH81x Tg mice. (d) Flow cytometry analysis of the phenotype of clonotype+ cells in the spleen of Taok3+/+
and Taok3–/– mice crossed to VH81x Tg mice. (e) Percentage of clonotype+ cells within B cells in the
spleen of Taok3+/+ and Taok3–/– mice crossed to VH81x Tg mice. (f) Flow cytometry analysis of ADAM10
expression on splenic CD93hi transitional B cells of Taok3+/+ and Taok3–/– mice crossed to VH81x Tg mice.
(g) Flow cytometry staining of the spleens of Rag-–/–recipients treated as in (f) for the presence of
CD21/35hi CD23lo marginal zone B cells and CD21/35- CD23hi Follicular B cells. (h) Immunofluorescent
staining of ADAM10+ CD45.2 (red) transferred T1B cells in the spleen of Rag–/– recipient mice. Green,
CD169+ metallophilic macrophages. Blue, DAPI nuclear counterstaining. Scale bar, 100 mm. *p<0.05
(Mann and Whitney test (c, e). Data are representative of at least two experiments. (c, e): mean + s.e.m.; n
= 4-12 mice per experiment.

serine/threonine MAP3 Kinase and is likely upstream of ERK, JNK or p38 Map Kinase
(unpublished observations and 40). Although it is likely therefore that lack of MAPK
mediated ADAM10 surface expression is the explanation for the MZB phenotype of
Taok3-deficient mice, we also need to consider additional effects of Taok3 on MAPK-
driven activation of NFκB, as canonical NFκB1 collaborates with Notch 2 in driving MZB
fate determination8. We found that the ADAM10 mediated cleavage of APP is also
reduced in Taok3–/– mice, suggesting that the BCR is not the only upstream regulator of
Taok3 and ADAM10 bioactivity. However, Taok3 is not absolutely or always required
for ADAM10 or Notch2 activation, as Taok3–/– mice do not phenocopy all aspects of
ADAM10 or Notch2 deficiency. Adam10–/– or Notch2–/– mice are embryonic lethal,
whereas Taok3–/– mice are not. Our in vitro findings using OP9-Dll1 cells to stimulate
transitional B cells suggest that the phenotype of Taok3–/– mice is the result of
disruption of short-lived Notch-Notch-ligand interactions. Such short-lived interactions
might occur on the MRC network of the B cell follicles, and might be restricted in time
due to the continuous shuttling behavior of MZB cells from the MZ to the B cell follicles5,
24.

Our data resolve a longstanding confusion in the field as to the precise identity of MZB
precursors. It was long held that MZB cells emanate from progenitors with a CD23+ T2B
phenotype that express CD1d and CD216, 10, 17, 41. Others have also proposed that FoB
cells can develop into MZB cells, particularly in immunodeficient and lymphopenic
hosts, a process that might be driven by homeostatic proliferation24, 41, 42. Data from
mice in which the repertoire of developing MZB cells was followed by measuring the
junctional diversity of the heavy chain CDR3 region revealed that T1B cells were the
most likely direct progenitors for MZB cells16. Our data using adoptive transfer, as well
as the strong correlations between ADAM10 expression on T1B cells and final

83
population size of the MZB pool suggest that indeed commitment of MZB cells is made
at the T1B cell stage, and that surface ADAM10 staining is a robust marker for
identifying cells in which MZB commitment is complete. However, we could show that
some T2B cells could still differentiate into MZB cells. We predict that also those T2B
cells upregulate ADAM10 surface expression after adoptive transfer in vivo10.

In conclusion, we have shown that signals from the BCR control MZB lineage choice by
regulating the surface expression of the metalloproteinase ADAM10 that has the
potential to cleave and activate Notch2 in T1 transitional B cells undergoing positive
selection.

ACKNOWLEDGEMENTS

H.H. is supported by Ghent University Grant (GOA 01G02817). M.V. is supported by a

FWO Fellowship (grant 3F023515W). P.P was supported by Marie Curie grant (MC
237581). S.J. is supported by FWO grant G085915N and a University of Ghent (MRP-
GROUP-ID). M.K. is supported by Swiss National Science Foundation (SNF 310030-
163443/1). J.F.K. was supported by NIAID AI14782 and AI100005. B.N.L. was
supported by an ERC consolidator grant 261231 and a University of Ghent
Multidisciplinary Research Platform (MRP-GROUP-ID) and Ghent University Grant (GOA
01G02817).

AUTHOR CONTRIBUTIONs

HH designed, performed and analyzed experiments and wrote the manuscript. MV

performed and analyzed experiments and wrote the manuscript. PP, KD, WT, KV, LV, SJ,
IR, SNS, RH, and KC performed and analyzed experiments. JJH constructed mice. MK,
BdS, JFK, and DHC provided crucial tools and analyzed experiments. BNL conceived the
study, designed and analyzed experiments and wrote the manuscript

COMPETING FINANCIAL INTERESTS

The authors declare no competing financial interests.

84
METHODS

Mice
We generated Taok3–/– mice from ES cells (clone CC0463 from the Sanger Institute
Gene Trap Resource (SIGTR), Cambridge, UK) in which a loxP flanked splicing acceptor
(SA) preceding a nuclear b-galactosidase–neomycin resistance cassette and poly A tail
was inserted as a gene trap in the intronic region between exon 1 and exon 2 of the
Taok3 gene, leading to premature transcriptional termination. ES cells were originally
made in Sv129 background, but germline transmitting mice were backcrossed for 10
generations to C57Bl/6 mice. Control animals for Taok3–/– mice were Taok3+/+
littermates from a heterozygous Taok3+/– breeding. Irf4-/- mice were obtained from Dr
W. Agace, Lund University, and were originally derived from The Jackson laboratory.
VH81x mice expressing the VH81x heavy chain, that pairs with a limited repertoire of
endogenous light chains through binding constraints have been previously described14 .
Mb1Cre mice, in which Cre recombinase expression is under the control of the Cd79a
promoter, active from the pre B cell stage onwards were obtained from The Jackson
laboratory and have been previously described27. Mice were maintained under specific
pathogen free conditions. All animal experiments and procedures were approved by
the local animal ethics committee of Ghent University.

Flow cytometry
Cells suspensions were obtained from the spleen, lymph nodes or the bone marrow of
6-8 week old mice through a 100 mm mesh and red blood cell lysis. Surface stainings
were performed in PBS at 4 °C using the following antibodies: CD19 (1D3;
eBiosciences), CD93 (AA4.1; ebiosciences), CD23 (B3B4; eBiosciences), CD21/35 (4E3;
eBiosciences), IgM (Il/41; BD biosciences), CD1d (1B1; BD Biosciences), CD3 (17A2;
eBiosciences), CD4 (RM4-5; ebiosciences), CD8 (53-6.7; eBiosciences), CD11c (N418;
eBiosciences), MHCII (M5/114.15.2; eBiosciences), Ly-6G (1A8; eBiosciences), Siglec F
(E50-2440; BD biosciences), NK1.1 (PK136; BD biosciences), ADAM10 (139712; R&D
systems). For live-dead cell discrimination, we used a fixable viability dye eFluor506
(eBiosciences). Only a single step labeling protocol was used, followed by washing in
PBS. For analysis of various B cell progenitors in the bone marrow, we stained red blood
cell lysed bone marrow cells with antibodies to B220 (RA3-6B2), CD43 (1B11), CD24
(M1/69), CD19 (1D3), lineage (CD3 (17A2), Ly6G (1A8), CD11b (M1/70), Ter119,
NK1.1 5PK136)), according to a staining panel obtained from the Immgen consortium.
Sample data were acquired with a 4 laser BD LSR Fortessa flow cytometer (BD
Biosciences) using the BD FACSDiva Software. Data analysis was performed using
FlowJo software (Treestar).

Immunizations
For experiments addressing humoral immune responses, mice (n=6 per immunization
and per genotype) were immunized with TNP-Ficoll (50 ug/ animal, i.v or i.p. in 200 ul
PBS) or TNP-KLH in alumunium hydroxide (100 mg/ animal, i.p. + 1 mg Alum in 200 ml

85
PBS) on day 0, followed by a booster of TNP-KLH of 100 mg at day 35. TNP-Ficoll and
KLH-TNP were from Biosearch Technologies and alumunium hydroxide. Immune
responses were read out at day 0 (baseline), 7 (for TNP-Ficoll), 35 (pre-booster) and 42
for TNP-KLH/alum by measuring serum immunoglobulin levels. TNP-specific IgG1,
IgG2, IgG3 and IgM antibodies were measured by commercially available ELISA
(Biosearch Technologies). For measuring the response to Streptococcus pneumoniae,
1x108 heat inactivated pneumococci (Strain D36) were injected intravenously into wild-
type (WT) and Taok3–/– mice and the IgM Ab response to the phosphorylcholine antigen
5 days later by ELISA. The IgM response to PC was measured by coating plates with 25
mg/ml of bovine serum albumin conjugated PC (Biosearch Technologies), followed by
detection of serum IgM to PC by anti-IgM antibodies, essentially as described43. Levels
of sCD23 were measured in the serum at day 5 after 2 injections of 20 mg of 19G5
antibody (directed against the stalk region of CD23) on day 0 and day2, by commercially
available sCD23 ELISA (R&D systems), as described10, 30. This 19G5 antibody causes
ADAM10-dependent cleavage of sCD23.

Cleavage of amyloid precursor protein

We generated mouse embryonic fibroblasts (MEF) from wild-type and Taok3–/– E13.5
embryos obtained from heterozygous breedings. The liver was removed from the
embryos before cell suspensions were made using dispase and trypsin. To immortalize
cells, MEFs were transiently transfected with a plasmid containing SV40 large T
oncogen cDNA and passaged multiple times to select for transfected cells. In some
experiments, subconfluent Taok3+/+ and Taok3–/– MEF were transduced with the
recombinant adenovirus Ad5/cytomegalovirus-APP (bearing human APP-695)44 for 5h
at 37 °C, after which they were kept overnight at 37°C in Dulbecco's modified Eagle's
medium supplemented with 10% fetal bovine serum. 24h post-infection the medium
was changed to minimal volume Dulbecco's modified Eagle's medium supplemented
with 0,2% fetal bovine serum and incubated at 37°C for 24h. The conditioned medium
was collected and 15µl was used for the determination of secreted APP fragments
(APPs) by SDS-PAGE and Western blotting with the polyclonal antibody 22C11
(eBioscience). 15µl was loaded on SDS-PAGE gel and followed by Western Blotting to
visualize specific APPsα with 4B4 antibody (generously provided by Stefan
Lichtenthaler)45. Cell lysates were made in STE (5mM Tris pH7,4; 250mM sucrose; 1mM
EGTA) + 1% TX-100 and were also analyzed by Western blot for the detection of APP
carboxy terminal fragments (CTFs) with B63 antibody.

Stimulation of transitional B cells on OP9-Dll1 cells

OP9-Dll1 or control OP9-GFP cells were grown to 80% confluence in 24 well plates in
optimized medium (a-MEM, supplemented with 20% fetal calf serum) as described in
detail46. .For enrichment of transitional B cells, red cell lysis was performed first on
spleen suspensions, cells washed and stained using CD93-APC. Cells were sorted using
magnetic bead enrichment with anti-APC beads and LS columns (both from Miltenyi).
Subsequently, 5x105 magnetically enriched CD93 (4AA.1)+ transitional B cells were

86
added to the OP9 cells. RNA was extracted at 4 and 18h after setting up the co-culture
and cDNA was made using a commercially available kit (Transcriptor High Fidelty Kit,
Roche). In some experiments, the B cell growth and survival factor BAFF (10 ng/ml;
R&D systems) was added according to a published protocol29, and cells were harvested
for flow cytometry 3 days later. Survival was analyzed using flow cytometry by adding
DAPI to the cells.

Confocal imaging
Confocal imaging was performed on spleen sections or magnetically enriched
transitional B cells. The following antibodies were used: ADAM10 (clone EPR5622,
rabbit polyclonal) was purchased from Abcam. IgM (Il/41), B220 (RA3-6B2), CD45.2
(142) and CD1d (1B1) were obtained from BD Biosciences. CD3 (17A2) was obtained
from ebiosciences. CD169 (MOMA-1) was obtained from Serotec Biorad. Briefly, 7mm
spleen frozen sections or purified CD93+ transitional cells were fixed for 5 minutes in
PFA 4%. After washing with PBS, sections were stained with the primary antibodies for
60 minutes at room temperature, followed by a 30 minute-incubation period with
secondary antibodies (all obtained from Jackson Immunoresearch). For visualizing the
NP-reactive pool of MZB cells, we injected FITC-labeled Ficoll (Biosearch Technologies)
intravenously. After two hours mice were euthanized and the distribution of FITC-
Ficoll on spleen sections, which were also stained for CD169 to delineate the marginal
zone. Sections were counterstained with DAPI. Images were acquired on a Zeiss
LSM710 confocal microscope equipped with 488nm, 561 nm and 633 nm lasers, and
with a tunable 2-photon laser. Images were analyzed on Imaris software.
To reveal the clonotypic B cells of VH81x Tg mice, we used in house generated
antibodies to the heavy chain of VH81x (clone MZ21, rat IgG2a, FITC labeled) and the
Vκ1C light chain (clone FO27, rat IgG2a, AF647-labeled)14.

Immunoblots
Antibodies to Notch2 (D76A6, Rabbit) and NF-kappaB2 p100/p52 were from Cell
Signaling Technologies. Antibodies Taok3 (Clone Ab70297), ADAM10 (EPR5622) were
from Abcam. For Western blotting, 500 ml cold RIPA lysis buffer supplemented with
protease and phosphatase inhibitors (Roche) were added to freshly collected spleens.
These were then homogenized using a rod homogenizer and centrifuged at 14,000 g 4°C
for 15min. Approximately 10 mg of protein was loaded on polyacrylamide gels for
Western blot.

PCRs
mRNA was extracted using the TRIreagent according to manufacturer’s specifications
(Roche Applied Sciences). cDNA was synthesized with 0,5 mg of mRNA using the High
Fidelity cDNA synthesis kit (Roche Applied Sciences). Real time PCR was the conducted
on the samples using specific primers and the Roche Syber Green master mix.
GAPDH: 5’: TGGTGCTTGTCTCACTGACC; 3’: TTCAGTATGTTCGGCTTCCC
L27: 5’: CATGAACTTGCCCATCTCG; 3’: TGAAAGGTTAGCGGAAGTGC

87
TAOK3: 5’: TTGCATGAAATTGGACATGGGA, 3’: CGATGGTGTTAGGATGCTTCAG
Deltex1: 5’: AGGCGGTGATGAGCAATC, 3’: ACCCAGGCAAGAAGTTCACA
Hes1: 5’:AAAGCCTATCATGGAGAAGAGGCG, 3’:GGAATGCCGGGAGCTATCTTTCTT
Hes5: 5’: AAAGCCTATCATGGAGAAGAGGCG,3’ GGAATGCCGGGAGCTATCTTTCTT
Hey1:5’: ACACTGCAGGAGGGAAAGGTT, 3’: CAAACTCCGATAGTCCATAGCCA

Adoptive transfer of T1B and T2B cells

Transitional B cells were first enriched from the spleens of CD45.2 VH81x Tg mice using
CD93-APC staining and magnetic bead enrichment using anti-APC beads and LS
columns (Miltenyi). DAPI- living cells were stained for CD23 and surface ADAM10 and
sorted using flow cytometry. Subsequently, 1x106 T1 ADAM10+ or ADAM10- subsets or
T2 cells were mixed with 10x106 CD45.1 splenocytes and injected intravenously into
CD45.1 Rag2-/- recipient mice in 200 ml of phosphate buffered saline. Analysis of
splenic B cell subsets was performed 5 days later by perfoming flow cytometry on cell
suspensions and donor cells were detected using antibodies to CD45.2 (clone A20 from
BD Biosciences). We also analyzed the distribution of CD45.2 injected cells on spleen
sections also stained for CD169 to delineate the marginal zone.

Generation of mixed bone marrow chimeric mice

Bone marrow cells were obtained from CD45.1 Taok3+/+ and CD45.2 Taok3-/- donor
mice. (CD45.1xCD45.2)F1 acceptor mice were irradiated using 10Gy, followed by the
intravenous injection of 2x106 Taok3+/+ cells and 2x106 Taok3-/- cells at least 4 hours
after the irradiation. We did not add antibiotics to the drinking water. Mice were
euthanized 6-8 weeks after reconstitution. Bone marrow and spleen were analyzed for
the presence of B cell progenitors and mature B cells, respectively. The reconstitution
was validated using the ratio of CD45.1 wild-type vs CD45.2. Taok3–/– cells.

88
SUPPLEMENTARY DATA

89
90
91
92
REFERENCES

1. Martin, F., Oliver, A.M. & Kearney, J.F. Marginal zone and B1 B cells unite in the early response against T-
independent blood-borne particulate antigens. Immunity 14, 617-629 (2001).
2. Tanigaki, K. et al. Notch-RBP-J signaling is involved in cell fate determination of marginal zone B cells. Nat
Immunol 3, 443-450 (2002).
3. Kuroda, K. et al. Regulation of marginal zone B cell development by MINT, suppressor of Notch/RBP-J
signaling pathway. Immunity 18, 301-312 (2003).
4. Tan, J.B. et al. Lunatic and manic fringe cooperatively enhance marginal zone B cell precursor competition
for delta-like 1 in splenic endothelial niches. Immunity 30, 254-263 (2009).
5. Fasnacht, N. et al. Specific fibroblastic niches in secondary lymphoid organs orchestrate distinct Notch-
regulated immune responses. J Exp Med 211, 2265-2279 (2014).
6. Saito, T. et al. Notch2 is preferentially expressed in mature B cells and indispensable for marginal zone B
lineage development. Immunity 18, 675-685 (2003).
7. Witt, C.M., Won, W.J., Hurez, V. & Klug, C.A. Notch2 haploinsufficiency results in diminished B1 B cells and a
severe reduction in marginal zone B cells. J Immunol 171, 2783-2788 (2003).
8. Moran, S.T. et al. Synergism between NF-kappa B1/p50 and Notch2 during the development of marginal
zone B lymphocytes. J Immunol 179, 195-200 (2007).
9. Wu, L., Maillard, I., Nakamura, M., Pear, W.S. & Griffin, J.D. The transcriptional coactivator Maml1 is required
for Notch2-mediated marginal zone B-cell development. Blood 110, 3618-3623 (2007).
10. Gibb, D.R. et al. ADAM10 is essential for Notch2-dependent marginal zone B cell development and CD23
cleavage in vivo. J Exp Med 207, 623-635 (2010).
11. Hampel, F. et al. CD19-independent instruction of murine marginal zone B-cell development by constitutive
Notch2 signaling. Blood 118, 6321-6331 (2011).
12. Simonetti, G. et al. IRF4 controls the positioning of mature B cells in the lymphoid microenvironments by
regulating NOTCH2 expression and activity. J Exp Med 210, 2887-2902 (2013).
13. Hozumi, K. et al. Delta-like 1 is necessary for the generation of marginal zone B cells but not T cells in vivo.
Nat Immunol 5, 638-644 (2004).
14. Martin, F. & Kearney, J.F. Positive selection from newly formed to marginal zone B cells depends on the rate
of clonal production, CD19, and btk. Immunity 12, 39-49 (2000).
15. Wen, L. et al. Evidence of marginal-zone B cell-positive selection in spleen. Immunity 23, 297-308 (2005).
16. Carey, J.B., Moffatt-Blue, C.S., Watson, L.C., Gavin, A.L. & Feeney, A.J. Repertoire-based selection into the
marginal zone compartment during B cell development. J Exp Med 205, 2043-2052 (2008).
17. Cariappa, A. et al. The follicular versus marginal zone B lymphocyte cell fate decision is regulated by Aiolos,
Btk, and CD21. Immunity 14, 603-615 (2001).
18. Pillai, S. & Cariappa, A. The follicular versus marginal zone B lymphocyte cell fate decision. Nat Rev Immunol
9, 767-777 (2009).
19. Boyce, K.J. & Andrianopoulos, A. Ste20-related kinases: effectors of signaling and morphogenesis in fungi.
Trends Microbiol 19, 400-410 (2011).
20. Chuang, H.C., Wang, X. & Tan, T.H. MAP4K Family Kinases in Immunity and Inflammation. Adv Immunol 129,
277-314 (2016).
21. Guinamard, R., Okigaki, M., Schlessinger, J. & Ravetch, J.V. Absence of marginal zone B cells in Pyk-2-deficient
mice defines their role in the humoral response. Nat Immunol 1, 31-36 (2000).
22. Balazs, M., Martin, F., Zhou, T. & Kearney, J. Blood dendritic cells interact with splenic marginal zone B cells
to initiate T-independent immune responses. Immunity 17, 341-352 (2002).
23. Kin, N.W., Crawford, D.M., Liu, J., Behrens, T.W. & Kearney, J.F. DNA microarray gene expression profile of
marginal zone versus follicular B cells and idiotype positive marginal zone B cells before and after
immunization with Streptococcus pneumoniae. J Immunol 180, 6663-6674 (2008).
24. Arnon, T.I., Horton, R.M., Grigorova, I.L. & Cyster, J.G. Visualization of splenic marginal zone B-cell shuttling
and follicular B-cell egress. Nature 493, 684-688 (2013).
25. Cinamon, G. et al. Sphingosine 1-phosphate receptor 1 promotes B cell localization in the splenic marginal
zone. Nat Immunol 5, 713-720 (2004).
26. Lu, T.T. & Cyster, J.G. Integrin-mediated long-term B cell retention in the splenic marginal zone. Science 297,
409-412 (2002).
27. Hobeika, E. et al. Testing gene function early in the B cell lineage in mb1-cre mice. Proc Natl Acad Sci U S A
103, 13789-13794 (2006).
28. Descatoire, M. et al. Identification of a human splenic marginal zone B cell precursor with NOTCH2-
dependent differentiation properties. J Exp Med 211, 987-1000 (2014).

93
29. Roundy, K.M., Jacobson, A.C., Weis, J.J. & Weis, J.H. The in vitro derivation of phenotypically mature and
diverse B cells from immature spleen and bone marrow precursors. Eur J Immunol 40, 1139-1149 (2010).
30. Weskamp, G. et al. ADAM10 is a principal 'sheddase' of the low-affinity immunoglobulin E receptor CD23.
Nat Immunol 7, 1293-1298 (2006).
31. De Strooper, B., Vassar, R. & Golde, T. The secretases: enzymes with therapeutic potential in Alzheimer
disease. Nat Rev Neurol 6, 99-107 (2010).
32. Gupta, N. et al. Quantitative proteomic analysis of B cell lipid rafts reveals that ezrin regulates antigen
receptor-mediated lipid raft dynamics. Nat Immunol 7, 625-633 (2006).
33. Karlsson, M.C. et al. Macrophages control the retention and trafficking of B lymphocytes in the splenic
marginal zone. J Exp Med 198, 333-340 (2003).
34. Moriyama, Y. et al. Delta-like 1 is essential for the maintenance of marginal zone B cells in normal mice but
not in autoimmune mice. Int Immunol 20, 763-773 (2008).
35. Chen, X., Martin, F., Forbush, K.A., Perlmutter, R.M. & Kearney, J.F. Evidence for selection of a population of
multi-reactive B cells into the splenic marginal zone. Int Immunol 9, 27-41 (1997).
36. Pillai, S., Cariappa, A. & Moran, S.T. Marginal zone B cells. Annu Rev Immunol 23, 161-196 (2005).
37. Casola, S. et al. B cell receptor signal strength determines B cell fate. Nat Immunol 5, 317-327 (2004).
38. Deng, W., Cho, S., Su, P.C., Berger, B.W. & Li, R. Membrane-enabled dimerization of the intrinsically
disordered cytoplasmic domain of ADAM10. Proc Natl Acad Sci U S A 111, 15987-15992 (2014).
39. Xu, P., Liu, J., Sakaki-Yumoto, M. & Derynck, R. TACE activation by MAPK-mediated regulation of cell surface
dimerization and TIMP3 association. Sci Signal 5, ra34 (2012).
40. Kapfhamer, D. et al. JNK pathway activation is controlled by Tao/TAOK3 to modulate ethanol sensitivity.
PLoS One 7, e50594 (2012).
41. Srivastava, B., Quinn, W.J., 3rd, Hazard, K., Erikson, J. & Allman, D. Characterization of marginal zone B cell
precursors. J Exp Med 202, 1225-1234 (2005).
42. Dammers, P.M., de Boer, N.K., Deenen, G.J., Nieuwenhuis, P. & Kroese, F.G. The origin of marginal zone B cells
in the rat. Eur J Immunol 29, 1522-1531 (1999).
43. Tortola, L. et al. IL-21 induces death of marginal zone B cells during chronic inflammation. Blood 116, 5200-
5207 (2010).
44. Tolia, A., Chavez-Gutierrez, L. & De Strooper, B. Contribution of presenilin transmembrane domains 6 and 7
to a water-containing cavity in the gamma-secretase complex. J Biol Chem 281, 27633-27642 (2006).
45. Kuhn, P.H. et al. ADAM10 is the physiologically relevant, constitutive alpha-secretase of the amyloid
precursor protein in primary neurons. EMBO J 29, 3020-3032 (2010).
46. Holmes, R. & Zuniga-Pflucker, J.C. The OP9-DL1 system: generation of T-lymphocytes from embryonic or
hematopoietic stem cells in vitro. Cold Spring Harb Protoc 2009, pdb prot5156 (2009).

94
 IV. Results part two

TAO-kinase 3 governs the terminal

differentiation of conventional dendritic cells
through Notch2 signaling

Vanderkerken M, Vandersarren L, Maes B, Pouliot P, Toussaint W, Deswarte K,

Vanheerswynghels M, Martens L, Van Gassen S, Kirkling ME, Reizis B, Conrad DH,
Stowell S, Hammad H, Lambrecht BN

Manuscript in preparation

Author contribution:
Matthias Vanderkerken performed and analyzed all experiments (except gene
expression analysis and FlowSOM analysis), created the figures and wrote the
manuscript.

95
ABSTRACT

Notch2-instruction determines the development and function of specific conventional

dendritic cell (cDC) subsets in the spleen and small intestine. However, it is unclear how
receptiveness to Notch-ligands is regulated in dendritic cells. We found that mice
deficient in the Ste20 kinase Thousand And One Kinase 3 (TAOK3) lack ESAM+ CD4+
type 2 cDCs (cDC2s). Furthermore, Taok3-/- cDC2s displayed an altered molecular
signature and failed to prime CD4+ T cells in an allogeneic red blood cell transfusion
model. Remarkably, Notch2-dependent CD103+ cDC2s in the small intestinal lamina
propria were not affected by Taok3 deletion. The role of TAOK3 in cDCs is cell-intrinsic
and independent of its role in marginal zone B cell development. Bone marrow-derived
dendritic cells from Taok3-/- mice responded less to Notch ligation in vitro. Conversely,
overexpression of the Notch intracellular domain in vivo rescued development of ESAM+
CD4+ cDC2s in Taok3-/- mice. In conclusion, TAOK3 controls terminal differentiation of
splenic cDC2s by regulating their receptivity for Notch2-instruction.

INTRODUCTION

Conventional dendritic cells (cDCs) are critical orchestrators of innate and adaptive
immune responses. Across species, two subtypes of cDCs are conserved that differ in
their molecular signature, function and the transcription factors needed for their
development (Guilliams et al., 2016). Type 1 dendritic cells (cDC1s) can be identified by
high surface levels of XCR1 and CD24 and additionally express CD8α or the integrin αE
(CD103) in lymphoid and nonlymphoid tissues, respectively. Exquisitely capable of
cross-presenting exogenous antigens on MHC-I, cDC1s are indispensable for the
generation of cytotoxic T cell responses against viruses and tumors. Type 2 dendritic
cells (cDC2s) are characterized by expression of signal regulatory protein α (SIRPα;
CD172a) and the integrin αM (CD11b). They preferentially activate CD4+ T cells in
response to extracellular antigens (Durai and Murphy, 2016; Merad et al., 2013; Mildner
and Jung, 2014).

Dendritic cells reside in almost all peripheral tissues, where they are constantly
replenished by circulating precursor cells (Kabashima et al., 2005; Kamath et al., 2000;
Naik et al., 2007). Commitment of pre-cDCs to the cDC1 or cDC2 fate begins in the bone
marrow but their terminal differentiation requires tissue-specific cues in the periphery.
While the transcriptional network controlling cDC1 development has been well
delineated and involves ID2, NFIL3, BATF3 and IRF8 (Aliberti et al., 2003; Grajales-
Reyes et al., 2015; Hildner et al., 2008; Jackson et al., 2011; Kashiwada et al., 2011;
Sichien et al., 2016), the factors that govern cDC2 fate specification remain poorly
characterized. The transcription factors IRF4, ZEB2 and KLF4 were shown to control
the development of specific cDC2 subsets in different organs, but cannot fully account

96
for the heterogeneity among terminally differentiated cDC2s (Persson et al., 2013;
Schlitzer et al., 2013; Scott et al., 2016; Tussiwand et al., 2015).

Interestingly, the development of specific cDC populations is controlled by Notch

signaling, an evolutionary conserved pathway that influences fate choices in immune
cells via micro-environmental cues. Membrane contacts between fibroblastic delta-like
ligand 1 (DLL1) and the Notch2 receptor instruct the terminal differentiation of CD8α+
cDC1s and a major subset of cDC2s that co-expresses CD4 and endothelial cell-selective
adhesion molecule (ESAM) in the spleen (Fasnacht et al., 2014; Lewis et al., 2011). The
absence of ESAM+ CD4+ cDC2s in mice that lack the Notch2 receptor or the downstream
transcriptional regulator RBPjκ in dendritic cells impaired CD4+ T cell priming in
response to blood-borne antigens (Briseno et al., 2018; Lewis et al., 2011). Notch2
signaling also controls the development of IL23-secreting CD103+ cDC2s in the
intestinal lamina propria, cells that are crucially required in the early defense against
attaching and effacing pathogens such as Citrobacter rodentium (Satpathy et al., 2013).
However, it is unclear how activation of the Notch pathway is regulated in cDCs. It is
remarkable that only a subset of cDCs becomes Notch-instructed, whereas all cDCs
express Notch2 (Caton et al., 2007). Furthermore, although Notch2-ligation markedly
improved the generation of mouse and human DC1s in vitro, this model could not
induce relevant levels of ESAM on bone marrow-derived DC2s (Balan et al., 2018;
Kirkling et al., 2018). These findings suggest that additional mechanisms might be in
place that control Notch activation.

Thousand-and-one (TAO) kinases constitute a subfamily of Ste20-like proteins. Ste20

was identified in yeast as a serine/threonine kinase that activates MAP3 kinases in
response to extracellular mating signals (Dan et al., 2001; Strange et al., 2006). Three
TAO kinase paralogs have been identified in mammals: TAOK1 (also known as
MAP3K16, PSK2 or MARKK), TAOK2 (MAP3K17 or PSK1) and TAOK3 (MAP3K18, JIK or
DPK) (Delpire, 2009). While the regulation and function of these kinases remain largely
unknown, their role in the immune system is starting to be unraveled. We have
previously reported how Taok3 controls commitment to the marginal zone B (MZB) cell
fate in transitional B cells by regulating the surface expression of the metalloprotease A
Disintegrin And Metalloprotease 10 (ADAM10) (Hammad et al., 2017). Upon Notch2
ligation, proteolytic cleavage by ADAM10 is required to allow the action of
intramembrane proteases such as γ secretase, and NICD translocation. While ADAM10
controls Notch2 activation in MZB cells, its role in cDC development is not known (Gibb
et al., 2010).

Here, we set out to address the role of Taok3 in conventional dendritic cells. We reveal a
cell-intrinsic function of Taok3 in splenic cDCs. The loss of Taok3 abrogated the terminal
differentiation of ESAM+ CD4+ cDC2s and impaired CD4+ T cell priming in response to
allogeneic red blood cells. Using both in vivo and in vitro approaches, we further
demonstrate that Notch2 signaling is defective in Taok3-deficient cDCs.

97
RESULTS

Taok3-/- mice lack a subset of cDC2s

During our study on B cell development, we noticed alterations in the number of CD11c+
cells in mice genetically deficient for Taok3 (Hammad et al., 2017). To identify more
precisely the effects of TAOK3 in the mononuclear phagocyte system in an unbiased
manner, we applied the unsupervised clustering algorithm FlowSOM on flow cytometry
datasets from Taok3-/- and Taok3+/+ splenocytes. FlowSOM clusters similar cells into
different nodes based on the expression of distinct markers and organizes all nodes in a
minimal spanning tree (Van Gassen et al., 2015). First, we gated out dead cells, NK cells,
B and T cells to enrich for myeloid cells. Automated clustering of remaining cells (Fig.
S1a) was compared with manual gating (Fig. S1b) and revealed robust identification of
cDC1s, cDC2s, classical monocytes, patrolling monocytes, macrophages and neutrophils
in different metaclusters. Among cDCs, the manually gated XCR1+ CD172a- cDC1
population corresponded to one metacluster containing two highly similar nodes,
slightly differing in their level of CD8α expression. In contrast, XCR1- CD172a+ cDC2s
were automatically clustered into 8 distinct nodes and 2 metaclusters, illustrating the
phenotypic heterogeneity of cDC2s. Whereas the first metacluster comprised ESAM+
CD4- cDC2s and ESAM+ CD4+ cDC2s, the second contained ESAMlo cDC2s. Strikingly, the
frequency of the ESAM+ CD4+ cDC2 nodes was reduced more than threefold in Taok3-/-
mice (Fig. 1a, Fig. S1c), while the other cDC2 and cDC1 clusters were not significantly
altered. Among other metaclusters, less pronounced yet significant reductions were
observed in the percentage of macrophages and classical monocytes. In line with our
previous findings, there was also a small increase in the percentage of splenic
neutrophils in Taok3-/- mice (Hammad et al., 2017).

To validate these findings, we analyzed the splenic cDC compartment using a manual
gating strategy (Fig. S2a). Indeed, ESAM+ CD4+ cDCs were almost completely eliminated
in Taok3-/- mice (Fig. 1b, 1c). Moreover, the number of CD172a+ cDC2s was significantly
decreased (Fig. 1g), implying that the effects of Taok3 are not limited to impaired
expression of these surface markers. Remaining cDC2s also displayed lower median
fluorescence intensities (MFIs) of F4/80, slightly decreased surface levels of MHCII,
CD172a and DCIR2 (33D1), and an increased MFI of CD11b (Fig. 1d, Fig. S3a). While
the number and percentage of XCR1+ cDC1s were not altered (Fig. 1b, 1g), Taok3-/-
cDC1s expressed significantly lower levels of CD8α and CD205 and had a slightly higher
MFI of CD24 compared to littermate controls (Fig 1d). Loss of Taok3 did not impact the
expression of the transcription factors IRF4 and IRF8 in cDC1s and cDC2s (Fig. 1d and
data not shown).

In secondary lymphoid organs, cDCs occupy specific anatomical locations crucial for
their function (Eisenbarth, 2019). In the spleen, cDC2s have been shown to reside in the
marginal zone and in so-called splenic bridging channels where the T cell zone contacts

98
Figure 1: Taok3-/- mice lack ESAM+ CD4+ splenic cDCs (legend on next page).

99
Figure 1: Taok3-/- mice lack ESAM+ CD4+ splenic cDCs.
(a) The FlowSOM algorithm was run on single, live, lineage (CD3e, CD19, NK1.1, Ter119)- splenocytes
from Taok3-/- mice and littermate controls. A minimal spanning tree with 49 nodes and 12 metaclusters
was generated by automated, unsupervised clustering. The average FlowSOM tree for all Taok3+/+ (left)
and all Taok3-/- (right) mice are displayed. Cluster size correlates with its frequency among all analyzed
cells, pie charts indicate the mean expression of the given surface markers for each node, background
colors indicate metaclusters. The metaclusters containing ESAM+ cDC2s is encircled in the Taok3-/- tree.
Two-tailed Mann-Whitney U test with Bonferroni correction for multiple testing (Mann-Whitney U tests
for all metaclusters are displayed in Fig. S1c). Data are representative of 2 biologically independent
experiments (n=8-9 mice per group). The FlowSOM algorithm was run 5 times to ensure reproducibility
of the results. (b) Representative contour plots of splenic cDCs from Taok3+/+ and Taok3-/- mice (gating
strategy see Fig. S2a). Numbers adjacent to the gates denote the percentage of the total cDC population.
(c) Percentage of ESAM+ CD4+ cells among splenic cDCs. Data pooled from 2 independent experiments
(n=8-9 mice per group), two-tailed Mann-Whitney U test. (d) Histograms indicating expression levels on
the surface of splenic cDC1s (top) and cDC2s (bottom) of Taok3+/+ (black) and Taok3-/- (blue) mice.
Conventional DC1s are defined as XCR1+ CD172a- cDCs, cDC2s as XCR1- CD172a+ cDCs. Data are
representative of at least 2 independent experiments (n=at least 8 mice per group). (e) Confocal image of
spleens from Taok3-/- and Taok3+/+ mice. Green: CD169, white: DCIR2, magenta: CD11c, blue: B220. Data
are representative of 3 independent experiments (n=13 mice per group). (f) Frequency of pre-cDCs
(gating strategy see Fig. S2b) among live cells in bone marrow and spleen of Taok3+/+ (black) and Taok3-/-
(blue) mice. Data pooled from two independent experiments (n=10 per group). Two-tailed Mann-
Whitney U test. (g) Absolute number of cDC1s and cDC2s in the spleen (top left), lung (top right), skin-
draining lymph node (SLN, bottom left) and liver (bottom right) of Taok3+/+ (black) and Taok3-/- (blue)
mice. ‘Res.’ and ‘Migr.’ indicate resident and migratory cDC populations, respectively. Data are
representative of at least 2 independent experiments (n=8-10 for spleen, lung and liver) or are
representative of 4-5 mice per group (SLN). Two-tailed Mann-Whitney U test. (h,i) cDNA microarray from
sorted cDC1s and cDC2s of Taok3-/- and Taok3+/+ mice. (h) Principal component analysis. (i) Heatmap of
the most differentially expressed genes between Taok3+/+ and Taok3-/- cDC1s (top) or cDC2s (bottom) for
the individual samples. Dendritic cells were sorted from 3 mice per group. Bars indicate mean ± SD. ns =
non-significant, * = p<0.05, ** = p<0.01, *** = p<0.001, **** = p<0.0001.

the red pulp (Yi and Cyster, 2013). Staining for CD11c and DCIR2 (33D1) allowed us to
assess the positioning of cDC2s in Taok3-/- mice (Fig. 1e). While DCIR2+ clusters were
readily identified at bridging channels in wild type littermates, they were not seen in
the spleen of Taok3-/- mice. Although the median DCIR2 expression was slightly lower
on Taok3-/- cDC2s (Fig. 1d), this could not sufficiently explain the drastic loss of DCIR2
signal on histology. Accordingly, CD11c+ cells were scattered throughout the red pulp
and did not aggregate in clear clusters at the bridging channels in Taok3-/- mice. In
contrast, the staining of CD11c+ cells residing in the T cell zone was preserved in
knockout mice, in line with the preserved cDC1 counts found by flow cytometry. Of note,
no B220 staining was identified outside of the ring of CD169+ marginal metallophilic
macrophages, consistent with the absence of MZB cells in Taok3-/- mice (Hammad et al.,
2017).

cDC development in Taok3-/- mice

We wondered whether the loss of ESAM+ CD4+ cDC2s in Taok3-/- mice reflected defects
in the development of cDCs from progenitors in the bone marrow. Committed cDC

100
precursors called pre-cDCs have been identified in bone marrow and spleen (Grajales-
Reyes et al., 2015; Schlitzer et al., 2015). In the bone marrow, the absence of Taok3 had
no effect on the frequency of bipotent SiglecH+ Ly6C- and SiglecH+ Ly6C+ pre-cDCs
(gating strategy Fig. S2b). SiglecH- Ly6C- CD24hi pre-cDC1s were equally represented in
the bone marrow, but increased in the spleen of Taok3-/- mice compared to littermate
controls. The percentage of SiglecH- Ly6C+ pre-cDC2s, the direct precursors for mature
cDC2s, was significantly elevated in bone marrow and spleen (Fig. 1f). These data
suggest that the decreased cDC2 numbers in Taok3-/- mice are not due to defective
development of dendritic cell progenitors. Rather, the increased frequency of Taok3-/-
pre-cDC2s might imply a failure to terminally differentiate, resulting in accumulation of
pre-cDC2s. In this regard, the unexpected increase in splenic pre-cDC1s supports subtle
derangements in the differentiation of mature cDC1s as well. Alternatively, these effects
could be the consequence of compensatory mechanisms serving to restore a missing
mature cDC population. In the latter case, however, one would also expect increased
numbers of early bone marrow pre-cDCs.

To study whether the loss of cDC2s was unique to spleen, we examined the frequency of
cDCs in other lymphoid and nonlymphoid organs. We found intact numbers of cDC2s in
the lung, liver, and skin-draining lymph node (SLN) of Taok3-/- mice (Fig. 1g). However,
subtle changes were noted in the phenotype of cDC2s across different tissues: the
expression levels of F4/80 and CD172a were consistently lower in spleen, lung, liver
and lymph nodes. In addition, the MFI of CD11b was significantly increased in the liver
but not in the lung of Taok3-/- mice (Fig. S3a-e). We found a small but significant
increase in the number of cDC1s in the lung, while resident cDC1s in the SLN were
decreased in knockout mice (Fig. 1g). Thus, loss of Taok3 is associated with an altered
cDC2 phenotype across different tissues.

Altered gene expression profile in Taok3-/- cDCs

To identify Taok3-regulated genes, we compared genome-wide expression profiles of
splenic cDC1s and cDC2s from Taok3-/- mice and littermate controls. Principal
component analysis demonstrated separation of cDC1 and cDC2 cell types along one
direction, and separation of Taok3+/+ and Taok3-/- cells along another (Fig. 1h). Gene
expression analysis revealed 8 differentially expressed (DE) genes in cDC1s and 101 DE
genes in cDC2s between wild type and knockout cells, while Taok3 was effectively
knocked out in both cDC subsets (Fig. 1i). These results are consistent with the minor
impact of Taok3 loss on cDC1 homeostasis compared to the effects observed in cDC2s.
The top DE genes for cDC1s and cDC2s are listed in figure 1i. Based on gene ontology
(GO) enrichment analysis, the processes associated with downregulated genes mainly
concern regulation of interleukin-2 (IL-2) biosynthesis and leukocyte differentiation. In
contrast, the main GO term associated with upregulated genes in Taok3-/- cDC2s is
regulation of cell adhesion. Indeed, many DE genes in cDC2s related to cell adhesion and
migration: L1cam, Epcam, Vim, Itgb5, Lgals1 and Lgals3 were upregulated, while Mmp9
and Mmp12 were downregulated in knockout cDC2s. Differentially expressed genes in

101
cDC2s also comprised type 1 interferon induced genes such as Ifitm1 and Oasl2. In
addition, the expression of several genes associated with small GTPases was increased
(Rhob) or decreased (Arhgap6 and Rasgrp3) in Taok3-/- cDC2s.

Functional responses of Taok3-/- cDCs

We reasoned that the defects in terminal differentiation might affect the maturation and
function of Taok3-deficient cDCs. To assess their ability to mature and migrate in
response to pathogen-associated molecular patterns, we injected mice with the TLR4
ligand lipopolysaccharide (LPS). Upon exposure to LPS, cDC1s (not shown) and cDC2s
from Taok3-/- mice upregulated the costimulatory markers CD40, CD80 and CD86 to the
same extent as littermate controls (Fig. 2a, 2b). Mature Taok3-/- cDC2s had also
upregulated the chemokine receptor CCR7, albeit not to the same level as wild type
cDC2s. To verify whether the difference in surface expression levels of CCR7 impacted
the ability of type 2 dendritic cells to migrate to the T cell zone, we assessed the position
of CD11c+ dendritic cells in the spleen 8 hours after LPS exposure in vivo. In wild type
mice, the CD11c+ clusters in the bridging channels had disappeared and dendritic cells
had relocated to the T cell zone upon LPS injection. In Taok3-/- mice, similar
accumulation of CD11c+ cells in the T cell zone was observed (Fig. 2c). Thus, Taok3
deficiency did not impair the normal maturation and migration of dendritic cells to the
T cell zone.

Next, we wanted to probe the capacity of Taok3-/- cDCs to prime CD4+ T cell immune
responses in vivo. The role of splenic cDCs in the generation of CD4+ T cell responses
against blood-borne antigens is well established (Briseno et al., 2018; Gatto et al., 2013;
Yi and Cyster, 2013). More specifically, efficient CD4+ T cell priming crucially depended
on CD4+ cDC2s in a model of allogeneic red blood cell transfusion (Calabro et al., 2016a).
Taok3-/- mice and littermate controls were transfused with stored RBCs from congenic
B6 mice that express a membrane-bound fusion protein containing antigenic epitopes
of hen egg lysozyme (HEL), OVA and Duffyb blood group antigen (HOD RBCs). Storage of
these HOD-RBCs renders them immunogenic, although the underlying mechanism
remains unknown. Transfusion of HOD-RBCs induced significantly less proliferation of
transferred OVA-specific CD4+ T cells in Taok3-/- mice compared to littermate controls
(Fig. 2d-f). This effect might be caused by the absence of ESAM+ CD4+ cDC2s, but could
also be due to mislocalisation or functional impairment of remaining cDC2s in Taok3-/-
mice. In summary, these data suggest a crucial role for Taok3 in the generation of CD4+
T cell responses to allogeneic RBCs by cDC2s.

Taok3 is required intrinsically in cDC2s

Normal cDC2 development relies on multiple environmental factors such as RELB-
expressing stromal cells and DLL1 on fibroblastic cells (Briseno et al., 2017; Fasnacht et
al., 2014). To address whether cDC homeostasis required Taok3 expression in
hematopoietic or stromal cells, we crossed Taok3-/- mice to Vav-cre mice that express
the Cre recombinase in all hematopoietic cells. Because the gene trap construct of

102
Figure 2: Loss of Taok3 hampers CD4+ T cell priming to allogeneic red blood cells
(a) Histograms of surface CD40, CD86 and CCR7 expression on splenic cDC2s from Taok3+/+ (black) and
Taok3-/- mice (blue). Mice were injected intraperitoneally with 20µg LPS in PBS (dotted line) or PBS only
(full line) and sacrificed 8h later. (b) Median fluorescence intensities of CD40, CD86 and CCR7 on cDC2s
from (a). One-way ANOVA with post-hoc Tukey’s test. (c) Confocal images of spleens from Taok3-/- mice
and littermate controls treated with LPS or PBS as in (a). B220 in blue, CD169 in white, CD11c in orange.
Data in (a-c) are representative of 2 independent experiments (n=7-8 per group in total). (d-f) Taok3-/-
and littermate control mice were injected with 5x105 CTV-labeled OTII-specific CD45.1+ CD4+ T cells and
transfused with 100μl of HOD-RBCs (see materials and methods) 1 day later. (e) Representative dot plots
of OTII cells in the spleen and inguinal lymph node (iLN) of Taok3+/+ (left) and Taok3-/- (right) acceptor
mice 72h after HOD-RBC transfusion. Plots are pre-gated on live, CD19- NK1.1- Ter119- CD3e+ CD4+
CD45.1+ cells. Numbers adjacent to the gates indicate the percentage of OTII cells. (f) Proliferation and
expansion index of OTII cells in the spleen of the acceptors, according to the Flowjo proliferation
algorithm. Two-tailed Mann-Whitney U test. Data in (e) and (f) are representative of 2 independent
experiments (n=14-15 acceptor mice per group in total). Bars indicate mean ± SD. ns = non-significant, * =
p<0.05, ** = p<0.01, *** = p<0.001, **** = p<0.0001.

103
Taok3-/- mice contains loxP sites, TAOK3 expression can be restored selectively by Cre-
mediated recombination. We found that the percentage of ESAM+ CD4+ cDC2s was
almost completely restored in the spleen of Taok3-/- x Vav-cre+ mice, excluding a role for
stromal Taok3 in cDC2 development (Fig. S4a, S4b). Conceivably, the observed cDC
phenotype of Taok3-/- mice could be secondary to their MZB cell deficit. Marginal zone B
cells do occupy a similar niche as bridging channel cDCs, and regulatory roles of B cells
in cDC homeostasis have been described (Kabashima et al., 2005). To test this
hypothesis, we crossed Taok3-/- mice to Mb1-cre mice allowing selective gene trap
reversal in B cells (Hobeika et al., 2006). While the number of CD21/35+ CD23lo MZB
cells was partially restored in Taok3-/- x Mb1-cre+ mice, the frequency and phenotype of
splenic cDCs did not differ from littermate Taok3-/- x Mb1-cre- controls (Fig. 3a, 3b). In
order to address more directly whether the role of Taok3 in cDCs was cell-intrinsic, we
generated Taok3-/- x CD11c-cre mice. Strikingly, the percentage of ESAM+ CD4+ cDC2s in
Taok3-/- x CD11c-cre+ mice was restored to wild type levels (Fig. 3a, 3b). As expected,
MZB cells were still completely absent in these mice. Thus, selective restoration of
Taok3 expression in CD11c+ cells was sufficient for ESAM+ CD4+ cDC2 development. We
validated these findings in Taok3-/- x CD11c-cre-GFP+ mice, where the Cre is expressed
only after the pre-cDC stage. While this still restored cDC2 numbers, it rescued ESAM
and CD4 expression less efficiently than the earlier acting CD11c-cre, suggesting a role
for Taok3 early in the commitment of pre-cDC2s to cDC2s (Fig. S4c, S4d). Rescue of
Taok3 expression in dendritic cells was also sufficient to restore clustering of DCIR2+
dendritic cells at marginal zone bridging channels (Fig. 3c). Together, these data
demonstrate independent and cell-intrinsic roles of Taok3 in dendritic cells and MZB
cells. Furthermore, they prove that the presence and bridging channel location of
splenic cDC2s does not depend on the presence of marginal zone B cells.

To confirm the cell-intrinsic role of Taok3 in cDC2s, we generated mixed bone marrow
chimeric mice. Lethally irradiated wild type mice were reconstituted with CD45.2
Taok3-/- and CD45.1 wild type bone marrow in equal parts (Fig. 3d). After
reconstitution, MZB cells were almost uniquely derived from the WT compartment,
while bone marrow of both genotypes contributed equally to the follicular B cell
population in the spleen (Fig. 3e), confirming earlier results (Hammad et al., 2017).
Interestingly, Taok3 expression was not intrinsically required for the generation of
cDC1s or of dendritic cell progenitors in spleen and bone marrow. In contrast, wild type
ESAM+ CD4+ cDC2s outcompeted Taok3-/- counterparts by a ratio of more than 10:1.
This effect was less pronounced on the total cDC2 population because Taok3-/-
preferentially gave rise to ESAMlo cDC2s.

To further delineate the dendritic cell-specific role of Taok3, we created conditional

Taok3fl/fl mice and crossed them to CD11c-cre mice to generate Taok3ΔDC mice. We
confirmed a significant decrease in the number and percentage of cDC2s in the spleen of
Taok3ΔDC mice (Fig. 3f-h). Again, this reduction was specific for a subpopulation of
cDC2s that expressed high levels of CD4, ESAM and F4/80. While the frequency of

104
cDC1s was comparable to littermate controls, the MFI of CD8α and CD205 was slightly
lower in Taok3ΔDC mice (Fig. 3g), supporting our earlier findings. Based on these
results, we conclude that Taok3 plays a nonredundant, cell-intrinsic role in the terminal
differentiation of ESAM+ CD4+ cDCs.

Impaired Notch signaling in Taok3-/- cDCs

The homeostasis of ESAM+ CD4+ cDC2s requires homeostatic signals through the
lymphotoxin β receptor (LTβR) (Kabashima et al., 2005; Lewis et al., 2011; Satpathy et
al., 2013). We hypothesized that TAOK3 might regulate signaling through the LTβR in
dendritic cells. However, surface expression of LTβR was comparable between cDCs
from Taok3-/- mice and littermate controls (data not shown). In addition, treatment of
Taok3-/- mice with an agonistic antibody against the LTβR could not restore cDC2
numbers or rescue expression of ESAM and CD4 (data not shown). Therefore, it seems
unlikely that LTβR signaling functions downstream of TAOK3.
On the other hand, the loss of ESAM+ cDC2s in Taok3-deficient mice is highly
reminiscent of the phenotype of mice lacking Notch signaling in dendritic cells (Lewis et
al., 2011; Satpathy et al., 2013). Moreover, we previously demonstrated how Notch
instruction is hampered in Taok3-/- transitional B cells due to the lack of surface
ADAM10 (Hammad et al., 2017). To assess whether loss of Taok3 also affected the
Notch2 pathway in cDC2s, we compared the transcriptional profile of Taok3-/- cDC2s
with published gene expression profiles of Notch2-/- cDC2s (Briseno et al., 2018;
Satpathy et al., 2013). Intriguingly, we found several genes to be significantly
upregulated in both datasets: Crip1, Ctsd, Dapk1, Fcgr3, Klra2, L1cam, Lgals1, Lgals3,
Ltb4r1, Mafb, Pilra, Pilrb1, Pilrb2, Plin3, Pltp, Smpdl3a, Tlr8, Tmem106a and Vim. We
confirmed elevated surface levels of L1CAM on Taok3-deficient cDC2s at the protein
level (data not shown). Similarly, we identified several common downregulated genes:
Arhgap6, Ccr6, Cdh17, Clec4a4, Cyp4f37, Frmd5, Mmp9, Rab30, Rasgrp3, Scn3a, Slc6a12
and Tcf7. Clec4a4 encodes DCIR2, which we found earlier to be decreased on Taok3-/-
cDC2s at the protein level.

It has been reported that Notch2 also controls the expression of ESAM in a fraction of
cDC1s in the spleen and in resident cDC2s in SLN and mesenteric lymph nodes (MLN)
(Satpathy et al., 2013). Upon careful examination, we could confirm ESAM expression on
±20% of cDC1s in wild type spleen. Remarkably, this fraction was significantly reduced
in Taok3-/- mice (data not shown) and Taok3ΔDC mice. Similarly, the percentage of ESAM-
expressing cells was reduced in resident cDC2s in SLN and MLN (Fig. S5a-d). Thus,
Taok3 and Notch2 target highly similar cDC populations, supporting a common
regulatory pathway.

Expression of the Notch2 receptor was intact in Taok3-/- cDCs at protein and mRNA
level (data not shown). However, activation of the canonical Notch pathway is regulated
by specific proteolytic cleavage events that occur upon ligand binding (Kopan and

105
Figure 3: The role of Taok3 in splenic cDC2s is cell-intrinsic (legend on next page).

106
Figure 3: The role of Taok3 in splenic cDC2s is cell-intrinsic
(a) Representative contour plots of cDCs (upper row) and marginal zone B cells (lower row) in the spleen of
Taok3-/-, Taok3+/+, Taok3-/- x CD11ccre and Taok3-/- x Mb1cre mice. The ESAM+ CD4+ cDC gate was pre-gated on
cDCs, the MZB cell gate was pre-gated on live, CD19+, CD93- cells. (b) Frequency of the indicated cDC subsets
among all cDCs (top) and among all live splenocytes (middle). Quantification of marginal zone B cells (MZB) as
a percentage of all live splenocytes (bottom). One-way ANOVA with Tukey’s post-hoc test. Data in (a) and (b)
are representative of 2 independent experiments (n= 7-8 per group). (c) Representative confocal images of
spleens from Taok3+/+ (left), Taok3-/- (middle), and Taok3-/- x CD11ccre mice. White: DCIR2, green: B220, red:
CD3e. Data are representative of 2 independent experiments (n=4-8 per group). (d) Schematic representation
of CD45.1 WT:CD45.2 Taok3-/- mixed bone marrow (BM) chimeras. (e) Contribution of donor cells to immune
cell populations 10 weeks after transplantation in BM and spleen. Results are expressed as the ratio between
CD45.1 (WT) and CD45.2 (Taok3-/-) cells. Ratios of CD45.1/CD45.2 origin along cDC1 along cDC1 and cDC2
development are compared to follicular (FoB) and marginal zone (MZB) B cells. Unless otherwise indicated, cell
populations originate from spleen. The same Ly6C- SiglecH+ and Ly6C+ SiglecH+ BM pre-cDC populations are
depicted twice. One-way ANOVA with Bonferroni correction for multiple testing. Data representative of 2
independent experiments (n= 13 mice). (f) Representative contour plot of ESAM and CD4 expression on splenic
cDCs from Taok3ΔDC mice. (g) Histograms illustrate surface expression of the indicated markers on splenic
cDC1s (top) and cDC2s (bottom) from Taok3ΔDC mice (blue) and Cre-negative littermate controls (black). (h)
Quantification of splenic cDC subsets as a percentage of all cDCs (left) or as a percentage of all live cells (right).
Two-tailed Mann-Whitney U test. Data in (f) and (h) are representative of 4 mice per group. Numbers adjacent
to the gates (a, f) indicate the percentage of the parent population. Bars indicate mean ± SD. ns = non-
significant, * = p<0.05, ** = p<0.01, *** = p<0.001, **** = p<0.0001.

Ilagan, 2009). ADAM10 appears to be the crucial Notch2 protease in vivo and controls
the development of marginal zone B cells (Gibb et al., 2010). In analogy with transitional
B cells, we wondered whether surface expression of ADAM10 was altered in Taok3-/-
dendritic cells or their progenitors. However, despite extensive efforts, we failed to
demonstrate detectable levels of surface ADAM10 in wild type dendritic cells or pre-
cDCs by flow cytometry (data not shown). To assess a potential role of ADAM10 in the
development of Taok3- and Notch2-dependent dendritic cells, we crossed conditional
Adam10fl/fl mice to CD11c-cre mice. In the spleen of these Adam10ΔDC mice, ESAM+ CD4+
cDC2s were completely absent (Fig. 4a, 4b) and clustering of DCIR2+ cDC2s at the
bridging channels was impaired (Fig. 4c). Similarly, ESAM expression was abrogated in
Adam10-deficient cDC1s (Fig. S5e, S5f). To address whether the role of ADAM10 in
cDCs was cell-intrinsic, we generated mixed bone marrow chimeras. We lethally
irradiated wild type mice and reconstituted them with CD45.1 WT and CD45.2 Adam10
ΔDC bone marrow in a 1:1 ratio (Fig. 4d). After reconstitution, we found equal

contributions of both donors to the follicular B cell pool, while MZB cells were
preferentially of WT origin (Fig. 4e). This suggests off target effects of CD11c-cre in
MZB cells, a finding supported by the reduced MZB rim in Adam10 ΔDC found on
histology (Fig. 4c). Along the developmental continuum of cDC2s, a step-up in the
WT/Adam10 ΔDC ratio was seen from the pre-cDC2 stage to the mature cDC2 stage. An
additional step-up was seen in the ratio of ESAM+ CD4+ cDC2s that were almost
exclusively derived from Adam10-sufficient bone marrow. In contrast, mature cDC1s
and pre-cDCs were equally derived from wild type and Adam10ΔDC bone marrow. In
conclusion, ADAM10 plays a cell-intrinsic role in ESAM+ CD4+ cDC2s and the highly
resembling phenotypes of Taok3-, Adam10- and Notch2-conditional knockout mice
suggest regulation by a common pathway conserved in MZB cells and cDC2s.

107
We reasoned that if Taok3 controlled Notch2 signaling in cDC2s, forced overexpression
of the notch intracellular domain (NICD) might rescue the loss of Taok3. To test this
hypothesis, we crossed Rosa26-LSL-NICD mice onto a Taok3+/+ and Taok3-/- background.
In these mice, NICD is overexpressed from the Rosa26 locus after Cre-mediated
recombination of a transcriptional stop site. Because the original Taok3-/- mice
contained a floxed gene trap, we first generated another Taok3 knockout mouse by
removing exon 6 with Crispr-Cas9 technology. We confirmed that the cDC phenotype of
these Taok3ex6-/- mice was identical to that of the original Taok3-/- mice (Fig. S6a-c). In
wild type littermates, CD11c-cre driven NICD overexpression increased the fraction of
ESAM+ CD4+ and ESAM+ CD4- cDC2s, confirming the role of Notch2 signaling in these
cDC subsets. Overexpression of NICD in Taok3ex6-/- cDCs significantly rescued the
differentiation of cDC2s. In particular, the percentage of cDCs expressing ESAM and CD4
increased more than five-fold (Fig. 4f, 4h). In cDC1s, NICD overexpression restored the
expression of CD8α to levels comparable to wild type mice (Fig. 4g, 4i). Accordingly, a
normal proportion of cDC1s expressed ESAM again in Taok3ex6-/- NICDCD11c mice (Fig.
S5g, S5h).

Recently, a novel method was described to generate CD8α+ cDC1s in vitro on DLL1-
expressing fibroblasts, in a Notch2-dependent manner (Balan et al., 2018; Kirkling et al.,
2018). To further assess the ability of Taok3-/- (pre-)cDCs to undergo Notch-instruction,
we cultured Taok3+/+ and Taok3-/- bone marrow-derived dendritic cells (BMDCs) on
OP9 cells transduced with Dll1 or with a GFP construct as a control. We confirmed that
WT BMDCs cultured on OP9-DLL1 fibroblasts generated higher numbers of DC1s and
DC2s and lower numbers of plasmacytoid dendritic cells (pDCs) than OP9-GFP
cocultures (data not shown). In line with previous findings (Kirkling et al., 2018), OP9-
DLL1 strongly induced the expression of CD8α and CD205 on BM-derived cDC1s but
failed to induce relevant levels of ESAM expression on DCs. Compared to littermate
controls, cocultures of Taok3-/- BMDCs with OP9-DLL1 fibroblasts generated less XCR1+
DC1s and CD172a+ DC2s (data not shown). Additionally, loss of Taok3 markedly
reduced the expression of levels of CD8α and CD205 on DC1s (Fig. 4j, 4k). Based on the
results from in vivo and in vitro experiments, we conclude that Taok3 controls the
responsiveness of cDCs to Notch ligation.

Taok3 is redundant for the development of CD103+ CD11b+ cDC2s in the small
intestine
Notch2 conditional knockout mice also lack a subset of type 2 dendritic cells in the small
intestine that expresses CD103. Given the importance of Taok3 for the development of
Notch2-dependent splenic cDC2s, we further investigated whether Taok3 is required for
CD103+ CD11b+ cDCs in the small intestine. The percentage of CD103+ cDC2s in
mesenteric lymph nodes and in the lamina propria of the small intestine was unaltered
in Taok3-/- mice (Fig. S7a-c) and Taok3ΔDC mice (Fig. S7d) compared to littermate
controls. Next, we analyzed the contribution of Taok3-sufficient and Taok3-/- bone
marrow to intestinal cDC subsets in mixed bone marrow chimeras. The WT: Taok3ΔDC

108
Figure 4: Defective Notch signaling in Taok3-/- cDCs
(a) Representative contour plot of ESAM and CD4 expression on splenic cDCs from Adam10ΔDC mice and
Cre-negative littermate controls. (b) Absolute number of splenic cDC2s (top) and percentage of ESAM+
CD4+ cells (bottom) among cDCs in Adam10ΔDC mice. Two-tailed Mann-Whitney U test. Data in (a) and (b)
are representative of 4 independent experiments (n=15-16 per group). (c) Confocal image of spleens
from Adam10ΔDC (right) and littermate control (left) mice. White: DCIR2, green: B220, magenta: CD11c.
(d) Schematic representation of WT: Adam10ΔDC mixed bone marrow chimeras. (e) Contribution of donor
cells to several cell populations along cDC1 and cDC2 development, 12 weeks after mixed BM

109
transplantation. Results are expressed as the ratio between CD45.1 (WT) and CD45.2 (Adam10ΔDC) cells.
Unless otherwise indicated, cell populations originate from spleen. Ly6C- SiglecH+ and Ly6C+ SiglecH+ BM
pre-cDCs are depicted twice. Data are representative of 5 acceptor mice. (f, g) Representative contour
plots of splenic cDCs from Taok3+/+ NICDTg/+ CD11ccre+/+ (Taok3+/+ control), Taok3+/+ NICDTg/+
CD11ccreTg/+ (Taok3+/+ NICDCD11c), Taok3-/- NICDTg/+ CD11ccre+/+ (Taok3-/- control) and Taok3-/- NICDTg/+
CD11ccreTg/+ (Taok3-/- NICDCD11ccre) mice. (h) Frequency of ESAM+ CD4+ cells among cDC2s in spleen. (i)
Frequency of CD8α+ cells among cDC1s in spleen. One-way ANOVA with Tukey’s post-hoc test (h, i). Data
in (f-j) are representative of 2 independent experiments (n=6-9 per group). (j) Representative contour
plots of bone marrow-derived dendritic cells from Taok3-/- mice or littermate controls co-cultured with
OP9-GFP or OP9-DL1 cells. Plots are pre-gated on live, GFP- lineage- B220- CD64- CD11c+ MHCII+ cells. (k)
Median fluorescence intensity of CD8α and CD205 on the surface of cDC1s from Taok3-/- or Taok3+/+ BM-
derived DCs co-cultured with OP-DL1 fibroblasts. Two-tailed Man-Whitney U test. Data in (j) and (k) are
representative of 2 independent experiments (9 biological replicates per group). Numbers adjacent to the
gates in a, f, g and j represent the percentage of total cDCs. Bars indicate mean ± SD. * = p<0.05, ** =
p<0.01, *** = p<0.001, **** = p<0.0001.

origin ratio for CD103+ CD11b+ cDC2s was not different from the ratio of pre-cDC2s in
bone marrow (Fig. S7e). Together, these data suggest that Taok3 is redundant for the
development of CD103+ cDC2s in the small intestine.

DISCUSSION

In contrast to cDC1s, type 2 dendritic cells harbor striking heterogeneity. The functional
and phenotypic diversity of cDC2s is determined at least partially by tissue-specific
imprints, allowing cDC2s to acquire specific functions adapted to their environment
(Durai and Murphy, 2016; Mildner and Jung, 2014). One such factor that dictates tissue-
specific differentiation in cDCs is the Notch pathway. In the spleen, Notch signaling
instructs a specific transcriptional program in cDC2s that is essential for their unique
function in capturing and processing blood-borne antigens and presenting them to CD4+
T cells (Briseno et al., 2018; Caton et al., 2007; Lewis et al., 2011; Satpathy et al., 2013).
However, the Notch2 receptor is widely expressed by cDCs and it is unclear how the
Notch pathway is selectively activated only in specific subsets of dendritic cells (Caton
et al., 2007). Here, we report a crucial and cell-intrinsic role for the Sterile-20 related
kinase TAOK3 in the terminal differentiation of Notch2-dependent cDCs.

The role of Taok3 upstream of the Notch2 pathway is based on several lines of evidence.
First, we found that both genes similarly control the development of a population of
splenic cDC2s that is characterized by high levels of ESAM, CD4 and F4/80.
Interestingly, loss of Taok3 affected the expression of F4/80 and CD172a across
different tissues, implying that it might have effects, to different extents, in all cDC2s.
Whether the expression of F4/80 and CD172a is also affected by Notch2-deficiency, is
not known. Second, Taok3 controlled ESAM expression in a fraction of cDC1s and
resident lymph node cDC2s, subpopulations that were selectively ablated in Notch2-
deficient mice (Satpathy et al., 2013). Among the 101 differentially regulated genes in

110
Taok3-/- cDC2s, 31 were also regulated by RBPjκ or Notch2 (Briseno et al., 2018; Lewis
et al., 2011). Next, Taok3-/- bone marrow generated less DC1s and DC2s in an in vitro
model of Notch-driven dendritic cell differentiation. Moreover, surface expression of
CD8α and CD205, target genes of the Notch pathway, was lower in Taok3-/- cDC1s in
vitro and in vivo. Finally, we demonstrated how conditional overexpression of the Notch
intracellular domain rescued development of ESAM-expressing cDC1s and cDC2s in
Taok3-/- mice, indicating a function of Taok3 upstream of NICD.

Surprisingly, loss of Taok3 did not affect the differentiation of Notch2-dependent

CD103+ cDC2s in the small intestine. Similarly, Taok3-/- cDC1s did not exactly
phenocopy the effect of Notch2-deficiency (Lewis et al., 2011). In contrast to the
pronounced and cell-intrinsic deficit of cDC1s in Notch2fl/fl x CD11ccre mice, the number
of CD8α+ cDC1s was not significantly decreased in Taok3-/- or Taok3ΔDC mice. In
addition, no intrinsic effects were observed in competitive bone marrow chimeras.
These findings imply that Taok3 might differentially regulate Notch signaling in
different subtypes of cDCs. Interestingly, no effect on splenic cDC1s and intestinal
CD103+ cDC2s was observed in RBPjκfl/fl x CD11ccre mice, suggesting that these effects
might be mediated by a non-canonical Notch pathway that is not controlled by Taok3.
Nonredundant roles for non-canonical Notch signaling have previously been
demonstrated in immune cells (Dongre et al., 2014; Gentle et al., 2012; Shin et al., 2014).
To validate the coexistence of both pathways in cDCs, it would be interesting to assess
whether RBPjκ controls ESAM expression in cDC1s. If Taok3 only regulates certain
aspects of the Notch2 transcriptional program, that could explain why key Notch target
genes such as Hes1 and Deltex1 were not differentially expressed in Taok3-/- cDC2s.
Alternatively, these target genes might be expressed only during a specific
developmental window before the mature cDC stage. Nevertheless, the fact that Notch2-
/- mice are embryonically lethal (Hamada et al., 1999) while Taok3-/- mice are not

further supports the idea that Taok3 only controls Notch signaling under specific
conditions.

Furthermore, this dichotomy could elucidate why the effects of Notch signaling on
cDC2s in vivo are not fully replicated in OP9-DL1 BMDC cocultures. Whereas ESAM+
cDC2s are the population most reliant on Notch signaling in vivo, the coculture system
primarily promotes DC1 development (Balan et al., 2018; Kirkling et al., 2018).
Moreover, expression of ESAM is barely induced on DCs generated in vitro (Kirkling et
al., 2018). We hypothesize that, apart from Notch2-ligation, the generation of true
ESAM+ cDC2s requires additional signals in a Taok3-dependent manner. The factors that
induce the function of TAOK3 in dendritic cells require further investigation, but
lymphotoxin α1β2, oxysterols and retinoic acid derivatives are attractive possibilities
given their important role in the homeostasis of splenic cDC2s (Beijer et al., 2013; Gatto
et al., 2013; Kabashima et al., 2005; Klebanoff et al., 2013; Lu et al., 2017; Satpathy et al.,
2013; Yi and Cyster, 2013).

111
The role of Taok3 and Notch signaling seems to concentrate on the terminal
differentiation of cDCs. Pre-cDC1s and pre-cDC2s numbers were increased in the spleen
of Taok3-/- mice, implying that the transition to mature cDCs requires Taok3. However,
the effects of CD11c-cre mediated gene trap reversal were more pronounced than those
of the later acting CD11c-cre-GFP, indicating that Taok3 activity in pre-cDCs might
already determine their late differentiation. In line with this, Notch also controlled
target gene expression in ESAM- cDC2s (Satpathy et al., 2013). In analogy with our
findings in MZB cell development (Hammad et al., 2017), we hypothesized that Taok3
controlled the action of ADAM10 by regulating its translocation to the cell surface.
However, we failed to identify a convincing ADAM10 signal on cDCs or pre-cDCs by flow
cytometry. The reasons behind this observation remain unclear. It is possible that the
surface levels on (pre-)cDCs are lower than those on transitional B cells, or that
ADAM10 is more rapidly downregulated ex vivo. Alternatively, ADAM10 might only be
expressed transiently during a brief developmental window. However, the drastic
reduction in Notch2-dependent cDC1s and cDC2s in conditional ADAM10 knockout
mice display a strongly suggests that ADAM10 is the essential Notch2 sheddase in
dendritic cells. In support of our findings, a functional role of dendritic cell-ADAM10 in
the generation of TH2 responses was previously reported (Damle et al., 2018). Though
off-target effects of the CD11c-driven Cre have been described (Abram et al., 2014;
Srinivas et al., 2001), our mixed bone marrow experiments clearly confirmed the effects
of ADAM10 to be cell-intrinsic. To advance our understanding of the role of ADAM10 in
cDCs, the development of additional tools such as reporter mice and in vivo assays for
proteolytic activity is warranted.

Using an allotransfusion model, we demonstrated an impaired capacity of Taok3-/-

cDC2s to induce CD4+ T cell responses to allogeneic red blood cells. This was not caused
by an inability of cDC2s to undergo normal maturation and migrate to the T cell zone
upon TLR activation. Conceivably, remaining Taok3-/- cDCs might be unable to process
or present antigens efficiently. However, Notch2-deficient and wild type ESAM- cDC2s
are better inducers of CD4+ proliferation than wild type ESAM+ cDC2s in vitro,
suggesting they are intrinsically capable of priming CD4+ T cells (Briseno et al., 2018).
Alternatively, the defective clustering of cDC2s in marginal zone bridging channels
could have limited the exposure of Taok3-deficient cDC2s to antigens. The location of
cDC2 in bridging channels is controlled by sphingosine-1-phosphate (S1P) signaling and
by gradients of 7α, 25-dihydroxycholesterol and 7α, 27-dihydroxycholesterol sensed by
the chemokine receptor EBI2 (GPR183) (Czeloth et al., 2007; Gatto et al., 2013; Lu et al.,
2017; Yi and Cyster, 2013). Indeed, cDC localization in bridging channels was shown to
be required for efficient T cell priming to circulating particulate antigens (Calabro et al.,
2016b; Gatto et al., 2013; Yi and Cyster, 2013). It is unclear whether the EBI2 pathway
and Notch instruction are interrelated or simply control different aspects of cDC2
function. It was reported that Gpr183-/- cDC2s still express ESAM and can be rescued by
treatment with an LTβR-agonist (Yi and Cyster, 2013). These findings sharply contrast

112
with the phenotype of Taok3-/- cDCs, implying that the effects of Taok3 cannot be
completely explained by interference with EBI2 signaling.

Taken together, Taok3 appears to control the acquisition of a Notch transcriptional

program in specialized cDC and B cell subsets in the marginal zone. Whether Taok3
controls ADAM10 activity directly or through indirect effects remains elusive. Marginal
zone B cells shuttle continuously between the follicle and the marginal zone, in a
manner dependent on integrins, S1P1, CXCR5 and shear flow (Arnon et al., 2013;
Cinamon et al., 2008; Lu and Cyster, 2002; Tedford et al., 2017). Moreover, they require
continuous exposure to DLL1 to maintain their identity (Moriyama et al., 2008; Sekine
et al., 2009). We cannot exclude that a similar mechanism might apply to cDC2s.
Intravascular labeling reportedly only stained 60% of cDC2s, suggesting that almost half
of all cDC2s were not situated in regions accessible to circulating antibody such as the
white pulp (Calabro et al., 2016b). Intriguingly, many genes controlled by Taok3 in
dendritic cells relate to adhesion and migration. Adhesion molecules such as Epcam and
L1cam were upregulated, while genes that facilitate migration through extracellular
matrix such as Mmp9, Mmp12, Lgals1 and Lgals3 were downregulated in Taok3-/- cDC2s.
Thus, TAOK3 might enable correct shuttling of cDC2s in steady state, allowing regular
access to DLL1 in the white pulp. However, these hypotheses require further
investigation.

In conclusion, we demonstrate how the Ste20 kinase TAOK3 controls Notch2-

instruction in conventional dendritic cells and identify ADAM10 as the relevant Notch2
sheddase in cDCs in vivo.

ACKNOWLEDGEMENTS
We thank Julie Van Duyse and Gert Van Isterdael from the Flow Core of the
Inflammation Research Center for their technical assistance. M.V. is supported by a
fellowship grant from FWO (grant 3F023515W). L.V. was supported by an IWT
fellowship (111581). B.M. is supported by a fellowship grant from FWO (grant
11U0116N). H.H. is supported by a Ghent University Grant (GOA 01G02817). B.N.L. was
supported by an ERC consolidator grant 261231, the University of Ghent
Multidisciplinary Research Platform (MRP-GROUP-ID) and a Ghent University grant
(GOA 01G02817).

AUTHOR CONTRIBUTIONS
M.V., B.M., L.V., P.P., K.D., M.K. and M.V.H. performed experiments. M.V., M.K., L.M. and
S.V.G analyzed data. B.R., S.S. and D.H.C. provided key reagents or mice. H.H. and B.N.L.
designed the research. M.V., H.H. and B.N.L. wrote the manuscript.

113
MATERIALS AND METHODS

Mice
Wild type (WT) C57Bl/6J mice were obtained from the Janvier laboratory. Taok3-/- mice
were described previously (Hammad et al., 2017). In brief, a loxP-flanked splicing
acceptor (SA) was inserted as a gene trap in the intronic region between exon 1 and
exon 2 of the Taok3 gene, leading to premature transcriptional termination. Conditional
Taok3fl/fl mice were generated in house with the Easi-crispr method (Quadros et al.,
2017). gRNAs were designed using the CRISPOR webtool to target a region in intron 5
(5‘CCGTCGTTGACTCTGCACAT 3’) and a region in intron 6 (5’
GGAGGCTGAGGCGGAACCAA 3’) in order to introduce loxP sites flanking exon 6
(ENSMUSE00000373662.1) of the Taok3 gene. A long single stranded DNA repair
template containing exon6, upstream and downstream intronic region with loxP sites
inserted, was co-injected with the RNP complexes in zygotes of C57BL/6J mice. F1
offspring of germline transmitting founders were interbred until mice were
homozygous for the floxed allele. Taok3 -/- mice were generated by injecting C57BL/6J
zygotes with RNP complexes with gRNAs targeting a region in intron 5 (5’
GGGTAACTGTGGTGACTTTG 3’) and a region in intron 6 (5’ GGAGGCTGAGGCGGAACCAA
3’), resulting in the deletion of exon 6. Adam10fl/fl mice (Gibb et al., 2010) and HOD mice
(Desmarets et al., 2009) have been previously described. In brief, HOD mice express a
triple fusion protein under the control of the human β globin promotor. The fusion
protein consists of hen egg lysozyme (HEL), a fragment of ovalbumin that contains both
OT-I and OT-II-specific epitopes, and the Duffyb blood group antigen. Mb1-cre mice
(Hobeika et al., 2006) were kindly provided by Michael Reth from the Max Plack
Institute of immunobiology and epigenetics, Freiburg, Germany. CD11c-cre (B6.Cg-Tg
(Itgax-cre)1-1 Reiz/J) mice, CD11c-cre-GFP (B6.Cg-Tg(Itgax-cre,-EGFP)4097Ach/J)
mice, OTII mice (B6.Cg-Tg(TcraTcrb)425Cbn/J), CD45.1 mice (B6.SJL-
Ptprc Pepc /BoyJ)
a b and Rosa26-lox-stop-lox-NICD-EGFP mice
(Gt(ROSA)26Sor tm1(Notch1)Dam /J) mice were acquired from the Jackson Laboratory. All
experimental mice were on a C57Bl/6J background and were housed in the VIB-UGent
animal facility under specific pathogen free conditions. Housing conditions entailed
individually ventilated cages in a controlled day-night cycle and food and water intake
ad libitum. Both male and female mice were used, between 6-10 weeks of age. Mice
were age- and sex-matched and randomly assigned to experimental groups. All animal
experiments and procedures were approved by the local animal ethics committee of
Ghent University and were in accordance with the Belgian animal protection law.

Isolation of tissue leukocytes

Cell suspensions from spleen and bone marrow were obtained through mechanical
disruption and filtering over a 70µm mesh. Lung, liver and lymph nodes were cut with
scissors, digested for 30 min in RPMI-1640 (Gibco) containing 20 μg/ml Liberase TM
(Roche) and 10 U/ml DNAse I (Roche) at 37°C and filtered through a 70µm nylon sieve.
Intestines were soaked in RPMI-1640 with 2% fetal calf serum (Bodinco), opened

114
longitudinally and cut into 5mm sections. These segments were incubated twice in PBS
containing 2mM EDTA at 37°C for 20 min after which the supernatant epithelial cells
were discarded. Remaining tissue was digested in RPMI-1640 containing 0.6mg/ml
Collagenase VIII (Sigma) and 10U/ml DNAse I (Roche) at 37°C with shaking for 15 min,
and passed through a 40µm filter afterwards. Samples were incubated in ammonium
chloride buffer (10mM KHCO3, 155 mM NH4Cl, 0.1 mM EDTA in milliQ water) for
erythrocyte lysis. Total cell counts were obtained by adding counting beads (123count
eBeads, Thermo Fisher Scientific).

Flow cytometry
Single cell suspensions were incubated with a fixable viability dye (eFluor506 or
eFluor780 from eBioscience) to identify dead cells and with an FcγRII/III antibody
(2.4G2) for 30 minutes to limit aspecific binding. After washing, the cells were
incubated for 30 min at 4°C with a mixture of fluorochrome- or biotin-labeled
antibodies. Following antibodies were used: B220 (RA3-6B2), BST2 (120G8), CCR7
(4B12), CD3 (145-2C11), CD4 (GK1.5), CD4 (RM4-5), CD8α (53-6.7), CD11b (M1/70),
CD11c (N418), CD19 (1D3), CD21/35 (4E3), CD23 ((B3B4), CD24 (M1/69), CD26
(H194-112), CD40 (3/23), CD43 (S6), CD45 (30-F11), CD64 (X54-5/7.1), CD80 (16-
10A1), CD86 (PO3), CD93 (AA4.1), CD103 (2-E7), CD135 (A2F10), CD172a (P84),
CD205 (205yekta), DCIR2 (33D1), ESAM (1G8), F4/80 (BM8), IgM (II/41), LTβR (3C8),
Ly6C (HK1.4 or AL-21), Ly6G (1A8), MHCII (M5/114.15.2), NK1.1 (PK136), Siglec H
(ebio440c), Ter119 (TER-119), XCR1 (ZET). When using biotin-labeled primary
antibodies, a second surface-staining step with BV786- or BV605-coupled streptavidin
(BD Biosciences) was performed. For intracellular staining of the transcription factors
IRF4 and IRF8, cells were fixed using the FoxP3 fixation/permeabilisation kit
(eBioscience) according to the manufacturer’s protocol. Cells were incubated with
PerCP-efluor710-labeled anti-IRF8 (V3GYWCH) or unlabeled goat anti-mouse IRF4 (M-
17). Unlabeled IRF4 was detected with an AF647-labeled donkey anti goat antibody
(Invitrogen) in an additional intracellular staining step. Samples were acquired on an
LSRFortessa or BD Symphony cytometer with FACSDiva software (BD Biosciences) and
analyzed using FlowJo software (LLC, Ashland, OR). Cell sorting was performed using a
FACSAria II (BD Biosciences).

FlowSOM
Splenocytes from Taok3-/- and Taok3+/+ mice were stained and acquired as described
above. Automated analysis of flow cytometry samples was done with the FlowSOM
algorithm (Van Gassen et al., 2015). First, the compensated data were manually gated
on single, live, Ter119- NK1.1- CD3e- CD19- splenocytes. The remaining cells from all
samples were concatenated and assigned to a Self-Organizing Map with a 7x7 grid.
Similar clusters of cells were thus grouped into 49 nodes. After mapping the data onto
the grid, a Minimal Spanning Tree was built that connects all nodes based on similarities
in the expression of surface markers. A meta-clustering of the nodes into 12
metaclusters was calculated. Subsequently, Taok3-/- and Taok3+/+ samples were mapped

115
individually to the same Self-Organizing Map and Minimal Spanning Tree. Because of
the random initialization of the FlowSOM grid, each run of the algorithm will have
slightly different results. Therefore, we confirmed that the results of the automated
analysis varied only minimally in five runs of the algorithm.

Bone marrow chimeras

Bone marrow cells were obtained from CD45.1 wild type and CD45.2 Taok3−/− or CD45.2
Adam10fl/fl CD11ccreTg/+ donor mice. CD45.1 × CD45.2 F1 wild type acceptor mice were
irradiated using 9.5 Gy, followed by the intravenous injection of 2 × 106 WT cells and 2 ×
106 Taok3−/− or Adam10fl/fl CD11ccreTg/+ cells in endotoxin-free PBS at least 4 h after the
irradiation. Acceptor mice were euthanized 8-10 weeks after reconstitution. Bone
marrow, spleen and mesenteric lymph nodes were analyzed for the presence of cDC cell
progenitors and mature cDCs. The reconstitution was validated using the ratio of
CD45.1 wild-type to CD45.2 knockout cells in follicular B cells.

Treatment protocols
For in vivo maturation and migration assays, mice were intravenously injected with
20µg lipopolysacharide (LPS) in PBS or with PBS only. After 8 hours, the mice were
sacrificed and their spleen was analyzed by flow cytometry and confocal imaging. For
LTβR agonism in Taok3-/- mice, 100μg of agonistic rat anti-mouse LTβR antibody (anti-
LTβR, clone 4H8, kindly provided by Carl Ware) was injected intraperitoneally on days
0, 3, 7 and 10 followed by euthanasia on day 12. Control mice were injected with PBS
only.

Allotransfusion assay
OVA-specific naive CD4+ T cells were sorted from the spleen and lymph nodes of OTII-
mice. The OTII cells were labeled with Cell Proliferation Dye eFluor450 (eBioscience) at
100µM, according to the manufacturer’s instructions. Approximately 5 x 105 OTII cells
were injected i.v. into recipient mice. Erythrocytes were collected from HOD transgenic
mice in 12% CPDA-1 (citrate phosphate dextrose adenine) anticoagulant and
leukoreduced with a Pall neonatal filter, followed by 4°C storage for 12 days. Before
transfusion, the RBCs were washed by centrifugation and the packed RBCs were diluted
1:15 with sterile PBS. One day after OTII transfer, 100µl of diluted HOD RBCs was
transfused i.v. into recipient mice. Three days after erythrocyte transfer, the spleen and
lymph nodes of acceptor mice were harvested and analyzed by flow cytometry.

Confocal microscopy
Spleens were harvested, embedded in OCT Tissue Tek medium (Takura) and
immediately frozen by liquid nitrogen. Ten µm thick cryostat sections were fixed for 10
min by paraformaldehyde 2% in PBS, washed with PBS and permeabilized/blocked
with PBS containing 0,1% triton X-100 (Roche) and 2% BSA (Sigma) (PBT buffer) for 30
min. The following antibodies were used: FITC-conjugated CD169 (MOMA-1) was
obtained from Serotec Bio-Rad. PE-conjugated DCIR2 (33D1), eFluor660-conjugated

116
CD11c (N418) and eFluor450-conjugated B220 (RA3-6B2) antibodies were obtained
from eBioscience. APC-conjugated CD3e (145-2C11) was obtained from BD Biosciences.
Sections were stained with the primary antibodies diluted in PBT buffer for 60 min at
room temperature. Sections were counterstained with DAPI, washed, and coverslipped
with antifading polyvinyl alcohol containing DABCO (Sigma-Aldrich). Images were
acquired on a Zeiss LSM710 confocal microscope equipped with 488-nm, 561-nm and
633-nm lasers and with a tunable two-photon laser. Emissions were recorded in three
separate channels. Digital pictures at 1024 x 1024 pixel density and 8-bit depth were
acquired with ZEN software (Zeiss). Images were analyzed with Imaris software.

RNA extraction and qRT-PCR

Whole spleen tissue was homogenized with the Tissuelyser II (Qiagen) prior to RNA
extraction using TriPure isolation reagent (Roche). Splenic cDC1s and cDC2s were
sorted directly into RLT buffer and RNA was extracted with the Rneasy Mini kit
(Qiagen) according to the manufacturer’s protocol. cDNA was synthesized with the
Sensifast cDNA synthesis kit (Bioline). qRT-PCR reactions were conducted in a
LightCycler 480 (Roche) using the SensiFAST sybr no-ROX mix (Bioline). The following
primer pairs were used: Gadph (5’:TGGTGCTTGTCTCACTGACC, 3’:
TTCAGTATGTTCGGCTTCCC), Ubc (5’: AAAGCCCCTCAATCTCTGGAC, 3’:
TGCATCGTCTCTCTCACGGA), Hmbs (5’:
GAAACTCTGCTTCGCTGCATT, 3’: TGCCCATCTTTCATCACTGTATG), Actb (5’:
GCTTCTAGGCGGACTGTTACTGA, 3’: GCCATGCCAATGTTGTCTCTTAT), Sdha (5’:
TTTCAGAGACGGCCATGATCT, 3’: TGGGAATCCCACCCATGTT), Taok3 (5’:
TTGCATGAAATTGGACATGGGA, 3 ′ :CGATGGTGTTAGGATGCTTCAG), Notch2 (5’:
GACTGCCAATACTCCACCTCT, 3’: CCATTTTCGCAGGGATGAGAT. Data were analyzed
with qBase+ software (Biogazelle).

Culture of bone marrow-derived dendritic cells on OP9-DL1 cells

OP9 cell lines transduced with GFP- or DL1-encoding retroviruses were cultured in
MEMα medium containing 20% FCS (Bodinco), 56µg/ml gentamycin (Gibco) and 1%
GlutaMAX (Gibco) at 37°C with 5% CO2. Single cell suspensions of bone marrow were
generated by crushing tibias and femurs and filtering them over a 70µm nylon sieve.
After erythrocyte lysis, cells were resuspended in complete RPMI-1640 medium
supplemented with 10% FCS (Bodinco), 1,1 mg/ml β-mercaptoethanol, 56µg/ml
gentamycin, 1% GlutaMAX and 250 ng/ml hFLT3L. The BM cells were plated at 2 x 106
per well in 24-well plates. On day 3 of the primary BM culture, half of the cells from
every well was transferred to a single well that contained a monolayer of OP9-GFP or
OP9-DL1 cells. OP9 cells were pretreated with 10µg/ml mitomycin C for 2 hours at 37°C
and plated on 24-well plates one day before the start of the coculture. Cell cultures were
analyzed on day 7.

117
Microarray
Dendritic cells were sorted from the spleen of Taok3-/- and Taok3+/+ animals by FACS.
Within the conventional dendritic cell gate, cDC1s were defined as XCR1+ CD172a- and
cDC2s as XCR1- CD172a+. RNA was extracted using the QIAshredder spin column
(Qiagen) and RNeasy Mini Kit (Qiagen) according to the manufacturer’s protocol and
sent to the Nucleomics facility, VIB Leuven, Belgium where the microarrays were
performed using the GeneChip Mouse Gene 1.0 ST arrays (Affymetrix). Samples were
subsequently analyzed using R/Bioconductor. All samples passed quality control, and
the Robust Multi-array Average (RMA) procedure was used to normalize data within
arrays (probeset summarization, background correction and log2-transformation) and
between arrays (quantile normalization). Only probesets that mapped uniquely to one
gene were kept, and for each gene, the probeset with the highest expression level was
kept. Differentially expressed genes were identified with the Limma package from
Bioconductor, and the resulting P values were corrected for multiple testing with the
Benjamini-Hochberg procedure for control of the false-discovery rate. Probe sets with a
corrected P value <0.05 and an increase or decrease in expression of at least 2 fold were
considered differentially expressed. PCA plot was created using the 15% of genes with
the most variable expression. The heatmaps show the relative expression per gene,
calculated by per gene subtracting the mean normalised value of each normalised log2
value.

Statistical analysis
Statistical analyses were performed with Graphpad Prism 7.0 (Graphpad Software, La
Jolla, CA). For comparisons between two groups, a two-tailed Mann-Whitney test was
used. For comparisons between three or more groups, a one-way ANOVA with Tukey’s
correction for multiple comparisons was used. Bonferroni correction was used when
testing multiple selected comparisons. The significance level α was 0.05. Error bars
represent mean ± standard deviation of the mean. Significant differences are indicated
as * = p<0.05, ** = p<0.01, *** = p<0.001, ****=p<0.0001.

118
SUPPLEMENTARY DATA

Supplementary figure 1: Identification of a cDC2 deficit in Taok3-/- mice using the unsupervised
clustering algorithm FlowSOM
(a) FlowSOM tree generated from concatenated data from Taok3+/+ and Taok3-/- splenocytes as described
in Figure 1a and in materials and methods. Here, equal cluster sizes are depicted to allow identification of
all clusters. Similar clusters were automatically grouped together into 12 metaclusters, denoted by the
background color of the nodes. (b) FlowSOM tree displaying the overlap between the manual gating
strategy and automated clustering. Pie charts indicate the percentage of each node that falls in the
manually predefined gates listed in the legend on the left. (c) Table of Mann-Whitney U test results when
comparing the frequency of each metacluster between Taok3+/+ and Taok3-/- mice. P values were adjusted
using the Bonferroni correction for multiple comparisons. Direction: 1 indicates higher in wild type, -1
indicates lower in wild type. Data are representative of 2 biologically independent experiments (n=8-9
mice per group). The FlowSOM algorithm was run 5 times to ensure reproducibility of the results.

119
Supplementary figure 2: Gating strategy for cDCs and pre-cDCs.
(a) Gating strategy for conventional dendritic cells. In the spleen, cDCs are defined as single, live, CD64-
lineage (CD3e/CD19/Ter119/NK1.1)- MHCII+ CD11chi CD26+ cells. Type 1 cDCs (cDC1s) are defined as
XCR1+ CD172a-, type 2 cDCs (cDC2s) as XCR1- CD172a+. Alternative gating of cDC subsets based on CD4,
CD8α and ESAM expression is shown. The gating strategy for cDCs lung, liver, small intestine and lymph
nodes is similar but includes a CD45+ gate. In lymph nodes, resident cDCs are identified as CD11chi
MHCII+, migratory cDCs as CD11c+ MHCIIhi (b) Gating strategy for pre-cDCs. Pre-cDCs are defined as
single, live, lineage (CD3e/CD19/Ter119/NK1.1)- B220- Ly6G- CD11cint/+ MHCIIlo CD172lo/int CD135+ cells.
Within pre-cDCs, SiglecH+ Ly6C- and SiglecH+ Ly6C+ cells contain both cDC1 and cDC2 potential. Pre-
cDC1s are defined as SiglecH- Ly6C- CD24hi pre-cDCs, pre-cDC2s as SiglecH- Ly6C+ pre-cDCs. Gating is
shown for bone marrow but is similar for pre-cDCs in spleen.

120
Supplementary figure 3: Loss of Taok3 alters the cDC2 phenotype across different tissues.
(a-e) Surface expression levels of F4/80, CD172a and CD11b on splenic cDC2s in the spleen (a), lung (b),
liver (c), mesenteric lymph nodes (d) and skin-draining lymph nodes (e) from Taok3-/- mice (blue) or
littermate controls (black), as determined by flow cytometry. ‘Res.’ and ‘Migr.’ denote resident and
migratory cDC populations, respectively. Data representative of at least 2 independent experiments (n=8-
12 per group, a-d) or representative of 4-5 mice per group (e). Two-tailed Mann-Whitney U test. Bars
indicate mean ± SD. * = p<0.05, ** = p<0.01, *** = p<0.001, **** = p<0.0001.

121
Supplementary figure 4: Selective gene trap reversal of Taok3 in dendritic cells rescues the
differentiation of cDC2s.
(a) Representative contour plots of splenic cDCs from Taok3+/+ Vav cre-, Taok3+/+ Vav cre+, Taok3-/- Vav
cre- and Taok3-/- Vav cre+ mice. (b) Percentage of ESAM+ CD4+ cells among cDCs. One-way ANOVA with
Tukey’s post-hoc test. Data in (a) and (b) are representative of 3-4 mice per group. (c) Representative
contour plots of splenic cDCs from Taok3+/+ CD11c cre-GFP-, Taok3-/- CD11c cre-GFP- and Taok3-/- CD11c
cre-GFP+ mice. (d) Percentage of ESAM+ CD4+ cells among cDCs. One-way ANOVA with Tukey’s post-hoc
test. Data are representative of at least 2 independent experiments (n=8 per group). Numbers within the
gate in (a) and (c) denote the percentage of the total cDC population. Bars indicate mean ± SD. * = p<0.05,
** = p<0.01, *** = p<0.001, **** = p<0.0001.

122
Supplementary figure 5: Taok3 controls ESAM expression in resident lymphoid tissue dendritic
cells.
(a, b) Representative flow plots of resident cDCs (gated as single, live, CD45+ lineage
(CD3e/CD19/Ter119/NK1.1)- CD64- CD11chi MHCII+ cells) from skin-draining lymph nodes (SLN) (a) and
mesenteric lymph nodes (MLN) (b) in Taok3-/- mice and littermate controls. Data are representative of 5
mice per group. Numbers adjacent to the gates indicate the percentage of all cDC2s. (c) Representative
plots of CD8α and ESAM expression on XCR1+ splenic cDC1s from Taok3ΔDC mice and littermate controls.
An FMO control for ESAM staining is depicted. (d) Percentage of splenic cDC1s that express ESAM. (c, d)
Data representative of 4 mice per group. Two sided Mann-Whitney U test. (e) Representative plots of
XCR1+ splenic cDC1s from ADAM10ΔDC mice and littermate controls. (f) Percentage of splenic cDC1s that
express ESAM. (e, f) Data representative of at least 2 independent experiments (n=8 per group). Two
sided Mann-Whitney U test. (g) Representative plots of XCR1+ splenic cDC1s from Taok3-/- NICDCD11ccre
mice and littermate controls. (h) Percentage of splenic cDC1s that express ESAM. (g, h) Data
representative of at least 2 independent experiments (n=8 per group). Two sided Mann-Whitney U test.
Numbers within the gates indicate the percentage of all cDC1s (c, e, g). Bars indicate mean ± SD. * =
p<0.05, ** = p<0.01, *** = p<0.001, **** = p<0.0001.

123
Supplementary figure 6: The phenotype of splenic cDCs in Taok3ex6-/- mice is similar to that of
Taok3-/- mice
(a) Representative contour plots of splenic cDCs from Taok3ex6-/- mice and littermate controls. Numbers
adjacent to the gates indicate the frequency among total cDCs. (b) Quantification of number of cDC1s and
cDC2s (top) and of ESAM+ CD4+ cDCs (bottom) in spleen. Two-sided Mann-Whitney U test. Bars indicate
mean ± SD. (c) Histograms of surface marker expression in cDC1s (top) and cDC2s (bottom) from
Taok3ex6-/- mice (blue lines) and littermate controls (black lines). Data in (a-c) are representative of at
least 2 independent experiments (n=8 mice per group). * = p<0.05, ** = p<0.01, *** = p<0.001, **** =
p<0.0001.

124
Supplementary figure 7: Development of intestinal CD103+ CD11b+ cDCs does not rely on Taok3.
(a) Representative contour plots of cDCs from mesenteric lymph nodes of Taok3+/+ and Taok3-/- mice.
Plots are pre-gated on live, CD45+ lineage (CD3e, CD19, NK1.1)- CD64- CD11c+ MHCII+ cells. (b, c)
Frequency of cDC1 and cDC2 among live cells (left) and frequency of CD103+ CD11b-, CD103+ CD11b+ and
CD103- CD11b+ cells among cDCs (right) in mesenteric lymph node (MesLN) and in the lamina propria of
the small intestine (SI). Two-tailed Mann-Whitney U test. Data in (a) and (b) are representative of 3
independent experiments (n=10-12 per group), data in (c) are representative of 2 independent
experiments (n=7-8). Numbers adjacent to gates indicate the percentage of all cDCs. (d) Frequency of cDC
subsets among CD45+ cells (left) and among all cDCs (right) in mesenteric lymph node of Taok3ΔDC mice
and littermate controls. Two-tailed Mann-Whitney U test. Data are representative of 3-4 mice per group.
(e) Wild type acceptor mice were lethally irradiated and transplanted with mixed CD45.1 WT:CD45.2
Taok3-/- bone marrow in a 1:1 ratio. Contributions of CD45.1 and CD45.2 bone marrow to BM pre-cDCs
and mature cDCs in mesenteric lymph nodes of acceptor mice 10 weeks after reconstitution are
presented as CD45.1/CD45.2 ratios. One-way ANOVA with Bonferroni correction for multiple testing.
Data are representative of 5 acceptor mice. Bars indicate mean ± SD. * = p<0.05, ** = p<0.01, *** =
p<0.001, **** = p<0.0001.

125
REFERENCES

Abram, C.L., G.L. Roberge, Y. Hu, and C.A. Lowell. 2014. Comparative analysis of the efficiency and specificity of
myeloid-Cre deleting strains using ROSA-EYFP reporter mice. J Immunol Methods 408:89-100.
Aliberti, J., O. Schulz, D.J. Pennington, H. Tsujimura, C. Reis e Sousa, K. Ozato, and A. Sher. 2003. Essential role for
ICSBP in the in vivo development of murine CD8alpha + dendritic cells. Blood 101:305-310.
Arnon, T.I., R.M. Horton, I.L. Grigorova, and J.G. Cyster. 2013. Visualization of splenic marginal zone B-cell shuttling
and follicular B-cell egress. Nature 493:684-688.
Balan, S., C. Arnold-Schrauf, A. Abbas, N. Couespel, J. Savoret, F. Imperatore, A.C. Villani, T.P. Vu Manh, N. Bhardwaj,
and M. Dalod. 2018. Large-Scale Human Dendritic Cell Differentiation Revealing Notch-Dependent Lineage
Bifurcation and Heterogeneity. Cell Rep 24:1902-1915 e1906.
Beijer, M.R., R. Molenaar, G. Goverse, R.E. Mebius, G. Kraal, and J.M. den Haan. 2013. A crucial role for retinoic acid in
the development of Notch-dependent murine splenic CD8- CD4- and CD4+ dendritic cells. Eur J Immunol
43:1608-1616.
Briseno, C.G., M. Gargaro, V. Durai, J.T. Davidson, D.J. Theisen, D.A. Anderson, 3rd, D.V. Novack, T.L. Murphy, and K.M.
Murphy. 2017. Deficiency of transcription factor RelB perturbs myeloid and DC development by
hematopoietic-extrinsic mechanisms. Proc Natl Acad Sci U S A 114:3957-3962.
Briseno, C.G., A.T. Satpathy, J.T.t. Davidson, S.T. Ferris, V. Durai, P. Bagadia, K.W. O'Connor, D.J. Theisen, T.L. Murphy,
and K.M. Murphy. 2018. Notch2-dependent DC2s mediate splenic germinal center responses. Proc Natl Acad
Sci U S A 115:10726-10731.
Calabro, S., A. Gallman, U. Gowthaman, D. Liu, P. Chen, J. Liu, J.K. Krishnaswamy, M.S. Nascimento, L. Xu, S.R. Patel, A.
Williams, C.A. Tormey, E.A. Hod, S.L. Spitalnik, J.C. Zimring, J.E. Hendrickson, S.R. Stowell, and S.C.
Eisenbarth. 2016a. Bridging channel dendritic cells induce immunity to transfused red blood cells. J Exp Med
213:887-896.
Calabro, S., D. Liu, A. Gallman, M.S. Nascimento, Z. Yu, T.T. Zhang, P. Chen, B. Zhang, L. Xu, U. Gowthaman, J.K.
Krishnaswamy, A.M. Haberman, A. Williams, and S.C. Eisenbarth. 2016b. Differential Intrasplenic Migration
of Dendritic Cell Subsets Tailors Adaptive Immunity. Cell Rep 16:2472-2485.
Caton, M.L., M.R. Smith-Raska, and B. Reizis. 2007. Notch-RBP-J signaling controls the homeostasis of CD8- dendritic
cells in the spleen. J Exp Med 204:1653-1664.
Cinamon, G., M.A. Zachariah, O.M. Lam, F.W. Foss, Jr., and J.G. Cyster. 2008. Follicular shuttling of marginal zone B cells
facilitates antigen transport. Nat Immunol 9:54-62.
Czeloth, N., A. Schippers, N. Wagner, W. Muller, B. Kuster, G. Bernhardt, and R. Forster. 2007. Sphingosine-1
phosphate signaling regulates positioning of dendritic cells within the spleen. J Immunol 179:5855-5863.
Damle, S.R., R.K. Martin, C.L. Cockburn, J.C. Lownik, J.A. Carlyon, A.D. Smith, and D.H. Conrad. 2018. ADAM10 and
Notch1 on murine dendritic cells control the development of type 2 immunity and IgE production. Allergy
73:125-136.
Dan, I., N.M. Watanabe, and A. Kusumi. 2001. The Ste20 group kinases as regulators of MAP kinase cascades. Trends
Cell Biol 11:220-230.
Delpire, E. 2009. The mammalian family of sterile 20p-like protein kinases. Pflugers Arch 458:953-967.
Desmarets, M., C.M. Cadwell, K.R. Peterson, R. Neades, and J.C. Zimring. 2009. Minor histocompatibility antigens on
transfused leukoreduced units of red blood cells induce bone marrow transplant rejection in a mouse
model. Blood 114:2315-2322.
Dongre, A., L. Surampudi, R.G. Lawlor, A.H. Fauq, L. Miele, T.E. Golde, L.M. Minter, and B.A. Osborne. 2014. Non-
Canonical Notch Signaling Drives Activation and Differentiation of Peripheral CD4(+) T Cells. Front Immunol
5:54.
Durai, V., and K.M. Murphy. 2016. Functions of Murine Dendritic Cells. Immunity 45:719-736.
Eisenbarth, S.C. 2019. Dendritic cell subsets in T cell programming: location dictates function. Nat Rev Immunol
19:89-103.
Fasnacht, N., H.Y. Huang, U. Koch, S. Favre, F. Auderset, Q. Chai, L. Onder, S. Kallert, D.D. Pinschewer, H.R. MacDonald,
F. Tacchini-Cottier, B. Ludewig, S.A. Luther, and F. Radtke. 2014. Specific fibroblastic niches in secondary
lymphoid organs orchestrate distinct Notch-regulated immune responses. J Exp Med 211:2265-2279.
Gatto, D., K. Wood, I. Caminschi, D. Murphy-Durland, P. Schofield, D. Christ, G. Karupiah, and R. Brink. 2013. The
chemotactic receptor EBI2 regulates the homeostasis, localization and immunological function of splenic
dendritic cells. Nat Immunol 14:446-453.
Gentle, M.E., A. Rose, L. Bugeon, and M.J. Dallman. 2012. Noncanonical Notch signaling modulates cytokine responses
of dendritic cells to inflammatory stimuli. J Immunol 189:1274-1284.

126
Gibb, D.R., M. El Shikh, D.J. Kang, W.J. Rowe, R. El Sayed, J. Cichy, H. Yagita, J.G. Tew, P.J. Dempsey, H.C. Crawford, and
D.H. Conrad. 2010. ADAM10 is essential for Notch2-dependent marginal zone B cell development and CD23
cleavage in vivo. J Exp Med 207:623-635.
Grajales-Reyes, G.E., A. Iwata, J. Albring, X. Wu, R. Tussiwand, W. Kc, N.M. Kretzer, C.G. Briseno, V. Durai, P. Bagadia, M.
Haldar, J. Schonheit, F. Rosenbauer, T.L. Murphy, and K.M. Murphy. 2015. Batf3 maintains autoactivation of
Irf8 for commitment of a CD8alpha(+) conventional DC clonogenic progenitor. Nat Immunol 16:708-717.
Guilliams, M., C.A. Dutertre, C.L. Scott, N. McGovern, D. Sichien, S. Chakarov, S. Van Gassen, J. Chen, M. Poidinger, S. De
Prijck, S.J. Tavernier, I. Low, S.E. Irac, C.N. Mattar, H.R. Sumatoh, G.H.L. Low, T.J.K. Chung, D.K.H. Chan, K.K.
Tan, T.L.K. Hon, E. Fossum, B. Bogen, M. Choolani, J.K.Y. Chan, A. Larbi, H. Luche, S. Henri, Y. Saeys, E.W.
Newell, B.N. Lambrecht, B. Malissen, and F. Ginhoux. 2016. Unsupervised High-Dimensional Analysis Aligns
Dendritic Cells across Tissues and Species. Immunity 45:669-684.
Hamada, Y., Y. Kadokawa, M. Okabe, M. Ikawa, J.R. Coleman, and Y. Tsujimoto. 1999. Mutation in ankyrin repeats of
the mouse Notch2 gene induces early embryonic lethality. Development 126:3415-3424.
Hammad, H., M. Vanderkerken, P. Pouliot, K. Deswarte, W. Toussaint, K. Vergote, L. Vandersarren, S. Janssens, I.
Ramou, S.N. Savvides, J.J. Haigh, R. Hendriks, M. Kopf, K. Craessaerts, B. de Strooper, J.F. Kearney, D.H.
Conrad, and B.N. Lambrecht. 2017. Transitional B cells commit to marginal zone B cell fate by Taok3-
mediated surface expression of ADAM10. Nat Immunol 18:313-320.
Hildner, K., B.T. Edelson, W.E. Purtha, M. Diamond, H. Matsushita, M. Kohyama, B. Calderon, B.U. Schraml, E.R. Unanue,
M.S. Diamond, R.D. Schreiber, T.L. Murphy, and K.M. Murphy. 2008. Batf3 deficiency reveals a critical role for
CD8alpha+ dendritic cells in cytotoxic T cell immunity. Science 322:1097-1100.
Hobeika, E., S. Thiemann, B. Storch, H. Jumaa, P.J. Nielsen, R. Pelanda, and M. Reth. 2006. Testing gene function early in
the B cell lineage in mb1-cre mice. Proc Natl Acad Sci U S A 103:13789-13794.
Jackson, J.T., Y. Hu, R. Liu, F. Masson, A. D'Amico, S. Carotta, A. Xin, M.J. Camilleri, A.M. Mount, A. Kallies, L. Wu, G.K.
Smyth, S.L. Nutt, and G.T. Belz. 2011. Id2 expression delineates differential checkpoints in the genetic
program of CD8alpha+ and CD103+ dendritic cell lineages. EMBO J 30:2690-2704.
Kabashima, K., T.A. Banks, K.M. Ansel, T.T. Lu, C.F. Ware, and J.G. Cyster. 2005. Intrinsic lymphotoxin-beta receptor
requirement for homeostasis of lymphoid tissue dendritic cells. Immunity 22:439-450.
Kamath, A.T., J. Pooley, M.A. O'Keeffe, D. Vremec, Y. Zhan, A.M. Lew, A. D'Amico, L. Wu, D.F. Tough, and K. Shortman.
2000. The development, maturation, and turnover rate of mouse spleen dendritic cell populations. J
Immunol 165:6762-6770.
Kashiwada, M., N.L. Pham, L.L. Pewe, J.T. Harty, and P.B. Rothman. 2011. NFIL3/E4BP4 is a key transcription factor
for CD8alpha(+) dendritic cell development. Blood 117:6193-6197.
Kirkling, M.E., U. Cytlak, C.M. Lau, K.L. Lewis, A. Resteu, A. Khodadadi-Jamayran, C.W. Siebel, H. Salmon, M. Merad, A.
Tsirigos, M. Collin, V. Bigley, and B. Reizis. 2018. Notch Signaling Facilitates In Vitro Generation of Cross-
Presenting Classical Dendritic Cells. Cell Rep 23:3658-3672 e3656.
Klebanoff, C.A., S.P. Spencer, P. Torabi-Parizi, J.R. Grainger, R. Roychoudhuri, Y. Ji, M. Sukumar, P. Muranski, C.D. Scott,
J.A. Hall, G.A. Ferreyra, A.J. Leonardi, Z.A. Borman, J. Wang, D.C. Palmer, C. Wilhelm, R. Cai, J. Sun, J.L. Napoli,
R.L. Danner, L. Gattinoni, Y. Belkaid, and N.P. Restifo. 2013. Retinoic acid controls the homeostasis of pre-
cDC-derived splenic and intestinal dendritic cells. J Exp Med 210:1961-1976.
Kopan, R., and M.X. Ilagan. 2009. The canonical Notch signaling pathway: unfolding the activation mechanism. Cell
137:216-233.
Lewis, K.L., M.L. Caton, M. Bogunovic, M. Greter, L.T. Grajkowska, D. Ng, A. Klinakis, I.F. Charo, S. Jung, J.L.
Gommerman, Ivanov, II, K. Liu, M. Merad, and B. Reizis. 2011. Notch2 receptor signaling controls functional
differentiation of dendritic cells in the spleen and intestine. Immunity 35:780-791.
Lu, E., E.V. Dang, J.G. McDonald, and J.G. Cyster. 2017. Distinct oxysterol requirements for positioning naive and
activated dendritic cells in the spleen. Sci Immunol 2:
Lu, T.T., and J.G. Cyster. 2002. Integrin-mediated long-term B cell retention in the splenic marginal zone. Science
297:409-412.
Merad, M., P. Sathe, J. Helft, J. Miller, and A. Mortha. 2013. The dendritic cell lineage: ontogeny and function of
dendritic cells and their subsets in the steady state and the inflamed setting. Annu Rev Immunol 31:563-604.
Mildner, A., and S. Jung. 2014. Development and function of dendritic cell subsets. Immunity 40:642-656.
Moriyama, Y., C. Sekine, A. Koyanagi, N. Koyama, H. Ogata, S. Chiba, S. Hirose, K. Okumura, and H. Yagita. 2008. Delta-
like 1 is essential for the maintenance of marginal zone B cells in normal mice but not in autoimmune mice.
Int Immunol 20:763-773.
Naik, S.H., P. Sathe, H.Y. Park, D. Metcalf, A.I. Proietto, A. Dakic, S. Carotta, M. O'Keeffe, M. Bahlo, A. Papenfuss, J.Y.
Kwak, L. Wu, and K. Shortman. 2007. Development of plasmacytoid and conventional dendritic cell subtypes
from single precursor cells derived in vitro and in vivo. Nat Immunol 8:1217-1226.

127
Persson, E.K., H. Uronen-Hansson, M. Semmrich, A. Rivollier, K. Hagerbrand, J. Marsal, S. Gudjonsson, U. Hakansson, B.
Reizis, K. Kotarsky, and W.W. Agace. 2013. IRF4 transcription-factor-dependent CD103(+)CD11b(+)
dendritic cells drive mucosal T helper 17 cell differentiation. Immunity 38:958-969.
Quadros, R.M., H. Miura, D.W. Harms, H. Akatsuka, T. Sato, T. Aida, R. Redder, G.P. Richardson, Y. Inagaki, D. Sakai, S.M.
Buckley, P. Seshacharyulu, S.K. Batra, M.A. Behlke, S.A. Zeiner, A.M. Jacobi, Y. Izu, W.B. Thoreson, L.D. Urness,
S.L. Mansour, M. Ohtsuka, and C.B. Gurumurthy. 2017. Easi-CRISPR: a robust method for one-step
generation of mice carrying conditional and insertion alleles using long ssDNA donors and CRISPR
ribonucleoproteins. Genome Biol 18:92.
Satpathy, A.T., C.G. Briseno, J.S. Lee, D. Ng, N.A. Manieri, W. Kc, X. Wu, S.R. Thomas, W.L. Lee, M. Turkoz, K.G. McDonald,
M.M. Meredith, C. Song, C.J. Guidos, R.D. Newberry, W. Ouyang, T.L. Murphy, T.S. Stappenbeck, J.L.
Gommerman, M.C. Nussenzweig, M. Colonna, R. Kopan, and K.M. Murphy. 2013. Notch2-dependent classical
dendritic cells orchestrate intestinal immunity to attaching-and-effacing bacterial pathogens. Nat Immunol
14:937-948.
Schlitzer, A., N. McGovern, P. Teo, T. Zelante, K. Atarashi, D. Low, A.W. Ho, P. See, A. Shin, P.S. Wasan, G. Hoeffel, B.
Malleret, A. Heiseke, S. Chew, L. Jardine, H.A. Purvis, C.M. Hilkens, J. Tam, M. Poidinger, E.R. Stanley, A.B.
Krug, L. Renia, B. Sivasankar, L.G. Ng, M. Collin, P. Ricciardi-Castagnoli, K. Honda, M. Haniffa, and F. Ginhoux.
2013. IRF4 transcription factor-dependent CD11b+ dendritic cells in human and mouse control mucosal IL-
17 cytokine responses. Immunity 38:970-983.
Schlitzer, A., V. Sivakamasundari, J. Chen, H.R. Sumatoh, J. Schreuder, J. Lum, B. Malleret, S. Zhang, A. Larbi, F. Zolezzi,
L. Renia, M. Poidinger, S. Naik, E.W. Newell, P. Robson, and F. Ginhoux. 2015. Identification of cDC1- and
cDC2-committed DC progenitors reveals early lineage priming at the common DC progenitor stage in the
bone marrow. Nat Immunol 16:718-728.
Scott, C.L., B. Soen, L. Martens, N. Skrypek, W. Saelens, J. Taminau, G. Blancke, G. Van Isterdael, D. Huylebroeck, J.
Haigh, Y. Saeys, M. Guilliams, B.N. Lambrecht, and G. Berx. 2016. The transcription factor Zeb2 regulates
development of conventional and plasmacytoid DCs by repressing Id2. J Exp Med 213:897-911.
Sekine, C., Y. Moriyama, A. Koyanagi, N. Koyama, H. Ogata, K. Okumura, and H. Yagita. 2009. Differential regulation of
splenic CD8- dendritic cells and marginal zone B cells by Notch ligands. Int Immunol 21:295-301.
Shin, H.M., M.E. Tilahun, O.H. Cho, K. Chandiran, C.A. Kuksin, S. Keerthivasan, A.H. Fauq, T.E. Golde, L. Miele, M. Thome,
B.A. Osborne, and L.M. Minter. 2014. NOTCH1 Can Initiate NF-kappaB Activation via Cytosolic Interactions
with Components of the T Cell Signalosome. Front Immunol 5:249.
Sichien, D., C.L. Scott, L. Martens, M. Vanderkerken, S. Van Gassen, M. Plantinga, T. Joeris, S. De Prijck, L. Vanhoutte, M.
Vanheerswynghels, G. Van Isterdael, W. Toussaint, F.B. Madeira, K. Vergote, W.W. Agace, B.E. Clausen, H.
Hammad, M. Dalod, Y. Saeys, B.N. Lambrecht, and M. Guilliams. 2016. IRF8 Transcription Factor Controls
Survival and Function of Terminally Differentiated Conventional and Plasmacytoid Dendritic Cells,
Respectively. Immunity 45:626-640.
Srinivas, S., T. Watanabe, C.S. Lin, C.M. William, Y. Tanabe, T.M. Jessell, and F. Costantini. 2001. Cre reporter strains
produced by targeted insertion of EYFP and ECFP into the ROSA26 locus. BMC Dev Biol 1:4.
Strange, K., J. Denton, and K. Nehrke. 2006. Ste20-type kinases: evolutionarily conserved regulators of ion transport
and cell volume. Physiology (Bethesda) 21:61-68.
Tedford, K., M. Steiner, S. Koshutin, K. Richter, L. Tech, Y. Eggers, I. Jansing, K. Schilling, A.E. Hauser, M. Korthals, and
K.D. Fischer. 2017. The opposing forces of shear flow and sphingosine-1-phosphate control marginal zone B
cell shuttling. Nat Commun 8:2261.
Tussiwand, R., B. Everts, G.E. Grajales-Reyes, N.M. Kretzer, A. Iwata, J. Bagaitkar, X. Wu, R. Wong, D.A. Anderson, T.L.
Murphy, E.J. Pearce, and K.M. Murphy. 2015. Klf4 expression in conventional dendritic cells is required for T
helper 2 cell responses. Immunity 42:916-928.
Van Gassen, S., B. Callebaut, M.J. Van Helden, B.N. Lambrecht, P. Demeester, T. Dhaene, and Y. Saeys. 2015. FlowSOM:
Using self-organizing maps for visualization and interpretation of cytometry data. Cytometry A 87:636-645.
Yi, T., and J.G. Cyster. 2013. EBI2-mediated bridging channel positioning supports splenic dendritic cell homeostasis
and particulate antigen capture. Elife 2:e00757.

128
 V. Results part three

Innate lymphoid cells control the homeostasis of

ESAM+ type 2 splenic dendritic cells through the
lymphotoxin β receptor

Matthias Vanderkerken, Satoshi Fukuyama, Charlotte L. Scott, Paula S. Norris, Gerard

Eberl, James P. Di Santo, Eric Vivier, Hamida Hammad, Carl F. Ware, Bart N. Lambrecht*,
Carl De Trez*

*These authors contributed equally to this work

Manuscript submitted

Author contribution:
Matthias Vanderkerken performed and analyzed the experiments in figures 1A-H, 2A-D,
3A-C, 4A-C, 4E-H, created the figures and wrote the manuscript.

129
ABSTRACT

Splenic type 2 dendritic cells (cDC2s) are crucial for the initiation of adaptive immune
responses against blood-borne antigens. The homeostasis of Notch2-dependent cDC2s
depends on cell-intrinsic activation of the lymphotoxin β receptor (LTβR) by
membrane-bound lymphotoxin α1β2. Although B cells are a major source of
lymphotoxin, we report here that B-cells are dispensable for the terminal differentiation
of splenic ESAM-positive cDC2s. Rather, innate lymphoid cells (ILCs) maintain cDC2
homeostasis in an LTβR-dependent manner. In Rag2-/- Il2rg-/- and Rag2-/- Rorc-/- mice,
loss of ILC3s selectively reduced the frequency of Notch2-dependent cDC2s. In contrast,
NK cells were not required for normal cDC2 development or differentiation. Agonism of
the LTβR or adoptive transfer of ILCs restored the terminal differentiation of ESAM-
positive cDC2s in Rag-/- Il2rg-/- mice. We conclude that lymphotoxin α1β2-expressing
ILC3s are necessary and sufficient to drive the terminal differentiation of Notch2-
dependent splenic cDC2s.

INTRODUCTION

Through their ability to recognize pathogens and present antigens to T cells, dendritic
cells (DCs) bridge innate and adaptive immunity. Conventional DCs (cDCs) are derived
from pre-cDC progenitors in the bone marrow that circulate in the blood stream to seed
peripheral tissues and secondary lymphoid organs (Grajales-Reyes et al., 2015; Liu et
al., 2009; Naik et al., 2007; Schlitzer et al., 2015). Pre-cDCs then further differentiate
into XCR1+ cDC1s that excel at cross-presenting exogenous antigens to CD8+ T cells or
into CD172a (SIRPα)-expressing cDC2s that preferentially prime CD4+ T cells (Guilliams
et al., 2016; Mildner and Jung, 2014). The instructive factors regulating the divergent
lineage commitment of cDCs are not fully understood, but clear roles have been
demonstrated for the transcription factors ID2, NFIL3, BATF3 and IRF8 in cDC1
development (Aliberti et al., 2003; Grajales-Reyes et al., 2015; Hildner et al., 2008;
Jackson et al., 2011; Kashiwada et al., 2011; Luda et al., 2016; Schiavoni et al., 2002).
Conversely, cDC2 development depends on the transcription factors IRF2, IRF4, RUNX3,
RELB, KLF4 and ZEB2 (Bajana et al., 2016; Briseno et al., 2017; Dicken et al., 2013;
Ichikawa et al., 2004; Persson et al., 2013; Schlitzer et al., 2013; Scott et al., 2016;
Tussiwand et al., 2015; Wu et al., 1998; Wu et al., 2016). In the spleen, a subpopulation
of terminally differentiated cDC2s was identified that expresses CD4 and endothelial
specific adhesion molecule (ESAM) and depends on activation of the NOTCH2 receptor
and its downstream transcription factor RBP-jΚ (Briseno et al., 2018; Caton et al., 2007;
Kirkling et al., 2018; Lewis et al., 2011; Satpathy et al., 2013). NOTCH2-dependent
cDC2s localize in marginal zone bridging channels through expression of the oxysterol-
sensitive chemokine receptor EBI2 (GPR183) and have a pivotal role in CD4+ T cell
priming to blood-borne particulate antigen (Briseno et al., 2018; Gatto et al., 2013; Yi
and Cyster, 2013).

130
One of the instructive signals for cDC2 development and homeostasis in the spleen is
the membrane-expressed heterotrimeric lymphotoxin α1β2 (LTα1β2). Blocking
lymphotoxin β receptor (LTβR) signaling was shown to decrease local proliferation of
cDCs, resulting in decreased CD4+ cDC2 numbers in the spleen (Abe et al., 2003;
Kabashima et al., 2005; Wang et al., 2005; Wu et al., 1999). Despite the fact that LTβR is
expressed on many cell types including stromal cells, mixed bone marrow chimeras and
conditional Ltbr knockout mice have established a cell-intrinsic requirement for LTβR
expression on cDC2s (De Trez et al., 2008; Kabashima et al., 2005; Satpathy et al., 2013).
The precise cellular interactions that provide cDC2 progenitors with membrane-bound
lymphotoxin have not been characterized. Transplantation of wild-type bone marrow
into lethally irradiated Lta-/- recipients restores splenic cDC numbers, suggesting that
the source of lymphotoxin is within the hematopoietic compartment (Wu et al., 1999).
In steady state, LTα1β2 is expressed by many different immune cells including B and T
lymphocytes, innate lymphoid cells (ILCs) as well as DCs (Kim et al., 2014; Kumar et al.,
2015; Ware et al., 1992). It has been proposed that lymphotoxin-expressing B cells are
crucial for the homeostasis of cDC2s through LTβR engagement on DCs (Kabashima et
al., 2005), but other reports have questioned this finding, showing gross abnormalities
in splenic architecture in lymphotoxin-deficient states and even higher frequencies of
cDC2s in Rag1-/- mice lacking B and T cells (Crowley et al., 1999; Fu et al., 1998;
Gonzalez et al., 1998; Moseman et al., 2012; Wu et al., 1999; Zindl et al., 2009).

Innate lymphoid cells are increasingly implicated in tissue homeostasis and constitute
an alternative source of lymphotoxin ligand (Vivier et al., 2018). Group 1 ILCs secrete
IFNγ and TNF in response to interleukin-12 (IL-12), IL-15 or IL-18, and resemble
NKp46+ natural killer (NK) cells phenotypically. Group 2 ILCs secrete IL-5 and IL-13
when activated by epithelial cytokines in mucosal barrier sites (Klose and Artis, 2016;
Zook and Kee, 2016). Group 3 ILCs are more heterogenous. During fetal life, RORγt-
dependent lymphoid tissue inducer (LTi) cells instruct the formation of lymphoid
organs by providing lymphotoxin to stromal cells (van de Pavert and Mebius, 2010). In
adult mice, RORγt+ ILC3s are mostly subdivided in NKp46- CD4+ CCR6+ LTi-like ILC3s
and NKp46+ CD4- CCR6- ILC3s (Melo-Gonzalez and Hepworth, 2017; Vivier et al., 2018).
Expression of LTα1β2 is a hallmark of all non-NK ILCs, but the highest levels are found in
murine and human ILC3s (Bjorklund et al., 2016; Gury-BenAri et al., 2016; Magri et al.,
2014; Reboldi et al., 2016).

Here, we addressed the hypothesis that innate lymphoid cells constitute a crucial source
of lymphotoxin that controls homeostasis of Notch2-dependent splenic cDCs. We
demonstrate that in the absence of B cells, the terminal differentiation of splenic
Notch2-dependent cDC2s is fully preserved in an LTα1β2-dependent manner. In
contrast, mice that lack ILC3s display a complete loss of ESAM+ CD4+ cDC2s. Defective
differentiation of cDC2s in Rag-/- Il2rg-/- mice that lack all T-, B- and innate lymphocytes
was partially restored by administration of LTβR-agonist or by adoptive transfer of

131
ILCs. These experiments reveal a crucial role for ILCs in controlling the homeostasis of
cDC2s in the spleen.

RESULTS

Notch2-dependent splenic cDC2s continuously require LTβR signaling

Previous studies addressing the role of lymphotoxin in DC homeostasis have relied on
separation of splenic cDC subsets based on CD4 and CD8α expression (Kabashima et al.,
2005; Wu et al., 1999). Using a more universal gating strategy that can be applied to all
tissues (Supplementary figure 1)(Guilliams et al., 2016), we re-addressed the role of
LTα1β2 signaling in the development of different cDC subsets. Wild type (WT) C57Bl/6
mice received an i.p. injection of LTβR-Fc fusion protein, a decoy receptor that blocks
the biological activity of LTα1β2. While the number of XCR1+ cDC1s remained
unaffected, the CD172a+ cDC2 count was significantly reduced in the spleen of treated
mice (Figure 1A, B). Within the CD172a+ cDC2 population, the largest decrease was
seen in ESAM+ CD4+ cells, although both CD4+ and CD4- numbers were affected (Suppl.
figure 2A). These findings are in line with previous reports in which ESAM+ DCs
depended most on the presence of the LTβR (Lewis et al., 2011; Satpathy et al., 2013).
These newer gating approaches also revealed that CD4 and CD8α can be almost
interchangeably used with ESAM and XCR1 to identify terminally differentiated cDC2
and cDC1, respectively. It should be noted that the antagonist also inhibits the activity of
LIGHT, another LTβR ligand (Mauri et al., 1998). However, LIGHT expression was not
required for cDC2 homeostasis (De Trez et al., 2008). Thus, LTα1β2 blockade targets all
cDC2s, but depletes terminally differentiated ESAM+ CD4+ cDC2s in particular.

B cells are not required to maintain Notch2-dependent DCs

The homeostasis of splenic cDC2s depends on triggering of the LTβR through direct cell-
cell contact. It has been proposed that the expression of LTα1β2 on B cells is essential for
the homeostasis of CD4+ splenic cDCs (Kabashima et al., 2005). To probe whether the
terminal differentiation of cDC2s indeed depends on the presence of B cells, we
compared the cDC populations of WT and Rag2-/- mice, that lack mature B and T cells
(Figure 1C, D). Although the total number of cDCs was decreased, we observed a
normal distribution of the CD8α+, CD4+ and CD8α- CD4- cDC subpopulations in the
spleen of Rag2-/- mice (Figure 2A, B). In particular, the percentage of terminally
differentiated ESAM+ CD4+ cDC2s was not decreased (Figure 1C, D). Surprisingly, we
found that cDC2s in Rag2-/- mice had upregulated CD4 expression. Though the
mechanisms regulating CD4 expression in cDCs have not been elucidated, we
hypothesized that the elevated CD4 expression might result from elevated cytokine
levels in the absence of competing T cells, as reported for interleukin-2 (IL-2) and IL-7
(Bando and Colonna, 2016). We therefore analyzed μMT mice, which carry a mutation
in the immunoglobulin μ chain gene leading to absence of mature B cells while
preserving T cell development. Similarly, the fractions of CD8α+ cDC1, and CD8α- CD4-

132
Figure 1: Splenic cDC2s are preserved in an LTβ-dependent manner in the absence of B cells
(legend on next page).

133
Figure 1: Splenic cDC2s are preserved in an LTβ-dependent manner in the absence of B cells.
(A) Representative contour plots of WT mice treated with mouse LTβR-Fc fusion protein (as described in
materials and methods) and PBS-treated littermates. Plots are pre-gated on splenic cDCs as illustrated in
Fig. S1. (B) Quantification of the indicated cDC subsets after LTβR-Fc treatment in WT mice as absolute
numbers and as percentage of total cDCs. (C) Representative flow plots of splenic cDCs from WT, Rag2-/-
and μMT mice. (D) Bar charts of ESAM+ CD4+ cDCs from WT, Rag2-/- and μMT spleens displaying cell
counts (top) or the percentage of cDCs (bottom). (E) Representative flow plot of ESAM and CD4
expression on splenic cDCs from Mb1cre x Rosa26-LSL-DTA mice and littermate controls. Cell counts and
cDC percentage of ESAM+ CD4+ cDCs are illustrated in (F). (G) Representative flow plots of splenic cDCs
from μMT mice treated with LTβR-Fc fusion protein or PBS (see materials and methods). (H)
Quantification of the ESAM+ CD4+ cDC subset in LTβR-Fc-treated and control μMT mice as cell counts and
percentage of cDCs. (A, B) Data are representative of 5 mice per group. (C, D) Data are pooled from 2
independent experiments (n=7-12 per group). (E, F) Data are representative of 2 independent
experiments (n=6-8 per group). (G, H) Data are representative of 2 independent experiments (n=6-7 per
group). The numbers adjacent to the gates in A, C, E and G represent the percentage of the total cDC
population. Bar graphs indicate mean ± SD. Two-tailed Mann-Whitney U test (B, D, F, H). * = p < 0.05, ** =
p < 0.01, *** = p < 0.001.

and CD4+ cDC2s were comparable between WT and µMT mice, effects also seen for
terminally differentiated cDC2s expressing both CD4 and ESAM (Figure 1C, D and data
not shown).

Under specific conditions, B cells can develop through a noncanonical pathway in

absence of functional surface IgM (Hasan et al., 2002; Macpherson et al., 2001).
Furthermore, μMT mice can still generate innate-like B1 cells (Ghosh et al., 2012). It is
therefore possible that residual or immature B cells still contribute to the homeostasis
of cDC2s in μMT mice. To generate mice that completely lack all peripheral B cells, we
crossed Mb1cre mice to Rosa26-lox-stop-lox-Diphteria toxin A (Mb1-DTA) mice. While no
residual CD19+ B cells were found in these Mb1-DTA mice (data not shown), the
distribution of their splenic cDC subsets did not differ from littermate control mice.
Again, despite lower cDC2 counts in the spleen of Mb1-DTA mice, the fraction of
remaining CD172a+ cDC2s that co-expressed ESAM and CD4 was unaltered (Figure 1E,
F). Combined, these data suggest that in absence of B lymphocytes, the development
and differentiation of Notch2-dependent dendritic cells is largely unaffected.

It is possible that, in the constitutive absence of B cells, increases in cytokine levels or

enhanced availability of certain ligands may have compensated for the absence of B-cell
derived lymphotoxin and maintained cDC2s through alternative pathways. To examine
whether cDCs still depended on LTα1β2 in the absence of B cells, we treated μMT-/- mice
with LTβR-Fc (Figure 1G, H). The number of ESAM+ CD4+ cDC2s was markedly
decreased in the spleen of LTβR-Fc treated μMT-/- mice compared to littermate controls.
In contrast, CD8α+ or XCR1+ cDCs and ESAM- CD4- cDC2s were not affected (Suppl.
figure 2B). Similarly, treatment of Rag2-/- mice with LTβR-Fc revealed a marked
decrease in cDC2 numbers through a specific loss of ESAM+ CD4+ dendritic cells

134
(Supplementary figure 3A-D). These findings imply that even in absence of B cells,
Notch2-dependent cDC2s require LTβR signaling.

ESAM+ CD4+ cDC2s depend on the common γ subunit of the IL-2R

Although B cells are the most abundant source of LTα1β2 in secondary lymphoid organs,
surface expression of LTα1β2 has been consistently observed on several other cell types.
Surface LTα1β2 can be induced by IL-2 and IL-7 in T cells and LTis (Ware et al., 1992;
Yoshida et al., 2002). Interleukin-2 family cytokines have also been implicated in the
development of various LTα1β2-expressing cells, such as T cells, NK cells and innate
lymphoid cells. For example, the survival of NK cells requires IL-15 (Kennedy et al.,
2000) whereas ILC development is IL-7 dependent (Zook and Kee, 2016). Mice deficient
in the common cytokine receptor γ chain (Il2rg-/-) lack functional receptors for IL-2, IL-
4, IL-7, IL-9, IL-15 and IL-21. To assess the role of the common γ chain in the
development or maintenance of splenic DCs, we compared the cDC populations of WT,
Rag2-/- and Rag2-/- Il2rg-/- mice in steady state (Figure 2A, B). While splenic cDC1 and
cDC2 numbers were reduced in both Rag2-/- and Rag2-/- Il2rg-/- mice compared to WT
mice, most cDC2s from Rag2-/- mice still expressed CD4 and ESAM, showing a full
competency of terminal differentiation. Compared to Rag2-/- mice, splenic cDC2
numbers were significantly lower in Rag2-/- Il2rg-/- mice. Strikingly, virtually none of the
remaining CD172a+ cDC2s expressed the terminal differentiation markers CD4 and
ESAM. In contrast to the drastic reduction in ESAM+ CD4+ cDC2s, CD8α+ cDC1 and CD4-
cDC2 numbers were comparable between Rag2-/- and Rag2-/- Il2rg-/- mice. We observed
increased levels of major histocompatibility complex II (MHCII) and a slight decrease in
CD24 surface expression on both cDC1s and cDC2s from Rag2-/- and Rag2-/- Il2rg-/- mice
compared to WT cDCs. The expression levels of CD11c, CD26, F4/80, IRF8 and IRF4 did
not differ between genotypes (Figure 2C and data not shown). Surface staining
demonstrated that LTβR expression was unaltered on cDC1s and slightly increased on
cDC2s from Rag2-/- and Rag2-/- Il2rg-/- mice compared to WT cDCs (Figure 2D). This
receptor increase may represent a lack of ligand-induced receptor downmodulation as
suggested in previous reports (Kabashima et al., 2005; Lewis et al., 2011; Yi and Cyster,
2013). Thus, the development and/or maintenance of terminally differentiated Notch-
dependent cDC2s is impaired in Rag-/- Il2rg-/- mice, lacking T-, B- and innate
lymphocytes.

Innate lymphoid cells control cDC2 homeostasis

Given our observations in Rag-/- Il2rg-/- mice, we hypothesized that Il2rg-dependent
ILCs can maintain cDC2s. The expression of LTα1β2 on LTis during embryonic
development is well documented and more recent single-cell RNA sequencing has also
confirmed expression of Ltb in ILC populations in adult humans and mice (Bjorklund et
al., 2016; Gury-BenAri et al., 2016; Kruglov et al., 2013; Reboldi et al., 2016). To
establish whether loss of NK cells or innate lymphoid cells impacts the homeostasis of

135
Figure 2: Rag2-/- Il2rg-/- mice lack ESAM+ CD4+ splenic cDC2s.
(A) Representative contour plots of splenic cDCs from WT, Rag2-/- and Rag2-/- Il2rg-/- mice in steady state.
Numbers adjacent to the gates indicate the percentage of all cDCs. (B) Quantification of splenic cDC
subsets from WT, Rag2-/- and Rag2-/- Il2rg-/- mice as absolute numbers (left) or as a percentage of all cDCs
(right). (C) MFI histograms for the indicated surface markers on XCR1+ (upper row) and CD172a+ (lower
row) splenic cDCs. (D) MFI histograms of surface LTβR on splenic XCR1+ and CD172a+ cDCs. An FMO for
LTβR staining is depicted in grey. (C, D) Blue lines represent WT, green lines represent Rag2-/- and orange
lines represent Rag2-/- Il2rg-/- cDCs. (A-C) Data representative of at least 4 independent experiments (n=
15-20 mice per group). (B) Data pooled from 2 independent experiments (n=7-8 mice). (D) Data are
representative of 2 independent experiments (n=8 per group). Bar graphs indicate mean ± SD. One-way
ANOVA with Tukey’s post-hoc test (B). * = p < 0.05, ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001.

splenic DCs, we treated Rag2-/- mice with depleting anti-CD90 and anti-NK1.1
antibodies. The depletion efficiency was validated using the gating strategy illustrated
in Supplementary figure 4F. Treatment with anti-CD90 almost completely eliminated
Tbet+ ILC1s, GATA3+ ILC2s and RORγt+ ILC3s while causing only a partial, nonsignificant
reduction in NKp46+ NK1.1+ NK cells in the spleen. Inversely, anti-NK1.1 efficiently
depleted NK cells but left other ILCs intact (Figure 4C).

136
In the spleens of Rag2-/- mice treated with anti-NK1.1 antibodies, the number of ESAM+
CD4+ cDC2s did not differ from isotype treated littermate controls (Figure 3A, B).
Because such depletion is temporary and incomplete, we also investigated the cDC
populations of IL-15-/- mice that constitutively lack most NK cells (Kennedy et al., 2000).
Terminally differentiated CD4+ cDC2s were intact in Rag-/- Il15-/- mice compared with
Rag-/- Il15+/+ littermate controls (Supplementary figure 4A, B). Others have reported
development of NK cells that could respond to murine cytomegalovirus through Ly49H
and IL-12 activation independently of IL-15 (Sun et al., 2009). To further confirm the
redundant role of NK cells in cDC homeostasis, we therefore crossed NKp46cre mice to
Rosa26-lox-stop-lox-Diphteria toxin A mice to generate mice completely devoid of NK
cells (Supplementary figure 4E). Of note, these mice also lack NKp46+ ILC1s and a
subset of CCR6- CD4- ILC3s (Cella et al., 2009; Cortez and Colonna, 2016; Luci et al.,
2009; Narni-Mancinelli et al., 2011; Sanos et al., 2009). In line with experiments in Il15-/-
mice, the absence of NK cells in NKp46-DTA mice did not impact the presence and
differentiation of LTβ-dependent cDC2s (Supplementary figure 4C, D). In contrast,
treatment of Rag2-/- mice with anti-CD90 antibodies significantly reduced CD172a+
cDC2 numbers compared to isotype-treated controls (Figure 3A, B). Remarkably, the
effect was again specific for ESAM+ CD4+ cDCs, sparing XCR1+ cDCs and CD4- cDC2s. We
concluded that ILCs, but not NK cells, are required for the homeostasis of Notch2-
dependent splenic cDCs.

Within the ILC family, ILC3s express the highest levels of Ltb on mRNA level (Bjorklund
et al., 2016; Gury-BenAri et al., 2016; Reboldi et al., 2016), and other groups have
reported the close anatomical proximity of RORγt+ ILC3s and cDC2 in the splenic
marginal zone and bridging channels (Hoorweg et al., 2015; Kim et al., 2007; Magri et
al., 2014). To validate cell-cell contacts between ILCs and cDCs, we performed
fluorescent imaging on splenic sections from Rag2-/- and Rag2-/- Il2rg-/- mice. IL7Rα
(CD127)+ cells were readily detected within CD11c+ clusters in Rag2-/- mice (Figure
3F). We identified prominent and intimate membrane contacts between IL7Rα+ CD4+
cells and CD11c+ cells (Figure 3G). Because the expression of CD4 on the surface of
ILC3s is higher (~7-10 times) than on cDC2s (based on mean fluorescence intensity
detected by flow cytometry, data not shown), the CD4 signal of cDC2s using immuno-
histofluorescence is much weaker. As expected, substantially fewer IL7Rα+ cells resided
in the CD11c+ cDC clusters in Rag2-/- Il2rg-/- mice (Figure 3F). To assess whether
absence of ILC3s specifically affected cDC2 homeostasis, we compared the cDC subsets
of Rag1-/- Rorc-/- and Rag1-/- mice (Figure 3D, E). While the frequencies of CD8α+ cDC1
and CD4- cDC2 did not differ between the experimental groups, CD4+ cDC2s were
significantly decreased in Rag1-/- Rorc-/- spleen. Taken together, these data suggest a
nonredundant role for RORγt-dependent ILC3s in the development or maintenance of
LTβ-dependent dendritic cells in lymphopenic mice.

137
Figure 3: RORγt+ ILC3s, but not NK cells, control the terminal differentiation of cDC2s in
lymphopenic mice (legend on next page).

138
Figure 3: RORγt+ ILC3s, but not NK cells, control the terminal differentiation of cDC2s in
lymphopenic mice.
(A) Representative contour plots of splenic cDCs from Rag2-/- mice treated with depleting anti-CD90,
anti-NK1.1 or isotype control antibodies as described in materials and methods. (B) Quantification of the
indicated cDC subsets in the spleen of treated Rag2-/- mice as absolute numbers (left) or as a percentage
of total cDCs (right). (C) Bar charts of NK cell, ILC1, ILC2 and ILC3 counts in the spleen of Rag2-/- mice
after treatment with anti-CD90 or anti-NK1.1. (D) Representative flow plots of splenic cDCs from Rag1-/-
and Rag1-/- Rorc-/- mice, quantified in (E) as a percentage of live cells (left) or of total cDCs (right). (F)
Immunofluorescence image of spleen sections from Rag2-/- and Rag2-/- Il2rg-/- mice. Sections were stained
for CD11c (green) and IL-7Rα (red). Magnification 20x. (G) Immunofluorescence image of Rag2-/- spleen
stained for CD11c (green), IL7Rα (blue) and CD4 (red). Left: 63x magnification, right: magnification of
white rectangle. (A, B) Data representative of 3 independent experiments (n=8-12). (C) Data pooled from
2 independent experiments (n=7-8 per group). (D, E) Data pooled from 2 independent experiments (n= 4-
5 per group). (F, G) Data representative of 2 independent experiments. The numbers adjacent to the gates
in A and D represent the percentage of the total cDC population for that gate. Bar graphs indicate mean ±
SD. Kruskal-Wallis test with post-hoc Dunn’s test (B, C), two-tailed Mann-Whitney U test (E). * = p < 0.05,
** = p < 0.01, *** = p < 0.001, **** = p < 0.0001.

LTβR-agonism restores ESAM+ CD4+ cDC2s in Rag2-/- Il2rg-/- mice

While the defective development of ILC3s in Rag2-/- Il2rg-/- mice and Rag2-/- Rorc-/- mice
could explain the lack of LTβ-dependent cDCs in these mice, it is possible that loss of the
common cytokine receptor γ chain affects cDCs through another mechanism. Il2rg-
dependent cytokines might stimulate LTα1β2 expression on a yet undiscovered cellular
source. Alternatively, loss of the common γ chain could impair the development or
survival of dendritic cells directly via an LTβR-independent pathway. In that case, cDC
development would be hampered irrespective of the presence of LTα1β2 in the micro-
environment. To probe the intrinsic ability of cDCs to respond to LTβR stimulation, we
made use of a monoclonal LTβR-antibody that has agonistic effects in vivo. Remarkably,
cDC2 numbers were almost completely restored to Rag2-/- levels when Rag2-/- Il2rg-/-
mice were treated with this LTβR-agonist. In particular, the frequency of ESAM+ CD4+
cDC2s increased significantly after LTβR stimulation (Figure 4 A, B). Treatment of
Rag2-/- Il2rg-/- mice with anti-LTβR did not restore NK or ILC3 numbers (Figure 4C).
Confocal microscopy of Rag2-/- spleens revealed clusters of CD11c+ dendritic cells
containing CD4+ and CD8α+ cells around central arterioles (Figure 4D), in agreement
with previous reports (Crowley et al., 1999). The size and number of CD11c+ DC clusters
were dramatically reduced in Rag2-/- Il2rg-/- mice. However, treatment with agonistic
anti-LTβR antibody rescued the formation of cDC clusters in Rag2-/- Il2rg-/- mice,
correlating with the expansion of CD8α- cDCs (Figure 4D). Together, these findings
demonstrate the intrinsic ability of cDCs from Rag2-/- Il2rg-/- mice to respond to LTβR-
agonism and develop into phenotypically normal, terminally differentiated cDCs. This
further supports the idea that the homeostasis of ESAM+ CD4+ cDC2s in Rag2-/- Il2rg-/- is
impaired through the absence of ILC-derived LTα1β2.

139
Transfer of ILCs into Rag2-/- Il2rg-/- mice rescues LTβ-dependent cDC2s
The use of agonist LTβR-antibody might have led to a non-physiological level of LTβR
agonism in Rag2-/- Il2rg-/- mice. We reasoned that if ILCs are required to maintain
ESAM+ CD4+ cDCs in Rag2-/- mice, transfer of ILCs into Rag2-/- Il2rg-/- mice might also be
sufficient to restore cDC2 homeostasis. To test this, we isolated CD45.1+ CD11c- NK1.1-
CD90hi CD127+ ILCs from Rag2-/- mice (Supplementary figure 5A and B) and
adoptively transferred the cells into Rag2-/- Il2rg-/- mice (Figure 4E). To boost ILC
proliferation in vivo prior to recovery, Rag2-/- donor mice were treated with
recombinant human interleukin-7 (rhIL-7) in complex with an anti-IL-7 antibody
beforehand. Two weeks after transfer to naïve Rag2-/- Il2rg-/- mice, we were able to
retrieve the donor cells in the spleens of acceptor mice. Notably, a significant proportion
of transferred ILCs (sorted as NK1.1-) had adopted an NK1.1+ NKp46+ phenotype
(Figure 4F) with upregulation of T-bet and downregulation of RORγt (data not shown).
This conversion of ILC3s to ILC1s has been reported previously (Klose et al., 2013;
Vonarbourg et al., 2010). Nevertheless, a significant number of RORγt+ ILC3s could also
be retrieved in the acceptor mice (Figure 4F). Transfer of ILCs to Rag2-/- Il2rg-/- mice
significantly increased the number of CD172a+ cDC2 and restored the expression levels
ESAM and CD4 expression (Figure 4G, H). In conclusion, innate lymphoid cells
constitute an essential source of LTα1β2 that is required and sufficient for the
homeostasis of Notch2-dependent splenic cDCs.

Figure 4: Transfer of ILCs in Rag2-/- Il2rg-/- mice restores ESAM+ CD4+ cDCs.
(A) Representative flow plots of splenic cDCs from Rag2-/- controls and Rag2-/- Il2rg-/- mice treated with
agonistic anti-LTβR antibody (as described in materials and methods) or PBS. The ESAM+ CD4+ subset of
cDCs is quantified in (B) as percentage of live cells and of total cDCs. (C) Bars depict the absolute number
of splenic NK cells (left) and ILC3s (right) in Rag2-/- mice and Rag2-/- Il2rg-/- mice treated with anti-LTβR
or PBS. (D) Immunofluorescence image of spleen sections from Rag2-/-, Rag2-/- Il2rg-/- control mice and
Rag2-/- Il2rg-/- mice treated with anti-LTβR. Sections were stained for CD11c (green), CD8α (blue) and
CD4 (red). Left: 5x magnification, right: 20x magnification of white rectangles. (E) Schematic
representation of the adoptive transfer protocol. (F) Representative flow plots of Rag2-/- Il2rg-/- spleens
14 days after adoptive transfer of ILCs or PBS alone. For the ILC3 gate, plots were pre-gated on live, CD3e-
CD19- Ter119- CD45.1+ NK1.1- CD90hi cells. For the NK1.1+ NKp46+ cell gate, plots were pre-gated on live,
CD3e- CD19- Ter119- CD45.1+ cells. Numbers adjacent to the gates indicate the percentage of the parent
population. (G) Splenic cDCs of Rag2-/- Il2rg-/- mice 14 days after adoptive transfer of ILCs versus PBS-
treated controls. (H) Quantification of ESAM+ CD4+ cDCs as absolute cell number (top) and as percentage
of cDCs (bottom left). Quantification of ESAM MFI on CD172a+ cDCs in Rag2-/- Il2rg-/- mice after ILC
transfer (bottom right). (A, B) Data representative of 4 independent experiments (n=4 for Rag2-/-, n=11-
13 for both Rag2-/- Il2rg-/- groups). (B, C) Data pooled from 2 independent experiments (n=3-7 per group).
(D) Data representative of 2 independent experiments. (F-H) Data pooled from 2 independent
experiments (n=6-11 per group). The numbers adjacent to the gates in A, F and G represent the
percentage of total cDCs. Bar graphs indicate mean ± SD. Kruskal-Wallis test with post-hoc Dunn’s test (B,
C), two-tailed Mann-Whitney U test (H). * = p < 0.05, ** = p < 0.01, *** = p < 0.001.

140
Figure 4: Transfer of ILCs in Rag2-/- Il2rg-/- mice restores ESAM+ CD4+ cDCs (legend on previous page).

141
DISCUSSION

The requirement of LTβR for cDC2 homeostasis has been appreciated for almost 15
years. However, how LTβR is activated in vivo has remained more enigmatic. It was
proposed that B cells constitute an indispensable source of LTa1β2 ligand promoting
cDC2 maintenance. The first basis of this argument has been that B cell-specific
lymphotoxin α (LTα) overexpression increases CD4+ cDC2s in the spleen. The second
was that irradiated WT mice reconstituted with a mixed µMT:Lta-/- bone marrow lack
lymphotoxin on B cells and have a defect in CD4+ cDC2s (Kabashima et al., 2005).

These findings are at odds with the conclusions from our report. We have used three
different genetic models (Rag2-/-, µMT and Mb1-DTA) that constitutively lack B cells, all
of which contained a relatively normal distribution of cDC subsets. Though a decrease in
the absolute number of the global cDC population was observed, a normal percentage of
cDCs was able to acquire a terminally differentiated ESAM+ CD4+ phenotype.
Kabashima et al. used irradiation and bone marrow chimerism, which might have
induced alterations in other immune or stromal cells that depend on LTβ. B-cell derived
LTa1β2 is indeed a trophic factor for lymphoid tissue stromal cells and serves as a tonic
signal to maintain normal follicular architecture (Dubey et al., 2017; Fu et al., 1998;
Gonzalez et al., 1998). Moreover, LTa1β2 is required for a normal marginal sinus
anatomy and absence of B cells disrupted the function of marginal zone reticular cells
(MRCs) and marginal zone macrophages in spleen and subcapsular macrophages in
lymph node (Crowley et al., 1999; Moseman et al., 2012; Zindl et al., 2009). Therefore,
the described effects of B-cell derived LTα1β2 on cDC2s might have resulted from
indirect effects rather than direct B cell–cDC interactions. A possibility that remains to
be explored is that irradiation of WT recipients led to damage to ILC3s, and that BM
transplantation did not sufficiently restore lymphotoxin to wild type levels in µMT:Lta-/-
chimeras. Although ILCs are relatively radioresistant, when damaged, it is known that
replacement with wild type ILCs can take a long time in chimeras (Vivier et al., 2018).
Therefore the µMT:Lta-/- chimeras analyzed at 8 weeks (Kabashima et al., 2005) might
have had defects in ILCs on top of a lack of lymphotoxin in B cells that could have
explained the profound defect in CD4+ cDC2s. Alternatively, redundancy might exist
between B cells and ILC3s, which would not be surprising given similar reports of
redundant functions between innate and adaptive lymphocytes (Fu et al., 1998; Kim et
al., 2007; Rankin et al., 2016; Vely et al., 2016).

Treatment of B-cell deficient mice with LTβR-Fc drastically reduced the ESAM+ cDC2
population, in agreement with previous observations of a non-B, non-T cell maintaining
DC numbers in the spleen (Wu et al., 1999). These results are also in line with previous
findings that LTβR and Notch signaling intersect in ESAM+ cDC2s. Ligation of the
NOTCH2 receptor on cDC2s by stromal cell-derived DLL1 triggers a differentiation
program required for optimal CD4+ T cell priming and induction of germinal center

142
responses. Competitive bone marrow chimeras and epistasis experiments revealed that
specifically these Notch-instructed cDC2s depended strongly on LTβR signaling
(Briseno et al., 2018; Fasnacht et al., 2014; Lewis et al., 2011; Satpathy et al., 2013).
Interestingly, LTβR-/- bone marrow was outcompeted by its wild type counterpart in
generating CD103+ cDC2s in the small intestine, another Notch2-dependent cDC
population (Satpathy et al., 2013). However, it is unclear whether or how impaired
Notch instruction affects the ability to encounter or respond to LTα1β2. In that regard, it
would be interesting to determine whether LTβR agonism can mitigate the effects of
Notch2 deletion in cDC2s. Inversely, LT signals might be required to become responsive
to Notch ligation.

By comparing Rag2-/-, Rag-/- Il2rg-/-, Rag-/- Rorc-/- mice and Rag-/- mice treated with anti-
CD90 antibodies, we now identified ILC3s as necessary and sufficient for the
development of ESAM+ CD4+ cDCs in lymphopenic mice. In animals lacking ILCs, LTβR
agonism led to a full restoration of ESAM+ CD4+ cDC numbers, suggesting that ILC3s
perform this task via provision of lymphotoxin. Although we have not formally shown
that lymphotoxin is intrinsically derived from ILC3s, previous reports support this
scenario. The expression of lymphotoxin on group 3 ILC3s is well known for its role in
lymphoid tissue development. Similarly, adult LTi-like ILC3s express high levels of
lymphotoxin and exert vital functions in immune system homeostasis (Kim et al., 2014;
Zhang et al., 2018). The interaction between LTα1β2 on ILC3s and LTβR on cDCs is
critically required for the production of IL-23 by dendritic cells in the intestine (Reboldi
et al., 2016; Tumanov et al., 2011). In the subepithelial dome of Peyer’s patches, ILC3s
and Batf3-independent CD11b+ cDC2s were also shown to interact in a LTβR-dependent
manner. Despite the close and long term interactions between LTα1β2-expressing B
cells and DCs in the Peyer’ patch, LTα1β2 from ILC3s was demonstrated to be critically
required to instruct DCs to induce isotype switching to IgA in B cells (Kruglov et al.,
2013; Reboldi et al., 2016). One possible explanation offered was the higher levels of
LTα1β2 on ILC3s compared with B cells.

Taken together, ILC3s and cDC2s seem to occupy a conserved niche where LTβR- and
Notch-ligands congregate and instruct the terminal differentiation of ESAM+ CD4+ cDCs.
Recently, some light was shed on potential factors that could regulate access to this
niche. The chemokine receptor EBI2 (Gpr183) is required for the localization of ILC3s in
intestinal cryptopatches and isolated lymphoid follicles (Chu et al., 2018; Emgard et al.,
2018). Interestingly, splenic cDC2s also required EBI2 expression for correct
positioning in splenic bridging channels and at the B:T zone interface (Gatto et al., 2013;
Yi and Cyster, 2013). Marginal reticular cells express high levels of Ch25h, the enzyme
that allows synthesis of the main EBI2 ligand 7α,25-dihydrocholesterol (Rodda et al.,
2018). Strikingly, the loss of CD4+ splenic cDC2s in EBI2-deficient mice could be rescued
by LTβR-agonism, and it has been suggested that EBI2 controls the access of cDCs to
sources of LTα1β2 (Yi and Cyster, 2013). Whether ILC3s use EBI2 to localize in the

143
splenic marginal zone is not known, but it is tempting to speculate that the EBI2
receptor might congregate cDC2s, ILC3s and B cells to a common niche for LTβR
activation. Alternatively, other chemotactic factors such as CCL20 could play a role. The
subepithelial dome of Peyer’s patches is rich in CCL20, the ligand for the chemokine
receptor CCR6 that is expressed by ILC3s, B cells and cDCs. Upregulation of CCL20 was
induced in follicle-associated epithelium by lymphotoxin (Rumbo et al., 2004).
Interestingly, CCR6 is also expressed on cDC2s in the spleen, though its functional
significance remains to be determined (Kucharzik et al., 2002).

In summary, defects in LTβR activation, Notch signaling or EBI2 appear to target a

common pathway of terminal cDC2 differentiation in the spleen. By identifying the ILC3
as a necessary and sufficient cellular partner providing lymphotoxin to differentiating
DCs, we believe it will not be too long before we fully understand the precise molecular
and cellular sequence of events that lead to differentiation of this unique subset of
dendritic cells.

144
MATERIALS AND METHODS

Mice
Wild type (WT) C57Bl/6J mice were obtained from the Janvier laboratory. All
transgenic mice and their corresponding littermate controls were housed in the VIB-
UGhent animal facility under specific pathogen free conditions. Housing conditions
entailed individually ventilated cages in a controlled day-night cycle and food and water
intake ad libitum. As no differences in the dendritic cell phenotype were observed
between male and female mice, both sexes were used. All mice were 6-10 weeks of age.
Mice were age- and sex-matched and blindly randomized into experimental groups. All
animal experiments performed were approved by the local animal ethics committee
(VIB-UGhent) and were performed in accordance with the Belgian animal protection
law.

Rag2-/-, Ighmtm1Cgn (μMT) and Rosa26-lox-STOP-lox-Diphteria toxin A mice (RosaDTA/DTA)

mice were acquired from Jackson. RosaDTA/DTA mice were crossed to Mb1Cre/Cre mice
(Hobeika et al., 2006) that were kindly provided by Michael Reth from the Max Plack
Institute of immunobiology and epigenetics (Freiburg, Germany). The B cell deficient
mice resulting from this intercross are designated as Mb1-DTA mice. RosaDTA/DTA mice
were also crossed to NcriCre/iCre mice (Narni-Mancinelli et al., 2011), kindly provided by
Eric Vivier from CIML (Marseille, France). NK cell-deficient offspring from this cross are
designated as NKp46-DTA mice. Rag1-/- IL15-/- and Rag1-/- Rorc-/- double knockout mice
were a kind gift of James P. Di Santo and Gerard Eberl from the Institut Pasteur (Paris,
France). Rag2-/- Il2rg-/- mice were acquired from Taconic and were on a mixed
C57BL/6J x C57BL/10SgSnAi background. All other mice were backcrossed on a
C57Bl/6 background for at least 10 generations.

Reagents and treatment protocols

To block LTα1β2 in vivo, mice were treated with an LTβR-Fc decoy receptor. This fusion
protein, generated from the human IgG1 Fc domain and the extracellular domain of the
mouse lymphotoxin β receptor, was shown to inhibit the biological activity of
lymphotoxin α1β2 (Browning et al., 1997). Mice were administered 100μg of LTβR-Fc
per injection. For LTβR agonism in Rag2-/- Il2rg-/- mice, 100μg of an agonistic rat anti-
mouse LTβR antibody (anti-LTβR, clone 4H8) was injected (Banks et al., 2005). To
deplete NK cells or ILCs, 300μg of rat anti-mouse NK1.1 (PK136) or 300μg of rat anti-
mouse CD90 (YTS154) were administered, respectively. Littermate control Rag2-/- mice
were treated with isotype control antibody (mouse IgG2a κ, clone C1.18.4, BioXCell).
Anti-NK1.1 and anti-CD90.2 antibodies were acquired from Bioceros. All reagents
described above were diluted in PBS. Mice were injected intraperitoneally on day
1,3,7,10 and euthanised on day 12-13. To expand ILCs in vivo, Rag2-/- mice were treated
intraperitoneally with a mixture of 2,5 μg recombinant human IL-7 (Stemcell

145
Technologies) and 15 μg mouse anti-human IL-7 antibody (BioXCell, clone M25) diluted
in PBS every other day for one week.

Isolation of tissue leukocytes

Spleens were perfused with RPMI-1640 (Gibco) containing collagenase III (0.35mg/ml,
CLS-3, Worthington) and incubated for 30 min at 37°C in RPMI-1640 medium
containing collagenase III (1.4mg/ml). Digestion was stopped by adding PBS with 2mM
EDTA and cell suspensions were passed through a 70µm nylon mesh filter. Samples
were incubated in ammonium chloride buffer (10mM KHCO3, 155 mM NH4Cl, 0.1 mM
EDTA in milliQ water) for erythrocyte lysis. Total cell counts were obtained by adding
counting beads (123count eBeads, Thermo Fisher Scientific). Mesenteric lymph nodes
were cut into small pieces, digested for 30 min in RPMI-1640 containing 20 μg/ml
Liberase TM (Roche) and 10 U/ml DNAse I (Roche) at 37°C and filtered through a 70µm
nylon sieve. Intestines were soaked in PBS 2% fetal calf serum (Bodinco), opened
longitudinally and cut into 5mm sections. These segments were incubated twice in PBS
with 2mM EDTA at 37°C for 20 min after which the supernatant epithelial cells were
discarded. Remaining tissue was digested in RPMI-1640 containing 0.6mg/ml
Collagenase VIII (Sigma) and 10U/ml DNAse I (Roche) at 37°C with shaking for 15 min,
and passed through a 40µm filter afterwards.

Flow cytometry
Single cell suspensions were first incubated with a fixable viability dye (eFluor506 or
eFluor780 from eBioscience) to identify dead cells and with an FcγRII/III antibody
(2.4G2) for 30 minutes to limit non-specific binding. After washing, the cells were
incubated with a mixture of fluorescently labeled antibodies for 30 min at 4°C.
Following antibodies were used: CD3 (145-2C11), CD19 (1D3), CD11c (N418), NK1.1
(PK136), MHCII (M5/114.15.2), CD127 (SB/199), CD127 (A7R34), Ly6G (1A8), CD11b
(M1/70), F4/80 (BM8), CD26 (H194-112), CD64 (X54-5/7.1), CD8α (53-6.7), CD24
(M1/69), CD172a (P84), XCR1 (ZET), Ter119 (TER-119), CD4 (GK1.5), CD90.2 (30-
H12), CD90.1 (OX-1), CD4 (RM4-5), ESAM (1G8), NKp46 (29A1.4), LTβR (3C8), CD45
(30-F11). When biotin-labeled primary antibodies were used, a second surface-staining
step with BV786- or BV605-coupled streptavidin (BD Biosciences) was performed. For
intracellular staining of the transcription factors IRF4, IRF8, RORγt, Tbet and Gata3,
cells were fixed using the FoxP3 fixation/permeabilisation kit (eBioscience) according
to the manufacturer’s protocol. Cells were incubated with PerCP-efluor710-labeled anti-
IRF8 (V3GYWCH), unlabeled goat anti-mouse IRF4 (M-17), PE-labeled RORγt (AFKJS-9)
or PE-CF594-labeled RORγt (Q31-378), PE-Cy7-labeled TBET (4B10) or APC-labeled
GATA3 (TWAJ). Unlabeled IRF4 was detected with an AF647-labeled donkey anti goat
antibody (Invitrogen) in an additional intracellular staining step. Samples were
acquired on an LSRFortessa cytometer with FACSDiva software (BD Biosciences) and
analyzed using FlowJo software (LLC, Ashland, OR). Cell sorting was performed using a
FACSAria II (BD Biosciences).

146
Immunofluorescence microscopy
Spleens were embedded in OCT compound (Sakura Finetek USA, Torrance, CA), frozen
by liquid N2 and stored at -80°C. Cryostat 6-µm sections were fixed in acetone for 10
minutes and rehydrated in PBS. Sections were incubated for 30 min with 1% H2O2 for
biotin-conjugated anti-CD11c (HL3; Pharmingen). To block endogenous biotin, the
sections for biotin-conjugated anti-CD11c were further treated with the Avidin/Biotin
block (Vector Laboratories, Burlingame, CA). Sections were treated for 30 min with PBS
containing 1% BSA (PBS-BSA) to block nonspecific binding, and stained overnight with
uncoupled, biotinylated or appropriate fluorescence-labeled primary antibody in PBS-
BSA. Primary antibodies used were: FITC-conjugated anti-CD11c (HL3; Pharmingen),
PE-conjugated anti-CD4 (RM4-5; Pharmingen), APC-conjugated anti-CD8a (53-6.7;
Pharmingen), biotin-conjugated CD11c (HL3; Pharmingen) and biotin-conjugated anti-
IL-7Rα (A7R34; eBioscience, San Diego, CA). Sections were washed in PBS and
incubated with the secondary detection reagents streptavidin-APC (Pharmingen) for IL-
7Rα staining or streptavidin-HRP (PerkinElmer, Boston, MA) for CD11c staining. The
Tyramid Signal Amplification (TSA) system (PerkinElmer) was used for CD11c
amplification in accordance with the product protocol. The slides were washed in PBS
and mounted in anti-fading GEL/MOUNT (Biomeda, Foster City, CA). Sections were
visualized using a Marianas fluorescence microscope with SlideBook software. Images
were processed with ImageJ (NIH software).

Adoptive transfer of ILCs

Rag2-/- donor mice were treated with rhIL-7:anti-IL-7 complex as described above on
day 1, 3 and 5. On day 6, their spleen, small intestine and mesenteric lymph nodes were
processed to single-cell suspensions and stained for fluorescence-activated cell sorting.
Innate lymphoid cells were sorted as single, live, CD45+ Ly6G- CD11c- autofluorescence-
lineage (CD3e/CD19/NK1.1)- CD90hi CD127+ cells. Approximately 400 000 ILCs were
adoptively transferred into Rag2-/- Il2rg-/- mice via intrasplenic injection. In this
procedure, acceptor mice are anesthetized with isoflurane, the flank skin and
peritoneum are opened and 100µl of PBS is injected directly into the spleen using a 29G
syringe. The peritoneum is closed and the skin is stapled. Buprenorphine analgesia (0,1
mg/kg) is administered subcutaneously. Littermate control mice underwent the same
procedure but received PBS without ILCs. Acceptor mice were euthanized 14 days after
the adoptive transfer.

Statistical analysis
Statistical analyses were performed with Graphpad Prism 7.0 (Graphpad Software, La
Jolla, CA). For comparisons between two groups, a two-tailed Mann-Whitney test was
used. For comparisons between three or more groups, a one-way ANOVA with Tukey’s
correction for multiple comparisons was used. If the assumptions for ANOVA were not
met, we performed the Kruskal-Wallis test for unpaired data, if significant followed by
Dunn’s post hoc test. The significance level α was 0.05. Error bars represent standard

147
deviation of the mean. Non-significant differences are not indicated, significant
differences are expressed as * = p<0.05, ** = p<0.01, *** = p<0.001, ****=p<0.0001.

AUTHOR CONTRIBUTIONS

MV performed and analyzed experiments and wrote the manuscript. SF and CLS
performed experiments. PSN, GE, JPDS and EV provided key reagents and mice. HH
designed and analyzed experiments. CFW designed experiments and provided reagents.
BNL and CDT conceptualized the research, designed experiments, analyzed data and
wrote the manuscript.

ACKNOWLEDGEMENTS

MV is supported by a fellowship grant from FWO (grant 3F023515W). C.L.S. is

supported by a Marie Curie Intra-European Fellowship (IEF) as part of Horizon 2020.
H.H. is supported by a Ghent University Grant (GOA 01G02817). B.N.L. was supported
by an ERC consolidator grant 261231, the University of Ghent Multidisciplinary
Research Platform (MRP-GROUP-ID) and a Ghent University grant (GOA 01G02817).

148
SUPPLEMENTARY MATERIALS

Supplementary figure 1: Gating strategy for cDCs.

Gating strategy for the identification of conventional dendritic cells (cDCs) in wild type spleen.
Conventional DCs are defined as single, live, CD64- lineage (CD3e/CD19/NK1.1/Ter119)- CD26+ CD11chi
MHCII+ cells. Conventional DC1s are defined as XCR1+ CD172a-, while cDC2s are XCR1- CD172a+.
Alternative gates based on CD8α, ESAM or CD4 surface expression are illustrated.

149
Supplementary figure 2: LTβ-signals are required for ESAM expression in cDC2s
(A) Number of splenic cDC2 subsets based on expression of ESAM and CD4. Data are representative of 5
mice per group. (B) Number of cDC1s, cDC2s and cDC2 subsets in spleen of µMT mice treated with LTβR-
Fc or with PBS. Data pooled from 2 independent experiments (n=6-7 mice per group). Bar graphs indicate
mean ± SD. Two-tailed Mann-Whitney U test (A, B). * = p < 0.05, ** = p < 0.01, *** = p < 0.001.

Supplementary figure 3: LTβ maintains cDC2 homeostasis in Rag-/- mice.

Rag2-/- mice were treated with LTβR-Fc fusion protein or PBS only (see materials and methods). (A)
Representative flow plots of ESAM and CD4 expression splenic cDCs from LTβR-treated and control Rag2-
/- mice. (B) Representative histograms of CD4 and ESAM surface expression on CD172a+ splenic cDCs of

Rag2-/- mice treated with LTβR-Fc fusion protein (green) or PBS (blue). (C) Bar charts displaying the
different cDC subsets as percentage of live cells (left) or total cDCs (right). (D) Quantification of the
number of splenic CD172a+ cDCs (left) and the MFI of ESAM on CD172a+ cDCs in LTβR-treated Rag2-/-
mice. (A-D) Data are pooled from 2 independent experiments (n=4-5 per group). The numbers adjacent to
the gates in (A) represent the percentage of the total cDC population. Bar graphs indicate mean ± SD.
Two-tailed Mann-Whitney U test (C, D). * = p < 0.05, ** = p < 0.01, *** = p < 0.001.

150
Supplementary figure 4: cDC2 homeostasis is not mediated by NK cells.
(A) Representative contour plots of splenic cDCs from Rag1-/- and Rag1-/- Il15-/- mice, with quantification
of the cDC subsets in (B) as percentages of live cells (left) and of all cDCs (right). (C) Representative flow
plots of splenic cDCs from NKp46cre x Rosa26-LSL-DTA mice and littermate controls. cDC subsets are
quantified in (D) as absolute cell counts and as a percentage of all cDCs. (E) Bar chart depicting NK cell
numbers in the spleen of NKp46-DTA mice. (F) Gating strategy for the identification of NK cells and ILCs
in the spleen of Rag2-/- mice. NK cells are defined as single, live, lineage (CD3e/CD19/Ter119)- NK1.1+
NKp46+ cells. Within the non-NK cell gate, ILC1s are identified as CD90.2hi RORγt- Gata3- Tbet+, ILC2s as
CD90.2hi RORγt- Gata3+ Tbet- and ILC3s as CD90.2hi RORγt+. (A, B) Representative of 3-4 mice per group.
(C-E) Representative of 2 independent experiments (7-8 mice per group). The numbers adjacent to the
gates (A and C) represent percentages of the total cDC population. Bar graphs indicate mean ± SD. Two-
tailed Mann-Whitney U test (B, D, E). * = p < 0.05, ** = p < 0.01.

151
Supplementary figure 5: Sorting strategy for ILC isolation.
(A) Fluorescence-activated cell sorting strategy for ILC isolation from the intestinal lamina propria of
Rag2-/- mice pre-treated with rhIL-7:IL7 receptor complex as described in materials and methods. Gating
is similar for splenic and mesenteric lymph node ILCs (not shown). ILCs are purified as single, live, CD45+,
Ly6G-, autofluorescence-, CD11c-, lineage (CD3e/CD19/NK1.1/Ter119)-, CD90.2hi CD127+. (B) Purity
check of ILCs sorted as in (B).

152
REFERENCES

Abe, K., F.O. Yarovinsky, T. Murakami, A.N. Shakhov, A.V. Tumanov, D. Ito, L.N. Drutskaya, K. Pfeffer, D.V. Kuprash, K.L.
Komschlies, and S.A. Nedospasov. 2003. Distinct contributions of TNF and LT cytokines to the development
of dendritic cells in vitro and their recruitment in vivo. Blood 101:1477-1483.
Aliberti, J., O. Schulz, D.J. Pennington, H. Tsujimura, C. Reis e Sousa, K. Ozato, and A. Sher. 2003. Essential role for
ICSBP in the in vivo development of murine CD8alpha + dendritic cells. Blood 101:305-310.
Bajana, S., S. Turner, J. Paul, E. Ainsua-Enrich, and S. Kovats. 2016. IRF4 and IRF8 Act in CD11c+ Cells To Regulate
Terminal Differentiation of Lung Tissue Dendritic Cells. J Immunol 196:1666-1677.
Bando, J.K., and M. Colonna. 2016. Innate lymphoid cell function in the context of adaptive immunity. Nat Immunol
17:783-789.
Banks, T.A., S. Rickert, C.A. Benedict, L. Ma, M. Ko, J. Meier, W. Ha, K. Schneider, S.W. Granger, O. Turovskaya, D.
Elewaut, D. Otero, A.R. French, S.C. Henry, J.D. Hamilton, S. Scheu, K. Pfeffer, and C.F. Ware. 2005. A
lymphotoxin-IFN-beta axis essential for lymphocyte survival revealed during cytomegalovirus infection. J
Immunol 174:7217-7225.
Bjorklund, A.K., M. Forkel, S. Picelli, V. Konya, J. Theorell, D. Friberg, R. Sandberg, and J. Mjosberg. 2016. The
heterogeneity of human CD127(+) innate lymphoid cells revealed by single-cell RNA sequencing. Nat
Immunol 17:451-460.
Briseno, C.G., M. Gargaro, V. Durai, J.T. Davidson, D.J. Theisen, D.A. Anderson, 3rd, D.V. Novack, T.L. Murphy, and K.M.
Murphy. 2017. Deficiency of transcription factor RelB perturbs myeloid and DC development by
hematopoietic-extrinsic mechanisms. Proc Natl Acad Sci U S A 114:3957-3962.
Briseno, C.G., A.T. Satpathy, J.T.t. Davidson, S.T. Ferris, V. Durai, P. Bagadia, K.W. O'Connor, D.J. Theisen, T.L. Murphy,
and K.M. Murphy. 2018. Notch2-dependent DC2s mediate splenic germinal center responses. Proc Natl Acad
Sci U S A 115:10726-10731.
Browning, J.L., I.D. Sizing, P. Lawton, P.R. Bourdon, P.D. Rennert, G.R. Majeau, C.M. Ambrose, C. Hession, K. Miatkowski,
D.A. Griffiths, A. Ngam-ek, W. Meier, C.D. Benjamin, and P.S. Hochman. 1997. Characterization of
lymphotoxin-alpha beta complexes on the surface of mouse lymphocytes. J Immunol 159:3288-3298.
Caton, M.L., M.R. Smith-Raska, and B. Reizis. 2007. Notch-RBP-J signaling controls the homeostasis of CD8- dendritic
cells in the spleen. J Exp Med 204:1653-1664.
Cella, M., A. Fuchs, W. Vermi, F. Facchetti, K. Otero, J.K. Lennerz, J.M. Doherty, J.C. Mills, and M. Colonna. 2009. A
human natural killer cell subset provides an innate source of IL-22 for mucosal immunity. Nature 457:722-
725.
Chu, C., S. Moriyama, Z. Li, L. Zhou, A.L. Flamar, C.S.N. Klose, J.B. Moeller, G.G. Putzel, D.R. Withers, G.F. Sonnenberg,
and D. Artis. 2018. Anti-microbial Functions of Group 3 Innate Lymphoid Cells in Gut-Associated Lymphoid
Tissues Are Regulated by G-Protein-Coupled Receptor 183. Cell Rep 23:3750-3758.
Cortez, V.S., and M. Colonna. 2016. Diversity and function of group 1 innate lymphoid cells. Immunol Lett 179:19-24.
Crowley, M.T., C.R. Reilly, and D. Lo. 1999. Influence of lymphocytes on the presence and organization of dendritic cell
subsets in the spleen. J Immunol 163:4894-4900.
De Trez, C., K. Schneider, K. Potter, N. Droin, J. Fulton, P.S. Norris, S.W. Ha, Y.X. Fu, T. Murphy, K.M. Murphy, K. Pfeffer,
C.A. Benedict, and C.F. Ware. 2008. The inhibitory HVEM-BTLA pathway counter regulates lymphotoxin
receptor signaling to achieve homeostasis of dendritic cells. J Immunol 180:238-248.
Dicken, J., A. Mildner, D. Leshkowitz, I.P. Touw, S. Hantisteanu, S. Jung, and Y. Groner. 2013. Transcriptional
reprogramming of CD11b+Esam(hi) dendritic cell identity and function by loss of Runx3. PLoS One
8:e77490.
Dubey, L.K., P. Karempudi, S.A. Luther, B. Ludewig, and N.L. Harris. 2017. Interactions between fibroblastic reticular
cells and B cells promote mesenteric lymph node lymphangiogenesis. Nat Commun 8:367.
Emgard, J., H. Kammoun, B. Garcia-Cassani, J. Chesne, S.M. Parigi, J.M. Jacob, H.W. Cheng, E. Evren, S. Das, P.
Czarnewski, N. Sleiers, F. Melo-Gonzalez, E. Kvedaraite, M. Svensson, E. Scandella, M.R. Hepworth, S. Huber,
B. Ludewig, L. Peduto, E.J. Villablanca, H. Veiga-Fernandes, J.P. Pereira, R.A. Flavell, and T. Willinger. 2018.
Oxysterol Sensing through the Receptor GPR183 Promotes the Lymphoid-Tissue-Inducing Function of
Innate Lymphoid Cells and Colonic Inflammation. Immunity 48:120-132 e128.
Fasnacht, N., H.Y. Huang, U. Koch, S. Favre, F. Auderset, Q. Chai, L. Onder, S. Kallert, D.D. Pinschewer, H.R. MacDonald,
F. Tacchini-Cottier, B. Ludewig, S.A. Luther, and F. Radtke. 2014. Specific fibroblastic niches in secondary
lymphoid organs orchestrate distinct Notch-regulated immune responses. Journal of Experimental Medicine
211:2265-2279.

153
Fu, Y.X., G. Huang, Y. Wang, and D.D. Chaplin. 1998. B lymphocytes induce the formation of follicular dendritic cell
clusters in a lymphotoxin alpha-dependent fashion. J Exp Med 187:1009-1018.
Gatto, D., K. Wood, I. Caminschi, D. Murphy-Durland, P. Schofield, D. Christ, G. Karupiah, and R. Brink. 2013. The
chemotactic receptor EBI2 regulates the homeostasis, localization and immunological function of splenic
dendritic cells. Nat Immunol 14:446-453.
Ghosh, S., S.A. Hoselton, and J.M. Schuh. 2012. mu-chain-deficient mice possess B-1 cells and produce IgG and IgE, but
not IgA, following systemic sensitization and inhalational challenge in a fungal asthma model. J Immunol
189:1322-1329.
Gonzalez, M., F. Mackay, J.L. Browning, M.H. Kosco-Vilbois, and R.J. Noelle. 1998. The sequential role of lymphotoxin
and B cells in the development of splenic follicles. J Exp Med 187:997-1007.
Grajales-Reyes, G.E., A. Iwata, J. Albring, X. Wu, R. Tussiwand, W. Kc, N.M. Kretzer, C.G. Briseno, V. Durai, P. Bagadia, M.
Haldar, J. Schonheit, F. Rosenbauer, T.L. Murphy, and K.M. Murphy. 2015. Batf3 maintains autoactivation of
Irf8 for commitment of a CD8alpha(+) conventional DC clonogenic progenitor. Nat Immunol 16:708-717.
Guilliams, M., C.A. Dutertre, C.L. Scott, N. McGovern, D. Sichien, S. Chakarov, S. Van Gassen, J. Chen, M. Poidinger, S. De
Prijck, S.J. Tavernier, I. Low, S.E. Irac, C.N. Mattar, H.R. Sumatoh, G.H.L. Low, T.J.K. Chung, D.K.H. Chan, K.K.
Tan, T.L.K. Hon, E. Fossum, B. Bogen, M. Choolani, J.K.Y. Chan, A. Larbi, H. Luche, S. Henri, Y. Saeys, E.W.
Newell, B.N. Lambrecht, B. Malissen, and F. Ginhoux. 2016. Unsupervised High-Dimensional Analysis Aligns
Dendritic Cells across Tissues and Species. Immunity 45:669-684.
Gury-BenAri, M., C.A. Thaiss, N. Serafini, D.R. Winter, A. Giladi, D. Lara-Astiaso, M. Levy, T.M. Salame, A. Weiner, E.
David, H. Shapiro, M. Dori-Bachash, M. Pevsner-Fischer, E. Lorenzo-Vivas, H. Keren-Shaul, F. Paul, A.
Harmelin, G. Eberl, S. Itzkovitz, A. Tanay, J.P. Di Santo, E. Elinav, and I. Amit. 2016. The Spectrum and
Regulatory Landscape of Intestinal Innate Lymphoid Cells Are Shaped by the Microbiome. Cell 166:1231-
1246 e1213.
Hasan, M., B. Polic, M. Bralic, S. Jonjic, and K. Rajewsky. 2002. Incomplete block of B cell development and
immunoglobulin production in mice carrying the muMT mutation on the BALB/c background. Eur J
Immunol 32:3463-3471.
Hildner, K., B.T. Edelson, W.E. Purtha, M. Diamond, H. Matsushita, M. Kohyama, B. Calderon, B.U. Schraml, E.R. Unanue,
M.S. Diamond, R.D. Schreiber, T.L. Murphy, and K.M. Murphy. 2008. Batf3 deficiency reveals a critical role for
CD8alpha+ dendritic cells in cytotoxic T cell immunity. Science 322:1097-1100.
Hobeika, E., S. Thiemann, B. Storch, H. Jumaa, P.J. Nielsen, R. Pelanda, and M. Reth. 2006. Testing gene function early in
the B cell lineage in mb1-cre mice. Proc Natl Acad Sci U S A 103:13789-13794.
Hoorweg, K., P. Narang, Z. Li, A. Thuery, N. Papazian, D.R. Withers, M.C. Coles, and T. Cupedo. 2015. A Stromal Cell
Niche for Human and Mouse Type 3 Innate Lymphoid Cells. J Immunol 195:4257-4263.
Ichikawa, E., S. Hida, Y. Omatsu, S. Shimoyama, K. Takahara, S. Miyagawa, K. Inaba, and S. Taki. 2004. Defective
development of splenic and epidermal CD4+ dendritic cells in mice deficient for IFN regulatory factor-2.
Proc Natl Acad Sci U S A 101:3909-3914.
Jackson, J.T., Y. Hu, R. Liu, F. Masson, A. D'Amico, S. Carotta, A. Xin, M.J. Camilleri, A.M. Mount, A. Kallies, L. Wu, G.K.
Smyth, S.L. Nutt, and G.T. Belz. 2011. Id2 expression delineates differential checkpoints in the genetic
program of CD8alpha+ and CD103+ dendritic cell lineages. EMBO J 30:2690-2704.
Kabashima, K., T.A. Banks, K.M. Ansel, T.T. Lu, C.F. Ware, and J.G. Cyster. 2005. Intrinsic lymphotoxin-beta receptor
requirement for homeostasis of lymphoid tissue dendritic cells. Immunity 22:439-450.
Kashiwada, M., N.L. Pham, L.L. Pewe, J.T. Harty, and P.B. Rothman. 2011. NFIL3/E4BP4 is a key transcription factor
for CD8alpha(+) dendritic cell development. Blood 117:6193-6197.
Kennedy, M.K., M. Glaccum, S.N. Brown, E.A. Butz, J.L. Viney, M. Embers, N. Matsuki, K. Charrier, L. Sedger, C.R. Willis,
K. Brasel, P.J. Morrissey, K. Stocking, J.C. Schuh, S. Joyce, and J.J. Peschon. 2000. Reversible defects in natural
killer and memory CD8 T cell lineages in interleukin 15-deficient mice. J Exp Med 191:771-780.
Kim, M.Y., F.M. McConnell, F.M. Gaspal, A. White, S.H. Glanville, V. Bekiaris, L.S. Walker, J. Caamano, E. Jenkinson, G.
Anderson, and P.J. Lane. 2007. Function of CD4+CD3- cells in relation to B- and T-zone stroma in spleen.
Blood 109:1602-1610.
Kim, T.J., V. Upadhyay, V. Kumar, K.M. Lee, and Y.X. Fu. 2014. Innate lymphoid cells facilitate NK cell development
through a lymphotoxin-mediated stromal microenvironment. J Exp Med 211:1421-1431.
Kirkling, M.E., U. Cytlak, C.M. Lau, K.L. Lewis, A. Resteu, A. Khodadadi-Jamayran, C.W. Siebel, H. Salmon, M. Merad, A.
Tsirigos, M. Collin, V. Bigley, and B. Reizis. 2018. Notch Signaling Facilitates In Vitro Generation of Cross-
Presenting Classical Dendritic Cells. Cell Rep 23:3658-3672 e3656.
Klose, C.S., and D. Artis. 2016. Innate lymphoid cells as regulators of immunity, inflammation and tissue homeostasis.
Nat Immunol 17:765-774.

154
Klose, C.S., E.A. Kiss, V. Schwierzeck, K. Ebert, T. Hoyler, Y. d'Hargues, N. Goppert, A.L. Croxford, A. Waisman, Y.
Tanriver, and A. Diefenbach. 2013. A T-bet gradient controls the fate and function of CCR6-RORgammat+
innate lymphoid cells. Nature 494:261-265.
Kruglov, A.A., S.I. Grivennikov, D.V. Kuprash, C. Winsauer, S. Prepens, G.M. Seleznik, G. Eberl, D.R. Littman, M.
Heikenwalder, A.V. Tumanov, and S.A. Nedospasov. 2013. Nonredundant function of soluble LTalpha3
produced by innate lymphoid cells in intestinal homeostasis. Science 342:1243-1246.
Kucharzik, T., J.T. Hudson, 3rd, R.L. Waikel, W.D. Martin, and I.R. Williams. 2002. CCR6 expression distinguishes
mouse myeloid and lymphoid dendritic cell subsets: demonstration using a CCR6 EGFP knock-in mouse. Eur
J Immunol 32:104-112.
Kumar, V., D.C. Dasoveanu, S. Chyou, T.C. Tzeng, C. Rozo, Y. Liang, W. Stohl, Y.X. Fu, N.H. Ruddle, and T.T. Lu. 2015. A
dendritic-cell-stromal axis maintains immune responses in lymph nodes. Immunity 42:719-730.
Lewis, K.L., M.L. Caton, M. Bogunovic, M. Greter, L.T. Grajkowska, D. Ng, A. Klinakis, I.F. Charo, S. Jung, J.L.
Gommerman, Ivanov, II, K. Liu, M. Merad, and B. Reizis. 2011. Notch2 receptor signaling controls functional
differentiation of dendritic cells in the spleen and intestine. Immunity 35:780-791.
Liu, K., G.D. Victora, T.A. Schwickert, P. Guermonprez, M.M. Meredith, K. Yao, F.F. Chu, G.J. Randolph, A.Y. Rudensky,
and M. Nussenzweig. 2009. In vivo analysis of dendritic cell development and homeostasis. Science 324:392-
397.
Luci, C., A. Reynders, Ivanov, II, C. Cognet, L. Chiche, L. Chasson, J. Hardwigsen, E. Anguiano, J. Banchereau, D.
Chaussabel, M. Dalod, D.R. Littman, E. Vivier, and E. Tomasello. 2009. Influence of the transcription factor
RORgammat on the development of NKp46+ cell populations in gut and skin. Nat Immunol 10:75-82.
Luda, K.M., T. Joeris, E.K. Persson, A. Rivollier, M. Demiri, K.M. Sitnik, L. Pool, J.B. Holm, F. Melo-Gonzalez, L. Richter,
B.N. Lambrecht, K. Kristiansen, M.A. Travis, M. Svensson-Frej, K. Kotarsky, and W.W. Agace. 2016. IRF8
Transcription-Factor-Dependent Classical Dendritic Cells Are Essential for Intestinal T Cell Homeostasis.
Immunity 44:860-874.
Macpherson, A.J., A. Lamarre, K. McCoy, G.R. Harriman, B. Odermatt, G. Dougan, H. Hengartner, and R.M. Zinkernagel.
2001. IgA production without mu or delta chain expression in developing B cells. Nat Immunol 2:625-631.
Magri, G., M. Miyajima, S. Bascones, A. Mortha, I. Puga, L. Cassis, C.M. Barra, L. Comerma, A. Chudnovskiy, M. Gentile, D.
Llige, M. Cols, S. Serrano, J.I. Arostegui, M. Juan, J. Yague, M. Merad, S. Fagarasan, and A. Cerutti. 2014. Innate
lymphoid cells integrate stromal and immunological signals to enhance antibody production by splenic
marginal zone B cells. Nat Immunol 15:354-364.
Mauri, D.N., R. Ebner, R.I. Montgomery, K.D. Kochel, T.C. Cheung, G.L. Yu, S. Ruben, M. Murphy, R.J. Eisenberg, G.H.
Cohen, P.G. Spear, and C.F. Ware. 1998. LIGHT, a new member of the TNF superfamily, and lymphotoxin
alpha are ligands for herpesvirus entry mediator. Immunity 8:21-30.
Melo-Gonzalez, F., and M.R. Hepworth. 2017. Functional and phenotypic heterogeneity of group 3 innate lymphoid
cells. Immunology 150:265-275.
Mildner, A., and S. Jung. 2014. Development and function of dendritic cell subsets. Immunity 40:642-656.
Moseman, E.A., M. Iannacone, L. Bosurgi, E. Tonti, N. Chevrier, A. Tumanov, Y.X. Fu, N. Hacohen, and U.H. von Andrian.
2012. B cell maintenance of subcapsular sinus macrophages protects against a fatal viral infection
independent of adaptive immunity. Immunity 36:415-426.
Naik, S.H., P. Sathe, H.Y. Park, D. Metcalf, A.I. Proietto, A. Dakic, S. Carotta, M. O'Keeffe, M. Bahlo, A. Papenfuss, J.Y.
Kwak, L. Wu, and K. Shortman. 2007. Development of plasmacytoid and conventional dendritic cell subtypes
from single precursor cells derived in vitro and in vivo. Nat Immunol 8:1217-1226.
Narni-Mancinelli, E., J. Chaix, A. Fenis, Y.M. Kerdiles, N. Yessaad, A. Reynders, C. Gregoire, H. Luche, S. Ugolini, E.
Tomasello, T. Walzer, and E. Vivier. 2011. Fate mapping analysis of lymphoid cells expressing the NKp46 cell
surface receptor. Proc Natl Acad Sci U S A 108:18324-18329.
Persson, E.K., H. Uronen-Hansson, M. Semmrich, A. Rivollier, K. Hagerbrand, J. Marsal, S. Gudjonsson, U. Hakansson, B.
Reizis, K. Kotarsky, and W.W. Agace. 2013. IRF4 transcription-factor-dependent CD103(+)CD11b(+)
dendritic cells drive mucosal T helper 17 cell differentiation. Immunity 38:958-969.
Rankin, L.C., M.J. Girard-Madoux, C. Seillet, L.A. Mielke, Y. Kerdiles, A. Fenis, E. Wieduwild, T. Putoczki, S. Mondot, O.
Lantz, D. Demon, A.T. Papenfuss, G.K. Smyth, M. Lamkanfi, S. Carotta, J.C. Renauld, W. Shi, S. Carpentier, T.
Soos, C. Arendt, S. Ugolini, N.D. Huntington, G.T. Belz, and E. Vivier. 2016. Complementarity and redundancy
of IL-22-producing innate lymphoid cells. Nat Immunol 17:179-186.
Reboldi, A., T.I. Arnon, L.B. Rodda, A. Atakilit, D. Sheppard, and J.G. Cyster. 2016. IgA production requires B cell
interaction with subepithelial dendritic cells in Peyer's patches. Science 352:aaf4822.
Rodda, L.B., E. Lu, M.L. Bennett, C.L. Sokol, X. Wang, S.A. Luther, B.A. Barres, A.D. Luster, C.J. Ye, and J.G. Cyster. 2018.
Single-Cell RNA Sequencing of Lymph Node Stromal Cells Reveals Niche-Associated Heterogeneity.
Immunity 48:1014-1028 e1016.

155
Rumbo, M., F. Sierro, N. Debard, J.P. Kraehenbuhl, and D. Finke. 2004. Lymphotoxin beta receptor signaling induces
the chemokine CCL20 in intestinal epithelium. Gastroenterology 127:213-223.
Sanos, S.L., V.L. Bui, A. Mortha, K. Oberle, C. Heners, C. Johner, and A. Diefenbach. 2009. RORgammat and commensal
microflora are required for the differentiation of mucosal interleukin 22-producing NKp46+ cells. Nat
Immunol 10:83-91.
Satpathy, A.T., C.G. Briseno, J.S. Lee, D. Ng, N.A. Manieri, W. Kc, X. Wu, S.R. Thomas, W.L. Lee, M. Turkoz, K.G. McDonald,
M.M. Meredith, C. Song, C.J. Guidos, R.D. Newberry, W. Ouyang, T.L. Murphy, T.S. Stappenbeck, J.L.
Gommerman, M.C. Nussenzweig, M. Colonna, R. Kopan, and K.M. Murphy. 2013. Notch2-dependent classical
dendritic cells orchestrate intestinal immunity to attaching-and-effacing bacterial pathogens. Nat Immunol
14:937-948.
Schiavoni, G., F. Mattei, P. Sestili, P. Borghi, M. Venditti, H.C. Morse, 3rd, F. Belardelli, and L. Gabriele. 2002. ICSBP is
essential for the development of mouse type I interferon-producing cells and for the generation and
activation of CD8alpha(+) dendritic cells. J Exp Med 196:1415-1425.
Schlitzer, A., N. McGovern, P. Teo, T. Zelante, K. Atarashi, D. Low, A.W. Ho, P. See, A. Shin, P.S. Wasan, G. Hoeffel, B.
Malleret, A. Heiseke, S. Chew, L. Jardine, H.A. Purvis, C.M. Hilkens, J. Tam, M. Poidinger, E.R. Stanley, A.B.
Krug, L. Renia, B. Sivasankar, L.G. Ng, M. Collin, P. Ricciardi-Castagnoli, K. Honda, M. Haniffa, and F. Ginhoux.
2013. IRF4 transcription factor-dependent CD11b+ dendritic cells in human and mouse control mucosal IL-
17 cytokine responses. Immunity 38:970-983.
Schlitzer, A., V. Sivakamasundari, J. Chen, H.R. Sumatoh, J. Schreuder, J. Lum, B. Malleret, S. Zhang, A. Larbi, F. Zolezzi,
L. Renia, M. Poidinger, S. Naik, E.W. Newell, P. Robson, and F. Ginhoux. 2015. Identification of cDC1- and
cDC2-committed DC progenitors reveals early lineage priming at the common DC progenitor stage in the
bone marrow. Nat Immunol 16:718-728.
Scott, C.L., B. Soen, L. Martens, N. Skrypek, W. Saelens, J. Taminau, G. Blancke, G. Van Isterdael, D. Huylebroeck, J.
Haigh, Y. Saeys, M. Guilliams, B.N. Lambrecht, and G. Berx. 2016. The transcription factor Zeb2 regulates
development of conventional and plasmacytoid DCs by repressing Id2. J Exp Med 213:897-911.
Sun, J.C., A. Ma, and L.L. Lanier. 2009. Cutting edge: IL-15-independent NK cell response to mouse cytomegalovirus
infection. J Immunol 183:2911-2914.
Tumanov, A.V., E.P. Koroleva, X. Guo, Y. Wang, A. Kruglov, S. Nedospasov, and Y.X. Fu. 2011. Lymphotoxin controls the
IL-22 protection pathway in gut innate lymphoid cells during mucosal pathogen challenge. Cell Host Microbe
10:44-53.
Tussiwand, R., B. Everts, G.E. Grajales-Reyes, N.M. Kretzer, A. Iwata, J. Bagaitkar, X. Wu, R. Wong, D.A. Anderson, T.L.
Murphy, E.J. Pearce, and K.M. Murphy. 2015. Klf4 expression in conventional dendritic cells is required for T
helper 2 cell responses. Immunity 42:916-928.
van de Pavert, S.A., and R.E. Mebius. 2010. New insights into the development of lymphoid tissues. Nat Rev Immunol
10:664-674.
Vely, F., V. Barlogis, B. Vallentin, B. Neven, C. Piperoglou, M. Ebbo, T. Perchet, M. Petit, N. Yessaad, F. Touzot, J.
Bruneau, N. Mahlaoui, N. Zucchini, C. Farnarier, G. Michel, D. Moshous, S. Blanche, A. Dujardin, H. Spits, J.H.
Distler, A. Ramming, C. Picard, R. Golub, A. Fischer, and E. Vivier. 2016. Evidence of innate lymphoid cell
redundancy in humans. Nat Immunol 17:1291-1299.
Vivier, E., D. Artis, M. Colonna, A. Diefenbach, J.P. Di Santo, G. Eberl, S. Koyasu, R.M. Locksley, A.N.J. McKenzie, R.E.
Mebius, F. Powrie, and H. Spits. 2018. Innate Lymphoid Cells: 10 Years On. Cell 174:1054-1066.
Vonarbourg, C., A. Mortha, V.L. Bui, P.P. Hernandez, E.A. Kiss, T. Hoyler, M. Flach, B. Bengsch, R. Thimme, C. Holscher,
M. Honig, U. Pannicke, K. Schwarz, C.F. Ware, D. Finke, and A. Diefenbach. 2010. Regulated expression of
nuclear receptor RORgammat confers distinct functional fates to NK cell receptor-expressing
RORgammat(+) innate lymphocytes. Immunity 33:736-751.
Wang, Y.G., K.D. Kim, J. Wang, P. Yu, and Y.X. Fu. 2005. Stimulating lymphotoxin beta receptor on the dendritic cells is
critical for their homeostasis and expansion. J Immunol 175:6997-7002.
Ware, C.F., P.D. Crowe, M.H. Grayson, M.J. Androlewicz, and J.L. Browning. 1992. Expression of surface lymphotoxin
and tumor necrosis factor on activated T, B, and natural killer cells. J Immunol 149:3881-3888.
Wu, L., A. D'Amico, K.D. Winkel, M. Suter, D. Lo, and K. Shortman. 1998. RelB is essential for the development of
myeloid-related CD8alpha- dendritic cells but not of lymphoid-related CD8alpha+ dendritic cells. Immunity
9:839-847.
Wu, Q., Y. Wang, J. Wang, E.O. Hedgeman, J.L. Browning, and Y.X. Fu. 1999. The requirement of membrane
lymphotoxin for the presence of dendritic cells in lymphoid tissues. J Exp Med 190:629-638.
Wu, X., C.G. Briseno, G.E. Grajales-Reyes, M. Haldar, A. Iwata, N.M. Kretzer, W. Kc, R. Tussiwand, Y. Higashi, T.L.
Murphy, and K.M. Murphy. 2016. Transcription factor Zeb2 regulates commitment to plasmacytoid
dendritic cell and monocyte fate. Proc Natl Acad Sci U S A 113:14775-14780.

156
Yi, T., and J.G. Cyster. 2013. EBI2-mediated bridging channel positioning supports splenic dendritic cell homeostasis
and particulate antigen capture. Elife 2:e00757.
Yoshida, H., A. Naito, J. Inoue, M. Satoh, S.M. Santee-Cooper, C.F. Ware, A. Togawa, S. Nishikawa, and S. Nishikawa.
2002. Different cytokines induce surface lymphotoxin-alphabeta on IL-7 receptor-alpha cells that
differentially engender lymph nodes and Peyer's patches. Immunity 17:823-833.
Zhang, Y., T.J. Kim, J.A. Wroblewska, V. Tesic, V. Upadhyay, R.R. Weichselbaum, A.V. Tumanov, H. Tang, X. Guo, H. Tang,
and Y.X. Fu. 2018. Type 3 innate lymphoid cell-derived lymphotoxin prevents microbiota-dependent
inflammation. Cell Mol Immunol 15:697-709.
Zindl, C.L., T.H. Kim, M. Zeng, A.S. Archambault, M.H. Grayson, K. Choi, R.D. Schreiber, and D.D. Chaplin. 2009. The
lymphotoxin LTalpha(1)beta(2) controls postnatal and adult spleen marginal sinus vascular structure and
function. Immunity 30:408-420.
Zook, E.C., and B.L. Kee. 2016. Development of innate lymphoid cells. Nat Immunol 17:775-782.

157
158
 VI. Results part four

IRF8-dependent dendritic cells are required to

control myeloproliferation

Matthias Vanderkerken, Cedric Bosteels, Dorine Sichien, Kim Deswarte, Manon

Vanheerswynghels, Justine Van Moorleghem, Farzaneh Fayazpour, Martin Guilliams,
Hamida Hammad, Tessa Kerre* and Bart N. Lambrecht*

* These authors contributed equally to this work

Manuscript in preparation

Author contribution:
Matthias Vanderkerken performed and analyzed all experiments and wrote the
manuscript.

159
ABSTRACT

Absence of dendritic cells (DCs) is linked with the onset of a myeloproliferative disorder
(MPD) in humans and mice. It has been suggested that the MPD observed in Irf8-/- mice
develops in response to the deficiency of type 1 DCs (cDC1s), although the underlying
mechanism remains enigmatic. Here, we found that the development of MPD in Irf8-/-
mice requires two hits, whereby loss of Irf8 in DCs disrupts an extrinsic
myelosuppressive effect that restricts neutrophil production by Irf8-deficient myeloid
progenitors. Using a mixed bone marrow approach and different conditional Cre-deleter
strains, we demonstrate that the observed effects are not due to off-target activity of
CD11c-cre in B-, T- or NK cells. In addition, we exclude a role for plasmacytoid DCs and
cDC1s and show that mere absence of cDC1s is not sufficient to trigger MPD. Because
IRF8 down-regulation is a pivotal step in human BCR-ABL+ chronic myeloid leukemia
(CML), further dissection of the precise cellular interactions by which the immune
system surveils myelopoiesis might open new therapeutic avenues for patients resistant
to current treatment regimens.

INTRODUCTION

Dendritic cells (DCs) are specialized in the uptake, processing and presentation of
foreign and self-antigens to adaptive lymphocytes. Through a variety of pattern
recognition receptors, DCs sense pathogenic molecules and discriminate those from
endogenous antigens. Once thought of as a unique cell expressing high levels of CD11c
and MHCII, it is now clear that several subsets of CD11c+ DCs exist that differ in
development, transcription factor dependency, phenotype, migratory behavior and
function. Apart from the generation of protective responses against foreign antigens,
conventional dendritic cells (cDCs) are crucial mediators of T cell tolerance to
innocuous self-antigens1-3. In recent years, accumulating evidence has also revealed an
unexpected role for dendritic cells in the regulation of myelopoiesis. Constitutive
ablation of dendritic cells in mice was associated with the onset of a myeloproliferative
disorder (MPD) characterized by splenomegaly and accumulation of neutrophils4, 5.
Similarly, depletion of dendritic cells using CD11c- or Zbtb46-driven expression of the
diphtheria toxin receptor (DTR) triggered neutrophilia6-9. In line, deletions of fms like
tyrosine kinase 3 (FLT3), basic leucine zipper ATF-like transcription factor 3 (BATF3),
core-binding factor subunit beta (CBTB) or transforming growth factor β -activated
kinase 1 (TAK1) in mice were all associated with deficiency of DCs and concomitant
increases in granulocyte numbers8, 10, 11. In a recent report, Liston et al. reported a
comparable myeloproliferative disorder in mice that lacked MHCII expression in DCs,
illustrating that not only numerical, but also functional alterations in DCs may be at the
basis of disturbed myelopoiesis12.

160
A combined DC deficiency and MPD have also been observed in mice that lack the
transcription factor interferon regulatory factor 8 (IRF8)13, 14. IRF8, formerly known as
interferon consensus binding protein (ICSBP), is member of a family of nine
transcription factors induced by interferon. Often in cooperation with other
transcription factors such as PU.1, SPI-B or BATF proteins, IRF8 binds regulatory DNA
sequences and can act both as an activator and repressor of transcription. The role of
IRF8 in hematopoiesis has been extensively investigated and important functions were
identified in both lymphoid and myeloid branches15-21. During myelopoiesis, IRF8
promotes development of monocytes and dendritic cells while repressing the
neutrophil fate22, 23. Accordingly, Irf8-/- mice lack normal numbers of dendritic cells and
monocytes but develop splenomegaly and extramedullary hematopoiesis driven by
granulocyte expansion, resembling human myeloproliferative disorders like chronic
myeloid leukemia (CML)13, 14. Remarkably, immunodeficient patients that carry IRF8
mutations also manifest an MPD-like phenotype24. The mechanism underlying the
connection between DC deficiency and increased granulopoiesis remains enigmatic. It
has been proposed that an empty dendritic cell niche might trigger elevated levels of
growth factors such as Flt3L, as a compensatory effort to differentiate myeloid
progenitors into DCs5. Indeed, the homeostasis of dendritic cells is controlled by effects
of Flt3L on hematopoietic stem cells via a feedback loop, while overactive Flt3-signaling
is also associated with myeloid leukemias25-27.

Within the cDC population, loss of Irf8 precludes the development and terminal
differentiation of one specific subset of dendritic cells specifically28-30. This subset,
named cDC1 by convention, can be identified by expression of XCR1 and lack of CD172a
across species and tissues31. It is crucial for the generation of protective T cell responses
against intracellular pathogens and tumors, as well as for the maintenance of peripheral
tolerance. In contrast, XCR1- CD172a+ cDC2s, which are implicated mainly in CD4+ T cell
priming to exogenous antigens, still develop in absence of Irf81-3, 32. Because of the
selective absence of cDC1s in Irf8-/- mice, it was proposed that cDC1s are the relevant
subtype that is required to keep normal myelopoiesis in check5. However, direct
interrogation of this hypothesis has long been impeded by the pleiotropic effects of
IRF8 in distinct cell types. Now, with the advent of novel genetic tools33-35 to study the
cDC1 subset we decided to further delineate the mechanism that drives MPD in Irf8-/-
mice. Using a mixed bone marrow chimera approach and different conditional Irf8
knockout mice, we demonstrate here that absence of cDC1s alone does not trigger
increased granulopoiesis. In contrast, expression of Irf8 in cDC2s or monocyte-derived
cells is necessary and sufficient to prevent MPD driven by Irf8-/- hematopoietic stem
cells.

161
RESULTS

Irf8-/- mice develop transplantable MPD that is prevented by transfer of wild type
BM cells
Others have reported the onset of a myeloproliferative disorder (MPD) in mice that lack
a functional IRF8 protein13, 14. To validate whether a similar phenotype was present in
the Irf8-/- mice generated in our lab, we analyzed the cellular composition of the bone
marrow and spleens from these mice. We confirmed splenomegaly and an increased
cellularity in the spleen and bone marrow (BM) of Irf8 knockout mice (figure 1A). This
was mainly caused by the expansion of Ly6Ghi neutrophils and of immature
granulocytes that are characterized by a Ly6Gint CD11b+ or Ly6G- Ly6C+ CD11b+
phenotype (figure 1B, 1C). The latter phenotypically and functionally resemble
myeloid-derived suppressor cells (MDSCs), in agreement with the finding that IRF8-
downregulation in granulocyte precursors is a crucial step in the development of MDSCs
under physiological conditions36. Confocal microscopy revealed a massive infiltration of
CD11b+ cells in the splenic red pulp of Irf8-/- mice as compared to the punctuate pattern
observed in littermate controls (figure 1D). Of note, monocytes and type 1 conventional
dendritic cells (cDC1s, data not shown) were absent in Irf8-/- mice, in line with previous
observations, reported by us and others21, 23, 28, 29. A hallmark of myeloproliferative
neoplasms is the occurrence of extramedullary hematopoiesis, whereby myeloid
progenitor cells are found outside of the bone marrow. Accordingly, we identified
increased numbers of lineage- cKit+ Sca1+ (LSK) hematopoietic stem cells (HSCs) and
granulocyte-macrophage progenitors (GMPs) in the BM and spleen of Irf8-/- mice
(figure 1E-G). These data demonstrate the presence of extramedullary hematopoiesis
in Irf8-/- mice and suggest that deranged myelopoiesis originates from effects at the
earliest stages of hematopoiesis, even before the GMP stage.

Jung et al. previously showed that this form of MPD could be transferred to other mice
by bone marrow transplantation, and that development of disease was prevented by co-
transferring wild type bone marrow. To confirm these findings in our mice and our
animal facility, we generated mixed bone marrow chimeras (figure 1H). First, we
confirmed that myeloproliferative disease was reliably reproduced upon
transplantation of Irf8-/- bone marrow to wild type acceptor mice (figure 1J).
Remarkably, co-transplantation of wild type (WT) BM together with Irf8-/- BM in a 1:1
ratio fully inhibited the development of MPD, with LSK and neutrophil counts in BM and
spleen comparable to acceptor mice that received WT BM only.

CD11c+ cells lacking IRF8 fail to control MPD of transplanted Irf8-/- progenitors
To assess whether this protective effect depended on wild type dendritic cells, we
generated mice that express the Diphtheria toxin A after CD11c-cre mediated
recombination of a floxed transcriptional stop cassette (CD11c-DTA mice). In these
mice, Itgax gene induction and CD11c expression is rapidly followed by apoptosis,
leading to absence of CD11c-positive dendritic cells. Strikingly, the dominant

162
Figure 1: Wild type hematopoietic cells prevent MPD
(A) Splenic weight in Irf8-/- mice and littermate controls. (B) Representative flow cytometry plots of
splenic classical monocytes, immature granulocytes and neutrophils in Irf8-/- and Irf8+/+ mice. Neutrophils
are gated as single, live, lineage (CD3e/CD19/NK1.1/Ter119)- Ly6G+ CD11b+ cells. Classical monocytes

163
are gated as live, lineage- Ly6G- Ly6Chi CD11b+ cells, immature granulocytes as live, lineage- Ly6G- Ly6Cint
CD11b+ cells. The number of neutrophils and Ly6Gint granulocytes in spleen is quantified in (C). (D)
Confocal image from the spleen of Irf8-/- and Irf8+/+ mice. B220 is depicted in green, DCIR2 in yellow,
CD11b in magenta. (E, F) Representative flow plots of LSK and GMP cells in bone marrow (E) and spleen
(F) of Irf8-/- mice. LSK cells are gated as single, live, lineage- CD11c- SSClo cKit+ Sca1+ cells, GMP cells as
single, live, lineage- CD11c- SSClo cKit+ Sca1- CD16/32+ CD34+ cells. (G) Quantification of the absolute
number of LSK and GMP cells in the spleen (top) and of the frequency of GMPs among lineage- cKit+ Sca1-
cells in BM and spleen (bottom) of Irf8-/- mice. (H) Schematic representation of mixed bone marrow
chimeras. Wild type acceptor mice were lethally irradiated and transplanted with WT BM only, Irf8-/- BM
only, or a 50:50% mixture of Irf8-/- BM with WT BM or CD11c-DTA BM. Acceptors were analyzed 12
weeks after BM transplantation. (I) Histogram of IRF8 expression in cDC1s, B cells and NK cells from BM
chimeric mice after reconstitution. Cells retrieved in mice transplanted with WT BM only are depicted in
black, cells from mice transplanted with Irf8-/- BM only in red, and cells from Irf8-/-:WT mixed BM
acceptors in blue. (J) Splenic weight, splenic LSK and neutrophil numbers in BM chimeras 12 weeks after
reconstitution. Data in (A, B) and (E-G) are representative of at least 3 independent experiments (n= 10
per group). Data in (D) are representative of 4 mice per group. Data in (I, J) are representative of 2
independent experiments (n= 9-11 mice per group). (B, E, F) Numbers adjacent to the gates indicate
percentage of the parent population. (A, C, G, J) Two-tailed Mann-Whitney U test. Bars indicate mean ± SD.
ns = non-significant, * = p<0.05, ** = p<0.01, *** = p<0.001, **** = p<0.0001.

suppressive effect of WT BM on development of MPD in Irf8-/- bone marrow grafts

required CD11c-expressing cells, as transplantation of a 1:1 mixture of Irf8-/- and
CD11c-DTA BM resulted in myeloproliferation. Moreover, the combination of Irf8-/- and
CD11c-DTA BM even led to an exaggerated phenotype (figure 1J). The correct
reconstitution of acceptor mice was validated using an IRF8 staining. As expected,
cDC1s did not develop from Irf8-/- BM or Irf8-/-:CD11c-DTA BM, but were generated by
Irf8-sufficient cells in Irf8-/-:WT BM chimeras. In contrast, other hematopoietic cell types
such as B- and natural killer (NK) cells from Irf8-/-:WT and Irf8-/-:CD11c-DTA BM
revealed a bimodal distribution of IRF8 expression, demonstrating contributions of
both components of the BM mixture to the reconstituted pool (figure 1I).

We reasoned that if the MPD driven by Irf8-/- BM can be prevented or contained by a WT

CD11c+ cell, this cell type must be affected in Irf8-/- mice. Thus, we hypothesized that the
bone marrow of Irf8flfl x CD11ccre mice would lack the ability to suppress MPD. We
validated that transplantation of a 1:1 mixture of Irf8-/- and WT BM did, but
cotransplantation of Irf8-/- and Irf8flfl x CD11ccre BM did not reconstitute the cDC1 pool
in wild type acceptor mice (figure 2A). As expected, the protective effect on the
accumulation of LSK cells, GMPs and neutrophils in the spleen was lost when the WT
BM compartment was replaced by Irf8flfl x CD11ccre BM (figure 2B, C). Together, these
data highlight the ability of BM-derived cells to control the development of MPD from
Irf8-deficient hematopoietic stem cells. Moreover, this effect depended on the
expression of Irf8 in CD11c+ cells.

Next, we wondered whether the absence of IRF8 expression in CD11c+ cells in itself was
sufficient to trigger myeloproliferative disease, in the absence of intrinsic germline
defects of IRF8 in all hematopoietic cells. In unmanipulated Irf8flfl x CD11ccre mice,

164
Figure 2: Suppression of granulopoiesis requires expression of Irf8 in CD11c+ cells
(A) cDC1 count in mixed BM chimeric mice after reconstitution with the indicated bone marrow. All
acceptor mice are WT. (B) Representative contour plots of GMPs, Ly6Cint granulocytes and neutrophils in

165
the spleen Irf8-/-:WT > WT and Irf8-/-:Irf8flfl CD11ccre+ > WT BM chimeric mice. Data are quantified as
absolute cell numbers in (C). (A-C) Data are representative of at least 3 independent experiments (n=12-
15 per group). Control groups are the same as in figure 1J. Two-tailed Mann-Whitney U test. (D)
Representative contour plot of GMPs, neutrophils and Ly6Cint granulocytes in the spleen of intact Irf8flfl
CD11ccre+ mice and Cre-negative littermates. (E) Confocal image of the spleens of Irf8flfl CD11ccre+ mice
and littermate controls. B220 is depicted in green, DCIR2 in yellow, CD11b in magenta and Ki67 in red.
(F) Quantification of the number of LSK cells, GMPs, Ly6Ghi neutrophils and Ly6Cint granulocytes in bone
marrow (BM, left) or spleen (middle) and of the frequency of GMPs among lineage- cKit+ Sca1- cells
(right). (D, F) Data are representative of two independent experiments (n=8 mice per group). Two-tailed
Mann-Whitney U test. (E) Data representative of 4 mice per group. (B, D) Numbers adjacent to the gates
indicate percentage of the parent population. Bars indicate mean ± SD. ns = non-significant, * = p<0.05, **
= p<0.01, *** = p<0.001, **** = p<0.0001.

lacking IRF8 only in CD11c expressing cells, the number of LSK and GMP cells in BM and
spleen were not significantly increased (figure 2D, 2F). Compared to littermate
controls, the spleen of Irf8flfl x CD11ccre mice contained an increased number of Ly6Cint
granulocytes, though the absolute numbers were far below the levels seen in Irf8-/-
mice. In line, the distribution pattern of CD11b+ cells in the red pulp of transgenic mice
was very similar to that of control mice, apart from a few small CD11b+ clusters (figure
2E). Overall, these findings imply that loss of Irf8 expression in CD11c+ cells is not
sufficient to trigger the development of a full-blown MPD, but that development of MPD
in Irf8-/- mice requires a double hit: 1. loss of IRF8 in hematopoietic progenitor cells and
2. lack of control of the IRF8-deficient progenitors by loss of IRF8 in a CD11c+
population.

CD11c+ T cells and NK cells lacking IRF8 still control MPD of transplanted Irf8-/-
progenitors
Based on experiments with CD11c-DTA mice, the inhibitory effects of WT BM on the
development of MPD have been attributed to dendritic cells5. Supporting this, Irf8 is a
key transcription factor controlling the development and homeostasis of cDC1s28-30.
However, CD11c-cre activity has been observed in other cell types such as
macrophages, natural killer (NK) cells, plasmablasts and activated T lymphocytes37-39.
The anti-leukemic effects of NK cells and cytotoxic T cells are well-documented40-42.
Furthermore, functional roles for Irf8 in NK cells and T cells have been demonstrated.
Biallelic mutations in IRF8 were shown to impair the homeostasis and cytotoxic
potential of NK cells humans and mice43. Similarly, Irf8 is expressed by activated T cells
and might control cytotoxic immune responses44, 45. Thus, we reasoned that the
protective effects derived from WT BM could stem from Irf8-dependent functions of NK
cells or T cells. To test this hypothesis, we transplanted a 1:1 mixture of Irf8-/- BM with
either WT, Rag2-/- or NKp46-DTA BM into lethally irradiated WT hosts. We confirmed
that NKp46-DTA mice, generated by crossing NKp46-cre mice with Rosa-lox-stop-lox-
DTA mice, are completely devoid of NK cells (data not shown). Again, reconstitution of
chimeric mice was validated using an IRF8 staining (figure 3A). While acceptors that
received Irf8-/-:WT BM harbored both Irf8-/- and Irf8-sufficient B cells and NK cells, Irf8-
/-:Rag2-/- and Irf8-/-:NKp46-DTA chimeric mice only generated Irf8-deficient B/T cells

and NK cells, respectively. Remarkably, loss of Irf8-sufficient B-, T- or NK cells did not

166
impair the ability to suppress myeloproliferative disease (figure 3B), suggesting that
the absence of protective effects from CD11c-DTA bone marrow cells on the
development of MPD from Irf8-/- progenitors was unlikely to be caused by lack of IRF8-
expressing B, T or NK cells.

XCR1+ cDC1s lacking IRF8 still control MPD of transplanted Irf8-/- progenitors
Having excluded B-, T- and NK cells, we concluded that the Irf8-dependent suppressive
effect must reside in the DC or mononuclear phagocyte system. The role of IRF8 as a
lineage determining transcription factor for cDC1 development is well-established and
it has been suggested that absence of cDC1s in Irf8-/- mice prompts the development of
MPD5, 28-30. By crossing XCR1-cre mice with conditional Irf8flfl mice, we generated mice
that lack IRF8 expression uniquely in cDC1s. Using the same BM chimera approach, we
assessed whether bone marrow from Irf8flfl x XCR1cre mice was able to confer the same
protection as wild type BM when transplanted together with Irf8-/- BM. Whereas Irf8-/- :
WT BM chimeras sufficiently generated XCR1+ cDC1s, these were completely absent in

Figure 3: IRF8 expression in B-, T- and NK cells is redundant for the inhibition of MPD
Wild type acceptor mice were lethally irradiated and transplanted with either WT BM, Irf8-/- BM, or an
equal mixture of Irf8-/- BM with WT, NKp46-DTA or Rag2-/- BM. (A) Representative histograms displaying
the modal distribution of IRF8 expression in cDCs, B cells and NK cells derived from acceptor spleens 12
weeks after transplantation with the indicated donor BM. (B) Numbers of LSK cells, GMPs and Ly6Ghi
neutrophils in the spleen of mixed BM chimeric mice. Data are representative of 2 independent
experiments (n=8-11 per group). Two-tailed Mann-Whitney U test. Bars indicate mean ± SD. ns = non-
significant, * = p<0.05, ** = p<0.01, *** = p<0.001, **** = p<0.0001.

167
mice transplanted with Irf8-/- : Irf8flfl x XCR1cre or Irf8-/- : Irf8flfl x CD11ccre BM (figure
4A). Importantly, while all CD172a+ cDC2s in Irf8-/- : Irf8flfl x CD11ccre acceptors were
Irf8-deficient, at least half of the cDC2s generated from Irf8-/- : Irf8flfl x XCR1cre BM
expressed low levels of Irf8, similar to wild type cDC2s (figure 4B). Strikingly, the
ability of Irf8flfl x XCR1cre BM to suppress MPD was comparable to that of WT BM and
contrasted clearly with the loss of protection from Irf8flfl x CD11ccre marrow (figure
4C). As expected, the absence of cDC1s alone also did not significantly impact the
frequency of LSK cells, GMPs, Ly6Gint granulocytes or neutrophils in the BM and spleen
of intact Irf8flfl XCR1cre mice (figure 4D, 4E). Taken together, these data suggest that
cDC1s are redundant for the control of myeloproliferation in IRF8-deficient progenitors.

CD64+ cells lacking IRF8 fail to control MPD of Irf8-/- progenitors

We next addressed a potential role for pDCs in controlling extramedullary
hematopoiesis. Plasmacytoid dendritic cells do not require Irf8 for their development,
but Irf8-deficient pDCs display an altered transcriptional profile and fail to produce
IFNα upon stimulation by CpG30. Moreover, pDCs are also targeted by Cre expressed
from the CD11c promotor30. Twelve weeks after reconstitution of lethally irradiated WT
mice with mixed Irf8-/-:WT BM, we treated acceptor mice with a pDC-depleting (clone
120G8) or an isotype control antibody for two weeks (figure 5A)46. While this
treatment significantly reduced the number of pDCs (figure 5B, 5C), it did not hamper
the containment of myeloproliferation (figure 5D).

Having ruled out contributions by cDC1s and pDCs, we next sought to identify the role
of cDC2s. While development of cDC2s does not require IRF8, recent findings from our
lab indicate that IRF8 might play a functional role in cDC2s under certain circumstances
(Bosteels et al., unpublished data). However, no mouse model has been identified to
date that allows reliable targeting of all cDC2s. Loss of thousand-and-one kinase 3
(Taok3) or the transcription factor Zeb2 was shown to affect the development of cDC2s,
albeit incompletely (unpublished data and 47, 48). Of note, Zeb2flfl x CD11ccre mice also
lack the majority of pDCs47, 48. To assess the role of cDC2s, we cotransplanted Irf8-/- BM
with WT, Taok3-/- or Zeb2flfl x CD11ccre BM into wild type acceptors. However, WT,
Taok3-/- and Zeb2flfl x CD11ccre BM components were equally capable of containing
granulocyte expansion (figure 5E). Some macrophages, monocytes and monocyte-
derived dendritic cells can also express the integrin CD11c. To assess the role of these
cells, we crossed Irf8flfl mice with CD64-cre. CD64, the high-affinity IgG receptor FcγRI, is
highly expressed on macrophages and distinguishes conventional DCs from monocyte-
derived DCs49, 50. Strikingly, the Irf8flfl CD64cre BM compartment failed to adequately
suppress the accumulation of GMPs, Ly6Cint granulocytes (not shown) and neutrophils
in the spleen of Irf8-/- : Irf8flfl CD64cre > WT chimeras. Rather, these mice developed a
myeloproliferative syndrome comparable to that seen in Irf8-/- : Irf8flfl CD11ccre > WT
chimeric mice (figure 5E). These findings, while requiring further confirmation, suggest
a potential role for monocyte-derived cells or macrophages in the control of Irf8-/--
driven MPD.

168
Figure 4: cDC1s do not suppress myeloproliferation
(A-C) Wild type mice were lethally irradiated and transplanted with either WT BM, Irf8-/- BM, or an equal
mixture of Irf8-/- BM with WT, Irf8flfl CD11ccre+ or Irf8flfl XCR1cre+ BM. (A) Representative contour plots of
splenic cDCs from BM chimeric mice after reconstitution. cDC1s are gated as XCR1+ CD172a-, cDC2s as
XCR1- CD172a+. (B) Histogram of IRF8 expression in splenic cDC2s from acceptor mice. Gate for IRF8
expression is based on Irf8-/->WT cDC2s. (C) Absolute number of LSK cells, GMPs and neutrophils in the
spleen of acceptor mice. Data are representative of 4-6 mice per group, 2 independent experiments. One-
way ANOVA with post-hoc Tukey’s test. (D) Representative contour plot of Ly6Ghi neutrophils and Ly6Cint
granulocytes in the spleen of intact Irf8flfl XCR1cre+ mice and littermate controls. (E) Quantification of the
number of LSK cells, GMPs, Ly6Cint granulocytes and Ly6Ghi neutrophils in Irf8flfl XCR1cre+ mice. Two-
tailed Mann-Whitney U test. Numbers next to the gates in (A, B, D) indicate percentages of the parent
population. Bars indicate mean ± SD. * = p<0.05, ** = p<0.01, *** = p<0.001, **** = p<0.0001.

169
Figure 5: Plasmacytoid dendritic cells do not suppress myeloproliferation
(A) Schematic representation of pDC depletion in BM chimeric mice. Wild type mice were lethally
irradiated and transplanted with an equal mixture of Irf8-/- and WT BM. Twelve weeks after
reconstitution, acceptor mice were treated every 48 hours with 250μg 120G8 or isotype control antibody
for two consecutive weeks. (B) Representative flow cytometry plot of splenic pDCs in 120G8- and
isotype-treated mice. Plasmacytoid dendritic cells were defined as live, lineage (Ter119/CD3e/CD19)-
B220+ SiglecH+ cells. Surface expression intensity of the 120G8 antigen BST2 (CD317) is color-coded.
Numbers adjacent to the gate indicate the percentage of pDCs within the parent population. The absolute
number of pDCs is quantified in (C). Effect of antibody-mediated depletion on the number of LSK cells,
GMPs and neutrophils in the spleen of mixed bone marrow chimeric mice. Data in (B-D) are
representative of 2 independent experiments (n=10). Two-tailed Mann-Whitney U test. (E) Number of
GMPs and neutrophils in the spleen of WT mice that were transplanted 12 weeks earlier with either WT
BM only, Irf8-/- BM only or an equal mixture of Irf8-/- BM with WT, Irf8flfl x CD11ccre+, Taok3-/-, Zeb2flfl x
CD11ccre+ or Irf8flfl x CD64cre+ BM. Data are representative of 5 mice per group. One-way ANOVA with
post-hoc Tukey’s test. Bars indicate mean ± SD. ns = non-significant, * = p<0.05, ** = p<0.01, *** =
p<0.001, **** = p<0.0001.

170
DISCUSSION

Several observations support an unconventional role of dendritic cells in the regulation

of myelopoiesis. Constitutive or induced ablation of dendritic cells in CD11c-DTA mice,
CD11c-DTR or Zbtb46-DTR mice triggers a myeloproliferative disorder characterized
mainly by accumulation of mature neutrophils and their developmental precursors4-9.
In line with these observations, the MPD that arises in mice that lack the transcription
factor IRF8 has been attributed to their cDC1 deficit. It was proposed that increased
myelopoiesis might serve as a compensatory mechanism trying to restore a missing DC
compartment5.

The uncontrolled granulocyte production in the bone marrow and at extramedullary

sites in DC-deficient mice is very reminiscent of the manifestations of
myeloproliferative neoplasms (MPN) in humans. In chronic myeloid leukemia (CML), a
prototypical MPN, reciprocal translocation between chromosomes (t9;22) leads to the
formation of the oncogenic fusion protein BCR-ABL. Chronic myeloid leukemia is
generally considered to develop in a cell-intrinsic manner from hematopoietic stem
cells containing the BCR-ABL fusion protein, a concept illustrated by the therapeutic
success of tyrosine kinase inhibitors that target this oncoprotein51, 52. However, several
lines of evidence suggest that additional mechanisms might be involved in CML
development. In contrast to untargeted approaches, conditional knockin of BCR-ABL
was not sufficient to trigger MPD in mice53, 54. Furthermore, not all human carriers of
the BCR-ABL translocation develop disease55-58. In addition, the drivers of progression
from the chronic phase of CML to the accelerated phase preceding blast crises remain
unknown59. It is well documented that NK and T cell responses against leukemic cells
are hampered in CML patients, and that immunotherapy heralds great therapeutic
potential60-69. IRF8 is down-regulated in BCR-ABL+ cells, explaining why CML patients
harbor an abnormal dendritic cell population70, 71. More importantly, IRF8
overexpression restored DC differentiation in myeloid precursors from CML patients
and improved the generation of cytotoxic immune responses72. Rather than a side effect,
the dendritic cell deficit in CML patients may thus constitute a driving force that
promotes disease development through defective priming of tumor-specific T cells.

Here, we present data that support a function of IRF8-dependent dendritic cells in the
control of myelopoiesis. Using a mixed bone marrow chimera approach, we found that
co-transplantation of wild type bone marrow could inhibit the development of MPD by
Irf8-/- BM, while CD11c-DTA or Irf8flfl x CD11ccre BM could not. Interestingly, intact Irf8flfl
x CD11ccre mice did not develop full-blown MPD, aside from subtle increases in splenic
granulocyte numbers. First, this suggests that the absence of DCs is not sufficient to
trigger disease. Rather, it impairs the generation of an inhibitory effect that puts a brake
on the uncontrolled proliferation of Irf8-deficient GMPs. Second, these observations
seem to set the phenotype of Irf8-/- mice apart from that of CD11c-DTA, CD11c-DTR,
TAK1- and MHCII-conditional knockout mice, that develop overt MPD even without an

171
intrinsic effect on hematopoietic progenitor cells4, 5, 10, 12. Possibly, loss of Irf8 in DCs
might impact the suppressive mechanism to a lesser extent than the complete absence
of DCs. Indeed, the neutrophilia in Irf8-/-:Irf8flfl x CD11ccre BM chimeras was less
pronounced than that of Irf8-/-:CD11c-DTA chimeras.

It remains uncertain to what extent the observed effects of DCs in the control of
myelopoiesis represent direct or T-cell mediated effects. Despite the notion that thymic
negative selection and generation of peripheral regulatory T cells were largely intact in
CD11c-DTA mice, these mice generated more Th17 cells and developed autoimmune
disease4, 5. It is conceivable that absence or alterations of DCs could lead to enhanced
Th17 polarization, either through lack of tolerogenic functions, or indirectly through
changes in the microbiome73-75. As Th17 can stimulate neutrophil recruitment, an MPD-
like phenotype and autoimmunity might therefore be different ends of the same
spectrum76, 77. Mice that lacked MHCII expression in dendritic cells only developed
neutrophilia on a Rag-sufficient background, supporting a requirement for T cells to
mediate the myelosuppressive effect. In contrast, MPD in conditional TAK1-knockout
mice did not depend on T cells. In our model of Irf8-/- driven MPD, it is possible that the
protective effect from Rag2-/- bone marrow was caused by interactions between Rag2-/-
dendritic cells and Irf8-/- T cells. However, we did exclude a direct role for Irf8 in T cells
for the inhibition of MPD, arguing against the proposition that the MPD in Irf8-/- mice is
secondary to T-cell intrinsic effects78. Interestingly, Irf8 was found to suppress Th17
differentiation, possibly in a cell-extrinsic manner79, 80. Hypothetically, Irf8-deficient DCs
might promote Th17 differentiation, or fail to promote alternative fates, leading to
increased cytokine production by an over-activated Th17 compartment. Hence, further
investigation of the T cell compartment and the presence of autoimmune pathology in
Irf8flfl x CD11ccre mice is warranted.

Previous work often relied on transgenic mouse models in which all CD11c+ cells are
ablated. However, CD11c is not entirely DC-specific, and many other cell types are
affected in these mice37-39. Using Rag-/- and NKp46-DTA bone marrow, we prove that the
Irf8-dependent CD11c+ cells suppressing MPD are not B-, T- or NK cells. Nevertheless,
this does not preclude a (Irf8-independent) contribution by these lymphocytes in the
containment of MPD, as mentioned before. Importantly, we demonstrate that protection
from MPD was independent of cDC1s, that did not develop in Irf8-/-: Irf8flfl XCR1cre BM
chimeras. Further negating the theory that a deficit in type 1 dendritic cells triggers
MPD development, hematopoietic stem cell and neutrophil numbers were normal in
intact Irf8flfl XCR1cre mice. By antibody-mediated depletion and through the use of
Zeb2flfl CD11ccre BM, we also demonstrated the redundancy of plasmacytoid dendritic
cells. Together, these findings suggested that the myelosuppressive function of DCs lies
with cDC2s or monocyte-derived cells. Here, we were confronted with the inadequacy
of genetic tools to address cDC2s in a specific and complete manner. Taok3-/- and Zeb2flfl
x CD11ccre BM were fully capable of suppressing MPD, but the number of cDC2s is only
partially decreased in these models. Further investigation is warranted to examine a

172
role for cDC2s, especially because preliminary findings suggest that Irf8 might play a
previously undiscovered, functional role in these cells (Bosteels et al, unpublished
results). Surprisingly, we found that MPD developed in Irf8-/-:Irf8flfl x CD64cre BM
chimeras. CD64 expression characterizes monocytes, tissue-resident macrophages, and
monocyte-derived dendritic cells49, 50. Whereas several studies support a role for IRF8
in the development and function of macrophages, little is known about the function of
Irf8 in mature monocytes81-84. These results require confirmation, and the targeting
efficiency of CD64cre in different hematopoietic cell types should be carefully examined
using IRF8 staining. Through the use of fluorescent reporter mice, it was recently
demonstrated that CD64cre also targets 10-20% of cDC2s and monocytes, as well as
minor fractions of cDC1s, B and T cells in spleen47. Therefore, excision of Irf8 in cDC2s
from Irf8flfl x CD64cre mice should be excluded before drawing conclusions about a role
of monocyte-derived cells in myelopoiesis. Moreover, it should be examined whether
intact Irf8flfl x CD64cre mice develop MPD or only display subtle effects as in Irf8flfl x
CD11ccre mice.

In conclusion, we found that the absence of type 1 dendritic cells does not trigger the
development of MPD. Rather, IRF8 controls a hitherto unidentified function of CD64+
monocyte-derived cells or potentially cDC2s in the regulation of hematopoiesis.
Impaired function of dendritic cells might explain why CML cells escape
immunosurveillance, and uncovering the precise mechanisms by which DCs control
hematopoiesis could open up new therapeutic avenues.

173
MATERIALS AND METHODS

Mice
Rag2-/-, Rosa26-lox-STOP-lox-Diphteria toxin A mice (RosaDTA/DTA) and CD11c-cre
(B6.Cg-Tg (Itgax-cre)1-1 Reiz/J) mice were acquired from Jackson. RosaDTA/DTA mice
were crossed to NcriCre/iCre mice85, kindly provided by Eric Vivier from CIML (Marseille,
France). Taok3-/-, Irf8fl/fl and Irf8-/- mice were generated in house and have been
described before30, 86. Zeb2fl/fl mice, also described previously, were kindly provided by
Geert Berx from the Cancer Research Institute Ghent87. Xcr1-IRES-iCre-2A-mTFP1 gene-
targeted (XCR1-cre) mice and CD64-cre mice were kindly provided by Bernard Malissen
from CIML (Marseille, France). Wild type (WT) C57Bl/6J mice were obtained from the
Janvier laboratory. All experimental mice were on a C57Bl/6J background and were
housed in the VIB-UGent animal facility under specific pathogen free conditions.
Housing conditions entailed individually ventilated cages in a controlled day-night cycle
and food and water intake ad libitum. Mice were age- and sex-matched and blindly
randomized into experimental groups. All animal experiments performed were
approved by the local animal ethics committee of Ghent University and were performed
in accordance with the Belgian animal protection law.

pDC depletion
To deplete plasmacytoid dendritic cells in vivo, mice were treated with an anti-BST2
antibody (clone 120G8) or an isotype control antibody (rat IgG1κ, clone HRPN,
BioXCell), diluted in PBS. Mice were administered 250 μg per injection and injected
intraperitoneally every 2 days for 14 days. Treatment of bone marrow chimeric mice
was started 12 weeks after transplantation, and mice were euthanized 2 days after the
last treatment.

Bone marrow chimeras

Wild type acceptor mice were irradiated using 9.5 Gy, followed by intravenous injection
of 4 × 106 donor cells in endotoxin-free PBS at least 4 h after the irradiation. Donor cells
consisted either of 100% Irf8-/- or Irf8+/+ bone marrow cells, or of 50:50 mixes of Irf8-/-
cells with a genotype of interest. Acceptor mice were euthanized 12 weeks after
reconstitution. Bone marrow and spleen were analyzed for the presence of
hematopoietic progenitor cells and mature myeloid cells. The reconstitution was
validated using flow cytometric analysis of IRF8 expression in various hematopoietic
cell types.

Cell isolation and flow cytometry

Cell suspensions from spleen and bone marrow were obtained through mechanical
disruption and filtering over a 70µm mesh. Samples were incubated in ammonium
chloride buffer (10mM KHCO3, 155 mM NH4Cl, 0.1 mM EDTA in milliQ water) for
erythrocyte lysis. Total cell counts were obtained by adding counting beads (123count
eBeads, Thermo Fisher Scientific). Single cell suspensions were first incubated with a

174
fixable viability dye (eFluor506 or eFluor780 from eBioscience) to identify dead cells
and with an FcγRII/III antibody (2.4G2) for 30 minutes to limit non-specific binding,
except for CD16/32 stainings. After washing, the cells were incubated with a mixture of
fluorescently labeled antibodies for 30 min at 4°C. Following antibodies were used: CD3
(145-2C11), CD19 (1D3), CD11c (N418), NK1.1 (PK136), MHCII (M5/114.15.2), Ly6G
(1A8), CD11b (M1/70), F4/80 (BM8), CD26 (H194-112), CD64 (X54-5/7.1), CD24
(M1/69), CD172a (P84), XCR1 (ZET), Ter119 (TER-119), NKp46 (29A1.4), B220 (RA3-
6B2), BST2 (120G8), CD135 (A2F10), Ly6C (HK1.4 or AL-21), SiglecH (ebio440c),
CD16/32 (93), c-Kit (2B8), Sca-1 (D7), CD34 (RAM34), CD48 (HM48-1), CD150 (TC15-
12F12.2). When biotin-labeled primary antibodies were used, a second surface-staining
step with BV786- or BV605-coupled streptavidin (BD Biosciences) was performed. For
intracellular staining of the transcription factor IRF8, cells were fixed using the FoxP3
fixation/permeabilisation kit (eBioscience) according to the manufacturer’s protocol.
Cells were then incubated with PerCP-efluor710-labeled anti-IRF8 (V3GYWCH) for 30
min at 4°C. Samples were acquired on an LSRFortessa cytometer with FACSDiva
software (BD Biosciences) and analyzed using FlowJo software (LLC, Ashland, OR).

Confocal microscopy
Spleens were harvested, embedded in OCT Tissue Tek medium (Takura) and
immediately frozen by liquid nitrogen. Ten µm thick cryostat sections were fixed for 10
min by paraformaldehyde 2% in PBS, washed with PBS and permeabilized/blocked
with PBS containing 0,1% triton X-100 (Roche) and 2% BSA (Sigma) (PBT buffer) for 30
min. The following antibodies were used: FITC-conjugated B220 (RA3-6B2) and BV421-
conjugated CD11b (M1/70) were obtained from BD Biosciences. PE-conjugated DCIR2
(33D1) was obtained from eBioscience. Unconjugated polyclonal rabbit anti-Ki67
antibody was purchased from Abcam (ab15580). Cy5-conjugated donkey anti-rabbit
IgG was obtained from Jackson ImmunoResearch (711-175-152). Sections were stained
with the primary or secondary antibodies diluted in PBT buffer for 60 min at room
temperature and washed in between. Sections were counterstained with DAPI, washed,
and coverslipped with antifading polyvinyl alcohol containing DABCO (Sigma-Aldrich).
Images were acquired on a Zeiss LSM710 confocal microscope equipped with 488-nm,
561-nm and 633-nm lasers and with a tunable two-photon laser. Emissions were
recorded in three separate channels. Digital pictures at 1024 x 1024 pixel density and 8-
bit depth were acquired with ZEN software (Zeiss). Images were analyzed with Imaris
software.

Statistical analysis
Statistical analyses were performed with Graphpad Prism 7.0 (Graphpad Software, La
Jolla, CA). For comparisons between two groups, a two-tailed Mann-Whitney U test was
used. For comparisons between three or more groups, a one-way ANOVA with Tukey’s
correction for multiple comparisons was used. The significance level α was 0.05. Error
bars represent standard deviation of the mean. Non-significant differences are not

175
indicated or denoted as ns, significant differences are expressed as * = p<0.05, ** =
p<0.01, *** = p<0.001, ****=p<0.0001.

AUTHOR CONTRIBUTIONS

M.V. performed and analyzed experiments and wrote the manuscript. C.B., D.S., K.D.,
M.VH., J.V.M. and F.F. performed experiments. M.G. provided mice. H.H., T.K. and B.N.L.
conceptualized the research, analyzed results and wrote the manuscript.

ACKNOWLEDGEMENTS

M.V., C.B. and D.S. were supported by an FWO fellowship. H.H. is supported by a Ghent
University Grant (GOA 01G02817). B.N.L. was supported by an ERC consolidator grant
261231, the University of Ghent Multidisciplinary Research Platform (MRP-GROUP-ID)
and a Ghent University grant (GOA 01G02817).

176
REFERENCES

1. Merad M, Sathe P, Helft J, Miller J, Mortha A. The dendritic cell lineage: ontogeny and function of dendritic
cells and their subsets in the steady state and the inflamed setting. Annu Rev Immunol. 2013;31:563-604.
2. Mildner A, Jung S. Development and function of dendritic cell subsets. Immunity. 2014;40(5):642-56.
3. Durai V, Murphy KM. Functions of Murine Dendritic Cells. Immunity. 2016;45(4):719-36.
4. Ohnmacht C, Pullner A, King SB, Drexler I, Meier S, Brocker T, et al. Constitutive ablation of dendritic cells
breaks self-tolerance of CD4 T cells and results in spontaneous fatal autoimmunity. J Exp Med. 2009;206(3):549-59.
5. Birnberg T, Bar-On L, Sapoznikov A, Caton ML, Cervantes-Barragan L, Makia D, et al. Lack of conventional
dendritic cells is compatible with normal development and T cell homeostasis, but causes myeloid proliferative
syndrome. Immunity. 2008;29(6):986-97.
6. Loschko J, Rieke GJ, Schreiber HA, Meredith MM, Yao KH, Guermonprez P, et al. Inducible targeting of cDCs
and their subsets in vivo. J Immunol Methods. 2016;434:32-8.
7. Meredith MM, Liu K, Darrasse-Jeze G, Kamphorst AO, Schreiber HA, Guermonprez P, et al. Expression of the
zinc finger transcription factor zDC (Zbtb46, Btbd4) defines the classical dendritic cell lineage. J Exp Med.
2012;209(6):1153-65.
8. Jiao J, Dragomir AC, Kocabayoglu P, Rahman AH, Chow A, Hashimoto D, et al. Central role of conventional
dendritic cells in regulation of bone marrow release and survival of neutrophils. J Immunol. 2014;192(7):3374-82.
9. Tittel AP, Heuser C, Ohliger C, Llanto C, Yona S, Hammerling GJ, et al. Functionally relevant neutrophilia in
CD11c diphtheria toxin receptor transgenic mice. Nat Methods. 2012;9(4):385-90.
10. Wang Y, Huang G, Vogel P, Neale G, Reizis B, Chi H. Transforming growth factor beta-activated kinase 1
(TAK1)-dependent checkpoint in the survival of dendritic cells promotes immune homeostasis and function. Proc
Natl Acad Sci U S A. 2012;109(6):E343-52.
11. Satpathy AT, Briseno CG, Cai X, Michael DG, Chou C, Hsiung S, et al. Runx1 and Cbfbeta regulate the
development of Flt3+ dendritic cell progenitors and restrict myeloproliferative disorder. Blood. 2014;123(19):2968-
77.
12. Humblet-Baron S, Barber JS, Roca CP, Lenaerts A, Koni PA, Liston A. Murine myeloproliferative disorder as a
consequence of impaired collaboration between dendritic cells and CD4 T cells. Blood. 2019;133(4):319-30.
13. Holtschke T, Lohler J, Kanno Y, Fehr T, Giese N, Rosenbauer F, et al. Immunodeficiency and chronic
myelogenous leukemia-like syndrome in mice with a targeted mutation of the ICSBP gene. Cell. 1996;87(2):307-17.
14. Turcotte K, Gauthier S, Tuite A, Mullick A, Malo D, Gros P. A mutation in the Icsbp1 gene causes susceptibility
to infection and a chronic myeloid leukemia-like syndrome in BXH-2 mice. J Exp Med. 2005;201(6):881-90.
15. Pathak S, Ma S, Shukla V, Lu R. A role for IRF8 in B cell anergy. J Immunol. 2013;191(12):6222-30.
16. Feng J, Wang H, Shin DM, Masiuk M, Qi CF, Morse HC, 3rd. IFN regulatory factor 8 restricts the size of the
marginal zone and follicular B cell pools. J Immunol. 2011;186(3):1458-66.
17. Carotta S, Willis SN, Hasbold J, Inouye M, Pang SH, Emslie D, et al. The transcription factors IRF8 and PU.1
negatively regulate plasma cell differentiation. J Exp Med. 2014;211(11):2169-81.
18. Ma S, Turetsky A, Trinh L, Lu R. IFN regulatory factor 4 and 8 promote Ig light chain kappa locus activation
in pre-B cell development. J Immunol. 2006;177(11):7898-904.
19. Lu R, Medina KL, Lancki DW, Singh H. IRF-4,8 orchestrate the pre-B-to-B transition in lymphocyte
development. Genes Dev. 2003;17(14):1703-8.
20. Kurotaki D, Yamamoto M, Nishiyama A, Uno K, Ban T, Ichino M, et al. IRF8 inhibits C/EBPalpha activity to
restrain mononuclear phagocyte progenitors from differentiating into neutrophils. Nat Commun. 2014;5:4978.
21. Sichien D, Lambrecht BN, Guilliams M, Scott CL. Development of conventional dendritic cells: from common
bone marrow progenitors to multiple subsets in peripheral tissues. Mucosal Immunol. 2017;10(4):831-44.
22. Tamura T, Kurotaki D, Koizumi S. Regulation of myelopoiesis by the transcription factor IRF8. Int J Hematol.
2015;101(4):342-51.
23. Becker AM, Michael DG, Satpathy AT, Sciammas R, Singh H, Bhattacharya D. IRF-8 extinguishes neutrophil
production and promotes dendritic cell lineage commitment in both myeloid and lymphoid mouse progenitors.
Blood. 2012;119(9):2003-12.
24. Bigley V, Maisuria S, Cytlak U, Jardine L, Care MA, Green K, et al. Biallelic interferon regulatory factor 8
mutation: A complex immunodeficiency syndrome with dendritic cell deficiency, monocytopenia, and immune
dysregulation. J Allergy Clin Immunol. 2018;141(6):2234-48.
25. Hochweller K, Miloud T, Striegler J, Naik S, Hammerling GJ, Garbi N. Homeostasis of dendritic cells in
lymphoid organs is controlled by regulation of their precursors via a feedback loop. Blood. 2009;114(20):4411-21.
26. Gilliland DG, Griffin JD. The roles of FLT3 in hematopoiesis and leukemia. Blood. 2002;100(5):1532-42.

177
27. Tsapogas P, Mooney CJ, Brown G, Rolink A. The Cytokine Flt3-Ligand in Normal and Malignant
Hematopoiesis. Int J Mol Sci. 2017;18(6).
28. Aliberti J, Schulz O, Pennington DJ, Tsujimura H, Reis e Sousa C, Ozato K, et al. Essential role for ICSBP in the
in vivo development of murine CD8alpha + dendritic cells. Blood. 2003;101(1):305-10.
29. Schiavoni G, Mattei F, Sestili P, Borghi P, Venditti M, Morse HC, 3rd, et al. ICSBP is essential for the
development of mouse type I interferon-producing cells and for the generation and activation of CD8alpha(+)
dendritic cells. J Exp Med. 2002;196(11):1415-25.
30. Sichien D, Scott CL, Martens L, Vanderkerken M, Van Gassen S, Plantinga M, et al. IRF8 Transcription Factor
Controls Survival and Function of Terminally Differentiated Conventional and Plasmacytoid Dendritic Cells,
Respectively. Immunity. 2016;45(3):626-40.
31. Guilliams M, Dutertre CA, Scott CL, McGovern N, Sichien D, Chakarov S, et al. Unsupervised High-
Dimensional Analysis Aligns Dendritic Cells across Tissues and Species. Immunity. 2016;45(3):669-84.
32. Murphy TL, Grajales-Reyes GE, Wu X, Tussiwand R, Briseno CG, Iwata A, et al. Transcriptional Control of
Dendritic Cell Development. Annu Rev Immunol. 2016;34:93-119.
33. Yamazaki C, Sugiyama M, Ohta T, Hemmi H, Hamada E, Sasaki I, et al. Critical roles of a dendritic cell subset
expressing a chemokine receptor, XCR1. J Immunol. 2013;190(12):6071-82.
34. Alexandre YO, Ghilas S, Sanchez C, Le Bon A, Crozat K, Dalod M. XCR1+ dendritic cells promote memory
CD8+ T cell recall upon secondary infections with Listeria monocytogenes or certain viruses. J Exp Med.
2016;213(1):75-92.
35. Ohta T, Sugiyama M, Hemmi H, Yamazaki C, Okura S, Sasaki I, et al. Crucial roles of XCR1-expressing
dendritic cells and the XCR1-XCL1 chemokine axis in intestinal immune homeostasis. Sci Rep. 2016;6:23505.
36. Netherby CS, Messmer MN, Burkard-Mandel L, Colligan S, Miller A, Cortes Gomez E, et al. The Granulocyte
Progenitor Stage Is a Key Target of IRF8-Mediated Regulation of Myeloid-Derived Suppressor Cell Production. J
Immunol. 2017;198(10):4129-39.
37. Jung S, Unutmaz D, Wong P, Sano G, De los Santos K, Sparwasser T, et al. In vivo depletion of CD11c+
dendritic cells abrogates priming of CD8+ T cells by exogenous cell-associated antigens. Immunity. 2002;17(2):211-
20.
38. Bennett CL, Clausen BE. DC ablation in mice: promises, pitfalls, and challenges. Trends Immunol.
2007;28(12):525-31.
39. Abram CL, Roberge GL, Hu Y, Lowell CA. Comparative analysis of the efficiency and specificity of myeloid-
Cre deleting strains using ROSA-EYFP reporter mice. J Immunol Methods. 2014;408:89-100.
40. Kolb HJ. Graft-versus-leukemia effects of transplantation and donor lymphocytes. Blood.
2008;112(12):4371-83.
41. Bollard CM, Barrett AJ. Cytotoxic T lymphocytes for leukemia and lymphoma. Hematology Am Soc Hematol
Educ Program. 2014;2014(1):565-9.
42. Handgretinger R, Lang P, Andre MC. Exploitation of natural killer cells for the treatment of acute leukemia.
Blood. 2016;127(26):3341-9.
43. Mace EM, Bigley V, Gunesch JT, Chinn IK, Angelo LS, Care MA, et al. Biallelic mutations in IRF8 impair human
NK cell maturation and function. J Clin Invest. 2017;127(1):306-20.
44. Sun L, St Leger AJ, Yu CR, He C, Mahdi RM, Chan CC, et al. Interferon Regulator Factor 8 (IRF8) Limits Ocular
Pathology during HSV-1 Infection by Restraining the Activation and Expansion of CD8+ T Cells. PLoS One.
2016;11(5):e0155420.
45. Miyagawa F, Zhang H, Terunuma A, Ozato K, Tagaya Y, Katz SI. Interferon regulatory factor 8 integrates T-
cell receptor and cytokine-signaling pathways and drives effector differentiation of CD8 T cells. Proc Natl Acad Sci U S
A. 2012;109(30):12123-8.
46. Asselin-Paturel C, Brizard G, Pin JJ, Briere F, Trinchieri G. Mouse strain differences in plasmacytoid dendritic
cell frequency and function revealed by a novel monoclonal antibody. J Immunol. 2003;171(12):6466-77.
47. Scott CL, T'Jonck W, Martens L, Todorov H, Sichien D, Soen B, et al. The Transcription Factor ZEB2 Is
Required to Maintain the Tissue-Specific Identities of Macrophages. Immunity. 2018;49(2):312-25 e5.
48. Wu X, Briseno CG, Grajales-Reyes GE, Haldar M, Iwata A, Kretzer NM, et al. Transcription factor Zeb2
regulates commitment to plasmacytoid dendritic cell and monocyte fate. Proc Natl Acad Sci U S A.
2016;113(51):14775-80.
49. Langlet C, Tamoutounour S, Henri S, Luche H, Ardouin L, Gregoire C, et al. CD64 expression distinguishes
monocyte-derived and conventional dendritic cells and reveals their distinct role during intramuscular
immunization. J Immunol. 2012;188(4):1751-60.

178
50. Gautier EL, Shay T, Miller J, Greter M, Jakubzick C, Ivanov S, et al. Gene-expression profiles and
transcriptional regulatory pathways that underlie the identity and diversity of mouse tissue macrophages. Nat
Immunol. 2012;13(11):1118-28.
51. Savona M, Talpaz M. Getting to the stem of chronic myeloid leukaemia. Nat Rev Cancer. 2008;8(5):341-50.
52. Apperley JF. Chronic myeloid leukaemia. Lancet. 2015;385(9976):1447-59.
53. Foley SB, Hildenbrand ZL, Soyombo AA, Magee JA, Wu Y, Oravecz-Wilson KI, et al. Expression of BCR/ABL
p210 from a knockin allele enhances bone marrow engraftment without inducing neoplasia. Cell Rep. 2013;5(1):51-
60.
54. Ross TS, Mgbemena VE. Re-evaluating the role of BCR/ABL in chronic myelogenous leukemia. Mol Cell
Oncol. 2014;1(3):e963450.
55. Boquett JA, Alves JR, de Oliveira CE. Analysis of BCR/ABL transcripts in healthy individuals. Genet Mol Res.
2013;12(4):4967-71.
56. Bayraktar S, Goodman M. Detection of BCR-ABL Positive Cells in an Asymptomatic Patient: A Case Report
and Literature Review. Case Rep Med. 2010;2010:939706.
57. Biernaux C, Loos M, Sels A, Huez G, Stryckmans P. Detection of major bcr-abl gene expression at a very low
level in blood cells of some healthy individuals. Blood. 1995;86(8):3118-22.
58. Bose S, Deininger M, Gora-Tybor J, Goldman JM, Melo JV. The presence of typical and atypical BCR-ABL
fusion genes in leukocytes of normal individuals: biologic significance and implications for the assessment of minimal
residual disease. Blood. 1998;92(9):3362-7.
59. Goldman JM, Melo JV. BCR-ABL in chronic myelogenous leukemia--how does it work? Acta Haematol.
2008;119(4):212-7.
60. Nieda M, Nicol A, Kikuchi A, Kashiwase K, Taylor K, Suzuki K, et al. Dendritic cells stimulate the expansion of
bcr-abl specific CD8+ T cells with cytotoxic activity against leukemic cells from patients with chronic myeloid
leukemia. Blood. 1998;91(3):977-83.
61. Held SA, Heine A, Mayer KT, Kapelle M, Wolf DG, Brossart P. Advances in immunotherapy of chronic myeloid
leukemia CML. Curr Cancer Drug Targets. 2013;13(7):768-74.
62. Pavlu J, Apperley JF. Allogeneic stem cell transplantation for chronic myeloid leukemia. Curr Hematol Malig
Rep. 2013;8(1):43-51.
63. Chen CI, Koschmieder S, Kerstiens L, Schemionek M, Altvater B, Pscherer S, et al. NK cells are dysfunctional
in human chronic myelogenous leukemia before and on imatinib treatment and in BCR-ABL-positive mice. Leukemia.
2012;26(3):465-74.
64. Rossignol A, Levescot A, Jacomet F, Robin A, Basbous S, Giraud C, et al. Evidence for BCR-ABL-dependent
dysfunctions of iNKT cells from chronic myeloid leukemia patients. Eur J Immunol. 2012;42(7):1870-5.
65. Cai A, Keskin DB, DeLuca DS, Alonso A, Zhang W, Zhang GL, et al. Mutated BCR-ABL generates immunogenic
T-cell epitopes in CML patients. Clin Cancer Res. 2012;18(20):5761-72.
66. Kijima M, Gardiol N, Held W. Natural killer cell mediated missing-self recognition can protect mice from
primary chronic myeloid leukemia in vivo. PLoS One. 2011;6(11):e27639.
67. Ohyashiki K, Katagiri S, Tauchi T, Ohyashiki JH, Maeda Y, Matsumura I, et al. Increased natural killer cells
and decreased CD3(+)CD8(+)CD62L(+) T cells in CML patients who sustained complete molecular remission after
discontinuation of imatinib. Br J Haematol. 2012;157(2):254-6.
68. Bocchia M, Gentili S, Abruzzese E, Fanelli A, Iuliano F, Tabilio A, et al. Effect of a p210 multipeptide vaccine
associated with imatinib or interferon in patients with chronic myeloid leukaemia and persistent residual disease: a
multicentre observational trial. Lancet. 2005;365(9460):657-62.
69. Pawelec G, Da Silva P, Max H, Kalbacher H, Schmidt H, Bruserud O, et al. Relative roles of natural killer- and
T cell-mediated anti-leukemia effects in chronic myelogenous leukemia patients treated with interferon-alpha. Leuk
Lymphoma. 1995;18(5-6):471-8.
70. Hjort EE, Huang W, Hu L, Eklund EA. Bcr-abl regulates Stat5 through Shp2, the interferon consensus
sequence binding protein (Icsbp/Irf8), growth arrest specific 2 (Gas2) and calpain. Oncotarget. 2016;7(47):77635-
50.
71. Waight JD, Banik D, Griffiths EA, Nemeth MJ, Abrams SI. Regulation of the interferon regulatory factor-8
(IRF-8) tumor suppressor gene by the signal transducer and activator of transcription 5 (STAT5) transcription factor
in chronic myeloid leukemia. J Biol Chem. 2014;289(22):15642-52.
72. Watanabe T, Hotta C, Koizumi S, Miyashita K, Nakabayashi J, Kurotaki D, et al. The transcription factor IRF8
counteracts BCR-ABL to rescue dendritic cell development in chronic myelogenous leukemia. Cancer Res.
2013;73(22):6642-53.
73. Omenetti S, Pizarro TT. The Treg/Th17 Axis: A Dynamic Balance Regulated by the Gut Microbiome. Front
Immunol. 2015;6:639.

179
74. Goto Y, Panea C, Nakato G, Cebula A, Lee C, Diez MG, et al. Segmented filamentous bacteria antigens
presented by intestinal dendritic cells drive mucosal Th17 cell differentiation. Immunity. 2014;40(4):594-607.
75. Loschko J, Schreiber HA, Rieke GJ, Esterhazy D, Meredith MM, Pedicord VA, et al. Absence of MHC class II on
cDCs results in microbial-dependent intestinal inflammation. J Exp Med. 2016;213(4):517-34.
76. Flannigan KL, Ngo VL, Geem D, Harusato A, Hirota SA, Parkos CA, et al. IL-17A-mediated neutrophil
recruitment limits expansion of segmented filamentous bacteria. Mucosal Immunol. 2017;10(3):673-84.
77. Disteldorf EM, Krebs CF, Paust HJ, Turner JE, Nouailles G, Tittel A, et al. CXCL5 drives neutrophil recruitment
in TH17-mediated GN. J Am Soc Nephrol. 2015;26(1):55-66.
78. Paschall AV, Zhang R, Qi CF, Bardhan K, Peng L, Lu G, et al. IFN regulatory factor 8 represses GM-CSF
expression in T cells to affect myeloid cell lineage differentiation. J Immunol. 2015;194(5):2369-79.
79. Ouyang X, Zhang R, Yang J, Li Q, Qin L, Zhu C, et al. Transcription factor IRF8 directs a silencing programme
for TH17 cell differentiation. Nat Commun. 2011;2:314.
80. Newman DM, Leung PS, Putoczki TL, Nutt SL, Cretney E. Th17 cell differentiation proceeds independently of
IRF8. Immunol Cell Biol. 2016;94(8):796-801.
81. Gupta M, Shin DM, Ramakrishna L, Goussetis DJ, Platanias LC, Xiong H, et al. IRF8 directs stress-induced
autophagy in macrophages and promotes clearance of Listeria monocytogenes. Nat Commun. 2015;6:6379.
82. Hagemeyer N, Kierdorf K, Frenzel K, Xue J, Ringelhan M, Abdullah Z, et al. Transcriptome-based profiling of
yolk sac-derived macrophages reveals a role for Irf8 in macrophage maturation. EMBO J. 2016;35(16):1730-44.
83. Wang IM, Contursi C, Masumi A, Ma X, Trinchieri G, Ozato K. An IFN-gamma-inducible transcription factor,
IFN consensus sequence binding protein (ICSBP), stimulates IL-12 p40 expression in macrophages. J Immunol.
2000;165(1):271-9.
84. Mass E, Ballesteros I, Farlik M, Halbritter F, Gunther P, Crozet L, et al. Specification of tissue-resident
macrophages during organogenesis. Science. 2016;353(6304).
85. Narni-Mancinelli E, Chaix J, Fenis A, Kerdiles YM, Yessaad N, Reynders A, et al. Fate mapping analysis of
lymphoid cells expressing the NKp46 cell surface receptor. Proc Natl Acad Sci U S A. 2011;108(45):18324-9.
86. Hammad H, Vanderkerken M, Pouliot P, Deswarte K, Toussaint W, Vergote K, et al. Transitional B cells
commit to marginal zone B cell fate by Taok3-mediated surface expression of ADAM10. Nat Immunol.
2017;18(3):313-20.
87. Scott CL, Soen B, Martens L, Skrypek N, Saelens W, Taminau J, et al. The transcription factor Zeb2 regulates
development of conventional and plasmacytoid DCs by repressing Id2. J Exp Med. 2016;213(6):897-911.

180
VII. General discussion

181
Taok3 and Notch signaling

Almost two decades ago, TAO-kinases were identified and classified as members of the
Ste20-like kinase family1-4. Since then, the function of these kinases in vivo has remained
largely enigmatic. Like other kinases of the GCK subfamily, TAOK3 was proposed to act
as an upstream regulator of MAPKs, based on in vitro studies1-4. In our lab, Taok3-/- mice
were initially generated to study the role of TAOK3 in the unfolded protein response
(UPR) and ER stress, though such a function could not be confirmed experimentally.
However, careful analysis of these mice revealed severe defects in the immune system.
In particular, we identified cell-intrinsic effects of TAOK3 in marginal zone B cells and
ESAM+ dendritic cells, two populations that require signals through the Notch2 receptor
for their development5-7. We found that TAOK3 controls Notch signaling activity in MZB
cells and cDCs using cocultures with DLL1-expressing fibroblasts, and demonstrated
that overexpression of the Notch intracellular domain rescued the terminal
differentiation of Taok3-deficient cDCs.

Nevertheless, Taok3 only controls certain aspects of Notch signaling or is involved only
under specific physiological conditions. This is foremost illustrated by the fact that
Notch2-/- mice die during embryogenesis while Taok3-/- mice do not8. Moreover, other
immune cells that require Notch2 instruction for their development, such as patrolling
monocytes and CD103+ CD11b+ cDCs in the small intestine7, 9, were unaffected by loss of
Taok3 (unpublished results and chapter IV). In this regard, the differential requirement
of Notch2 and the downstream transcription factor RBPjκ for dendritic cell
development is of particular interest. The frequency of splenic cDC1s and intestinal
CD103+ cDC2s was severely reduced in Notch2 knockout mice, but largely intact in mice
that lacked RBPjκ or Taok36. Interestingly, the populations that are affected in Taok3-/-
mice (ESAM+ cDC2s and MZB cells) also depend on RBPjκ activation downstream of
Notch210, 11. These findings suggest that Taok3 might specifically regulate the activation
of the canonical, RBPjκ-mediated Notch pathway. To validate the coexistence of both
canonical and non-canonical Notch signaling pathways in cDCs, it would be interesting
to assess whether RBPjκ also controls ESAM expression in cDC1s, as we have
demonstrated for Taok3. Based on the redundancy of Taok3 for the development of
patrolling monocytes (unpublished results), we speculate that these may also arise
through a non-canonical Notch pathway. Different mechanisms for RBPjκ-independent
Notch signaling have been demonstrated, yet how these apply to dendritic cell biology
remains uncertain12-15.

Little is known about how TAOK3 exerts its immune functions at the molecular level. Is
the kinase domain necessary for these activities, or is TAOK3 merely required as a
scaffold for other protein interactions? The development of mice that express a mutant
kinase-dead TAOK3 protein and the generation of specific kinase inhibitors will aid in
resolving this question. How is TAOK3 activated, and what are the negative regulators
controlling it? Furthermore, what are the direct substrates of TAOK3 in MZB cells and

182
dendritic cells and how do they control Notch signaling? These and many more
questions will need to be addressed in the future. Here, we will limit ourselves to
discuss the possibility that TAOK3 mediates its effects through A Disintegrin And
Metalloproteinases. Next, we will place the role of TAOK3 in perspective of other factors
implicated in the biology of MZB cells and dendritic cells.

Regulation of ADAM activity

The activation of Notch2 signaling is regulated through proteolytic shedding mediated

by metalloproteinases16. Whereas ADAM10 and ADAM17 were both able to cleave the
Notch2 receptor in vitro, ADAM10 was shown to be the relevant physiological sheddase
in B cells, and conditional ablation of Adam10 abrogated MZB cell development17. We
discovered that ADAM10 activity is controlled by TAOK3, a conclusion underpinned by
several observations. First, ADAM10 was not expressed on the surface of Taok3-
deficient T1 B cells. We did not find any differences in the mRNA and protein levels in
cell lysates from Taok3-/- and littermate control mice, suggesting that TAOK3 regulated
its localization to the cell membrane. Second, proteolytic cleavage of the low affinity IgE
receptor CD23, a substrate of ADAM10 in follicular B cells17, was decreased in Taok3-/-
mice. Finally, shedding of the amyloid precursor protein (APP), another well-known
substrate of ADAM10 involved in the pathogenesis of Alzheimer’s disease18, was equally
reduced in Taok3-/- MEF cells. We assessed whether ADAM10 also controlled Notch2
signaling in dendritic cells, and confirmed a drastic reduction in the number of CD8α+
cDC1s and ESAM+ cDC2s in the spleen of Adam10flfl CD11ccre mice. In support of our
findings, others have shown a functional role of ADAM10 in dendritic cells for the
generation of TH2 responses19. Though off-target effects of CD11c-cre activity have been
described20, 21, our mixed bone marrow experiments clearly confirmed the effects of
ADAM10 to be cell-intrinsic. However, we failed to identify ADAM10 surface expression
by dendritic cells and could therefore not prove directly whether Taok3 also controlled
ADAM10 localization in cDCs. The reason why ADAM10 could not be visualized remains
unclear. It might be that ADAM10 is more rapidly downregulated or digested in cDCs
upon ex vivo cell isolation, or that its visualization is hampered by higher background
fluorescence. Some have also proposed that Notch cleavage by ADAM10 might occur in
endosomal or exosomal compartments22-24, which would shield it from fluorescent
antibodies. Alternatively, ADAM10 might only be expressed transiently during a brief
developmental window, thus escaping our detection.

To study the regulation and biology of ADAM10 in vivo in more detail, the development
of novel tools is warranted. For example, generation of ADAM10-reporter mice could
enhance our understanding of the precise kinetics and subcellular localization of
ADAM10 expression along the developmental trajectory of MZB cells and dendritic cells.
In addition, crossing an ADAM10-reporter allele to the Taok3-/- background would allow
us to identify how and at what stage TAOK3 regulates its expression in different cell
types. Novel insights in how ADAM10 activity can be modulated will benefit not only the

183
field of immunology, as ADAM10 has been implicated in a wide array of diseases, from
Alzheimer’s disease to cancer25-28.

How exactly Taok3 controls ADAM10 bioactivity, will require further study. ADAM10 is
mainly regulated through posttranslational mechanisms, as discussed in depth in the
introduction16. Particularly intriguing is the regulation of ADAM10 by tetraspanins of
the C8 subgroup (TSPAN5, TSPAN10, TSPAN14, TSPAN15, TSPAN17 and TSPAN33),
which are named as such because their large extracellular loop contains eight cysteine
residues29-31. These proteins differentially influence the localization, activity and
substrate specificity of ADAM1032, 33. For example, Notch cleavage is promoted by
TSPAN5 and TSPAN14 but not by TSPAN15 or TSPAN3334. In contrast to the broad
expression pattern of ADAM10, tetraspanin C8 proteins have cell type-specific
expression patterns, which suggests that they can differentially modulate ADAM10
activity in different tissues35. Thus, activation of specific tetraspanin molecules is a
potential mechanism by which TAOK3 might influence the subcellular localisation and
substrate activity of ADAM10. An alternative hypothesis derives from the observation
that phosphorylation of Threonine 735 on ADAM17 regulates its dimerization and
enzymatic activity36. Similarly, ADAM10 contains a conserved threonine in position 719.
Given the presumed role of TAOK3 as a MAP3 kinase, it is very well possible that it
controls ADAM10 dimerization and bioactivity via the MAPK pathway37.

The marginal zone B cell fate choice: an integrative model

The development of splenic MZB cells from immature transitional B cells requires
integration of different signaling pathways. The crucial role of Notch2 signals in the
MZB lineage choice is illustrated by the fact that interference with the DLL1-Notch2
interaction abrogates MZB cell development, while Notch2 gain-of-function mutations
or overexpression drastically increase the number of MZB cells5, 11, 38-42. Second, the
quality and the strength of the BCR repertoire determine whether positive selection of
transitional B cells by self ligands or microbiome-derived ligands leads to deletion, FoB
cell or MZB development43-46. Finally, noncanonical NFκB signals through the BAFFR
maintain MZB homeostasis, while chemokine and integrin signals control their
retention and shuttling behavior47-51.

The prevailing model suggests that strong BCR signals in the follicle promote the
development of transitional B cells into follicular B cells by rendering them
unresponsive to Notch2 ligands. By contrast, weak BCR signaling is thought to promote
MZB development, possibly through enhancing the expression of one or more
components of the Notch2 signaling pathway51-54. However, the exact role of BCR
signaling remains a matter of debate, as positive selection also seems essential for the
generation of MZB cells38, 46. Moreover, at what developmental stage and how exactly
BCR and Notch pathways interact, is not well understood. In chapter III, we have
demonstrated that crosslinking of BCR receptors triggered surface expression of

184
ADAM10 in T1 B cells. ADAM10 was expressed on 5-10% of transitional B cells and
marked commitment to the MZB cell fate. While we show that some T2 B cells can still
differentiate into MZB cells, our findings support the concept that MZB cells mainly
emanate from T1 B cells, rather than from CD23+ T2 B cells or FoB cells5, 17, 43, 53, 55, 56.
Taok3-/- immature B cells failed to bring ADAM10 to the cell surface and did not give
rise to MZB cells, as they could not to respond to Notch2 ligation. In line, others have
shown that a failure to express ADAM10 abrogates MZB cell development17, 57.

The dependency of MZB cells on Notch2 signals is not limited to a single developmental
instruction. Antibody-mediated inhibition of the DLL1 - Notch2 interaction rapidly
depleted these long-lived cells, suggesting that regular Notch2 ligation is required to
maintain their identity58, 59. Confusion has arisen about where Notch2 binding occurs, as
the highest levels of DLL1-expression were detected in red pulp venules using a
reporter mouse41. However, conditional ablation of endothelial Dll1 did not affect the
homeostasis of MZBs, whereas Dll1 expression in white pulp fibroblasts was
indispensable42. Therefore, the shuttling behavior of MZB cells is likely coupled to their
transient exposure to Notch2-ligands in the follicle, although this requires experimental
validation. Our in vitro findings using OP9-DLL1 cells to stimulate transitional B cells
suggest that the MZB cell-deficit of Taok3–/– mice results from the disruption of short-
lived Notch-Notch ligand interactions. Such short-lived interactions might be restricted
in time due to the continuous shuttling behavior of MZB cells between the MZ and B cell
follicles42, 60. Intermittent, brief signals are indeed common in Notch-mediated
processes61. We hypothesize that Taok3 is required to allow rapid responsiveness to
Notch2 ligation, by inducing translocation of ADAM10 to the surface during migration to
the follicle.

Currently, we can only speculate about the upstream factors that activate TAOK3 in B
cells in vivo. Shuttling of MZB cells is controlled through the opposing actions of
sphingosine-1-phosphate (S1P), CXCL13 and shear flow48, 62. Expression of the integrins
LFA-1 and VLA-4 and cannabinoid receptor 2 (CB2) enables retention at the marginal
zone and prevents washing out of MZB cells into the red pulp, a feature that FoB cells do
not possess 47, 49, 50, 63, 64. How the cyclic alterations in integrin expression and
chemokine attraction are controlled, is unclear at present. This process involves the
activation of small GTPases by G proteins downstream of chemokine receptors such as
CXCR5 and S1P1. Activation of the Rho GTPases RhoA, Rac and Cdc42 by G-protein
coupled receptors (GPCR) leads to cytoskeletal rearrangements and is a general feature
of cell movement65-67. In line, mice that are deficient for RhoA, Rac or Cdc42, their
upstream G proteins or one of the many factors that regulate GTPase activity, all display
an abnormal marginal zone B cell compartment57, 68-75. Similarly, mice that lack
Dedicator of cytokinesis 8 (DOCK8), a Rho guanine exchange factor (RhoGEF), fail to
relocalize integrins upon BCR activation and have an MZB cell deficiency76. Members of
the PAK subfamily of Ste20-like kinases are well known effector proteins for the Rho
GTPases Cdc42 and Rac and have been linked to cell motility through the MAPK

185
Figure 1 | Role of TAOK3 in marginal zone B cell development. Binding of positively selecting ligands
to the BCR of immature T1 B cells leads to TAOK3-dependent localization of ADAM10 to the cell surface,
rendering positively selected B cells responsive to Notch2 activation by delta-like ligand 1 (DLL1)
expressed on marginal zone reticular cells in the B cell follicle of the white pulp. The Notch2 intracellular
domain released by ADAM10- and γ-secretase-mediated cleavage binds to the transcription factors
mastermind-like 1 (MAML1) and RBPjκ. Signals resulting from the BAFF-BAFFR interaction on B cells
activate nuclear factor-κB (NF-κB) through the canonical signaling pathway. Together, NF-κB and Notch2
intracellular domain–MAML1–RBPjκ synergistically induce the transcription of genes involved in MZB
cell fate77, 78. The sphingosine 1-phosphate receptor 1 (S1PR1) and integrin signaling contribute to the
shuttling of B cells between the red pulp and the white pulp, and to retention of B cells in the marginal
zone. TSPANC8, tetraspanin C8 family. Adapted, with permission, from ref.16

pathway79-81. GTPases Cdc42 and Rac and have been linked to cell motility through the
MAPK pathway79-81. Similarly, the germinal center kinase PSK is involved in actin
skeleton reorganisation82. However, there is no evidence to date to support activation of
TAOK3 by GPCRs or Rho GTPases. Interestingly, ADAM10 surface expression induced by
BCR crosslinking in transitional B cells depended on the G protein Gαi2, suggesting that
chemokines might control ADAM10 bioactivity57. Therefore, Rho GTPases could link
MZB cell motility and receptivity to Notch signals. Nevertheless, these concepts are
mainly hypothetical and require further investigation. In conclusion, retention and
migration signals might be interrelated with instructive signals, rather than represent

186
independent features of the MZB cell fate. Although it is likely that lack of MAPK
mediated ADAM10 surface expression is the explanation for the MZB phenotype of
Taok3-deficient mice, we also need to consider additional effects of Taok3 on MAPK-
driven activation of NFκB, given that canonical NFκB1 collaborates with Notch2 in
driving MZB fate determination77. Notch2 signals and NFκB1 signals synergize at the
transcriptional level in MZB cells, integrating Notch2 and BAFFR instruction77, 78.
Interestingly, canonical NFκB signals are equally required for the homeostasis of ESAM+
dendritic cells in a cell-intrinsic manner (unpublished results). Nevertheless, it is
unlikely that Taok3 affects the Notch transcriptional program through modulation of
NFκB signals, as p52 activation by BAFF was intact in Taok3-/- B cells.

Role of TAOK3 in dendritic cell development

In chapter IV, we demonstrated a role for Taok3 in the development of ESAM+ CD4+
cDC2s in the spleen. Using conditional gene-trap reversal, conditional knockout and
mixed BM chimera approaches, we proved the function of Taok3 in cDC2s to be cell-
intrinsic and independent of its role in MZB cell development. Interestingly, Taok3 also
controlled ESAM expression in a fraction of splenic cDC1s and in resident lymph node
cDC2s. In addition, the expression of F4/80 and CD172a in cDC2s was diminished in
different tissues in Taok3-/- mice. These findings suggest that Taok3 may exert a broader
function in cDCs across both lymphoid and non-lymphoid tissues. However, it remains
uncertain whether the phenotypic alterations of cDC2s in non-lymphoid tissues are also
cell-intrinsic, or how loss of Taok3 affects the function of these cells. The role of Taok3
seems to concentrate on the terminal differentiation of cDCs. Pre-cDC1s and pre-cDC2s
numbers were increased in the spleen of Taok3-/- mice, implying that the transition to
mature cDCs requires Taok3. Taken together, Taok3 controls the terminal
differentiation of dendritic cells across tissues.

The phenotype of Taok3-deficient cDCs overlaps considerably with mice that lack
expression of Dll1, Notch2 or RBPjκ6, 7, 42. We established that Notch-regulated genes
were affected in Taok3-/- cDCs in vivo, and demonstrated how Taok3-/- BMDCs failed to
respond to DLL1 in vitro. Conversely, overexpression of NICD rescued the development
of ESAM+ cDC2s in Taok3-/- mice. However, as we discussed above, the effect of Taok3-
deficiency is more restricted than that of complete Notch2 absence. One possible
explanation is that Taok3 only controls a part of the downstream action of Notch2 (cfr.
supra). Alternatively, Notch2 activation might require Taok3 only under certain
conditions, or in specific cDC subsets. We speculate here that splenic cDC2s might
shuttle between their bridging channel location and the T cell zone, similar to the
continuous migration of MZB cells. Like MZB cells, positioning of dendritic cells in the
marginal zone bridging channels depends on S1P, while cDCs can also express CXCR5 to
migrate to the white pulp83, 84. Remarkably, deletion of the Rho GTPase RhoA
specifically affected ESAM+ cDC2s in the spleen, suggesting that cell motility might be
required for the homeostasis of this cDC subset85. Intravascular labeling reportedly only

187
stained 60% of cDC2s, suggesting that almost half of all cDC2s were situated in regions
inaccessible to circulating antibody, such as the white pulp86. Intriguingly, many genes
controlled by Taok3 in dendritic cells relate to adhesion and migration. Adhesion
molecules such as Epcam and L1cam were upregulated, while genes that facilitate
migration through extracellular matrix such as Mmp9, Mmp12, Lgals1 and Lgals3 were
downregulated in Taok3-/- cDC2s. L1cam-deficient dendritic cells display impaired
transendothelial migration87. Thus, TAOK3 might be involved in shuttling of cDC2s,
allowing regular access or receptivity to Notch2 ligation in the white pulp. However,
shuttling of dendritic cells is hard to study in vivo, due to the short half-life of cDCs and
the lack of a cDC2-specific reporter88-90. One possible approach would be to perform a
dual antibody labeling experiment48. In such an experiment, intravascular labeling with
two cDC2-specific antibodies is performed with a time gap in between. If cDC2s
exchange continuously between the blood-exposed MZ and the antibody-secluded white
pulp, a proportion of cDC2s that bound the first antibody will no longer be exposed to
the second antibody, whereas sessile cDC2s will all be double positive.

Figure 2 | Molecular determinants of cDC differentiation and heterogeneity in the spleen.

cDC1 development requires the transcription factors IRF8, BATF3, NFIL3 and ID2. A subpopulation of
ESAM+ cDC1s also requires TAOK3 and NOTCH2 signals, but it is unclear to what degree NOTCH2
contributes to the development of ESAM- cDC1s. In the spleen, the development of terminally
differentiated ESAM+ cDC2s from pre-cDC2s depends on NOTCH2, RUNX3, LTβR and RELB. TAOK3 and
ADAM10 control activation of the Notch pathway in a cell-intrinsic manner.

188
Some discrepancies arise from the study of Notch signaling effects in dendritic cells in
vitro versus in vivo. Whereas ESAM+ cDC2s are the population most reliant on Notch
signaling in vivo, expression of ESAM is barely induced on DCs generated in vitro91. On
the other hand, Taok3 is required for the efficient production of cDC1s in vitro, but
Taok3-/- mice harbor normal numbers of splenic cDC1s. We hypothesize that in vivo, the
development of (ESAM-) cDC1s is largely independent of Taok3. Apart from Notch2-
ligation, the generation of true ESAM+ cDC2s probably requires additional signals in a
Taok3-dependent manner. Given their important role in the homeostasis of splenic
cDC2s, lymphotoxin α1β2, oxysterols and retinoic acid derivatives are all attractive
possibilities7, 89, 92-96. Preliminary experiments indeed suggest that lymphotoxin α1β2
and retinoic acid cooperatively stimulate development of ESAM+ cDC2s in OP9-DL1
cocultures (unpublished results). How these different pathways promote the
development of Notch2-dependent cDC2s, and whether Taok3 is required for these
effects, remain unanswered questions.

While type 1 dendritic cells are generally thought to be homogeneous in their ontogeny
and transcription factor dependence, evidence points to some heterogeneity within the
cDC1 lineage97, 98. Taok3 and Notch2 control the presence of ESAM+ cDC1s in spleen7.
However, it is unclear whether these cells truly represent a distinct subset, or merely
reflect a functional state of cDC1s. It is now clear that a subset of cDC1s, expressing
langerin, also populates the marginal zone98-100. It should be further investigated how
langerin and ESAM expression relate in cDC1s, whether ESAM+ cDC1s locate in the
marginal zone or the T cell zone, and whether Taok3 affects langerin+ cDC1s in the
marginal zone. Interestingly, it was reported that retinoic acid promotes development
of langerin+ cDC1s through the transcription factor RUNX3101. Retinoic acid and RUNX3
are equally involved in the development of ESAM+ cDC2s, suggesting a common
regulatory pathway95, 96, 102.

Role of LTβR in cDC2 homeostasis

In chapter V, we have demonstrated that antagonism of the LTβR in adult mice

selectively reduced ESAM+ cDC2s in the spleen. This is in agreement with epistasis
experiments showing that Notch2 and LTβR deletions target the same cDC population7.
We did not study whether other Notch2-dependent dendritic cells were affected in our
model. However, others have illustrated a role for LTβR in intestinal CD103+ cDC2s
using competitive BM chimerism7. While it is generally thought that trophic LTβR
promotes local proliferation of differentiated cDC2s89, 103, a developmental role cannot
be excluded. How the Notch and LTβR pathways intersect, remains unclear. In Notch2-/-
and Taok3-/- cDC2s, the surface expression of LTβR was preserved6. Of course, Notch2-
instruction might control the exposure to lymphotoxin α1β2. However, treatment with
an LTβR agonist could not mitigate the absence of ESAM+ cDC2s in Taok3-/- or Adam10flfl
CD11ccre mice (unpublished results). Inversely, LTβR stimulation could trigger

189
Figure 3 | Molecular interactions controlling the development and homeostasis of splenic cDC2s.
When circulating pre-cDC2s arrive in the spleen, they undergo terminal differentiation and functional
maturation in response to a variety of instructive signals. Activation of the Notch pathway relies on
binding of the NOTCH2 receptor to DLL1 expressed by stromal cells and subsequent receptor cleavage by
the metalloproteinase ADAM10, leading to nuclear translocation of NICD. Notch activation is controlled
by the kinase TAOK3, possibly through regulation of ADAM10 bioactivity. The transcription factor RUNX3
similarly controls the expression of canonical Notch target genes, probably through interactions with
RBPjκ. ILC3- or B cell-derived lymphotoxinα1β2 provides homeostatic signals through the LTβR, which
activates botch canonical (p50) and noncanonical (p52) NFκB pathways. The positioning of cDC2s in
proximity to DLL1 and LTα1β2 is controlled by the oxysterol receptor EBI2. How the Notch and LTβR
pathways interact, remains enigmatic. For the molecules highlighted in bold, a cell-intrinsic role in the
homeostasis of splenic cDC2s has been validated in vivo.

190
responsiveness to Notch ligation. To pursue this idea, it should be validated whether
supplementing OP9-DL1 BMDC cocultures with a LTβR agonist enhances the generation
of ESAM+ DC2s in vitro. Downstream, LTβR is known to induce both the canonical and
non-canonical NFκB pathway in stromal cells via the activation of NFκB-inducing kinase
(NIK)104, 105. We have confirmed that LTβR-agonism activates both p52 and p50 in
BMDCs. Moreover, we found that p50 (Nfkb1) is required for the development of ESAM+
cDC2s in a cell-intrinsic manner (unpublished results). Given the observation that p50
and Notch2 synergistically control MZB cell instruction77, it is tempting to speculate that
LTβR and Notch2 signals converge and cooperatively activate a transcriptional program
that controls the development of ESAM+ cDC2s.

Role of ILC3s in cDC2 homeostasis

Challenging the concept that B cells constitute an essential source of lymphotoxin, we

found that the differentiation of splenic cDCs was not altered in mice that constitutively
lack B cells. Despite a decrease in the absolute cDC number, a normal fraction of cDCs
was able to acquire a terminally differentiated ESAM+ CD4+ phenotype in Rag2-/-, mMT
and Mb1-DTA mice. Instead, we have revealed that ILC3s are essential and sufficient for
the homeostasis of ESAM+ cDC2s in lymphopenic mice. This is in apparent contradiction
with the findings of Kabashima et al., who found a reduction in CD4+ DCs in the spleen of
irradiated WT mice reconstituted with mixed μMT:Lta-/- bone marrow89. Potentially, the
use of a BM transplantation model and irradiation might have led to inadvertent
depletion of ILCs, whose recovery can take a long time106. Alternatively, the functions of
ILC3s might be redundant in mice that have a normal B cell population. Studying the
population of ESAM+ cDC2s in Rorc-cre x LTβflfl mice on both a wild type and Rag-/-
background, or in Rorγt-/- : Lta-/-> WT BM chimeras (analyzed at later time points after
reconstitution) might solve this issue.

We have not formally shown that the requirement for ILC3s was dependent on their
intrinsic expression of LTα1β2. Still, treatment with an LTβR-agonist rescued the
development of ESAM+ cDC2s in Rag-/- Il2rg-/- mice, and other actions of ILC3-derived
LTα1β2 support this idea. LTi-derived lymphotoxin instructs the development of
lymphoid tissues during embryogenesis and in postnatal life104, 107. In the intestinal
lamina propria, the interaction between LTα1β2 in ILC3s and LTβR in cDCs controlled
the production of IL-23 by dendritic cells in the intestine108, 109. Moreover, ILC3-derived
LTα1β2 was shown to be crucial to stimulate the LTβR on cDC2s in order to induce
isotype switching to IgA in B cells in Peyer’s patches109, 110. The presence of ILC3s in the
splenic marginal zone has been highlighted by others111. Thus, ILC3s and cDC2s occupy
a conserved niche in gut-associated lymphoid tissue and in the splenic marginal zone112.
Recently, some light was shed on potential factors that could regulate access to this
niche. The chemokine receptor EBI2 (Gpr183) is required for the localization of ILC3s in
intestinal cryptopatches and isolated lymphoid follicles113, 114. Interestingly, splenic

191
cDC2s also required EBI2 expression for correct positioning in splenic bridging
channels and at the B:T zone interface92, 93. Marginal reticular cells express high levels of
Ch25h, the enzyme required for the synthesis of the main EBI2 ligand, 7α, 25-
dihydrocholesterol 115. Strikingly, the loss of CD4+ splenic cDC2s in EBI2-deficient mice
could be rescued by LTβR-agonism, implying that EBI2 controls the access of cDCs to
sources of LTα1β2 92. Whether ILC3s use EBI2 to localize in the splenic marginal zone is
not known, but we hypothesize that the EBI2 receptor might congregate cDC2s and
ILC3s in a common niche for LTβR activation.

Intestinal ILC3s rely on activation of the retinoic acid receptor RARα in a cell-intrinsic
manner116. Similarly, fetal LTi cells were controlled by maternal retinoids117. While it is
unclear how retinoic acid affects splenic ILC3s, and whether it regulates LTβR signals,
this might provide an alternative explanation for the lack of ESAM+ cDC2s in mice
deprived of dietary retinoic acid.

Lost in translation: extrapolation to the human immune system

The microanatomical structure of the white pulp in human is poorly understood. The
differences between MZB cells in mice and humans in particular raise questions about
the translatability of our findings. In humans, the term MZB cell applies to a pool of
recirculating memory B cells that populate the outer border of follicles. In contrast,
murine MZB cells are unique to the spleen, circumscribe the whole white pulp and
harbor fewer mutations in immunoglobulin genes52, 118, 119. Gene expression analysis of
human IgM+ B cells derived from spleen and peripheral blood at the single cell level
could help to elucidate whether human MZB cells derived from different organs are
truly homogeneous. It is conceivable that a subset of B cells exists in human spleen that
does not recirculate and displays a more ‘naive’ phenotype, corresponding to murine
MZB cells. Remarkably, splenic marginal B cell lymphoma cells frequently harbor
mutations in NOTCH2 and genes related to the NFκB pathway120-122. Patients suffering
from Alagille syndrome, a rare inherited disease caused by NOTCH2 haploinsufficiency,
have a reduced number of MZB cells123. Moreover, a marginal zone B cell precursor
from human spleen differentiated into MZB-like cells when cultured in the presence of
DLL1123. While these findings imply a role for NOTCH2 in the development of human
MZB cells in the spleen, it is unknown whether the human ADAM10 and TAOK3
homologs are required for MZB development.

Despite the confusing nomenclature, the perifollicular zone in humans might be the true
equivalent of the rodent marginal zone. While the existence of a marginal sinus in
humans remains controversial, the perifollicular zone receives afferent blood at the
border between the white and red pulp124-129. The failure to identify a marginal sinus
might be due to the fact that some authors did not include the perifollicular region in
their histological analysis130. The MadCAM1+ stromal cells surrounding the follicle bear

192
resemblance to the MRC network in rodents and might be the source of DLL1 that is
expressed in this region100, 123. In addition, SIGN+ and SIGLEC1+ macrophages have been
identified in the perifollicular zone, though it is not clear whether they are organized in
a similar layered structure100. Similar to observations in mice, ILC3s and B-helper
neutrophils were found in the perifollicular zone in humans111, 131. We speculate that the
perifollicular dendritic cells identified by staining for CD205 correlate to langerin+
cDC1s in the rodent marginal zone98, 100. By contrast, little is known about the cDC2
compartment in human spleen. Whether human cDC2s express ESAM and rely on
signals through Notch2, LTβR and EBI2 receptors, remains enigmatic. Some reports
have also described a thin line of IgM+ IgD- CD27- B cells lining the PALS111. If future
studies should reveal a ‘rodent MZB-like’ subset within CD27+ IgM+ human MZB cells
that can be distinguished by novel surface markers derived from gene-expression
analyses, it would be interesting to assess whether this subset is preferentially localized
in the ‘marginal’ or perifollicular zone.

An unconventional role for dendritic cells in controlling myeloproliferation

Chronic myeloid leukemia (CML) is a myeloproliferative neoplasm that originates from

pluripotent hematopoietic stem cells carrying the BCR-ABL fusion gene. The
development of imatinib, a tyrosine kinase inhibitor (TKI) directed against the
constitutively active fusion protein, has drastically improved the prognosis of CML
patients132. However, lifelong treatment is necessary and many patients ultimately
develop resistance against TKIs, owing to the failure of these treatments to target
quiescent leukemic stem cells133, 134. The only curative therapy at present is allogeneic
stem cell transplantation, but its application is limited by the availability of suitable
donors and the risk of graft-versus-host disease135. Still, it demonstrates the potential of
immunotherapy to fully eradicate disease in CML patients. In chapter VI, we put
forward the hypothesis that impaired immunosurveillance in CML is crucial for the
development of disease. Indeed, CML cells are immunogenic, but fail to induce effective
cytotoxic responses in vivo136-138. We speculated that this might be due, at least in part,
to the abnormal dendritic cell compartment of these patients. Dendritic cells from CML
patients were less capable of capturing, processing and presenting antigen, and
displayed impaired migration ex vivo139. Accordingly, several reports have
demonstrated that BCR-ABL+ DCs fail to efficiently induce CD4+ or CD8+ T cell responses
in vitro140-143. While CD4+ T cells specific for a BCR-ABL fusion peptide were identied in
peripheral blood, they failed to recognize dendritic cells from patients with CML136.
These findings may explain why immunotherapy using autologous DCs yielded no
clinical responses144.

In line with previous reports, we found that mice lacking the transcription factor IRF8
develop a myeloproliferative disorder that strongly resembles CML145-147. In fact, the
effects of BCR-ABL appear to be mediated to a large extent by repression of IRF8148-150.
Moreover, overexpression of IRF8 in BCR-ABL-transduced myeloid precursor cells

193
arrested their growth in vitro and prohibited disease development in a mouse model of
CML151-153. IRF8 also governs the development, terminal differentiation and function of
cDCs in mice and humans154-158 and thus, the functional impairment of DCs in CML
patients could be explained by the suppression of IRF8. Whereas the phenotype,
function and transcriptome of BCR-ABL+ DCs and Irf8-/- DCs have not yet been
compared directly, it is likely that both genetic alterations lead to similar effects.

Using a mixed bone marrow chimera approach, we demonstrated that development of

MPD in Irf8-/- mice requires a double hit: 1) loss of Irf8 in myeloid progenitor cells; and
2) lack of control of the Irf8-deficient progenitors by loss of Irf8 in a CD11c+ population.
The fact that Irf8-deficiency in hematopoietic cells that do not express Itgax (CD11c) is
required for MPD to develop, may not seem surprising at first sight, as
myeloproliferative diseases are generally regarded to be caused by cell-intrinsic
derangements in pluripotent hematopoietic stem cells. It is well-accepted that IRF8
suppresses the development of neutrophils from myeloid precursors in a cell-intrinsic
manner159-162. In addition, the accumulation of neutrophils in Irf8-/- mice is enhanced by
the decreased susceptibility of Irf8-/- granulocytes to apoptosis151, 163-166. Nevertheless,
others have claimed that the onset of MPD originated from myeloid progenitor-extrinsic
effects. It was proposed that Irf8-/- mice develop MPD due to overproduction of GM-CSF
by Irf8-/- T cells167. In addition, conditional ablation of Irf8 in myeloid cells did not
increase neutrophil counts, although these experiments relied on the use of Lyz-cre,
which might act too late and generally displays a low targeting efficiency20, 156, 167, 168.
Another report suggested that MPD in Irf8-/- mice was caused by increased Flt3L levels
in response to their cDC1 deficit146. However, this hypothesis is not compatible with our
observation that intact Irf8flfl x CD11ccre mice, that completely lack cDC1s, do not
develop full-blown MPD.

The ability of syngeneic bone marrow to prevent MPD is remarkable, and goes against
the prevailing view that myeloproliferative diseases are caused solely by intrinsic
effects in myeloid precursors that lead to clonal hyperproliferation. The fact that CD11c-
DTA and Irf8flfl x CD11ccre BM did not prevent MPD, suggests that a functional or
numerical deficit in CD11c+ cells allows immune escape of abnormal granulocyte
progenitors. While these results implied a role for dendritic cells in the
immunosurveillance, the use of CD11c-cre comes with the caveat that off-target effects
on other immune cells, such as NK cells and T cells, are common20. Furthermore, the
cytotoxic responses of T cells and NK cells from CML patients are impaired, while
allogeneic NK cells have been shown to control the development of disease in a mouse
model of CML169-171. However, using Rag-/- and NKp46-DTA bone marrow, we
demonstrated that the Irf8-dependent CD11c+ cells suppressing MPD are not B-, T- or
NK cells, arguing against the proposition that the MPD in Irf8-/- mice is secondary to T-
cell intrinsic effects167. Nevertheless, this does not preclude a (Irf8-independent)
contribution by these lymphocytes in the containment of MPD. In our model of Irf8-/-
driven MPD, it is possible that the protective effect from Rag2-/- bone marrow was

194
caused by interactions between Rag2-/- dendritic cells and Irf8-/- T cells. The
development of MPD in mice that lack MHCII expression in DCs depends on the
presence of T cells, supporting such a mechanism172. Alternatively, Irf8-/- DCs might
hamper cytotoxic responses by NK cells. Dendritic cells have been shown to affect the
function of NK cells, though it is not known whether this requires IRF8173, 174.

Surprisingly, protection from MPD was independent of the presence of cDC1s. Neither
Irf8-/-:Irf8flfl XCR1cre BM chimeric mice nor unmanipulated Irf8flfl XCR1cre mice
developed myeloproliferative disease, providing further evidence against the
hypothesis that the mere absence of cDC1s can trigger MPD146. By antibody-mediated
depletion and through the use of Zeb2flfl CD11ccre BM, we also demonstrated the
redundancy of plasmacytoid dendritic cells for this myelosuppressive effect. These
findings are in line with the observation that BXH2 mice, that harbor a mutation in IRF8
that does not affect pDC development, still develop splenomegaly147, 175. However,
neither of the methods we used completely ablated pDCs. Ablation of pDCs through
knockin of the human BDCA2 promotor coupled to the diphtheria toxin receptor
(BDCA2-DTR) has been described, but this model may not be suited for long-term
depletion176, 177.

The role of cDC2s, macrophages and monocytes should be addressed further.

Preliminary data suggest a hitherto unappreciated role of IRF8 in the migration and
function of cDC2s (Bosteels et al., unpublished results) that could explain why loss of
Irf8 in CD11c+ XCR1- cells affects immunosurveillance. While Taok3-/- and Zeb2flfl x
CD11ccre BM were fully capable of suppressing MPD, the number of cDC2s is only
partially decreased in these models, and Irf8-dependent functions might have been
preserved in remaining cDC2s. Until genetic tools become available to target the cDC2
lineage more efficiently, pinpointing this function to cDC2s requires exclusion of all
alternative possibilities. The development of MPD in Irf8-/-:Irf8flfl x CD64cre BM
chimeras suggests the intriguing possibility that macrophages, monocytes or monocyte-
derived dendritic cells might keep abnormal myelopoiesis in check. While monocyte
development is arrested at the common monocyte precursor (cMoP) stage in Irf8-/-
mice, the little is known about the function of IRF8 in mature monocytes or monocyte-
derived dendritic cells156, 167, 178. By contrast, several studies support a role for IRF8 in
both the development and function of tissue-resident macrophages179-185. Mainly, IRF8
was implicated in the regulation of autophagy, antigen presentation and the production
of IL-12 in the defense against intracellular pathogens182, 183, 186. Interestingly,
conditional deletion of Irf8 in macrophages impaired their tumor-suppressive
capacity168. It has been reported that the use of CD64-cre confers significant off-target
effects187. The targeting efficiency of CD64-cre in different hematopoietic cell types in
our bone marrow chimeric model should be carefully examined using IRF8 staining.
Especially, off-target excision of Irf8 in cDC2s should be excluded before drawing
conclusions about a role of monocyte-derived cells in myelopoiesis. Moreover, it should
be examined whether intact Irf8flfl x CD64cre mice also develop MPD or only display

195
subtle effects as in Irf8flfl x CD11ccre mice. Also, further investigations are necessary to
delineate how the development and function of cells of the monocyte/macrophage
lineage is affected in Irf8flfl CD11ccre mice.

In conclusion, IRF8 controls an unexpected function of cDC2s, CD64+ monocyte-derived

cells (MCs) or macrophages in the regulation of hematopoiesis. Disruption of this
function might explain why CML cells escape immunosurveillance. The observation that
Irf8-overexpressing BCR-ABL transformed cells were able to induce cytotoxic T cell
responses, prevent and even reverse MPD in vivo, demonstrates the therapeutic
potential of modulating this antitumor immune response188. However, many questions
still need to be addressed. First, future studies should address whether co-
transplantation of wild type BM can also inhibit MPD development from BCR-ABL-
transduced mouse BM. Next, it should be assessed to what extent the mononuclear
phagocyte compartment of CML patients resembles that of Irf8-deficient mice. The
phenotype, function, and gene expression of BCR-ABL+ dendritic cells, monocyte-
derived cells and macrophages should be compared with Irf8-/- counterparts. Using
competitive BM chimerism and conditional knockout mice, the precise function of IRF8
in cDC2s, MCs and macrophages should be thoroughly investigated. To establish
whether the surveillance effects of cDCs/MCs are mediated by T or NK cells, Irf8-/- :
Irf8flfl x CD11ccre > WT mixed bone marrow chimeric mice can be treated with depleting
antibodies. By carefully dissecting the precise cellular interactions by which the
immune system surveils myelopoiesis, new therapeutic avenues might open up for
patients that are resistant to current treatment regimens.

196
REFERENCES

1. Yustein JT, Li D, Robinson D, Kung HJ. KFC, a Ste20-like kinase with mitogenic potential and capability to
activate the SAPK/JNK pathway. Oncogene. 2000;19(5):710-8.
2. Yustein JT, Xia L, Kahlenburg JM, Robinson D, Templeton D, Kung HJ. Comparative studies of a new
subfamily of human Ste20-like kinases: homodimerization, subcellular localization, and selective activation of MKK3
and p38. Oncogene. 2003;22(40):6129-41.
3. Tassi E, Biesova Z, Di Fiore PP, Gutkind JS, Wong WT. Human JIK, a novel member of the STE20 kinase family
that inhibits JNK and is negatively regulated by epidermal growth factor. J Biol Chem. 1999;274(47):33287-95.
4. Zhang W, Chen T, Wan T, He L, Li N, Yuan Z, et al. Cloning of DPK, a novel dendritic cell-derived protein
kinase activating the ERK1/ERK2 and JNK/SAPK pathways. Biochem Biophys Res Commun. 2000;274(3):872-9.
5. Saito T, Chiba S, Ichikawa M, Kunisato A, Asai T, Shimizu K, et al. Notch2 is preferentially expressed in
mature B cells and indispensable for marginal zone B lineage development. Immunity. 2003;18(5):675-85.
6. Lewis KL, Caton ML, Bogunovic M, Greter M, Grajkowska LT, Ng D, et al. Notch2 receptor signaling controls
functional differentiation of dendritic cells in the spleen and intestine. Immunity. 2011;35(5):780-91.
7. Satpathy AT, Briseno CG, Lee JS, Ng D, Manieri NA, Kc W, et al. Notch2-dependent classical dendritic cells
orchestrate intestinal immunity to attaching-and-effacing bacterial pathogens. Nat Immunol. 2013;14(9):937-48.
8. Hamada Y, Kadokawa Y, Okabe M, Ikawa M, Coleman JR, Tsujimoto Y. Mutation in ankyrin repeats of the
mouse Notch2 gene induces early embryonic lethality. Development. 1999;126(15):3415-24.
9. Gamrekelashvili J, Giagnorio R, Jussofie J, Soehnlein O, Duchene J, Briseno CG, et al. Regulation of monocyte
cell fate by blood vessels mediated by Notch signalling. Nat Commun. 2016;7:12597.
10. Caton ML, Smith-Raska MR, Reizis B. Notch-RBP-J signaling controls the homeostasis of CD8- dendritic cells
in the spleen. J Exp Med. 2007;204(7):1653-64.
11. Tanigaki K, Han H, Yamamoto N, Tashiro K, Ikegawa M, Kuroda K, et al. Notch-RBP-J signaling is involved in
cell fate determination of marginal zone B cells. Nat Immunol. 2002;3(5):443-50.
12. Shin HM, Minter LM, Cho OH, Gottipati S, Fauq AH, Golde TE, et al. Notch1 augments NF-kappaB activity by
facilitating its nuclear retention. EMBO J. 2006;25(1):129-38.
13. Hodkinson PS, Elliott PA, Lad Y, McHugh BJ, MacKinnon AC, Haslett C, et al. Mammalian NOTCH-1 activates
beta1 integrins via the small GTPase R-Ras. J Biol Chem. 2007;282(39):28991-9001.
14. Kwon C, Cheng P, King IN, Andersen P, Shenje L, Nigam V, et al. Notch post-translationally regulates beta-
catenin protein in stem and progenitor cells. Nat Cell Biol. 2011;13(10):1244-51.
15. Demehri S, Liu Z, Lee J, Lin MH, Crosby SD, Roberts CJ, et al. Notch-deficient skin induces a lethal systemic B-
lymphoproliferative disorder by secreting TSLP, a sentinel for epidermal integrity. PLoS Biol. 2008;6(5):e123.
16. Lambrecht BN, Vanderkerken M, Hammad H. The emerging role of ADAM metalloproteinases in immunity.
Nat Rev Immunol. 2018;18(12):745-58.
17. Gibb DR, El Shikh M, Kang DJ, Rowe WJ, El Sayed R, Cichy J, et al. ADAM10 is essential for Notch2-dependent
marginal zone B cell development and CD23 cleavage in vivo. J Exp Med. 2010;207(3):623-35.
18. De Strooper B, Vassar R, Golde T. The secretases: enzymes with therapeutic potential in Alzheimer disease.
Nat Rev Neurol. 2010;6(2):99-107.
19. Damle SR, Martin RK, Cockburn CL, Lownik JC, Carlyon JA, Smith AD, et al. ADAM10 and Notch1 on murine
dendritic cells control the development of type 2 immunity and IgE production. Allergy. 2018;73(1):125-36.
20. Abram CL, Roberge GL, Hu Y, Lowell CA. Comparative analysis of the efficiency and specificity of myeloid-
Cre deleting strains using ROSA-EYFP reporter mice. J Immunol Methods. 2014;408:89-100.
21. Srinivas S, Watanabe T, Lin CS, William CM, Tanabe Y, Jessell TM, et al. Cre reporter strains produced by
targeted insertion of EYFP and ECFP into the ROSA26 locus. BMC Dev Biol. 2001;1:4.
22. Mathews JA, Gibb DR, Chen BH, Scherle P, Conrad DH. CD23 Sheddase A disintegrin and metalloproteinase
10 (ADAM10) is also required for CD23 sorting into B cell-derived exosomes. J Biol Chem. 2010;285(48):37531-41.
23. Stoeck A, Keller S, Riedle S, Sanderson MP, Runz S, Le Naour F, et al. A role for exosomes in the constitutive
and stimulus-induced ectodomain cleavage of L1 and CD44. Biochem J. 2006;393(Pt 3):609-18.
24. Chastagner P, Rubinstein E, Brou C. Ligand-activated Notch undergoes DTX4-mediated ubiquitylation and
bilateral endocytosis before ADAM10 processing. Sci Signal. 2017;10(483).
25. Wetzel S, Seipold L, Saftig P. The metalloproteinase ADAM10: A useful therapeutic target? Biochim Biophys
Acta Mol Cell Res. 2017;1864(11 Pt B):2071-81.
26. Marcello E, Borroni B, Pelucchi S, Gardoni F, Di Luca M. ADAM10 as a therapeutic target for brain diseases:
from developmental disorders to Alzheimer's disease. Expert Opin Ther Targets. 2017;21(11):1017-26.

197
27. Atapattu L, Saha N, Chheang C, Eissman MF, Xu K, Vail ME, et al. An activated form of ADAM10 is tumor
selective and regulates cancer stem-like cells and tumor growth. J Exp Med. 2016;213(9):1741-57.
28. Moss ML, Stoeck A, Yan W, Dempsey PJ. ADAM10 as a target for anti-cancer therapy. Curr Pharm Biotechnol.
2008;9(1):2-8.
29. Prox J, Willenbrock M, Weber S, Lehmann T, Schmidt-Arras D, Schwanbeck R, et al. Tetraspanin15 regulates
cellular trafficking and activity of the ectodomain sheddase ADAM10. Cell Mol Life Sci. 2012;69(17):2919-32.
30. Dornier E, Coumailleau F, Ottavi JF, Moretti J, Boucheix C, Mauduit P, et al. TspanC8 tetraspanins regulate
ADAM10/Kuzbanian trafficking and promote Notch activation in flies and mammals. J Cell Biol. 2012;199(3):481-96.
31. Haining EJ, Yang J, Bailey RL, Khan K, Collier R, Tsai S, et al. The TspanC8 subgroup of tetraspanins interacts
with A disintegrin and metalloprotease 10 (ADAM10) and regulates its maturation and cell surface expression. J Biol
Chem. 2012;287(47):39753-65.
32. Matthews AL, Szyroka J, Collier R, Noy PJ, Tomlinson MG. Scissor sisters: regulation of ADAM10 by the
TspanC8 tetraspanins. Biochem Soc Trans. 2017;45(3):719-30.
33. Saint-Pol J, Eschenbrenner E, Dornier E, Boucheix C, Charrin S, Rubinstein E. Regulation of the trafficking
and the function of the metalloprotease ADAM10 by tetraspanins. Biochem Soc Trans. 2017;45(4):937-44.
34. Jouannet S, Saint-Pol J, Fernandez L, Nguyen V, Charrin S, Boucheix C, et al. TspanC8 tetraspanins
differentially regulate the cleavage of ADAM10 substrates, Notch activation and ADAM10 membrane
compartmentalization. Cell Mol Life Sci. 2016;73(9):1895-915.
35. Noy PJ, Yang J, Reyat JS, Matthews AL, Charlton AE, Furmston J, et al. TspanC8 Tetraspanins and A
Disintegrin and Metalloprotease 10 (ADAM10) Interact via Their Extracellular Regions: EVIDENCE FOR DISTINCT
BINDING MECHANISMS FOR DIFFERENT TspanC8 PROTEINS. J Biol Chem. 2016;291(7):3145-57.
36. Xu P, Liu J, Sakaki-Yumoto M, Derynck R. TACE activation by MAPK-mediated regulation of cell surface
dimerization and TIMP3 association. Sci Signal. 2012;5(222):ra34.
37. Kapfhamer D, King I, Zou ME, Lim JP, Heberlein U, Wolf FW. JNK pathway activation is controlled by
Tao/TAOK3 to modulate ethanol sensitivity. PLoS One. 2012;7(12):e50594.
38. Hampel F, Ehrenberg S, Hojer C, Draeseke A, Marschall-Schroter G, Kuhn R, et al. CD19-independent
instruction of murine marginal zone B-cell development by constitutive Notch2 signaling. Blood. 2011;118(24):6321-
31.
39. Yu J, Zanotti S, Walia B, Jellison E, Sanjay A, Canalis E. The Hajdu Cheney Mutation Is a Determinant of B-Cell
Allocation of the Splenic Marginal Zone. Am J Pathol. 2018;188(1):149-59.
40. Kuroda K, Han H, Tani S, Tanigaki K, Tun T, Furukawa T, et al. Regulation of marginal zone B cell
development by MINT, a suppressor of Notch/RBP-J signaling pathway. Immunity. 2003;18(2):301-12.
41. Tan JB, Xu K, Cretegny K, Visan I, Yuan JS, Egan SE, et al. Lunatic and manic fringe cooperatively enhance
marginal zone B cell precursor competition for delta-like 1 in splenic endothelial niches. Immunity. 2009;30(2):254-
63.
42. Fasnacht N, Huang HY, Koch U, Favre S, Auderset F, Chai Q, et al. Specific fibroblastic niches in secondary
lymphoid organs orchestrate distinct Notch-regulated immune responses. J Exp Med. 2014;211(11):2265-79.
43. Carey JB, Moffatt-Blue CS, Watson LC, Gavin AL, Feeney AJ. Repertoire-based selection into the marginal
zone compartment during B cell development. J Exp Med. 2008;205(9):2043-52.
44. Chen X, Martin F, Forbush KA, Perlmutter RM, Kearney JF. Evidence for selection of a population of multi-
reactive B cells into the splenic marginal zone. Int Immunol. 1997;9(1):27-41.
45. Martin F, Kearney JF. Positive selection from newly formed to marginal zone B cells depends on the rate of
clonal production, CD19, and btk. Immunity. 2000;12(1):39-49.
46. Wen L, Brill-Dashoff J, Shinton SA, Asano M, Hardy RR, Hayakawa K. Evidence of marginal-zone B cell-
positive selection in spleen. Immunity. 2005;23(3):297-308.
47. Cinamon G, Matloubian M, Lesneski MJ, Xu Y, Low C, Lu T, et al. Sphingosine 1-phosphate receptor 1
promotes B cell localization in the splenic marginal zone. Nat Immunol. 2004;5(7):713-20.
48. Cinamon G, Zachariah MA, Lam OM, Foss FW, Jr., Cyster JG. Follicular shuttling of marginal zone B cells
facilitates antigen transport. Nat Immunol. 2008;9(1):54-62.
49. Lu TT, Cyster JG. Integrin-mediated long-term B cell retention in the splenic marginal zone. Science.
2002;297(5580):409-12.
50. Muppidi JR, Arnon TI, Bronevetsky Y, Veerapen N, Tanaka M, Besra GS, et al. Cannabinoid receptor 2
positions and retains marginal zone B cells within the splenic marginal zone. J Exp Med. 2011;208(10):1941-8.
51. Pillai S, Cariappa A. The follicular versus marginal zone B lymphocyte cell fate decision. Nat Rev Immunol.
2009;9(11):767-77.
52. Pillai S, Cariappa A, Moran ST. Marginal zone B cells. Annu Rev Immunol. 2005;23:161-96.

198
53. Cariappa A, Tang M, Parng C, Nebelitskiy E, Carroll M, Georgopoulos K, et al. The follicular versus marginal
zone B lymphocyte cell fate decision is regulated by Aiolos, Btk, and CD21. Immunity. 2001;14(5):603-15.
54. Casola S, Otipoby KL, Alimzhanov M, Humme S, Uyttersprot N, Kutok JL, et al. B cell receptor signal strength
determines B cell fate. Nat Immunol. 2004;5(3):317-27.
55. Srivastava B, Quinn WJ, 3rd, Hazard K, Erikson J, Allman D. Characterization of marginal zone B cell
precursors. J Exp Med. 2005;202(9):1225-34.
56. Dammers PM, de Boer NK, Deenen GJ, Nieuwenhuis P, Kroese FG. The origin of marginal zone B cells in the
rat. Eur J Immunol. 1999;29(5):1522-31.
57. Hwang IY, Boularan C, Harrison K, Kehrl JH. Galphai Signaling Promotes Marginal Zone B Cell Development
by Enabling Transitional B Cell ADAM10 Expression. Front Immunol. 2018;9:687.
58. Sekine C, Moriyama Y, Koyanagi A, Koyama N, Ogata H, Okumura K, et al. Differential regulation of splenic
CD8- dendritic cells and marginal zone B cells by Notch ligands. Int Immunol. 2009;21(3):295-301.
59. Simonetti G, Carette A, Silva K, Wang H, De Silva NS, Heise N, et al. IRF4 controls the positioning of mature B
cells in the lymphoid microenvironments by regulating NOTCH2 expression and activity. J Exp Med.
2013;210(13):2887-902.
60. Arnon TI, Horton RM, Grigorova IL, Cyster JG. Visualization of splenic marginal zone B-cell shuttling and
follicular B-cell egress. Nature. 2013;493(7434):684-8.
61. Ambros V. Cell cycle-dependent sequencing of cell fate decisions in Caenorhabditis elegans vulva precursor
cells. Development. 1999;126(9):1947-56.
62. Tedford K, Steiner M, Koshutin S, Richter K, Tech L, Eggers Y, et al. The opposing forces of shear flow and
sphingosine-1-phosphate control marginal zone B cell shuttling. Nat Commun. 2017;8(1):2261.
63. Guinamard R, Okigaki M, Schlessinger J, Ravetch JV. Absence of marginal zone B cells in Pyk-2-deficient mice
defines their role in the humoral response. Nat Immunol. 2000;1(1):31-6.
64. Karlsson MC, Guinamard R, Bolland S, Sankala M, Steinman RM, Ravetch JV. Macrophages control the
retention and trafficking of B lymphocytes in the splenic marginal zone. J Exp Med. 2003;198(2):333-40.
65. Etienne-Manneville S, Hall A. Rho GTPases in cell biology. Nature. 2002;420(6916):629-35.
66. Ridley AJ, Schwartz MA, Burridge K, Firtel RA, Ginsberg MH, Borisy G, et al. Cell migration: integrating
signals from front to back. Science. 2003;302(5651):1704-9.
67. Hu J, Strauch P, Rubtsov A, Donovan EE, Pelanda R, Torres RM. Lsc activity is controlled by oligomerization
and regulates integrin adhesion. Mol Immunol. 2008;45(7):1825-36.
68. Rieken S, Sassmann A, Herroeder S, Wallenwein B, Moers A, Offermanns S, et al. G12/G13 family G proteins
regulate marginal zone B cell maturation, migration, and polarization. J Immunol. 2006;177(5):2985-93.
69. Chopin M, Quemeneur L, Ripich T, Jessberger R. SWAP-70 controls formation of the splenic marginal zone
through regulating T1B-cell differentiation. Eur J Immunol. 2010;40(12):3544-56.
70. Zhang S, Zhou X, Lang RA, Guo F. RhoA of the Rho family small GTPases is essential for B lymphocyte
development. PLoS One. 2012;7(3):e33773.
71. Guo F, Velu CS, Grimes HL, Zheng Y. Rho GTPase Cdc42 is essential for B-lymphocyte development and
activation. Blood. 2009;114(14):2909-16.
72. Croker BA, Tarlinton DM, Cluse LA, Tuxen AJ, Light A, Yang FC, et al. The Rac2 guanosine triphosphatase
regulates B lymphocyte antigen receptor responses and chemotaxis and is required for establishment of B-1a and
marginal zone B lymphocytes. J Immunol. 2002;168(7):3376-86.
73. Walmsley MJ, Ooi SK, Reynolds LF, Smith SH, Ruf S, Mathiot A, et al. Critical roles for Rac1 and Rac2 GTPases
in B cell development and signaling. Science. 2003;302(5644):459-62.
74. Girkontaite I, Missy K, Sakk V, Harenberg A, Tedford K, Potzel T, et al. Lsc is required for marginal zone B
cells, regulation of lymphocyte motility and immune responses. Nat Immunol. 2001;2(9):855-62.
75. Rubtsov A, Strauch P, Digiacomo A, Hu J, Pelanda R, Torres RM. Lsc regulates marginal-zone B cell migration
and adhesion and is required for the IgM T-dependent antibody response. Immunity. 2005;23(5):527-38.
76. Randall KL, Lambe T, Johnson AL, Treanor B, Kucharska E, Domaschenz H, et al. Dock8 mutations cripple B
cell immunological synapses, germinal centers and long-lived antibody production. Nat Immunol. 2009;10(12):1283-
91.
77. Moran ST, Cariappa A, Liu H, Muir B, Sgroi D, Boboila C, et al. Synergism between NF-kappa B1/p50 and
Notch2 during the development of marginal zone B lymphocytes. J Immunol. 2007;179(1):195-200.
78. Zhang X, Shi Y, Weng Y, Lai Q, Luo T, Zhao J, et al. The truncate mutation of Notch2 enhances cell
proliferation through activating the NF-kappaB signal pathway in the diffuse large B-cell lymphomas. PLoS One.
2014;9(10):e108747.
79. Bagrodia S, Cerione RA. Pak to the future. Trends Cell Biol. 1999;9(9):350-5.

199
80. Sells MA, Chernoff J. Emerging from the Pak: the p21-activated protein kinase family. Trends Cell Biol.
1997;7(4):162-7.
81. Rane CK, Minden A. P21 activated kinases: structure, regulation, and functions. Small GTPases. 2014;5.
82. Moore TM, Garg R, Johnson C, Coptcoat MJ, Ridley AJ, Morris JD. PSK, a novel STE20-like kinase derived from
prostatic carcinoma that activates the c-Jun N-terminal kinase mitogen-activated protein kinase pathway and
regulates actin cytoskeletal organization. J Biol Chem. 2000;275(6):4311-22.
83. Czeloth N, Schippers A, Wagner N, Muller W, Kuster B, Bernhardt G, et al. Sphingosine-1 phosphate signaling
regulates positioning of dendritic cells within the spleen. J Immunol. 2007;179(9):5855-63.
84. Eisenbarth SC. Dendritic cell subsets in T cell programming: location dictates function. Nat Rev Immunol.
2019;19(2):89-103.
85. Li S, Dislich B, Brakebusch CH, Lichtenthaler SF, Brocker T. Control of Homeostasis and Dendritic Cell
Survival by the GTPase RhoA. J Immunol. 2015;195(9):4244-56.
86. Calabro S, Liu D, Gallman A, Nascimento MS, Yu Z, Zhang TT, et al. Differential Intrasplenic Migration of
Dendritic Cell Subsets Tailors Adaptive Immunity. Cell Rep. 2016;16(9):2472-85.
87. Maddaluno L, Verbrugge SE, Martinoli C, Matteoli G, Chiavelli A, Zeng Y, et al. The adhesion molecule L1
regulates transendothelial migration and trafficking of dendritic cells. J Exp Med. 2009;206(3):623-35.
88. Kamath AT, Pooley J, O'Keeffe MA, Vremec D, Zhan Y, Lew AM, et al. The development, maturation, and
turnover rate of mouse spleen dendritic cell populations. J Immunol. 2000;165(12):6762-70.
89. Kabashima K, Banks TA, Ansel KM, Lu TT, Ware CF, Cyster JG. Intrinsic lymphotoxin-beta receptor
requirement for homeostasis of lymphoid tissue dendritic cells. Immunity. 2005;22(4):439-50.
90. Durai V, Murphy KM. Functions of Murine Dendritic Cells. Immunity. 2016;45(4):719-36.
91. Kirkling ME, Cytlak U, Lau CM, Lewis KL, Resteu A, Khodadadi-Jamayran A, et al. Notch Signaling Facilitates
In Vitro Generation of Cross-Presenting Classical Dendritic Cells. Cell Rep. 2018;23(12):3658-72 e6.
92. Yi T, Cyster JG. EBI2-mediated bridging channel positioning supports splenic dendritic cell homeostasis and
particulate antigen capture. Elife. 2013;2:e00757.
93. Gatto D, Wood K, Caminschi I, Murphy-Durland D, Schofield P, Christ D, et al. The chemotactic receptor EBI2
regulates the homeostasis, localization and immunological function of splenic dendritic cells. Nat Immunol.
2013;14(5):446-53.
94. Lu E, Dang EV, McDonald JG, Cyster JG. Distinct oxysterol requirements for positioning naive and activated
dendritic cells in the spleen. Sci Immunol. 2017;2(10).
95. Klebanoff CA, Spencer SP, Torabi-Parizi P, Grainger JR, Roychoudhuri R, Ji Y, et al. Retinoic acid controls the
homeostasis of pre-cDC-derived splenic and intestinal dendritic cells. J Exp Med. 2013;210(10):1961-76.
96. Beijer MR, Molenaar R, Goverse G, Mebius RE, Kraal G, den Haan JM. A crucial role for retinoic acid in the
development of Notch-dependent murine splenic CD8- CD4- and CD4+ dendritic cells. Eur J Immunol.
2013;43(6):1608-16.
97. Murphy TL, Grajales-Reyes GE, Wu X, Tussiwand R, Briseno CG, Iwata A, et al. Transcriptional Control of
Dendritic Cell Development. Annu Rev Immunol. 2016;34:93-119.
98. Backer RA, Diener N, Clausen BE. Langerin(+)CD8(+) Dendritic Cells in the Splenic Marginal Zone: Not So
Marginal After All. Front Immunol. 2019;10:741.
99. Idoyaga J, Suda N, Suda K, Park CG, Steinman RM. Antibody to Langerin/CD207 localizes large numbers of
CD8alpha+ dendritic cells to the marginal zone of mouse spleen. Proc Natl Acad Sci U S A. 2009;106(5):1524-9.
100. Pack M, Trumpfheller C, Thomas D, Park CG, Granelli-Piperno A, Munz C, et al. DEC-205/CD205+ dendritic
cells are abundant in the white pulp of the human spleen, including the border region between the red and white
pulp. Immunology. 2008;123(3):438-46.
101. Hashimoto-Hill S, Friesen L, Park S, Im S, Kaplan MH, Kim CH. RARalpha supports the development of
Langerhans cells and langerin-expressing conventional dendritic cells. Nat Commun. 2018;9(1):3896.
102. Dicken J, Mildner A, Leshkowitz D, Touw IP, Hantisteanu S, Jung S, et al. Transcriptional reprogramming of
CD11b+Esam(hi) dendritic cell identity and function by loss of Runx3. PLoS One. 2013;8(10):e77490.
103. De Trez C, Schneider K, Potter K, Droin N, Fulton J, Norris PS, et al. The inhibitory HVEM-BTLA pathway
counter regulates lymphotoxin receptor signaling to achieve homeostasis of dendritic cells. J Immunol.
2008;180(1):238-48.
104. van de Pavert SA, Mebius RE. New insights into the development of lymphoid tissues. Nat Rev Immunol.
2010;10(9):664-74.
105. Dejardin E, Droin NM, Delhase M, Haas E, Cao Y, Makris C, et al. The lymphotoxin-beta receptor induces
different patterns of gene expression via two NF-kappaB pathways. Immunity. 2002;17(4):525-35.
106. Vivier E, Artis D, Colonna M, Diefenbach A, Di Santo JP, Eberl G, et al. Innate Lymphoid Cells: 10 Years On.
Cell. 2018;174(5):1054-66.

200
107. Lee JS, Cella M, McDonald KG, Garlanda C, Kennedy GD, Nukaya M, et al. AHR drives the development of gut
ILC22 cells and postnatal lymphoid tissues via pathways dependent on and independent of Notch. Nat Immunol.
2011;13(2):144-51.
108. Tumanov AV, Koroleva EP, Guo X, Wang Y, Kruglov A, Nedospasov S, et al. Lymphotoxin controls the IL-22
protection pathway in gut innate lymphoid cells during mucosal pathogen challenge. Cell Host Microbe.
2011;10(1):44-53.
109. Reboldi A, Arnon TI, Rodda LB, Atakilit A, Sheppard D, Cyster JG. IgA production requires B cell interaction
with subepithelial dendritic cells in Peyer's patches. Science. 2016;352(6287):aaf4822.
110. Kruglov AA, Grivennikov SI, Kuprash DV, Winsauer C, Prepens S, Seleznik GM, et al. Nonredundant function
of soluble LTalpha3 produced by innate lymphoid cells in intestinal homeostasis. Science. 2013;342(6163):1243-6.
111. Magri G, Miyajima M, Bascones S, Mortha A, Puga I, Cassis L, et al. Innate lymphoid cells integrate stromal
and immunological signals to enhance antibody production by splenic marginal zone B cells. Nat Immunol.
2014;15(4):354-64.
112. Obata Y, Kimura S, Nakato G, Iizuka K, Miyagawa Y, Nakamura Y, et al. Epithelial-stromal interaction via
Notch signaling is essential for the full maturation of gut-associated lymphoid tissues. EMBO Rep. 2014;15(12):1297-
304.
113. Emgard J, Kammoun H, Garcia-Cassani B, Chesne J, Parigi SM, Jacob JM, et al. Oxysterol Sensing through the
Receptor GPR183 Promotes the Lymphoid-Tissue-Inducing Function of Innate Lymphoid Cells and Colonic
Inflammation. Immunity. 2018;48(1):120-32 e8.
114. Chu C, Moriyama S, Li Z, Zhou L, Flamar AL, Klose CSN, et al. Anti-microbial Functions of Group 3 Innate
Lymphoid Cells in Gut-Associated Lymphoid Tissues Are Regulated by G-Protein-Coupled Receptor 183. Cell Rep.
2018;23(13):3750-8.
115. Rodda LB, Lu E, Bennett ML, Sokol CL, Wang X, Luther SA, et al. Single-Cell RNA Sequencing of Lymph Node
Stromal Cells Reveals Niche-Associated Heterogeneity. Immunity. 2018;48(5):1014-28 e6.
116. Goverse G, Labao-Almeida C, Ferreira M, Molenaar R, Wahlen S, Konijn T, et al. Vitamin A Controls the
Presence of RORgamma+ Innate Lymphoid Cells and Lymphoid Tissue in the Small Intestine. J Immunol.
2016;196(12):5148-55.
117. van de Pavert SA, Ferreira M, Domingues RG, Ribeiro H, Molenaar R, Moreira-Santos L, et al. Maternal
retinoids control type 3 innate lymphoid cells and set the offspring immunity. Nature. 2014;508(7494):123-7.
118. Hendricks J, Bos NA, Kroese FGM. Heterogeneity of Memory Marginal Zone B Cells. Crit Rev Immunol.
2018;38(2):145-58.
119. Cerutti A, Cols M, Puga I. Marginal zone B cells: virtues of innate-like antibody-producing lymphocytes. Nat
Rev Immunol. 2013;13(2):118-32.
120. Kiel MJ, Velusamy T, Betz BL, Zhao L, Weigelin HG, Chiang MY, et al. Whole-genome sequencing identifies
recurrent somatic NOTCH2 mutations in splenic marginal zone lymphoma. J Exp Med. 2012;209(9):1553-65.
121. Martinez N, Almaraz C, Vaque JP, Varela I, Derdak S, Beltran S, et al. Whole-exome sequencing in splenic
marginal zone lymphoma reveals mutations in genes involved in marginal zone differentiation. Leukemia.
2014;28(6):1334-40.
122. Rossi D, Trifonov V, Fangazio M, Bruscaggin A, Rasi S, Spina V, et al. The coding genome of splenic marginal
zone lymphoma: activation of NOTCH2 and other pathways regulating marginal zone development. J Exp Med.
2012;209(9):1537-51.
123. Descatoire M, Weller S, Irtan S, Sarnacki S, Feuillard J, Storck S, et al. Identification of a human splenic
marginal zone B cell precursor with NOTCH2-dependent differentiation properties. J Exp Med. 2014;211(5):987-
1000.
124. Kusumi S, Koga D, Kanda T, Ushiki T. Three-dimensional reconstruction of serial sections for analysis of the
microvasculature of the white pulp and the marginal zone in the human spleen. Biomed Res. 2015;36(3):195-203.
125. Steiniger B, Barth P, Hellinger A. The perifollicular and marginal zones of the human splenic white pulp : do
fibroblasts guide lymphocyte immigration? Am J Pathol. 2001;159(2):501-12.
126. Steiniger B, Barth P, Herbst B, Hartnell A, Crocker PR. The species-specific structure of microanatomical
compartments in the human spleen: strongly sialoadhesin-positive macrophages occur in the perifollicular zone, but
not in the marginal zone. Immunology. 1997;92(2):307-16.
127. Schmidt EE, MacDonald IC, Groom AC. Comparative aspects of splenic microcirculatory pathways in
mammals: the region bordering the white pulp. Scanning Microsc. 1993;7(2):613-28.
128. Lewis SM, Williams A, Eisenbarth SC. Structure and function of the immune system in the spleen. Sci
Immunol. 2019;4(33).
129. Yamamoto K, Kobayashi T, Murakami T. Arterial terminals in the rat spleen as demonstrated by scanning
electron microscopy of vascular casts. Scan Electron Microsc. 1982(Pt 1):455-8.

201
130. Steiniger B, Bette M, Schwarzbach H. The open microcirculation in human spleens: a three-dimensional
approach. J Histochem Cytochem. 2011;59(6):639-48.
131. Puga I, Cols M, Barra CM, He B, Cassis L, Gentile M, et al. B cell-helper neutrophils stimulate the
diversification and production of immunoglobulin in the marginal zone of the spleen. Nat Immunol. 2011;13(2):170-
80.
132. Goldman JM, Melo JV. Chronic myeloid leukemia--advances in biology and new approaches to treatment. N
Engl J Med. 2003;349(15):1451-64.
133. Zhou H, Xu R. Leukemia stem cells: the root of chronic myeloid leukemia. Protein Cell. 2015;6(6):403-12.
134. Stagno F, Stella S, Spitaleri A, Pennisi MS, Di Raimondo F, Vigneri P. Imatinib mesylate in chronic myeloid
leukemia: frontline treatment and long-term outcomes. Expert Rev Anticancer Ther. 2016;16(3):273-8.
135. Barrett AJ, Ito S. The role of stem cell transplantation for chronic myelogenous leukemia in the 21st century.
Blood. 2015;125(21):3230-5.
136. Zorn E, Orsini E, Wu CJ, Stein B, Chillemi A, Canning C, et al. A CD4+ T cell clone selected from a CML patient
after donor lymphocyte infusion recognizes BCR-ABL breakpoint peptides but not tumor cells. Transplantation.
2001;71(8):1131-7.
137. Terme M, Borg C, Guilhot F, Masurier C, Flament C, Wagner EF, et al. BCR/ABL promotes dendritic cell-
mediated natural killer cell activation. Cancer Res. 2005;65(14):6409-17.
138. Biernacki MA, Marina O, Zhang W, Liu F, Bruns I, Cai A, et al. Efficacious immune therapy in chronic
myelogenous leukemia (CML) recognizes antigens that are expressed on CML progenitor cells. Cancer Res.
2010;70(3):906-15.
139. Dong R, Cwynarski K, Entwistle A, Marelli-Berg F, Dazzi F, Simpson E, et al. Dendritic cells from CML patients
have altered actin organization, reduced antigen processing, and impaired migration. Blood. 2003;101(9):3560-7.
140. Mumprecht S, Claus C, Schurch C, Pavelic V, Matter MS, Ochsenbein AF. Defective homing and impaired
induction of cytotoxic T cells by BCR/ABL-expressing dendritic cells. Blood. 2009;113(19):4681-9.
141. Bertazzoli C, Marchesi E, Passoni L, Barni R, Ravagnani F, Lombardo C, et al. Differential recognition of a
BCR/ABL peptide by lymphocytes from normal donors and chronic myeloid leukemia patients. Clin Cancer Res.
2000;6(5):1931-5.
142. Orsini E, Calabrese E, Maggio R, Pasquale A, Nanni M, Trasarti S, et al. Circulating myeloid dendritic cell
directly isolated from patients with chronic myelogenous leukemia are functional and carry the bcr-abl translocation.
Leuk Res. 2006;30(7):785-94.
143. Wang C, Al-Omar HM, Radvanyi L, Banerjee A, Bouman D, Squire J, et al. Clonal heterogeneity of dendritic
cells derived from patients with chronic myeloid leukemia and enhancement of their T-cells stimulatory activity by
IFN-alpha. Exp Hematol. 1999;27(7):1176-84.
144. Litzow MR, Dietz AB, Bulur PA, Butler GW, Gastineau DA, Hoering A, et al. Testing the safety of clinical-grade
mature autologous myeloid DC in a phase I clinical immunotherapy trial of CML. Cytotherapy. 2006;8(3):290-8.
145. Holtschke T, Lohler J, Kanno Y, Fehr T, Giese N, Rosenbauer F, et al. Immunodeficiency and chronic
myelogenous leukemia-like syndrome in mice with a targeted mutation of the ICSBP gene. Cell. 1996;87(2):307-17.
146. Birnberg T, Bar-On L, Sapoznikov A, Caton ML, Cervantes-Barragan L, Makia D, et al. Lack of conventional
dendritic cells is compatible with normal development and T cell homeostasis, but causes myeloid proliferative
syndrome. Immunity. 2008;29(6):986-97.
147. Turcotte K, Gauthier S, Tuite A, Mullick A, Malo D, Gros P. A mutation in the Icsbp1 gene causes susceptibility
to infection and a chronic myeloid leukemia-like syndrome in BXH-2 mice. J Exp Med. 2005;201(6):881-90.
148. Schmidt M, Nagel S, Proba J, Thiede C, Ritter M, Waring JF, et al. Lack of interferon consensus sequence
binding protein (ICSBP) transcripts in human myeloid leukemias. Blood. 1998;91(1):22-9.
149. Hjort EE, Huang W, Hu L, Eklund EA. Bcr-abl regulates Stat5 through Shp2, the interferon consensus
sequence binding protein (Icsbp/Irf8), growth arrest specific 2 (Gas2) and calpain. Oncotarget. 2016;7(47):77635-
50.
150. Waight JD, Banik D, Griffiths EA, Nemeth MJ, Abrams SI. Regulation of the interferon regulatory factor-8
(IRF-8) tumor suppressor gene by the signal transducer and activator of transcription 5 (STAT5) transcription factor
in chronic myeloid leukemia. J Biol Chem. 2014;289(22):15642-52.
151. Burchert A, Cai D, Hofbauer LC, Samuelsson MK, Slater EP, Duyster J, et al. Interferon consensus sequence
binding protein (ICSBP; IRF-8) antagonizes BCR/ABL and down-regulates bcl-2. Blood. 2004;103(9):3480-9.
152. Hao SX, Ren R. Expression of interferon consensus sequence binding protein (ICSBP) is downregulated in
Bcr-Abl-induced murine chronic myelogenous leukemia-like disease, and forced coexpression of ICSBP inhibits Bcr-
Abl-induced myeloproliferative disorder. Mol Cell Biol. 2000;20(4):1149-61.
153. Tamura T, Kong HJ, Tunyaplin C, Tsujimura H, Calame K, Ozato K. ICSBP/IRF-8 inhibits mitogenic activity of
p210 Bcr/Abl in differentiating myeloid progenitor cells. Blood. 2003;102(13):4547-54.

202
154. Salem S, Langlais D, Lefebvre F, Bourque G, Bigley V, Haniffa M, et al. Functional characterization of the
human dendritic cell immunodeficiency associated with the IRF8(K108E) mutation. Blood. 2014;124(12):1894-904.
155. Sontag S, Forster M, Qin J, Wanek P, Mitzka S, Schuler HM, et al. Modelling IRF8 Deficient Human
Hematopoiesis and Dendritic Cell Development with Engineered iPS Cells. Stem Cells. 2017;35(4):898-908.
156. Sichien D, Scott CL, Martens L, Vanderkerken M, Van Gassen S, Plantinga M, et al. IRF8 Transcription Factor
Controls Survival and Function of Terminally Differentiated Conventional and Plasmacytoid Dendritic Cells,
Respectively. Immunity. 2016;45(3):626-40.
157. Aliberti J, Schulz O, Pennington DJ, Tsujimura H, Reis e Sousa C, Ozato K, et al. Essential role for ICSBP in the
in vivo development of murine CD8alpha + dendritic cells. Blood. 2003;101(1):305-10.
158. Bajana S, Turner S, Paul J, Ainsua-Enrich E, Kovats S. IRF4 and IRF8 Act in CD11c+ Cells To Regulate
Terminal Differentiation of Lung Tissue Dendritic Cells. J Immunol. 2016;196(4):1666-77.
159. Scheller M, Foerster J, Heyworth CM, Waring JF, Lohler J, Gilmore GL, et al. Altered development and
cytokine responses of myeloid progenitors in the absence of transcription factor, interferon consensus sequence
binding protein. Blood. 1999;94(11):3764-71.
160. Becker AM, Michael DG, Satpathy AT, Sciammas R, Singh H, Bhattacharya D. IRF-8 extinguishes neutrophil
production and promotes dendritic cell lineage commitment in both myeloid and lymphoid mouse progenitors.
Blood. 2012;119(9):2003-12.
161. Tamura T, Nagamura-Inoue T, Shmeltzer Z, Kuwata T, Ozato K. ICSBP directs bipotential myeloid progenitor
cells to differentiate into mature macrophages. Immunity. 2000;13(2):155-65.
162. Li L, Jin H, Xu J, Shi Y, Wen Z. Irf8 regulates macrophage versus neutrophil fate during zebrafish primitive
myelopoiesis. Blood. 2011;117(4):1359-69.
163. Gabriele L, Phung J, Fukumoto J, Segal D, Wang IM, Giannakakou P, et al. Regulation of apoptosis in myeloid
cells by interferon consensus sequence-binding protein. J Exp Med. 1999;190(3):411-21.
164. Hu X, Yang D, Zimmerman M, Liu F, Yang J, Kannan S, et al. IRF8 regulates acid ceramidase expression to
mediate apoptosis and suppresses myelogeneous leukemia. Cancer Res. 2011;71(8):2882-91.
165. Yang J, Hu X, Zimmerman M, Torres CM, Yang D, Smith SB, et al. Cutting edge: IRF8 regulates Bax
transcription in vivo in primary myeloid cells. J Immunol. 2011;187(9):4426-30.
166. Huang W, Zhu C, Wang H, Horvath E, Eklund EA. The interferon consensus sequence-binding protein
(ICSBP/IRF8) represses PTPN13 gene transcription in differentiating myeloid cells. J Biol Chem. 2008;283(12):7921-
35.
167. Paschall AV, Zhang R, Qi CF, Bardhan K, Peng L, Lu G, et al. IFN regulatory factor 8 represses GM-CSF
expression in T cells to affect myeloid cell lineage differentiation. J Immunol. 2015;194(5):2369-79.
168. Twum DY, Colligan SH, Hoffend NC, Katsuta E, Cortes Gomez E, Hensen ML, et al. IFN regulatory factor-8
expression in macrophages governs an antimetastatic program. JCI Insight. 2019;4(3).
169. Dabholkar M, Tatake RJ, Advani S, Gangal SG. Studies on natural killer cells in chronic myeloid leukemia
patients in remission. Neoplasma. 1990;37(1):47-53.
170. Chiorean EG, Dylla SJ, Olsen K, Lenvik T, Soignier Y, Miller JS. BCR/ABL alters the function of NK cells and the
acquisition of killer immunoglobulin-like receptors (KIRs). Blood. 2003;101(9):3527-33.
171. Kijima M, Gardiol N, Held W. Natural killer cell mediated missing-self recognition can protect mice from
primary chronic myeloid leukemia in vivo. PLoS One. 2011;6(11):e27639.
172. Humblet-Baron S, Barber JS, Roca CP, Lenaerts A, Koni PA, Liston A. Murine myeloproliferative disorder as a
consequence of impaired collaboration between dendritic cells and CD4 T cells. Blood. 2019;133(4):319-30.
173. Andoniou CE, van Dommelen SL, Voigt V, Andrews DM, Brizard G, Asselin-Paturel C, et al. Interaction
between conventional dendritic cells and natural killer cells is integral to the activation of effective antiviral
immunity. Nat Immunol. 2005;6(10):1011-9.
174. Luu TT, Ganesan S, Wagner AK, Sarhan D, Meinke S, Garbi N, et al. Independent control of natural killer cell
responsiveness and homeostasis at steady-state by CD11c+ dendritic cells. Sci Rep. 2016;6:37996.
175. Tailor P, Tamura T, Morse HC, 3rd, Ozato K. The BXH2 mutation in IRF8 differentially impairs dendritic cell
subset development in the mouse. Blood. 2008;111(4):1942-5.
176. Mandl M, Drechsler M, Jansen Y, Neideck C, Noels H, Faussner A, et al. Evaluation of the BDCA2-DTR
Transgenic Mouse Model in Chronic and Acute Inflammation. PLoS One. 2015;10(8):e0134176.
177. Swiecki M, Gilfillan S, Vermi W, Wang Y, Colonna M. Plasmacytoid dendritic cell ablation impacts early
interferon responses and antiviral NK and CD8(+) T cell accrual. Immunity. 2010;33(6):955-66.
178. Kurotaki D, Yamamoto M, Nishiyama A, Uno K, Ban T, Ichino M, et al. IRF8 inhibits C/EBPalpha activity to
restrain mononuclear phagocyte progenitors from differentiating into neutrophils. Nat Commun. 2014;5:4978.
179. Kierdorf K, Erny D, Goldmann T, Sander V, Schulz C, Perdiguero EG, et al. Microglia emerge from
erythromyeloid precursors via Pu.1- and Irf8-dependent pathways. Nat Neurosci. 2013;16(3):273-80.

203
180. Hagemeyer N, Kierdorf K, Frenzel K, Xue J, Ringelhan M, Abdullah Z, et al. Transcriptome-based profiling of
yolk sac-derived macrophages reveals a role for Irf8 in macrophage maturation. EMBO J. 2016;35(16):1730-44.
181. Masuda T, Tsuda M, Yoshinaga R, Tozaki-Saitoh H, Ozato K, Tamura T, et al. IRF8 is a critical transcription
factor for transforming microglia into a reactive phenotype. Cell Rep. 2012;1(4):334-40.
182. Gupta M, Shin DM, Ramakrishna L, Goussetis DJ, Platanias LC, Xiong H, et al. IRF8 directs stress-induced
autophagy in macrophages and promotes clearance of Listeria monocytogenes. Nat Commun. 2015;6:6379.
183. Wang IM, Contursi C, Masumi A, Ma X, Trinchieri G, Ozato K. An IFN-gamma-inducible transcription factor,
IFN consensus sequence binding protein (ICSBP), stimulates IL-12 p40 expression in macrophages. J Immunol.
2000;165(1):271-9.
184. Mass E, Ballesteros I, Farlik M, Halbritter F, Gunther P, Crozet L, et al. Specification of tissue-resident
macrophages during organogenesis. Science. 2016;353(6304).
185. Xu H, Zhu J, Smith S, Foldi J, Zhao B, Chung AY, et al. Notch-RBP-J signaling regulates the transcription factor
IRF8 to promote inflammatory macrophage polarization. Nat Immunol. 2012;13(7):642-50.
186. Marquis JF, Kapoustina O, Langlais D, Ruddy R, Dufour CR, Kim BH, et al. Interferon regulatory factor 8
regulates pathways for antigen presentation in myeloid cells and during tuberculosis. PLoS Genet.
2011;7(6):e1002097.
187. Scott CL, T'Jonck W, Martens L, Todorov H, Sichien D, Soen B, et al. The Transcription Factor ZEB2 Is
Required to Maintain the Tissue-Specific Identities of Macrophages. Immunity. 2018;49(2):312-25 e5.
188. Deng M, Daley GQ. Expression of interferon consensus sequence binding protein induces potent immunity
against BCR/ABL-induced leukemia. Blood. 2001;97(11):3491-7.

204
 Curriculum vitae

PERSONALIA
Name: Matthias Vanderkerken
Date of birth: 29/05/1989
E-mail: matthias.vanderkerken@ugent.be

EDUCATION
2001-2007 Greek-Latin
Sint-Jozef-Klein-Seminarie
Sint-Niklaas, Belgium

2007-2010 Bachelor of Science in Medicine – with greatest distinction

Ghent University, Belgium

2010- 2014 Master of Science in Medicine – with greatest distinction

Ghent University, Belgium
2012 Master thesis “The value of whole body MRI, dynamic contrast-enhanced MRI and
diffusion weighted imaging in the diagnosis, prognosis and follow-up of multiple
myeloma.” - greatest distinction.

2015-2019: PhD in Health Sciences

Center for inflammation research, VIB-UGent
Doctoral thesis: ‘Molecular interactions controlling immune homeostasis in the marginal
zone.’
Promotors: Prof. Dr. Bart Lambrecht and Prof. Dr. Tessa Kerre
FWO-aspirant fellowship

2014-…: Residency in internal medicine

Ghent University, Belgium

PROFESSIONAL EXPERIENCE
2013: Erasmus-exchange to Friedrich-Schiller Universität, Jena, Germany.
3-month internship in internal medicine
2014: Internship in Butare and Kigali, Rwanda
3-month internship in internal medicine and pediatrics
2014-2015: Resident internal medicine at Elkerliek ZH, Helmond, The Netherlands

COURSES AND TRAININGS

2011 Certificate “Principles of electrocardiography”, Ghent University
2012 Certificate B1 of German language, UCT, Ghent University
2015 Certificate: Laboratory Animal Science Felasa C
2016 Advanced immunology course, American Association of Immunologists, Boston
2015-2019 Diverse trainings in flow cytometry, microscopy, image processing, qPCR,
biomedical engineering, statistics, FlowSOM
2019 Intensive review of internal medicine course, Harvard Medical School, Boston

205
LIST OF PUBLICATIONS

Lambrecht BN, Vanderkerken M, Hammad H. The emerging role of ADAM

metalloproteinases in immunity. Nat Rev Immunol. 2018;18(12):745-58.

Hammad H, Vanderkerken M, Pouliot P, Deswarte K, Toussaint W, Vergote K, et al.

Transitional B cells commit to marginal zone B cell fate by Taok3-mediated surface
expression of ADAM10. Nat Immunol. 2017;18(3):313-20.

Sichien D, Scott CL, Martens L, Vanderkerken M, Van Gassen S, Plantinga M, et al. IRF8
Transcription Factor Controls Survival and Function of Terminally Differentiated
Conventional and Plasmacytoid Dendritic Cells, Respectively. Immunity.
2016;45(3):626-40.

Dutoit JC, Vanderkerken M, Anthonissen J, Dochy F, Verstraete KL. The diagnostic value
of SE MRI and DWI of the spine in patients with monoclonal gammopathy of
undetermined significance, smouldering myeloma and multiple myeloma. Eur Radiol.
2014;24(11):2754-65.

Dutoit JC, Vanderkerken M, Verstraete KL. Value of whole body MRI and dynamic
contrast enhanced MRI in the diagnosis, follow-up and evaluation of disease activity and
extent in multiple myeloma. Eur J Radiol. 2013;82(9):1444-52.

Vanderkerken M, Fukuyama S, Scott CL, Norris PS, Eberl G, Di Santo JP, et al. Innate
lymphoid cells control the homeostasis of ESAM+ type 2 splenic dendritic cells through
the lymphotoxin β receptor. Manuscript submitted.

Vanderkerken M, Vandersarren L, Maes B, Toussaint W, Deswarte K, Vanheerswynghels

M, et al. TAO-kinase 3 governs the terminal differentiation of conventional dendritic
cells through Notch2 signaling. Manuscript in preparation.

Vanderkerken M, Bosteels C, Sichien D, Deswarte K, Vanheerswynghels M, Hammad H,

et al. IRF8-dependent dendritic cells keep myeloproliferation in check. Manuscript in
preparation.

Vandersarren L, Maes B, Vanderkerken M, Bosteels C, Pouliot P, Deswarte K, et al. The

Ste20 kinase TAOK3 controls development, homeostasis and activation of CD8+ T cells
by regulating IL-7R signaling. Manuscript submitted.

206
Dankwoord

207
Nooit tikte de klok zo snel als gedurende de voorbije vier jaren. Al durft het geheugen
selecteren en vertekenen, toch kijk ik bijzonder tevreden terug op mijn
doctoraatstraject. Dat is te danken aan vele collega’s, vrienden en familie, die hier hun
vermelding verdienen. In de eerste plaats gaat mijn dank uit naar Bart, Hamida en
Tessa. Bart, bedankt om me zoveel vertrouwen te geven. Je bezit een ongelofelijk
motiveringsvermogen en je onstuitbaar enthousiasme bij het zien van nieuwe data is
een drijfveer op zichzelf. Bedankt ook om me de kans te geven cursussen van
internationaal topniveau te volgen en om altijd open te staan voor samenwerkingen met
andere labo’s. Hamida, at times your calm realism was the perfect counterbalance to
Bart’s wildest suggestions. You were always ready to answer my questions, even if they
concerned infant reflux disease. Thank you both for allowing me a lot of autonomy, even
if I had never held a pipette before. Tessa, ik ben blij dat jij als hematoloog het project
mee kon begeleiden vanuit een andere invalshoek. Je was altijd heel toegankelijk en
steeds bereid te helpen waar mogelijk, ook al liep het onderzoeksproject soms anders
dan het werd geconcipieerd.

Carl, I want to thank you for our fruitful collaboration. You always showed a lot of
patience and respect, frequently endured traffic jams to attend our meetings, and
performed a lot of the practical work yourself. Hopefully our work will soon be
rewarded with publication.

Martin en Sophie, tijdens mijn doctoraat hebben jullie in een ijltempo eigen groepen
uitgebouwd die intussen, aangestoken door jullie gedrevenheid, als geoliede machines
functioneren. Bedankt voor jullie interesse en constructieve feedback! Charlie, I have
seen you progress from a disturbingly efficient postdoc to a team leader with a booming
career, but you were always prepared to help me with experiments if needed. It was an
honour to have you as a member of my examination committee.

Ruth, met een dromerige glimlach straal jij altijd en overal rust uit. Je beheert het lab als
een ‘goede huismoeder’ en weet als geen ander de dunne lijn tussen jovialiteit en
strengheid te bewandelen. Bedankt om mij door alle praktische beslommeringen te
loodsen in de eerste en laatste maanden.

Kim en Dorine, jullie waren gedroomde leermeesters, elk in geheel eigen stijl, en
hebben me opgeleid met meer geduld dan ik zelf zou kunnen opbrengen. Kim, ik kan
enkel jaloers zijn op de efficiëntie en rust waarmee jij bergen werk verzet, volgens mij
ontbreekt het jou aan stress-hormoonreceptoren. Het moet er wel uit: het zou West-
Vlamingen verboden moeten worden een mondmasker te dragen. Bedankt ook voor alle
opvoedkundige tips! Dorine, jij hebt talloze keren geprobeerd me voor mezelf te
behoeden en verbood me bijna verdere experimenten te doen, jammer genoeg ben ik
wat hardleers op dat vlak. Stilaan de laatste der Mohikanen, ben ik blij dat je nog even in
ons lab bent blijven plakken. Achter je façade van perfectionisme en schuine moppen zit
een superslim brein en een hart van goud.

208
Bastiaan, compagnon de route van het eerste uur. Ingeval je jezelf nog eens buitensluit
na middernacht, zal je nu een andere hulplijn moeten inschakelen. Ondanks alle
wetenschappelijke tegenslagen blijf je onverstoorbaar doorgaan, en je kent nu intussen
meer technieken dan gelijk welke doctoraatsstudent. Met Edith is in elk geval één
experiment geslaagd, al is een herhaling misschien geboden. Ik heb er alle vertrouwen
in dat jouw verdedigingen nu ook snel zullen volgen. Cédric, derde lid van het
voorschriftenhoekje, ook jij gaat de eindfase in, en geen haar op mijn hoofd dat
betwijfelt dat je je PhD schitterend zal afronden. Hopelijk kunnen we dan nog eens
‘Stapelei’ spelen.

Katrien, Julie, Nincy, het is onmogelijk niet met jullie overeen te komen, en Davos
behoort tot de leukste herinneringen van mijn PhD. Jullie positieve energie straalt af op
het hele lab. Beufke, veel succes met de laatste loodjes. Hetzelfde geldt voor Jessica. Al
ben je nu geëmigreerd, jouw vrolijke woordenvloed deed me altijd thuis voelen. Leen, je
hebt je in de babbelgang genesteld en doet die naam eer aan: jij krijgt zelfs een
doofstomme aan het praten. Bedankt voor de leuke momenten, en voor alle
kinderkleertjes die nu een tweede leven krijgen. Andrew, it appears that black plums
are as sweet as white ones. Best of luck with your projects! Helena, I won’t even try to
compete with your continuous stream of wordpuns. You are one-of-a-kind, and it was
really kind of you to travel all the way to San Francisco to perform that experiment for
me. Ines, Anneleen, Wouter, Freya, Johnny, Eva, Sofie and Victor, all of you are on
track with beautiful research projects. I wish you all the best in your future endeavours.
Jozefien, Antonio, Sjoerd and Stijn, you have mixed in quickly. Welcome in our lab!

Many people have come and gone in the past four years. My thanks go to Mary, Katrijn,
Martijn, Simon, Jonathan, Emma, Filipe and Hanne, who have all moved on in their
promising careers. All of you are excellent researchers, and I was truly impressed by
your knowledge and intelligence when I started in the lab. Lana en Eline, met jullie is
het lab twee schitterende collega’s armer. Eline, geniet ten volle van je reis down under,
en laat de wereldverbeteraar in je nooit temmen. Lana, na een moeizame start waren de
voorbije twee jaren bijzonder productief: intussen ben je een doctorstitel, een huis en
twee telgen rijker. We horen elkaar ongetwijfeld nog, al was het maar om elkaars kroost
tijdig van de crèche af te halen.

Special thanks go to all lab technicians: Justine, Manon, Sofie, Karl, Wendy, Farzaneh,
Bavo en Caroline. It is amazing how you keep the lab running smoothly despite the
rapid turnover of scientists. You were always kindly offering help and never declined or
complained when I asked assistance, even though it was often very last minute. Without
your help, I would not have been able to finish my PhD in time. The lab is truly blessed
having such experienced and helpful people aboard! In the same breath, I should
mention Gert and Julie. Jullie staan nooit in de schijnwerpers, maar zijn altijd
hulpvaardig en oprecht geïnteresseerd in ieders werk. Eigenhandig hebben jullie de
Flow Core uitgebouwd tot een unit waar menig instituut van droomt. Hoewel jullie een

209
verdiepje werden gepromoveerd, voelt het alsof jullie nog steeds deel van het uBla team
uitmaken!

Tenslotte gaat mijn dank uit naar mijn familie. Mama, papa, bedankt om me altijd de
vrijheid te geven om mijn hart te volgen en om me te steunen ongeacht welke keuzes ik
maak. Alles wat ik heb bereikt, heb ik aan jullie te danken. Carmen, Dany, Monique,
Griet, Lotte, Karel: bedankt om zo vaak in te springen als babysit of wanneer we het
druk hadden, zelfs met jullie eigen volle agenda’s.

Lieve Nele, samen kunnen we de wereld aan. Bedankt om zo begripvol en geduldig te

zijn wanneer een experiment uitliep of wanneer de deadlines naderden. Jouw openheid
en no-nonsense attitude zijn een voorbeeld voor mij. Ik kan me geen beter luisterend
oor, geen scherpere criticaster, geen liefhebbender vriendin, geen chaotischer
babbelkous, geen betere moeder voor Lukas indenken. Jullie zijn het beste dat me al is
overkomen.

Bedankt.

Matthias

210