Professional Documents
Culture Documents
RESEARCH ARTICLE
Regeneration of functional alveoli by adult
human SOX9+ airway basal cell transplantation
Qiwang Ma1, Yu Ma1,5, Xiaotian Dai2, Tao Ren3, Yingjie Fu4, Wenbin Liu1, Yufei Han1, Yingchuan Wu1,
Yu Cheng4, Ting Zhang5, Wei Zuo1,5,6&
1
Shanghai Pulmonary Hospital, School of Medicine, Tongji University, Shanghai 200433, China
2
Southwest Hospital, Third Military Medical University of PLA, Chongqing 400038, China
3
Shanghai Jiao Tong University Affiliated Sixth People’s Hospital, Shanghai 200233, China
4
The Institute for Biomedical Engineering and Nano Science, School of Medicine, Tongji University, Shanghai 200029, China
adult’s own lung—if can be identified, isolated and expanded marker KI67 (Figs. 1C and S1A). All of the P0 colonies were
—can be a new option for transplantation therapy. confirmed epithelium origin (E-cadherin+, Fig. S1A) and
In adult rodent, different populations of lung stem/pro- stained double positive for airway basal cell markers KRT5
genitor cells have been identified in last decade with capa- and P63 (Fig. 1C and 1D). We did not observe any P63
bility to reconstruct lung epithelium. Most of the mouse lung single positive colonies (Vaughan et al., 2015). Considering
stem/progenitor cells are facultative and can be induced to that BCs take for about 20% of total cell number in brushed
proliferate in response to injury as well as differentiate into samples of 3rd–4th order bronchus, it appeared that
one or more lung cell types (Kotton and Morrisey, 2014; Kim approximately 1% of the BCs in human airway could be
et al., 2005; Barkauskas et al., 2013; Hogan et al., 2014; clonogenic lung epithelium progenitors.
Desai et al., 2014). More recently, we and others found a We seeded one single BC onto feeder cells and grew them
rare population of p63+/Krt5+ distal airway stem cells into one single colony which was then picked up by cloning
(DASCs), which play essential role in murine lung repair after cylinders and passaged continually. The latest passage of
influenza-induced acute injury (Zuo et al., 2015; Vaughan BC clones had gone through 50 doublings (=1015 fold
et al., 2015). However in adult human, whether there are expansion) in our lab. The single cell-derived BC clones and
lung cells with regenerative capacity in vivo need to be their original brush-off tissue samples were analyzed by high-
explored. Given the huge differences between human vs. throughput RNA sequencing (RNA-Seq). On average, we
mouse of their respiratory systems in terms of develop- detected 16,230 genes and 25,223 transcripts. Thus, more
mental process, lung lobulation, branching pattern and cell than 60% of known human genes and transcripts were
composition, the identity of human lung progenitor cells need expressed in clonogenic BCs. Gene expression value cor-
to be rigorously evaluated. relation analysis showed that the clone transcriptome profiles
In the current work, we discovered the putative adult are distinct from their original brush-off tissues, but the two
human lung progenitor cells located at the bottom of “rugaes” clones from two independent persons share very similar
in airway epithelium, with a SOX9 marker to distinguish them transcriptome (Fig. 1E, Pearson correlation coefficient =
from other SOX9−/P63+/KRT5+ airway basal cells (BCs). 0.95). Single nucleotide polymorphism (SNP) analysis
From a trace amount of bronchoscopic brush-off lung tis- showed that BC clones have around 70% less polymorphism
sues, we isolated SOX9+ BCs and expanded them in vitro comparing to the brush-off tissues, which is in consistency
indefinitely. SOX9+ BCs transplanted into injured immune- with their single cell origination. High expression of BC
deficient mouse lung can regenerate functional lung markers (KRT5, P63, NGFR and S100A) and another puta-
epithelium with both human bronchiolar and alveolar tive mouse stem cell marker integrin α6β4 (Chapman et al.,
epithelium reconstituted. Most importantly, for the first time 2011) were observed in clones. In contrast, clonogenic BCs
we explored the clinical feasibility of autologous SOX9+ BC do not express other bronchial or alveolar lineage markers as
transplantation to treat two patients with chronic lung dis- shown by RNA-Seq and confirmed by immunostaining
eases. The clinical trial result is highly consistent with our (Figs. 1F and S1B). Protein-protein interaction analysis of
observation on mouse model, and making it a solid basis for overexpressed genes indicated three major signal molecule
future large-scale clinical study. networks including Notch1/2/3, FGF10/7 and Wnt7 ligand
and their downstream components. All three signaling
A B
Br Brush
Brush
Bronchoscopic Basal cell clones on Pedigree clone
brushing-off cells feeder cells from a single Feeder-free expansion
cells
C D
Br
Br
Br
GO term enrichment
FOXJ1
lo
lo
LRP6 LRP5
Br
Cell proliferation
MUC5AC
WNT7A WNT7B
0.8
0.6
0.4
0.2
Secretion
-2
-1
1
2
1
h-
h-
CTNNB1
ne
ne
Cilium
us
us
lo
lo
Br
Br
Figure 1. Isolation and characterization of BCs from SOX9+ human airway. (A) Diagram showing the process of clonogenic BCs
isolation and expansion. (B) Bronchoscopic image showing brushing of cells from human airway. (C) Left, BC colonies grown on
feeder cells; right, anti-KRT5 and anti-P63 immunostaining of BC colonies with nuclei counterstain. Human sample number n = 10.
Scale bar, 100 μm. (D) Left, BCs in human airway by anti-KRT5 and anti-P63 immunostaining. Inset, high magnification with club cell
(CC10+, cyan color) costaining; right, hematoxylin & eosin staining of the same section. Br, bronchus. Scale bar, 100 μm.
(E) Heatmap showing transcriptome profile correlation value of BC clones and brush-off tissues. (F) Expression heatmap of selected,
differentially expressed genes (P < 0.05) comparing BC clones and brush-off tissues. (G) Protein-protein interaction network of
selected genes with high expression level in BC clones. (H) Enriched gene ontology classes of BC clones versus brush-off tissues.
networks are previously known to play essential roles in 2013). Here we confirmed SOX9 expression in P63+/KRT5+
embryonic lung development (Bellusci et al., 1997; Rajagopal BC clones by immunostaining (Fig. 2A). Accordingly, by
et al., 2008; Tsao et al., 2008) (Fig. 1G). Gene ontology (GO) histological examination of human 2nd order (Fig. S2) and
term analysis demonstrated critical biological processes 3rd–4th order airway (Fig. 2B), we observed 1.3% ± 0.3%
enriched in BCs (Fig. 1H). and 1.7% ± 0.5% SOX9-expressing P63+ BCs, respectively.
The proportion of SOX9+ cells in total BCs is very close to
our estimation in clonogenic assay as mentioned above
Clonal analysis of SOX9+ BCs
(∼1%), suggesting SOX9 as a marker to distinguish clono-
Importantly, RNA-Seq data also showed that clonogenic BCs genic BCs vs. other non-clonogenic BCs. Interestingly, we
highly express SOX9 (Sex Determing Region Y- Box 9), a noticed that there are a few invaginations (rugaes) in 2–4
transcriptional factor known to be enriched in branching tips order human airway epithelium and the SOX9+ BCs are
of developmental lung. In embryonic development, SOX9 exclusively located near the base of the rugaes. There are
activity is required to maintain the undifferentiated status of averagely 3 (±1) SOX9+ BCs in each individual rugae. Fur-
distal lung progenitor and disruption of SOX9 function pre- ther immunofluorescent examination showed a very small
vents adult alveoli formation (Perl et al., 2005; Rockich et al., portion of them (<1%) are proliferative (KI67+) (Fig. 2C). Of
A B C
Br Br
Br
P63 KRT5 SOX9 SOX9 P63 P63 CC-10 SOX9 Ki-67 P63
D Feeder-free F H
es
es
Human lung
asg
sag
Relative expression
ng
1.5 Early passages
ass
n lu
Late passasges
pas
ep
ma
rly
1.0
Lat
Hu
Ea
0.5 P63
SOX9
0.1
KRT5
H 5
PX
LA B
G 3
M 1
1
Protein & Cell
SC MP
A
C
SP
SP
AQ
B1
U
O
MUC1
E G Acet-a-Tub
200
Relative expression
63
X9
19 20 21 22 X Y GAPDH
TP
KR
SO
Figure 2. Feeder-free expansion of SOX9+ BCs. (A) Immunostaining of SOX9+ BCs with anti-P63, anti-KRT5 and anti-SOX9
antibodies. (B) SOX9+ BCs in rugae of 3rd order human airway by anti-SOX9, anti-P63 and anti-CC10 immunostaining. Scale bar,
100 μm. (C) SOX9+ BCs in rugae of 3rd order human airway by anti-KI67 immunostaining. (D) BC colony cultured on feeder-free
condition. (E) Karyotyping of cultured BCs. (F) qPCR showing alveolar and bronchial epithelium marker gene expression of human
lung sample and SOX9+ BCs in early (P2) and late (P8) passages. n = 3, biological replicates. Error bars, S.E.M. (G) qPCR showing
progenitor cell marker (Krt5, P63 and SOX9) gene expression of human lung sample and SOX9+ BCs in early (P2) and late (P8)
passages. n = 3, biological replicates. Error bars, S.E.M. (H) Western blotting showing marker gene expression of human lung sample
and SOX9+ BCs in early (P2) and late (P8) passages.
note, SOX9+ BCs can also be isolated and expanded from We further analyzed SOX9+ BCs at single cell resolution.
those small airway (∼1 mm diameter) samples, which is 5 single cells from one person in normal group were selected
accessible by open-chest surgery or autopsy but not by at Passage 0 and expanded to Passage 1 and Passage 2.
bronchoscopy (data not shown). Great variation of their clonogenic capacity was observed at
The whole brushing sampling and SOX9+ BCs cloning Passage 1 (coefficient variation = 59.9%) and Passage 2
procedure was carried out on 15 individuals with a recovery (coefficient variation = 75.7%). Similar clonogenicity varia-
rate of 100%. Donors are from 4 different disease categories tion was observed in individuals from other disease cate-
including 5 normal healthy volunteers, 2 bronchiectasis gories and the average coefficient variation of all clones is
patients, 3 chronic COPD patients, and 5 interstitial lung 52.2% (Fig. S3C).
disease (ILD) patients with pulmonary fibrosis. The SOX9+ SOX9+ BCs grown on feeder cells can be transferred onto
BCs from different categories of diseases showed no petri dish pre-coated with collagen fibers for feeder-free
apparent difference in colony morphology (Fig. S3A) or culture. The feeder-free cultured SOX9+ BC can also form
marker expression (Fig. S3B). Their clonogenic efficiency colonies though their cell-cell contact within one colony is
seemed similar—but still need future investigation in much less tight comparing to those on feeders (Fig. 2D). The
larger cohort to get statistically meaningful conclusion. feeder-free cultured BCs are able to be passaged for at least
30 doublings with no obvious morphology change.
Karyotyping indicated their stable genetic characteristics fibroblast cells (data not shown) or human cervix-derived
along with passaging (Fig. 2E). Quantitative analysis of P63+/KRT5+/SOX9+ progenitor cells (Figs. 3G and S6E) can
progenitor markers (KRT5, P63 and SOX9) and lung barely incorporate into injured mouse lung either. This data
epithelium lineage markers at both RNA and protein level indicated the tissue specificity of different adult stem/pro-
indicated that there is no spontaneous differentiation of BCs genitor cells.
in the culture process (Fig. 2F–H).
Regenerated lung by SOX9+ BC transplantation
+
Xeno-transplanted SOX9 BCs give rise to human lung contributed to mouse pulmonary function
in vivo
Functional alveolar unit requires close epithelium-capillary
Next we examined whether the SOX9+ BC could differentiate interaction for exchange of gas, energy and other sub-
and regenerate lung tissue by transplanting such cells into stances. In the optically cleared mouse lung, we observed
mouse lung parenchyma. Firstly, immunodeficient NOD- branching major blood vessels in transplanted mouse lung
SCID mice were subjected to bleomycin intratracheal instil- (Fig. S7A). We also found that the thin, long-shape human
lation, which lead to rapid onset (8 days after bleomycin) AEC1 aligned together with microvascular vessels which are
damage of centrilobular and surrounding regions as shown positive for capillary endothelial markers CD34 and PECAM/
by microCT-scan and immunostaining. Masson trichrome CD31, with approximately 1 μm-thick integrinβ-1+ basement
A B
- SOX9+ BCs HuLamin GFP
+ SOX9+ BCs
C D E
GFP SOX9 HuLamin AQP5 SPC
AIv AIv
AQP5 HOPX
F +
SOX9 BC Chimera G
Protein & Cell
800 SOX9+ BC
Human gene expression
600
**
200
**
60
**
40
**
20
*
0
P5
PX
LA B
G 3
A1
BC from
SC MP
C
SP
AQ
B1
U
O
cervix
M
H
Figure 3. Transplantated SOX9+ BCs regenerate functional human lung in vivo. (A) Left, direct fluorescence image under
stereomicroscope showing NOD-SCID mouse lung without (upper panel) or with (lower panel) GFP-labeled SOX9+ BC
transplantation. Right, cryo-section and direct fluorescence imaging of transplanted GFP-labeled SOX9+ BCs in lung parenchyma.
Scale bar, 100 μm. (B) Immunofluorescence imaging of transplanted GFP-labeled SOX9+ BCs in lung parenchyma with human
specific Lamin A+C marker costaining. (C) Fully differentiated GFP+ cells lost SOX9 marker expression (arrowhead indicated). Scale
bar, 10 μm. (D) Confocal image with human specific Lamin A+C immunostaining (HuLamin) showing regenerated type I (AQP5+)
alveolar cells. No type II (SPC+) cells were observed. (E) Confocal image showing regenerated AEC1 (AQP5+ and HOPX+). AQP5 as
a membrane-bound protein distributes on surface of GFP+ cells. Arrowheads indicated the overlay of HOPX with GFP signal in
nucleus. Scale bar, 20 μm. (F) qPCR with human specific primers showing alveolar and bronchiolar epithelium marker gene
expression in SOX9+ BC transplanted chimeric lung (AEC1: AQP5 and HOPX; AEC2: SPB and LAMP3; bronchiolar cells: SCGB1A1
and MUC1). Biological replicates, n = 3. Error bars, S.E.M. (G) Left, clonogenic BCs isolated from human cervix epithelium obtained
by biopsy. Right, transplantation of equal numbers of BCs from lung and cervix indicated different incorporation efficiency.
TGF-β signaling modulates SOX9+ BC proliferation by TGF-β treatment together with mild change of some other
cell cycle-related genes (Fig. 6F). TGF-β had little effect on
To further improve the transplantation efficiency of SOX9+
the apoptosis of SOX9+ BCs (Fig. S8B). Collectively these
BCs, we screened multiple drugs and found Pirfenidone, an
experiments showed that the TGF-β/SMAD/P15 signaling
FDA approved anti-pulmonary fibrosis drug (King et al.,
axis could effectively modulate SOX9+ BC proliferation.
2014) could facilitate the SOX9+ BC transplantation effi-
Similar proliferation inhibitory effect of TGF-β/SMAD was
ciency significantly. Interestingly, transforming growth factor-
recently reported on TBC as well (Mou et al., 2016).
β (TGF-β) had the opposite effect (Figs. 6A and S8A). This
discovery prompted us to study the underlying molecular and
cellular mechanism. We found that Pirfenidone treatment Autologous SOX9+ BCs transplantation clinical trial
can abolish TGF-β-induced phosphorylation of SMAD2/ in bronchiectasis patients
SMAD3 (Fig. 6B). In turn, TGF-β treatment significantly
Bronchiectasis is a chronic lung disease radiographically
suppressed the clonogenicity and cell viability of SOX9+
characterized by permanent pathologic dilation of the small
BCs, which can be rescued by the SMAD2/SMAD3 inhibitor
and medium-sized bronchi, which may lead to respiratory
SB-431542 (Fig. 6C–E). Simutaneously, the expression of
failure and eventually to death. Patients with bronchiectasis,
p15(INK4B), a G1 cell cycle inhibitor, was strongly induced
Bv Bv
AIv AIv
C D E
GFP CD31 ITGB1 GFP CD31 E-Cad GFP ZO-1
AIv
AIv AIv
AIv
Bv
Figure 4. Regenerated alveoli with functional epithelium-capillary system. (A) Transplanted SOX9+ BCs (anti-GFP) and
capillary endothelium marker (anti-CD34). Scale bar, 100 μm. (B) Confocal image of SOX9+ BCs regenerated alveoli (Alv) and the
neighboring capillary blood vessel (Bv). Left, immunofluorescence; right, bright field. Scale bar, 20 μm. (C) Confocal image showing
the basement membrane (ITGB1+, white color, arrowhead indicated) between regenerated alveoli epithelium and capillary
endothelium (CD31+). Scale bar, 10 μm. (D) Confocal image showing the cell adherens junction (E-cadherin+, white color) between
regenerated alveoli epithelial cells. Scale bar, 20 μm. (E) Confocal image showing the cell tight junction (ZO-1+) between regenerated
alveoli epithelial cells. Scale bar, 20 μm. (F) Direct fluorescence image of the transplanted GFP-labeled SOX9+ BCs (green) and
bright-field image of tail vein delivered gold nanoparticles (AuNPs) of the same region (brown). Scale bar, 100 μm. (G) Direct
fluorescence image of the transplanted GFP-labeled SOX9+ BCs (green) and bright-field image of intratracheally delivered gold
nanoparticles (AuNPs) of the same region (brown). Scale bar, 100 μm.
if left untreated, will have a continual decrease of their pul- Two patients diagnosed as non-CF bronchiectasis were
monary function. Current pharmacological strategies to treat firstly enrolled for autologous SOX9+ BC transplantation on
bronchiectasis such as antibiotics, mucolytics and anti-in- April, 2016. Both patients are men in 50s, non-smokers.
flammatory agents could only control the disease exacer- Patient 1 was diagnosed as bronchiectasis 8 years ago with
bation but not improve the pulmonary function nor repair the productive cough and dyspnea on exertion symptom, which
damaged lung tissue (ten Hacken et al., 2007). To explore worsens continually under regular pharmacological treat-
the clinical feasibility of autologous SOX9+ BC transplanta- ment. CT scan shows multiple bronchial cylinder dilation and
tion, we conducted a pilot trial aiming to treat bronchiectasis patchy consolidation in his lung. Patient 2 was diagnosed as
by regenerating functional human lung. The general trial bronchiectasis and COPD decades ago, with productive
protocol and the cell manufacturer (Regend Therapeutics cough and dyspnea on exertion symptom, which worsens
Co.Ltd) were archived by China Food and Drug Adminis- continually under regular pharmacological treatment. CT
tration (CFDA) and National Health and Family Planning scan shows multiple bronchial cystic dilation, thicken bron-
Commission of China, and the trial was performed in national chial wall and patchy consolidation in his lung.
approved stem cell clinical research institute (Southwest For both patients, tissues were bronchoscopically col-
Hospital) after strict ethic commission review of preclinical lected from random region of left upper lobe and right upper
data (A part of but not all preclinical data was released in the lobe and transported to GMP (Good Manufacture Practices)
current manuscript). level tissue culture facility for SOX9+ BC isolation and
C D E
15 pCO2 20 pO2 100 * sO2
Fraction (%)
15 *
10
80
kPa
kPa
10
5
5 60
0 0
al
al
s
BC
BC
al
s
PB
PB
m
BC
PB
m
or
or
Protein & Cell
or
+
+
X9
X9
N
X9
eo
eo
N
SO
SO
eo
Bl
Bl
SO
Bl
+
+
eo
eo
eo
Bl
Bl
Bl
Figure 5. BC transplantation rescued mouse pulmonary function. (A) Injured mouse lung without or with GFP-labeled SOX9+
BCs transplantation by anti-GFP and anti-Fibronectin co-staining. Scale bar, 200 μm. (B) Left, immunofluorescence image of injured
mouse lung transplanted with GFP-labeled SOX9+ BCs; right, immunostaining on the same section showing exclusion of α-SMA+
myofibroblasts from GFP+ area. Scale bar, 200 μm. (C) CO2 partial pressure of mouse arterial blood before and 1 month after
bleomycin-induced injury with or without SOX9+ BCs transplantation. Each dot indicates an individual mouse. (D) O2 partial pressure
of mouse arterial blood 1 month after bleomycin-induced injury with or without SOX9+ BCs transplantation. Each dot indicates an
individual mouse. (E) O2 saturation of mouse arterial blood before and 1 month after bleomycin-induced injury with or without SOX9+
BCs transplantation. Each dot indicates an individual mouse.
expansion (Fig. 7A). Isolated SOX9+ BCs were cultured on spontaneously or with regular medicine, the recovery of
clinical-level feeder cells and then shifted to feeder-free patients suggested high probability that transplanted SOX9+
culture condition. Totally 1 × 106/kg body weight of SOX9+ BCs were able to regenerate functional lung in human, which
BCs were infused into distinct lobes of patients through is consistent with our observation in animal models. And we
bronchoscopy (Tzouvelekis et al., 2013) (Fig. 7B). Clinical will continue life-long observation on the two patients.
status of patients was evaluated 1 day before and 1, 3 and
12 months after cell transplantation. Although it is almost
DISCUSSION
impossible to directly track unlabelled transplanted cells in
human, we did observe regional repair of cystic dilation after In the current study, we revealed that a small population of
cell transplantation by high-resolution computed tomography SOX9+ BCs in adult airway can regenerate human lung
(HRCT) scan for Patient 2 (Fig. 7C). The thickened bronchial epithelium. A few SOX9+ BCs brushed off from human air-
wall also became thinner after cell therapy for Patient 2. ways can be expanded to sufficient number in feeder-free
Spirometry results indicated remarkable recovery of pul- condition and transplanted into injured mouse lung to
monary function in both patients after transplantation as regenerate human air exchanging units. In unperturbed lung,
measured by FEV1, FVC and DLCO/VA (Fig. 7D). Impor- the SOX9+ BCs are located in the base of airway epithelium
tantly, no aberrant cell growth or other related adverse invagination, which is reminiscent of intestinal stem cells
events were observed during the whole follow-up time. In the residing in gut crypt compartments. However unlike the
last follow-up (20 months after transplantation), Patient 1 intestinal stem cells which are constantly in active cell cycle
described improvement of dyspnea, improvement of exer- to replenish gut epithelium in a fast turn-over mode (Clevers,
cise capacity, less productive cough and less times of 2013), the SOX9+ BCs in airway are quiescent most of the
exacerbation after cell therapy; Patient 2 described less time, and could only be activated once a particular injury
productive cough and less times of exacerbation after cell signal from more distal lung was received. Like its counter-
therapy. As it is well documented that bronchiectasis is a part p63+/Krt5+ BCs in mouse, endogenous SOX9+ BCs
permanent, irreversible disease that cannot resolve might migrate towards inflamed parenchymal region and
A B
TGF-β - + - +
0.49 ± 0.024 0.00 ± 0.00 0.69 ± 0.042
Pirfenidone - - + +
P-Smad2/3
Smad2/3
C D
30 12 h
**
0.8 20
**
10 24 h
0.6 5
Relative expression
**
48 h
**
* *
**
**
TGF-β - + - +
**
***
SB-431542 - - + + -5
**
5
A1
E2
K4
A
P1
P2
N
D
PC
lin
lin
C
yc
yc
C
C
Figure 6. TGF-β signaling modulates SOX9+ BC proliferation. (A) Direct fluorescence image of mouse lung transplanted with
1 × 106 GFP-labeled SOX9+ BCs under dissection microscope. Each lung was from mouse with indicated treatment and harvested 7
days after transplantation. The left lobes were analyzed and the GFP+ cell numbers (×106) were counted by flow cytometry analysis.
Biological replicates, n = 3. PFD, Pirfenidone. (B) SOX9+ BCs were stimulated with 10 ng/mL TGF-β for 2 h, with or without 1 mg/mL
Pirfenidone treatment overnight. Western blotting of cell lysates with anti-phosphated-Smad2/3 and anti-total Smad2/3 antibodies
was performed to examine the activation of TGF-β pathway. (C) Direct fluorescence imaging of GFP-labeled SOX9+ BCs cultured in a
6-well plate in the absence or presence of 10 ng/mL TGF-β. Scale bar, 200 μm. (D) Quantification of clonogenicity of SOX9+ BCs in
the presence of 10 ng/mL TGF-β or 10 mmol SB. SB, TGF-β type I receptor inhibitor SB-431542. Technical replicates n = 3. (E) WST
viability assay of SOX9+ BCs treated by 10 ng/mL TGF-β or 5 mmol TGF-β inhibitor SB-431542, or their combination. Technical
replicates n = 3. (F) qPCR showing cell cycle-related gene expression level of SOX9+ BCs with 10 ng/mL TGF-β treatment for
indicated h. Biological replicates, n = 3.
rebuild the respiratory tree by proliferation and differentiation. whether SOX9+ BCs are the remains of previously docu-
We and previous studies found that there are P63+ and/or mented SOX9+ progenitors in embryonic lung. In embryonic
KRT5+ cells enriched in the damaged alveolar region of lung, SOX9+ embryonic progenitors are enriched in distal
many lung disease patients (including those with COPD, bud tips of respiratory tree at canalicular stage (Perl et al.,
pulmonary fibrosis and bronchiectasis) (Chilosi et al., 2002; 2005). Interestingly, previous report demonstrated that
Asano et al., 2011; Smirnova et al., 2016), which could be intravenous transplantation of human canalicular-stage
the progeny of endogenous SOX9+ BCs in the middle of embryonic lung cell mixture into NOD-SCID mice can give
differentiation process. Interestingly there was almost no rise to chimeric lung (Rosen et al., 2015). Similarly, a more
SPC+ AEC2 generated in our xeno-transplantation model, recent report showed mouse SOX9+ progenitors in embry-
which suggested the limitation of the SOX9+ progenitor onic lung can be grown in vitro as organoid and transplanted
potency and also the possibility that the traditional concept of to generate mouse alveoli (Nichane et al., 2017). Altogether
human AEC2 as the progenitor of AEC1 may not be correct these work supported the concept that SOX9+ cells are lung
in this circumstance. progenitors in both embryonic and adult lung.
The pattern that SOX9+ BCs adopt to regenerate the One important technical advance we bring out in this work
respiratory tree resembles the natural development process is the system to selectively expand SOX9+ BCs in a feeder-
of lung in gestation, which raises an open question as free condition. Previously through bronchoscopic brushing
A B
Patient 1 Patient 2
1.6/0.67
2.0/1.1
1.6/0.67
3.2/ -/0.67
1.6
C
Before
Protein & Cell
After
D
FEV1 (L) FVC (L)
Months post 0 1 3 12 0 1 3 12
transplantation
Patient 1 1.07 (36.6%) 0.94 1.40 1.29 2.56 (69.4%) 2.15 2.93 2.73
Patient 2 0.51 (20.8%) 0.85 0.81 0.71 0.73 (24.3%) 1.85 1.88 1.65
DLCO/VA (mmol/min/kPa/L)
Months post 0 1 3 12
transplantation
Patient 1 1.77 (128.2%) 1.88 1.66 1.60
Patient 2 0.21 (14.9%) 1.47 1.36 1.49
Figure 7. Autologous SOX9+ BC transplantation: a pilot clinical study. (A) Cultured SOX9+ BCs from two patients of
bronchiectasis. (B) Diagram showing the number of SOX9+ BCs (×107) transplanted into each lobe of Patient 1/Patient 2. Five lobes
are labeled with different colors. Note the Patient 1 undertook left lower lobectomy a decade ago. (C) Representative images of
consecutive CT scan show the regional recovery of bronchiectasis (yellow square) 1 year after autologous SOX9+ BC transplantation
in Patient 2. (D) Measurement of pulmonary function and exercise capacity in both patients: FEV1 (forced expiratory volume in 1
second), FVC (forced vital capacity), DLCO/VA (diffusing capacity of the lung for carbon monoxide adjusted by the alveoli volume)
were shown as absolute value and percentage to predicted level. For each parameter, percentage >80% was regarded as clinically
normal.
followed by routine basal cell culture, Crystal et al., obtained Most importantly, we showed that SOX9+ BCs cultured
P63+/KRT5+/SOX7/15/4+ basal cells but not the rare SOX9+ under GMP guidelines can be applied clinically in order to
subpopulation (Hackett et al., 2011) . Here from as few as reconstitute human lung for devastating chronic lung disease
one single SOX9+ BC, we can expand it to 5 × 107 purely treatment. To the best of our knowledge, this is the first
undifferentiated cells within 3–4 weeks. Therefore we can successful attempt to regenerate human large inner organ
acquire a homogeneous population of regenerative cells with based on cell replacement strategy. As demonstrated in
uniform characteristics, which is crucial for cell quality control hematopoietic, skin and corneal regeneration field, autolo-
in further clinical application. gous stem/progenitor cell transplantation strategy has been
successfully applied to treat multiple devastating diseases.
We show here that supplement of expandable SOX9+ BCs under patient’s consent and approved by Southwest Hospital Ethics
by autologous transplantation could repair the damaged lung Committee (Chongqing, China) and Shanghai East Hospital Ethics
in two bronchiectasis patients in both pulmonary structure Committee (Shanghai, China).
and function. Generally the bronchiectasis patients without
medical intervention will deteriorate over time, thus the Isolation and culture of human SOX9+ BCs
recovery of lung function and structure after cell therapy
To isolate the SOX9+ BCs, 2 mm brush with samples were cut with
suggested the efficacy of this strategy, and established a
scissors into 1 cm pieces. After removing sputum, the brush pieces
basis for future trials. As the SOX9+ BCs are derived from
were directly digested with dissociation buffer including DMEM/F12
patients’ own airway, it is a recapitulation and augmentation (Gibco, USA), 2 mg/mL protease XIV (Sigma, USA), 0.01% trypsin
of naturally occurring lung repair process. However after all, (Gibco, USA) and 10 ng/mL DNase I (Sigma, USA). Specimens
our pilot clinical trial is a very preliminary exploratory study were incubated at 37°C for an hour with gentle rocking. Alternatively,
so the safety and efficacy of the current strategy still need human small airway were dissected from a bulk of lung tissue and
additional verification in a much larger cohort. Although we digested in the same dissociation buffer at 37°C overnight. Disso-
have never observed aberrant growth of SOX9+ BCs in ciated cells were passed through 70-μm Nylon mesh (Falcon, USA)
NOD-SCID mouse model, longer time follow-up on the to remove aggregates and then washed twice with cold F12 medium.
patients is still necessary to fully eliminate the tumorigeneic Cell viability was assessed by exclusion of trypan blue dye. Cell
possibility. We have demonstrated the differentiation poten- pellets were collected by centrifuge of 200 ×g and plated onto mit-
EP1601Y, Thermo), P63 (deltaN, 1:200, 4A4, Abcam), E-cadherin To arrest SOX9+ BCs in mitosis metaphase, cells of 75% confluence
(1:200, H-108, Santa cruz), SOX9 (1:200, ERP14335-78, Abcam), were treated with 1 μg/mL colchicines for 7 h and digested into
SOX9 (1:200, AF3075, R&D); AEC markers: AQP5 (1:1000, single cells by 0.25% trypsin. Then the cells were incubated by 0.4%
EPR3747, Abcam), HOPX (1:500, E-1, Santa Cruz), PDPN (1:500, KCl at 37°C for 40 min and fixed by 10 mL fixation solution including
FL-162, Santa Cruz), SPC (1:200, M-20, Santa Cruz), SPC (1:200, methanol and glacial acetic acid (3:1) at room temperature for
FL-197, Santa Cruz), LAMP3 (1:200, 12632-1-AP, Proteintech); 30 min. Suspension with chromosomes was dropped and spread on
bronchiolar cell markers: CC10 (1:200, T-18, Santa Cruz), acety- slides. Samples on slides were treated by 0.0005% trypsin for 5 min
lated-α-Tubulin (1:1000, 6-11B-1, Abcam), MUC5AC (1:500, 45M1, and stained with 15% Giemsa (Sigma-Aldrich, USA). Banding pat-
Thermo), FOXJ1 (1:200, 2A5, eBioscience); vasculature markers: terns on chromosome spreads were checked for more than 15
CD31 (1:100, M-20, Santa Cruz), CD34 (1:1000, EP373Y, Abcam); mitotic phases and all of them are normal human cells. All cells used
myofibroblast marker: α-SMA (1:500, 1A4, DAKO), Fibronectin for clinical purpose were subjected to karyotyping in prior to
(1:500, F14, Abcam), others: KI67 (1:200, RM-9106, Thermo), GFP transplantation.
(1:200, B-2, Santa Cruz), GFP (1:200, FL, Santa Cruz), GFP (1:200,
T-19, Santa Cruz), ITGB1 (1:500, ERP16895, Abcam), Human Quantitative reverse transcription PCR
specific Lamin A+C (1:200, EPR4100, Abcam). Alexa Fluor-conju-
gated Donkey 488/594/647 (1:200, Life Technologies, USA) were Total RNA from tissues or cells were isolated using the RNeasy mini
used as secondary antibodies. For antibodies of low reactivity, Bio- kit with DNase digestion according to the manufacturer’s instructions
tin-Streptavidin signal amplification system (Life Technologies, USA) (Qiagen). RNA quality was determined by SimpliNano (GE Health-
Protein & Cell
was used. After counterstaining with DAPI (Roche, USA), samples care). 1 μg total RNA was reverse-transcribed into cDNA with Pri-
were treated with 0.1% Sudan Black (Sigma, USA) for 1 min to meScript™1st Strand cDNA synthesis Kit (TaKaRa). The real-time
remove autofluorescence and then mounted with VECTASHIELD® PCR assays were performed on an ABI 7500 real-time PCR system
Mounting medium (Vector labs, USA). Images were visualized under (Applied Biosystems) according to the instructions of SYBR® Premix
fluorescence microscope (Nikon 80i and Eclipse Ti, Nikon, Japan) or Ex Taq™II (Tli RNaseH Plus, Takara). qPCR reactions were set as
fluorescence stereomicroscope (MVX10, Olympus, Japan). Confo- following: 95°C for 2 min, then 40 cycles of 95°C for 10 s, and 60°C
cal images were taken under Nikon A1R microscope (Nikon, Japan). for 40 s. Melt curve stage was added after PCR amplification stage.
The threshold crossing value (Ct) of each transcript was normalized
RNA-sequencing and bioinformatics to reference genes (β-Actin or GAPDH). The relative expression
level of each genes was calculated using the 2−ΔΔCt method.
SOX9+ BCs isolated from two donors and their corresponding brush- Sequence of primer pairs for qPCR was listed in Supplementary
off specimens were subjected to RNA-Seq analysis. The total RNA Table.
concentration and RIN were measured by Agilent 2100 Bioanalyzer
(Agilent). For human SOX9+ BCs, 200 ng total RNA sample was Animal tissue histology
purified, and the first-strand cDNA was synthesized using first strand
master mix and super script II (Life Technologies). Second strand All animal experiments were conducted according to guidelines
master mix (Life Technologies) was then used to synthesize the approved by University Association for Laboratory Animal Science.
second-strand cDNA. After cDNA purification and adapter ligation, NOD-SCID mice (female or male, 6–10 weeks, The Jackson Lab-
PCR amplification was performed to enrich the cDNA fragments. For oratory, USA) were euthanized at proper time points and the dia-
brush-off samples, after RNA extraction and quality control, cDNA phragm was carefully cut open without touching the lung. In situ
was prepared using the SMARTer Ultra Low RNA Kit (Clontech) for fixation by injecting 3.7% formaldehyde (Sigma, USA) through tra-
Illumina sequencing. Low Input Library Prep Kit (Clontech) was then chea was performed using 29 G needle. Then the lung was dis-
used for library construction. The library quantity and quality was sected and fixed in 3.7% formaldehyde at 4°C overnight. For
verified by Agilent 2100 Bioanalyzer and real-time quantitative PCR. cryosection, the fixed lung was settled by 30% sucrose before
Then the library is sequenced using Illumina HiSeq 4000. Clean data embedding into the Tissue-Tek O.C.T compound (Sakura, Japan),
were acquired from raw data (fastq format) using the NGSQC Toolkit the 5–10 μm sections were cut using a cryotome (Leica microsys-
by removing low-quality reads. Clean RNA-seq reads were mapped tem, Germany). For paraffin section, the lung was dehydrated by
to the reference genome (Ensembl, GRCh37) using Tophat v2.0.049 gradient ethanol and processed in an automatic tissue processor,
using default settings. then embedded into the paraffin blocks. All the samples were sliced
With genome mapping result, gene expression level was calcu- into 5–7 μm thickness using microtome (Leica microsystem, Ger-
lated with RSEM software (v1.2.12). Transcript levels were quanti- many). Haematoxylin and eosin (H&E) staining was performed fol-
fied as fragments per kilobase of transcript per million mapped reads lowing standard protocol. Masson trichrome staining was performed
(FPKM). Pearson correlation coefficient between samples was cal- following the manual of Trichrome Staining Kit manual.
culated by R scripts (3.2.3). Heatmap was generated using R scripts.
The protein-protein interactions were retrieved from Human Protein Small animal micro-CT
Reference Database (HPRD, release 9) and visualized with Cytos-
Micro-CT was used to monitor mouse lung damage before trans-
cape (v3.3.0). SNPs were called by GATK (v3.4-0).
plantation. Mice were anesthetized through intraperitoneal injection
Karyotyping
with chloral hydrate and fixed with tap. Lung image was obtained the same slide was stained with LI Silver Enhancement Kit following
using a volumetric micro-CT scanner without respiratory gating manufacturer’s instruction (Thermo, USA). Brown color indicates the
(TriumphTM, Gamma medica-ideas, Northridge, USA). Scanning precipitation of AuNPs after reaction with silver
was performed at 70 kV, 350 μA. The number of projections were
512 slices and total acquire time was around 4.27 min. All data were Western Blotting
converted into digital imaging by TRIUMPH ‘X-O’ CT system
Cells were washed in cold PBS and harvested by plastic scraper.
software.
Collected lung tissues were washed in cold PBS, ground and lysed
by electric tissue grinder in RIPA buffer (150 mmol/L sodium chlo-
SOX9+ BC transplantation in mouse
ride, 0.5% Triton-X100, 0.5% sodium deoxycholate, 5 mmol/L EDTA,
Adult NOD-SCID mice (female or male, 6–10 weeks, The Jackson 0.1% SDS, 50 mmol/L Tris-HCl, pH 7.5) with protease inhibitors
Laboratory, USA) maintained in SPF animal facilities were used for cocktail (Roche, USA). Approximately 30 μg total protein from each
xeno-transplantation experiments. Mouse lung was injured by sample was loaded. Samples were separated on a 10% SDS poly-
intratracheally instilling with 3 U/kg body weight bleomycin (Sel- acrylamide gel and transferred to PVDF membranes (Roche, USA)
leckchem, USA) eight days prior to transplantation. Mouse lung was with electrophoresis blotting transfer apparatus. The membranes
monitored by small animal micro-CT before transplantation to verify were blocked with 5% dehydrated milk for 1.5 h and then incubated
lung damage. Then mice were anesthetized by I.P injection of 3% with primary antibodies overnight. The next day, the membranes
chloral hydrate and rested on a stand gesture. One million GFP- were incubated with horseradish peroxidase-conjugated secondary
This work was funded by The National Key Research and Develop- org/10.1021/ja801631c
ment Program of China (2017YFA0104600), Youth 1000 Talent Plan Chilosi M, Poletti V, Murer B, Lestani M, Cancellieri A, Montagna L,
of China to W. Zuo, Tongji University (Basic Scientific Research-In- Piccoli P, Cangi G, Semenzato G, Doglioni C (2002) Abnormal re-
terdisciplinary Fund and 985 Grant to W. Zuo), Shanghai Pulmonary epithelialization and lung remodeling in idiopathic pulmonary
Hospital (Annual Grant to W. Zuo), National Science Fund Committee fibrosis: the role of deltaN-p63. Laboratory investigation; a journal
of China (81570091 and 81770073 to W. Zuo), Academic Leadership of technical methods and pathology 82:1335–1345
Fund of Shanghai (16XD1403100 to T. Ren), Postdoc Grant of China Clevers H (2013) The intestinal crypt, a prototype stem cell
(2016M601651 to Y. Ma and 2016M591716 to Q. Ma) and Kiangnan compartment. Cell 154:274–284. https://doi.org/10.1016/j.cell.
Stem Cell Institute. We thank Prof. Zhongwei Li for comment on the 2013.07.004
manuscript, and thank Profs. Frank McKeon, Wa Xian and Nanshan Copelan EA (2006) Hematopoietic stem-cell transplantation. The
Zhong for helpful discussion of this work, and thank Profs. Xiaoqing New England journal of medicine 354:1813–1826. https://doi.org/
Zhang, Long Zhang and Huiping Li for reagents help. 10.1056/NEJMra052638
Desai TJ, Brownfield DG, Krasnow MA (2014) Alveolar progenitor
COMPLIANCE WITH ETHICS GUIDELINES and stem cells in lung development, renewal and cancer. Nature
507:190–194. https://doi.org/10.1038/nature12930
Qiwang Ma, Yu Ma, Xiaotian Dai, Tao Ren, Yingjie Fu, Wenbin Liu,
Gallico GG 3rd, O’Connor NE, Compton CC, Kehinde O, Green H
Yufei Han, Yingchuan Wu, Yu Cheng, Ting Zhang and Wei Zuo
(1984) Permanent coverage of large burn wounds with autolo-
declare that they have no conflict of interest. All institutional and
gous cultured human epithelium. The New England journal of
national guidelines for the care and use of laboratory animals were
medicine 311:448–451. https://doi.org/10.1056/NEJM198408163
followed. All procedures followed were in accordance with the ethical
110706
standards of the responsible committee on human experimentation
Hackett NR, Shaykhiev R, Walters MS, Wang R, Zwick RK, Ferris B,
(institutional and national) and with the Helsinki Declaration of 1975,
Witover B, Salit J, Crystal RG (2011) The human airway epithelial
as revised in 2000 (Mannino 2002). Informed consent was obtained
basal cell transcriptome. PloS one 6:e18378. https://doi.org/10.
from all patients for being included in the study.
1371/journal.pone.0018378
Hogan BL, Barkauskas CE, Chapman HA, Epstein JA, Jain R, Hsia
CC, Niklason L, Calle E, Le A, Randell SH, Rock J, Snitow M,
OPEN ACCESS
Krummel M, Stripp BR, Vu T, White ES, Whitsett JA, Morrisey EE
This article is distributed under the terms of the Creative Commons (2014) Repair and regeneration of the respiratory system:
Attribution 4.0 International License (http://creativecommons.org/ complexity, plasticity, and mechanisms of lung stem cell function.
licenses/by/4.0/), which permits unrestricted use, distribution, and Cell stem cell 15:123–138. https://doi.org/10.1016/j.stem.2014.
reproduction in any medium, provided you give appropriate credit to 07.012
the original author(s) and the source, provide a link to the Creative Huang SX, Islam MN, O’Neill J, Hu Z, Yang YG, Chen YW, Mumau
Commons license, and indicate if changes were made. M, Green MD, Vunjak-Novakovic G, Bhattacharya J, Snoeck HW
(2014) Efficient generation of lung and airway epithelial cells from
human pluripotent stem cells. Nature biotechnology 32:84–91.
REFERENCES https://doi.org/10.1038/nbt.2754
Ke MT, Fujimoto S, Imai T (2013) SeeDB: a simple and morphology-
Asano S, Takemura T, Katoh K, Taneda M, Kitagawa M (2011)
preserving optical clearing agent for neuronal circuit
Epithelial regeneration after diffuse alveolar damage in relation to
reconstruction. Nature neuroscience 16:1154–1161. https://doi. Rama P, Matuska S, Paganoni G, Spinelli A, De Luca M, Pellegrini G
org/10.1038/nn.3447 (2010) Limbal stem-cell therapy and long-term corneal regener-
Kim CF, Jackson EL, Woolfenden AE, Lawrence S, Babar I, Vogel S, ation. The New England journal of medicine 363:147–155. https://
Crowley D, Bronson RT, Jacks T (2005) Identification of doi.org/10.1056/NEJMoa0905955
bronchioalveolar stem cells in normal lung and lung cancer. Cell Rockich BE, Hrycaj SM, Shih HP, Nagy MS, Ferguson MA, Kopp JL,
121:823–835. https://doi.org/10.1016/j.cell.2005.03.032 Sander M, Wellik DM, Spence JR (2013) Sox9 plays multiple
King TE, Bradford WZ, Castro-Bernardini S, Fagan EA, Glaspole I, roles in the lung epithelium during branching morphogenesis.
Glassberg MK, Gorina E, Hopkins PM, Kardatzke D, Lancaster L, Proceedings of the National Academy of Sciences of the United
Lederer DJ, Nathan SD, Pereira CA, Sahn SA, Sussman R, States of America 110:E4456–4464. https://doi.org/10.1073/
Swigris JJ, Noble PW, Group AS (2014) A phase 3 trial of pnas.1311847110
pirfenidone in patients with idiopathic pulmonary fibrosis. The Rosen C, Shezen E, Aronovich A, Klionsky YZ, Yaakov Y, Assayag
New England journal of medicine 370:2083–2092. https://doi.org/ M, Biton IE, Tal O, Shakhar G, Ben-Hur H, Shneider D, Vaknin Z,
10.1056/nejmoa1402582 Sadan O, Evron S, Freud E, Shoseyov D, Wilschanski M,
Kotton DN, Morrisey EE (2014) Lung regeneration: mechanisms, Berkman N, Fibbe WE, Hagin D, Hillel-Karniel C, Krentsis IM,
applications and emerging stem cell populations. Nature medi- Bachar-Lustig E, Reisner Y (2015) Preconditioning allows
cine 20:822–832. https://doi.org/10.1038/nm.3642 engraftment of mouse and human embryonic lung cells, enabling
Mannino DM (2002) COPD: epidemiology, prevalence, morbidity lung repair in mice. Nature medicine 21:869–879. https://doi.org/
and mortality, and disease heterogeneity. Chest 121:121S–126S 10.1038/nm.3889
brush for the diagnosis of pulmonary infections. Chest 81:556– Zuo W, Zhang T, Wu DZ, Guan SP, Liew AA, Yamamoto Y, Wang X,
562 Lim SJ, Vincent M, Lessard M, Crum CP, Xian W, McKeon F
Zhang HY, Gharaee-Kermani M, Zhang K, Karmiol S, Phan SH (2015) p63(+)Krt5(+) distal airway stem cells are essential for
(1996) Lung fibroblast alpha-smooth muscle actin expression and lung regeneration. Nature 517:616–620. https://doi.org/10.1038/
contractile phenotype in bleomycin-induced pulmonary fibrosis. nature13903
The American journal of pathology 148:527–537
Protein & Cell
Abstract
Background: Gas exchange represents the key physiological function of the lung, and is dependent upon proper
formation of the delicate alveolar structure. Malformation or destruction of the alveolar gas-exchange regions are
key histopathological hallmarks of diseases such as bronchopulmonary dysplasia (BPD), chronic obstructive pulmonary
disease (COPD), and pulmonary fibrosis; all of which are characterized by perturbations to the alveolo-capillary barrier
structure. Impaired gas-exchange is the primary initial consequence of these perturbations, resulting in severe clinical
symptoms, reduced quality of life, and death. The pronounced morbidity and mortality associated with malformation
or destruction of alveoli underscores a pressing need for new therapeutic concepts. The re-induction of alveolarization
in diseased lungs is a new and exciting concept in a regenerative medicine approach to manage pulmonary diseases
that are characterized by an absence of alveoli.
Main text: Mechanisms of alveolarization first need to be understood, to identify pathways and mediators that may be
exploited to drive the induction of alveolarization in the diseased lung. With this in mind, a variety of candidate cell-types,
pathways, and molecular mediators have recently been identified. Using lineage tracing approaches and lung injury
models, new progenitor cells for epithelial and mesenchymal cell types – as well as cell lineages which are able to
acquire stem cell properties – have been discovered. However, the underlying mechanisms that orchestrate the complex
process of lung alveolar septation remain largely unknown.
Conclusion: While important progress has been made, further characterization of the contributing cell-types, the cell
type-specific molecular signatures, and the time-dependent chemical and mechanical processes in the developing, adult
and diseased lung is needed in order to implement a regenerative therapeutic approach for pulmonary diseases.
Keywords: Alveolarization, Neo-alveolarization, Regeneration
this regenerative approach remain a crucial topic to be Nature of alveolar epithelial cells during the saccular
discussed, and likely directions are highlighted in the last stage and alveolarization
paragraph of this review. During the saccular stage of lung development, distal parts
of the lung contain air sacs (sacculi or saccules) which are
Main text lined with an epithelial layer that originates from the fore-
Alveolarization gut endoderm, and consists of differentiated alveolar epi-
Alveolarization represents a process during lung develop- thelial type I cells (AECI) and alveolar epithelial type II
ment that leads to the formation and maturation of the cells (AECII) [5]. During the saccular stage, AECl and
distal parts of the lung: the alveoli. In rats and mice, alveo- AECII are derived from a common bipotent progenitor
larization takes place postnatally, whereas in humans, cell (Fig. 1) [21]. As described in a later section of this re-
alveolar development begins prior to birth [3, 15–20]. Due view, mechanical forces and fibroblast growth factor
to the limited availability of human tissue, most of the (FGF) 10 are amongst recently-identified regulators of this
studies dissecting the principles of alveolarization have differentiation process [22–24]. After alveolarization, and
been conducted in rodents. At birth, the murine lung is in in the adult lung, AECII acquire stem cell properties and
the saccular stage, which lasts from embryonic day (E)18.5 are capable of self-renewal to replace AECl after injury
until postnatal day (P)5 [5], and which is comparable to [21, 25]. Single-cell sequencing of distal lung epithelial
the stage of lung development in which most pre-term cells during the saccular stage confirmed a bipotent pro-
born human infants are undergoing at the time of prema- genitor for AECl and AECll and revealed further cell-type
ture rupture of membranes. As such, term-born mouse specific markers and subpopulations during the process of
and rats are often used to model the lung in pre-term differentiation [26]. Cuboidal AECII cells are capable of
born infants, with the important caveat that these lungs producing surfactant proteins and lipids which decrease
are perfectly competent for effective gas exchange at birth, the surface tension of the alveoli [5, 27]. The cell surface
contrasting with the situation in pre-term born human of squamous AECl expands drastically during the alveolar
infants. stage [28]. One AECl covers multiple alveoli during
Fig. 1 Alveolar epithelial cells during alveolarization. During the saccular stage, alveolar epithelial type I cells (AECI) and alveolar epithelial type II
cells (AECII) are derived from a common bipotent progenitor cell. After differentiation single AECl can cover multiple alveoli during alveolarization
and in the adult lung
Rodríguez-Castillo et al. Respiratory Research (2018) 19:148 Page 3 of 11
alveolarization and in the adult lung (Fig. 1) [28–30]. Due elastin further increases (Fig. 2) [5, 6, 33, 39]. One recent
to the close proximity of AECI to the capillary network, study has carefully dissected structural changes in the
and the comparatively larger surface area when compared developing alveoli, and elastin localization during alveolar-
to AECII, AECl represent the site of gas exchange [30]. ization, using 3D imaging techniques [14]. These analyses
pointed out that secondary crests arise as ridges into the
Role of mesenchymal cells and extracellular matrix during alveolar air-sac lumen. An organized network resembling
alveolarization a “fishnet” composed of αSMA and the ECM component
Further components of the alveolar air-sac walls (primary elastin runs within the ridges [14]. Quite similar observa-
septa) are endothelial cells that originate from the lung tions were made in a further study that analyzed the spatial
endoderm [8, 31, 32], and a variety of interstitial cell-types and temporal changes in elastin and laminin distribution
such as fibroblasts, which originate from the lung meso- during alveolarization [40]. That study demonstrated that
derm [8, 33–35]. In rat lung fibroblasts, an increase in ret- elastin fibers formed ring-like structures which were local-
inoic acid levels has been demonstrated during secondary ized to the saccular openings, and later on were intercon-
septation [36], and retinoic acid has been proposed to nected by further elastin fibers [40]. Crosslinking of ECM
impact elastin production [36]. These studies highlighted components has been demonstrated to be altered during
a potential regenerative activity of retinoids in the aberrant lung development [41]. The downstream signaling
lung. Retinoic acid administration to rats has been dem- molecule of the sonic hedgehog pathway, Gli-1, has been
onstrated to promote postnatal alveolarization, and to at- demonstrated to label a cell-lineage which gives rise to sec-
tenuate elastase-induced pulmonary emphysema [37, 38]. ondary crest myofibroblasts (Fig. 2) [42–44]. The depos-
During the saccular stage, elastin expression increases, ition of ECM components and the presence of alveolar
and elastin deposition by fibroblasts takes place [39]. In myofibroblasts seems to be an attribute for secondary sept-
rodents, by P4, so-called secondary septa appear in the ation and has been demonstrated to be dependent on
primary septa at sites of elastin deposition [5, 31]. At the platelet-derived growth factor (PDGF)-A signaling [6, 45].
tip (secondary crest) of these still-immature secondary The ligand PDGF-A is produced by epithelial cells and
septa, α-smooth muscle actin (αSMA)+ myofibroblasts ap- signals via the cognate receptor PDGF receptor (PDGFR)α,
pear, and the expression of ECM components such as which is expressed by mesenchymal cells [6, 46, 47]. In vivo
Fig. 2 Mesenchymal cell-types during alveolarization. Alveolar mesenchymal fibroblasts such as myofibroblasts and lipofibroblasts differentiate
from mesenchymal progenitor cells. Key cell lineages involved are the platelet-derived growth factor receptor (PDGFR)α lineage, the fibroblast
growth factor (FGF)10 lineage and the GLI-Kruppel family member (Gli-1) lineage. Elastin deposition by myofibroblasts takes place at the so called
“secondary crests” visualized in 2D lung sections
Rodríguez-Castillo et al. Respiratory Research (2018) 19:148 Page 4 of 11
labeling and lineage-tracing studies of PDGFRα+ cells have [53–57]. Further molecules such as peroxisome
demonstrated that pulmonary PDGFRα+ cells serve as pro- proliferator-activated receptor (PPAR)γ, cellular retinoic
genitor cells for peribronchial and alveolar myofibroblasts, acid binding protein (CRABP), and the transcription fac-
as well as for a proportion of the pulmonary lipofibroblasts, tor TCF21 are expressed by lung mesenchymal lipofibro-
and are present in the primary and secondary septa (Fig. 2) blasts (Fig. 3) [36, 58–61]. Lineage-tracing of FGF10+ cells
[6, 45, 48, 49]. Further studies have highlighted the pro- revealed labeling of a subset of the lipofibroblast population
genitor nature of PDGFRα+ cells for peribronchial smooth [62]. Furthermore, lipofibroblasts have been demonstrated
muscle cells, myofibroblasts and lipofibroblasts during pre- to support the synthesis of surfactant in AECII by providing
natal lung development and alveolarization using time-series triacylglycerols to AECII in a leptin- and stretch-dependent
RNA-Seq analyses and immunophenotyping [50, 51]. An- manner (Fig. 3) [53, 63]. During the period of secondary
other recent study has described the spatiotemporal distri- septation, the number of lipofibroblasts has been demon-
bution of PDGFRα and the two PDGF ligands PDGF-A and strated to increase and peak at the same time as the peak of
PDGF-C over the course of lung development [52]. Ex- secondary septation, at P7 [64]. Activation of PPARγ using
pression of the ligands was detected in epithelial and rosiglitazone (which promotes the lipofibroblast phenotype)
smooth muscle cells, whereas the expression of PDGFRα in rat pre- and post-natally has been demonstrated to be
was located to different mesenchymal cell populations protective against the structural changes that occur during
[52], which is consistent with the prevailing view that the development of hyperoxia-induced lung injury [65, 66].
epithelial-mesenchymal interactions are key mediators of However, the presence of lipofibroblasts in the human lung
lung development. remains controversial [54, 57, 67]. In contrast to rosiglita-
zone, which induces a lipogenic phenotype; Other reagents,
Possible role for lipogenic versus myogenic fibroblast stimuli and factors such as nicotine, mechanical forces,
phenotypes during alveolarization PDGFRα and transforming growth factor (TGF)-β have
Lipofibroblasts are also located within the primary and been demonstrated to induce a myogenic phenotype: nico-
secondary septa in close proximity to AECII [48, 49, 53]. tine treatment of isolated fibroblasts in vitro led to a myo-
Lipid droplets and the expression of adipocyte differenti- genic phenotype (differentiation from lipofibroblasts to
ation related protein (ADRP, encoded by the Plin2 gene) myofibroblasts) and could be reversed by rosiglitazone
represent phenotypic characteristics of lipofibroblasts treatment [68]. Furthermore it has been demonstrated that
Fig. 3 Lipogenic versus myogenic fibroblast phenotype. Lipogenic (lipofibroblast) and myogenic (myofibroblast) fibroblasts differentiate during
early lung development and the saccular stage. Lipofibrobasts support alveolar epithelial type II cells (AECII) cell function via an intercellular
crosstalk mediated by stretch, parathyroid hormone-related peptide (PTHRP), prostaglandin E2 (PGE2) and leptin while myofibroblasts produce
extracellular matrix molecules such as elastin. Activation of peroxisome proliferator activated receptor (PPAR)γ by Rosiglitazone promotes the
lipogenic phenotype. Stretch and transforming growth factor (TGF)-β induce the myogenic phenotype
Rodríguez-Castillo et al. Respiratory Research (2018) 19:148 Page 5 of 11
mechanical forces could stimulate the differentiation of fi- differentiation and function during alveolarization. The
broblasts towards myofibroblasts (Fig. 3) [69]. Expression formation of alveolospheres by alveolar epithelial cells
and activation of PDGFRα has also been demonstrated be (AEC) has been demonstrated to be supported by
involved in driving fibroblast differentiation towards a myo- PDGFRα+ cells in vitro [25]. In line with these findings, a
genic phenotype [6, 49, 70]. The expression of the PDGFRα mesenchymal cell population from the Axin2 cell-lineage
chain has been demonstrated to be regulated by TGF-β which expressed PDGFRα has been demonstrated to sup-
(Fig. 3) [71]. A very recent study has also carefully dissected port AECI and AECII differentiation and function [12]. In
the role of the ligand PDGF-A using mice carrying a floxed contrast, a rare population of AECII which express
Pdgfrα allele in combination with a Sftpc-Cre mouse strain Axin2 has been demonstrated to have alveolar stem cell
[72]. The authors demonstrated that PDGF-A was needed activity in the adult lung and to be supported by juxta-
for myofibroblast formation and proliferation as well as for crine Wnt signals from neighboring fibroblasts during
the regulation of AECII proliferation [72]. Taken together, homeostasis, and autocrine Wnt signals upon severe in-
tightly regulated differentiation of alveolar fibroblasts to- jury [80]. Further evidence for mesenchymal-epithelial
wards the myogenic or the lipogenic phenotype seems to interactions being key drivers of cell differentiation during
be relevant for secondary septation. Furthermore, neuropil- alveolar development has emerged from analyses compar-
lin 1 has been demonstrated to impact the PDGF-A axis ing two distinct mesenchymal cell populations of the
via activation of Src kinases and to be required for alveolar leucine-rich repeat-containing G protein-coupled receptor
mesenchymal cell migration [73]. Following this line, a re- (LGR)5 and LGR6 cell lineage: the LGR5 and LGR6 lin-
duced expression of PDGFRα and PDGFRβ has been dem- eages have been demonstrated to support cell function
onstrated in mesenchymal cells of infants who develop and differentiation of either alveolar or bronchiolar epi-
BPD [74]. Furthermore, the signaling via FGF members has thelial cells [13]. Furthermore human alveolar fibroblasts
been demonstrated to impact the formation of myofibro- have been demonstrated to exhibit direct intercellular
blasts from PDGFRα+ cells, as well as alveolar regeneration contact with AECl, AECll, capillaries and pericytes [81].
per se [75]. Deficiency of FGF10 has been demonstrated to This strategic localization might position fibroblasts to
be causative for the lethality in a mouse model of BPD [76]. mediate crosstalk between the epithelial and the capillary
Another mediator, Thy-1, which is expressed on lympho- endothelial cells, as well as epithelial-mesenchymal inter-
cytes and fibroblasts, has been demonstrated to severely actions which are key drivers of cellular differentiation
impact alveolarization: an arrest of alveolarization itself has and function during alveolarization.
been demonstrated upon global loss of Thy-1 (CD90) [7].
The glycoprotein Thy-1 inhibits TGF-β activation, which Significance of mechanical forces in alveolarization
leads to a reduced myogenic phenotype [7]. Since Thy-1 Further key drivers of cellular processes that promote
also is expressed on lipofibroblasts it might represent a alveolarization are mechanical forces. An understanding
molecule impacting on the balanced appearance of lipo- of lung alveolarization is based primarily on understand-
genic and myogenic fibroblasts during secondary septation ing processes such as cellular differentiation, localization,
(Fig. 3) [7, 48]. However, the role of leucocytes might be of production of ECM molecules and underlying signaling
relevance since inflammatory cells have been proposed to cascades. However, the impact of mechanical forces on
play a role – and to be present – during lung development these processes has been analyzed in the context of
[77], and resident alveolar macrophages have recently been intercellular crosstalk and the differentiation of alveolar epi-
implicated as master regulators of arrested lung devel- thelial cells [22–24, 63, 82]. Surfactant production of AECII
opment [78]. Apart from the role of lipogenic and myo- has been demonstrated to be stretch-dependent via a
genic phenotypes of fibroblasts in lung development, stretch-induced de novo synthesis of phosphatidyl choline
lipogenic and myogenic fibroblast phenotypes are also by AECII [63]. Furthermore, stretch-dependent interactions
involved in the progression and resolution of pulmon- with fibroblasts via parathyroid hormone-related peptide
ary fibrosis, which has been demonstrated in a murine (PTHRP), prostaglandin E2 (PGE2) and fibroblast-derived
model of bleomycin-induced lung fibrosis [79]. This leptin increased surfactant synthesis (Fig. 3) [63, 82]. A very
mechanism supports the hypothesis that understanding recent study carefully dissected the role of mechanical
the nature and development of lipogenic versus myogenic forces on alveolar epithelial cell differentiation using live
fibroblast phenotypic transformation might help to de- imaging techniques [23]. Mechanical forces generated
velop new therapeutic strategies for pulmonary diseases. by inhalation of amniotic fluid by prenatal breathing
movements were essential for the differentiation of AECI
Epithelial-mesenchymal interactions during [23]. Furthermore FGF10/FGFR2 signaling has been dem-
alveolarization onstrated to prevent flattening of alveolar progenitor cells,
The Interaction between mesenchymal and epithelial protecting AECII differentiation [23]. Mechanical forces
cells has been demonstrated to be essential for cellular have also been demonstrated to impact airway tube
Rodríguez-Castillo et al. Respiratory Research (2018) 19:148 Page 6 of 11
receptor (EGFR) resulting in expansion of epithelial pro- cell-type specific and inducible lineage tracing, cell and
genitor cells [92]. Taken together, growth of the vascular gene modulation, as well as cell ablation based on the
system, and vascular mediators and growth factors such as CreERT2/loxP system [101]. Differentiation processes,
SDF-1 and VEGF-A, are further drivers of epithelial cell stem cell features, and morphological appearance can be
function during lung regrowth after PNX. analyzed in combination with high quality imaging ap-
proaches. Even cell-lineage and single-cell sequencing
Impact of mesenchymal cells, mechanical forces and analyses can be performed at different time-points of
innate lymphoid cells on neo-alveolarization alveolarization. However, a key limitation of studying
Further key drivers of lung growth after PNX have been alveolarization in mice is the need to validate identified
identified in mesenchymal lineages. There is evidence candidate cells and genes in human tissue, which has to
for the participation of PDGFRα+ cells in compensatory be performed to ensure the ranslational value of the
lung regrowth [93, 94]. Gene expression profiling of pathway identified. Furthermore, it remains to be discov-
postnatal and adult mouse lungs undergoing PNX re- ered which candidates relevant for alveolarization in
vealed concordantly as well as variably regulated genes postnatal mice might also be relevant for lung regener-
[95]. There is a growing body of evidence that features ation. Taken together, analyzing alveolarization in mice
of bulk alveolarization occur during compensatory lung in vivo represents a suitable model, but validation of
growth of the adult lung, but alternative mechanisms candidate factors in human lung tissue and adult mouse
need to be considered in addition. Cell proliferation, lung tissue has to be considered.
change in mechanical forces, ECM remodeling and the
activation of different signaling pathways have been Elucidating neo-alveolarization using the PNX model
demonstrated in response to PNX [86, 92, 95–97]. Fi- Exploitation of the model of compensatory lung
nally, myeloid cells and type 2 innate lymphoid cells growth after PNX will provide a better understanding
(ICL2) have been demonstrated to hold regenerative of the cellular and molecular mechanism driving neo-
capacity during compensatory lung growth after PNX alveolarization. Beneficial aspects of this model are
[11]. A variety of cellular and molecular factors impact- represented by the possibility that the advantages of
ing on neo-alveolarization after PNX has been identi- transgenic mice can be used as well. Furthermore, alveolar
fied. However, target molecules capable to initiate neo- regrowth can be studied in the adult lung, which might be
alveolarization have not been recognized. Further more directly relevant to lung regeneration of the adult
combined lineage tracing, lineage ablation and cell lung. Like studying alveolarization in mice, a disadvantage
type specific loss or gain of function studies are needed to of studying regrowth after PNX in mice also remains, that
identify hierarchies and functions of cell types and driving factors might be different in mice and humans.
signaling cascades driving neo-alveolarization in the Furthermore, it has been demonstrated that strongest
adult lung. regrowth takes place in the cardiac lobe [86]. Therefore,
it might be necessary to restrict analyses to the
Models to approach lung regeneration cardiac lobe.
The induction of neo-alveolarization in the diseased lung
requires a detailed knowledge of the conditions which Perspectives
are necessary for alveolar septum formation. Therefore, A possible way to approach lung regeneration is to
the identification of suitable models to study alveolar better understand the composition of contributing cell
septum formation is essential. Transgenic tools [98–100] types, the molecular signature and plasticity of cells, and
and genome-wide screening approaches [83] have re- the ECM composition, which together drive alveolariza-
vealed a variety of contributing cellular and molecular tion and neo-alveolarization in order to identify possible
candidates and thus represent suitable tools to elucidate candidates for the induction of lung regeneration. Pro-
relevant candidate drivers for alveolarization. Therefore, genitor cells need to be identified. Single-cell genome-wide
alveolariaztion and neo-alveolarization after PNX are screening approaches lead to a broader and more complex
currently analyzed primarily in mice in vivo. However, understanding of cell lineages of different tissue compart-
some disadvantages of both models remain, and these ments [26, 102]. Better characterization of endodermal and
are discussed in the following sections. mesenchymal cell lineages including molecular signatures
might provide new insights into the cellular hierarchy of
Analyzing alveolarization alveolarization, and might reveal progenitor cell candidates.
Mice represent a valuable tool to analyze the process of Detailed time-dependent lineage tracing and cell depletion
alveolarization since alveolarization largely takes place approaches covering prenatal and postnatal periods is
postnatally in mice. Advantages of analyzing alveolariza- needed to discover cell type-specific commitment and func-
tion in vivo in mice are the possibility to perform tion during alveolarization. Furthermore, there is need to
Rodríguez-Castillo et al. Respiratory Research (2018) 19:148 Page 8 of 11
understand if factors driving bulk alveolarization in modulation of telomerase activity during lung devel-
the developing lung might also be capable of inducing opment might reveal possible target candidates that
neo-alveolarization in the adult lung. Cellular and impact lung regeneration. Furthermore, the identifica-
molecular candidates ensuring the homeostasis of the tion of possible targets to lung-specifically modulate
adult lung tissue additionally might be relevant for the process of aging might reveal targets for future
neo-alveolarization. A further aspect to approach al- therapeutic concepts. Regulation of the cell cycle rep-
veolar regeneration might include the impact of aging resents a key mechanism for tissue homeostasis and
and senescence on lung homeostasis and repair. As development. Signal transduction programs of cellular
an example, telomerase activity, which impacts on senescence cause an irreversible cell cycle arrest and
cellular senescence, has been demonstrated to be tis- thus might crucially impact lung regeneration [104].
sue specifically regulated in mice during development Apart from processes which drive alveolarization and
and aging [103]. This is relevant because lung-specific neo-alveolarization, processes which drive pulmonary
Fig. 4 Approaches for lung regeneration. Identification of progenitors or stem cells and mediators driving cellular differentiation during
alveolarization and neo-alveolarization might be used to identify possible target candidates to induce regeneration during pulmonary diseases
such as emphysema/chronic obstructive lung dsease and fibrosis. Promising target candidates are represented by platelet-derived growth factor
receptor (PDGFR)α+ cells and fibroblast growth factor (FGF)10 during alveolarization and bone morphogenic protein (BMP))/SMAD family member
(SMAD) signaling and stromal cell derived factor (SDF)-1 during neo-alveolarization
Rodríguez-Castillo et al. Respiratory Research (2018) 19:148 Page 9 of 11
diseases have to be considered. Finally, target candidates Center SFB1213 (to REM), Clinical Research Unit KFL309 (to REM) and
identified in mice, need to be validated in human tissue. individual research grant Mo 1879/1 (to REM).
20. Madurga A, et al. Recent advances in late lung development and the 49. McGowan SE, et al. Platelet-derived growth factor receptor-alpha-expressing
pathogenesis of bronchopulmonary dysplasia. Am J Physiol Lung Cell Mol cells localize to the alveolar entry ring and have characteristics of
Physiol. 2013;305(12):L893–905. myofibroblasts during pulmonary alveolar septal formation. Anat Rec
21. Desai TJ, Brownfield DG, Krasnow MA. Alveolar progenitor and stem cells (Hoboken). 2008;291(12):1649–61.
in lung development, renewal and cancer. Nature. 2014;507(7491):190–4. 50. Endale M, et al. Dataset on transcriptional profiles and the developmental
22. Tang Z, et al. Mechanical forces program the orientation of cell division characteristics of PDGFRalpha expressing lung fibroblasts. Data Brief. 2017;
during airway tube morphogenesis. Dev Cell. 2018;44(3):313–325 e5. 13:415–31.
23. Li J, et al. The strength of mechanical forces determines the differentiation 51. Endale M, et al. Temporal, spatial, and phenotypical changes of PDGFRalpha
of alveolar epithelial cells. Dev Cell. 2018;44(3):297–312 e5. expressing fibroblasts during late lung development. Dev Biol. 2017;425;161–75.
24. Hogan BLM. Integrating mechanical force into lung development. Dev Cell. 52. Gouveia L, Betsholtz C, Andrae J. Expression analysis of platelet-derived
2018;44(3):273–5. growth factor receptor alpha and its ligands in the developing mouse lung.
25. Barkauskas CE, et al. Type 2 alveolar cells are stem cells in adult lung. J Clin Physiol Rep. 2017;5(6):1–12.
Invest. 2013;123(7):3025–36. 53. McGowan SE, Torday JS. The pulmonary lipofibroblast (lipid interstitial cell) and
26. Treutlein B, et al. Reconstructing lineage hierarchies of the distal lung its contributions to alveolar development. Annu Rev Physiol. 1997;59:43–62.
epithelium using single-cell RNA-seq. Nature. 2014;509(7500):371–5. 54. Rehan VK, et al. Evidence for the presence of lipofibroblasts in human lung.
27. Whitsett JA, Wert SE, Weaver TE. Alveolar surfactant homeostasis and the Exp Lung Res. 2006;32(8):379–93.
pathogenesis of pulmonary disease. Annu Rev Med. 2010;61:105–19. 55. Vaccaro C, Brody JS. Ultrastructure of developing alveoli. I. The role of the
28. Yang J, et al. Development and plasticity of alveolar type 1 cells. interstitial fibroblast. Anat Rec. 1978;192(4):467–79.
Development. 2016;143:54–65. 56. Imamura M, et al. ADRP stimulates lipid accumulation and lipid droplet
29. Weibel ER. The mystery of “non-nucleated plates” in the alveolar epithelium formation in murine fibroblasts. Am J Physiol Endocrinol Metab. 2002;283(4):
of the lung explained. Acta Anat (Basel). 1971;78(3):425–43. E775–83.
30. Weibel ER. On the tricks alveolar epithelial cells play to make a good lung. 57. Ahlbrecht K, McGowan SE. In search of the elusive lipofibroblast, in human
Am J Respir Crit Care Med. 2015;191(5):504–13. lungs. Am J Physiol Lung Cell Mol Physiol. 2014;307:L605–8.
31. Tschanz SA, et al. Rat lungs show a biphasic formation of new alveoli 58. Varisco BM, et al. Thy-1 signals through PPARgamma to promote
during postnatal development. J Appl Physiol (1985). 2014;117(1):89–95. lipofibroblast differentiation in the developing lung. Am J Respir Cell Mol
32. Yamamoto H, et al. Epithelial-vascular cross talk mediated by VEGF-A and Biol. 2012;46(6):765–72.
HGF signaling directs primary septae formation during distal lung 59. McGowan SE, McCoy DM. Regulation of fibroblast lipid-storage and
morphogenesis. Dev Biol. 2007;308(1):44–53. myofibroblast phenotypes during alveolar septation in mice. Am J Physiol
33. Chao CM, et al. A breath of fresh air on the mesenchyme: impact of Lung Cell Mol Physiol. 2014;307:L618–31.
impaired mesenchymal development on the pathogenesis of 60. McGowan SE, et al. Peroxisome proliferators alter lipid acquisition and
bronchopulmonary dysplasia. Front Med (Lausanne). 2015;2:27. elastin gene expression in neonatal rat lung fibroblasts. Am J Phys. 1997;
34. Ruiz-Camp J, Morty RE. Divergent fibroblast growth factor signaling 273(6 Pt 1):L1249–57.
pathways in lung fibroblast subsets: where do we go from here? Am J 61. Acharya A, et al. Efficient inducible Cre-mediated recombination in Tcf21
Physiol Lung Cell Mol Physiol. 2015;309(8):L751–5. cell lineages in the heart and kidney. Genesis. 2011;49(11):870–7.
35. Rinkevich Y, et al. Identification and prospective isolation of a mesothelial 62. El Agha E, et al. Fgf10-positive cells represent a progenitor cell
precursor lineage giving rise to smooth muscle cells and fibroblasts for population during lung development and postnatally. Development.
mammalian internal organs, and their vasculature. Nat Cell Biol. 2012;14(12): 2014;141(2):296–306.
1251–60. 63. Torday JS, Rehan VK. Stretch-stimulated surfactant synthesis is coordinated
36. McGowan SE, Harvey CS, Jackson SK. Retinoids, retinoic acid receptors, and by the paracrine actions of PTHrP and leptin. Am J Physiol Lung Cell Mol
cytoplasmic retinoid binding proteins in perinatal rat lung fibroblasts. Am J Physiol. 2002;283(1):L130–5.
Phys. 1995;269(4 Pt 1):L463–72. 64. Maksvytis HJ, Vaccaro C, Brody JS. Isolation and characterization of the lipid-
37. Massaro GD, Massaro D. Postnatal treatment with retinoic acid increases the containing interstitial cell from the developing rat lung. Lab Investig. 1981;
number of pulmonary alveoli in rats. Am J Phys. 1996;270(2 Pt 1):L305–10. 45(3):248–59.
38. Massaro GD, Massaro D. Retinoic acid treatment abrogates elastase-induced 65. Rehan VK, et al. Rosiglitazone, a peroxisome proliferator-activated receptor-
pulmonary emphysema in rats. Nat Med. 1997;3(6):675–7. gamma agonist, prevents hyperoxia-induced neonatal rat lung injury in
39. Mizikova I, Morty RE. The extracellular matrix in bronchopulmonary vivo. Pediatr Pulmonol. 2006;41(6):558–69.
dysplasia: target and source. Front Med (Lausanne). 2015;2:91. 66. Rehan VK, et al. Antenatally administered PPAR-gamma agonist
40. Luo Y, et al. Spatial and temporal changes in extracellular elastin and laminin rosiglitazone prevents hyperoxia-induced neonatal rat lung injury. Am J
distribution during lung alveolar development. Sci Rep. 2018;8(1):8334. Physiol Lung Cell Mol Physiol. 2010;299(5):L672–80.
41. Mizikova I, et al. Collagen and elastin cross-linking is altered during aberrant 67. Tahedl D, et al. How common is the lipid body-containing interstitial cell in the
late lung development associated with hyperoxia. Am J Physiol Lung Cell mammalian lung? Am J Physiol Lung Cell Mol Physiol. 2014;307(5):L386–94.
Mol Physiol. 2015;308(11):L1145–58. 68. Rehan VK, et al. Mechanism of nicotine-induced pulmonary fibroblast
42. Li C, et al. Progenitors of secondary crest myofibroblasts are developmentally transdifferentiation. Am J Physiol Lung Cell Mol Physiol. 2005;289(4):L667–76.
committed in early lung mesoderm. Stem Cells. 2015;33(3):999–1012. 69. Balestrini JL, et al. The mechanical memory of lung myofibroblasts. Integr
43. Kugler MC, et al. Sonic hedgehog signaling regulates Myofibroblast function Biol (Camb). 2012;4(4):410–21.
during alveolar septum formation in murine postnatal lung. Am J Respir 70. McGowan SE, McCoy DM. Fibroblasts expressing PDGF-receptor-alpha
Cell Mol Biol. 2017;57(3):280–93. diminish during alveolar septal thinning in mice. Pediatr Res. 2011;
44. Ahlfeld SK, Perl AK. A “GLI-tch” in alveolar Myofibroblast differentiation. Am J 70(1):44–9.
Respir Cell Mol Biol. 2017;57(3):261–2. 71. Gronwald RG, Seifert RA, Bowen-Pope DF. Differential regulation of
45. Lindahl P, et al. Alveogenesis failure in PDGF-A-deficient mice is coupled to expression of two platelet-derived growth factor receptor subunits by
lack of distal spreading of alveolar smooth muscle cell progenitors during transforming growth factor-beta. J Biol Chem. 1989;264(14):8120–5.
lung development. Development. 1997;124(20):3943–53. 72. Gouveia L, Betsholtz C, Andrae J. PDGF-A signaling is required for secondary
46. Bostrom H, Gritli-Linde A, Betsholtz C. PDGF-A/PDGF alpha-receptor alveolar septation and controls epithelial proliferation in the developing
signaling is required for lung growth and the formation of alveoli but not lung. Development. 2018;145(7).
for early lung branching morphogenesis. Dev Dyn. 2002;223(1):155–62. 73. McGowan SE, McCoy DM. Neuropilin-1and platelet-derived growth factor
47. Andrae J, et al. Characterization of platelet-derived growth factor-a receptors cooperatively regulate intermediate filaments and mesenchymal
expression in mouse tissues using a lacZ knock-in approach. PLoS One. cell migration during alveolar septation. Am J Physiol Lung Cell Mol Physiol.
2014;9(8):e105477. 2018;315:L102–15.
48. Ntokou A, et al. Characterization of the platelet-derived growth factor 74. Popova AP, et al. Reduced platelet-derived growth factor receptor
receptor alpha-positive cell lineage during murine late lung development. expression is a primary feature of human bronchopulmonary dysplasia. Am
Am J Physiol Lung Cell Mol Physiol. 2015;309(9):L942–58. J Physiol Lung Cell Mol Physiol. 2014;307(3):L231–9.
Rodríguez-Castillo et al. Respiratory Research (2018) 19:148 Page 11 of 11
75. Perl AK, Gale E. FGF signaling is required for myofibroblast differentiation
during alveolar regeneration. Am J Physiol Lung Cell Mol Physiol. 2009;
297(2):L299–308.
76. Chao CM, et al. Fgf10 deficiency is causative for lethality in a mouse model
of bronchopulmonary dysplasia. J Pathol. 2016;241:91–103.
77. Guilliams M, et al. Alveolar macrophages develop from fetal monocytes that
differentiate into long-lived cells in the first week of life via GM-CSF. J Exp
Med. 2013;210(10):1977–92.
78. Kalymbetova TV, et al. Resident alveolar macrophages are master regulators
of arrested alveolarization in experimental bronchopulmonary dysplasia. J
Pathol. 2018;245(2):153–9.
79. El Agha E, et al. Two-way conversion between Lipogenic and myogenic
fibroblastic phenotypes marks the progression and resolution of lung
fibrosis. Cell Stem Cell. 2016;20:261–73.
80. Nabhan AN, et al. Single-cell Wnt signaling niches maintain stemness of
alveolar type 2 cells. Science. 2018;359(6380):1118–23.
81. Sirianni FE, Chu FS, Walker DC. Human alveolar wall fibroblasts directly link
epithelial type 2 cells to capillary endothelium. Am J Respir Crit Care Med.
2003;168(12):1532–7.
82. Torday JS, et al. Leptin mediates the parathyroid hormone-related protein
paracrine stimulation of fetal lung maturation. Am J Physiol Lung Cell Mol
Physiol. 2002;282(3):L405–10.
83. Beauchemin KJ, et al. Temporal dynamics of the developing lung
transcriptome in three common inbred strains of laboratory mice reveals
multiple stages of postnatal alveolar development. PeerJ. 2016;4:e2318.
84. Butler JP, et al. Evidence for adult lung growth in humans. N Engl J Med.
2012;367(3):244–7.
85. Fehrenbach H, et al. Neoalveolarisation contributes to compensatory lung
growth following pneumonectomy in mice. Eur Respir J. 2008;31(3):515–22.
86. Voswinckel R, et al. Characterisation of post-pneumonectomy lung growth
in adult mice. Eur Respir J. 2004;24(4):524–32.
87. Hsia CC, et al. Compensatory lung growth occurs in adult dogs after right
pneumonectomy. J Clin Invest. 1994;94(1):405–12.
88. Chapman HA, et al. Integrin alpha6beta4 identifies an adult distal lung
epithelial population with regenerative potential in mice. J Clin Invest. 2011;
121(7):2855–62.
89. Chung MI, et al. Niche-mediated BMP/SMAD signaling regulates lung alveolar
stem cell proliferation and differentiation. Development. 2018;145(9)
90. Hogan B. Stemming lung disease? N Engl J Med. 2018;378(25):2439–40.
91. Ackermann M, et al. Sprouting and intussusceptive angiogenesis in
postpneumonectomy lung growth: mechanisms of alveolar
neovascularization. Angiogenesis. 2014;17(3):541–51.
92. Ding BS, et al. Endothelial-derived angiocrine signals induce and sustain
regenerative lung alveolarization. Cell. 2011;147(3):539–53.
93. Chen L, et al. Dynamic regulation of platelet-derived growth factor receptor
alpha expression in alveolar fibroblasts during realveolarization. Am J Respir
Cell Mol Biol. 2012;47(4):517–27.
94. Green J, et al. Diversity of interstitial lung fibroblasts is regulated by
PDGFRalpha kinase activity. Am J Respir Cell Mol Biol. 2016;54:532–45.
95. Wolff JC, et al. Comparative gene expression profiling of post-natal and
post-pneumonectomy lung growth. Eur Respir J. 2010;35(3):655–66.
96. Kaza AK, et al. Epidermal growth factor augments postpneumonectomy
lung growth. J Thorac Cardiovasc Surg. 2000;120(5):916–21.
97. Kaza AK, et al. Keratinocyte growth factor enhances post-pneumonectomy
lung growth by alveolar proliferation. Circulation. 2002;106(12 Suppl 1):I120–4.
98. Swonger JM, et al. Genetic tools for identifying and manipulating fibroblasts
in the mouse. Differentiation. 2016;92(3):66–83.
99. Ruiz-Camp J, et al. Tamoxifen dosing for Cre-mediated recombination in
experimental bronchopulmonary dysplasia. Transgenic Res. 2017;26:165–70.
100. Ntokou A, et al. A novel mouse Cre-driver line targeting Perilipin 2-
expressing cells in the neonatal lung. Genesis. 2017;
101. Rawlins EL, Perl AK. The a “MAZE”ing world of lung-specific transgenic mice.
Am J Respir Cell Mol Biol. 2012;46(3):269–82.
102. Rinkevich Y, et al. Skin fibrosis. Identification and isolation of a dermal lineage
with intrinsic fibrogenic potential. Science. 2015;348(6232):aaa2151.
103. Prowse KR, Greider CW. Developmental and tissue-specific regulation of
mouse telomerase and telomere length. Proc Natl Acad Sci U S A. 1995;
92(11):4818–22.
104. Kumar M, Seeger W, Voswinckel R. Senescence-associated secretory
phenotype and its possible role in chronic obstructive pulmonary disease.
Am J Respir Cell Mol Biol. 2014;51(3):323–33.