You are on page 1of 68

19 Feb 2005 14:11 AR AR239-IY23-21.

tex XMLPublishSM (2004/02/24) P1: JRX


10.1146/annurev.immunol.23.021704.115707

Annu. Rev. Immunol. 2005. 23:683–747


doi: 10.1146/annurev.immunol.23.021704.115707
Copyright  c 2005 by Annual Reviews. All rights reserved
First published online as a Review in Advance on January 19, 2005

IMMUNOLOGY OF MULTIPLE SCLEROSIS∗


Mireia Sospedra and Roland Martin
Cellular Immunology Section, Neuroimmunology Branch, National Institute of
Neurological Disorders and Stroke, National Institutes of Health, Bethesda, Maryland
20892-1400; email: martinr@ninds.nih.gov; sospedrm@ninds.nih.gov
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org
by University of Missouri - Columbia on 03/18/13. For personal use only.

Key Words autoimmunity, autoimmune mechanisms, neuroimmunology,


demyelinating dieseases, EAE
■ Abstract Multiple sclerosis (MS) develops in young adults with a complex pre-
disposing genetic trait and probably requires an inciting environmental insult such as
a viral infection to trigger the disease. The activation of CD4+ autoreactive T cells
and their differentiation into a Th1 phenotype are crucial events in the initial steps,
and these cells are probably also important players in the long-term evolution of the
disease. Damage of the target tissue, the central nervous system, is, however, most
likely mediated by other components of the immune system, such as antibodies, com-
plement, CD8+ T cells, and factors produced by innate immune cells. Perturbations
in immunomodulatory networks that include Th2 cells, regulatory CD4+ T cells, NK
cells, and others may in part be responsible for the relapsing-remitting or chronic pro-
gressive nature of the disease. However, an important paradigmatic shift in the study
of MS has occurred in the past decade. It is now clear that MS is not just a disease
of the immune system, but that factors contributed by the central nervous system are
equally important and must be considered in the future.

INTRODUCTION
Multiple sclerosis (MS) is an inflammatory disease that affects the central nervous
system (CNS), i.e., the brain and spinal cord, and usually starts between 20 and
40 years of age (1, 2).1, 2 At least 350,000 individuals in the United States alone are
affected with MS. It leads to substantial disability through deficits of sensation and
of motor, autonomic, and neurocognitive function. The disease is usually not life


The U.S. Government has the right to retain a nonexclusive, royalty-free license in and to
any copyright covering this paper.
1
Owing to space restrictions, additional references for each section of this review are acces-
sible in the Supplemental Material. Follow the Supplemental Material link from the Annual
Reviews home page at http://www.annualreviews.org.
2
See Appendix for a full list of abbreviations used.

0732-0582/05/0423-0683$14.00 683
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

684 SOSPEDRA  MARTIN

shortening, but its socioeconomic importance is second only to trauma in young


adults (1, 2). There are two major forms of MS. Relapsing-remitting (RR)-MS is the
most frequent (85%–90%) and affects women about twice as often as men. Most
RR-MS patients later develop secondary progressive (SP)-MS (Figure 1). About
10%–15% of patients present with insidious disease onset and steady progression,
termed primary progressive (PP)-MS. It is not clear which factors are responsible
for the different courses. There is also heterogeneity in morphological alterations
of the brain found by magnetic resonance imaging (MRI) (3) or histopathological
evaluation (4, 5), as well as in clinical presentation, e.g., which CNS system and
areas are primarily affected and whether a patient responds to treatment. The
factors underlying this heterogeneity are not completely understood but include
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

a complex genetic trait that translates into different immune abnormalities and/or
by University of Missouri - Columbia on 03/18/13. For personal use only.

increased vulnerability of CNS tissue to inflammatory insult or reduced ability to


repair damage (Figure 1).
MS is still considered a CD4+ Th1-mediated autoimmune disease (6, 7). This
view is based on the cellular composition of brain and cerebrospinal fluid
(CSF)-infiltrating cells and data from experimental allergic (autoimmune) en-
cephalomyelitis (EAE) (8). In the EAE model, the injection of myelin components
into susceptible animals leads to a CD4+ -mediated autoimmune disease that shares
similarities with MS (6, 8) and can be adoptively transferred by encephalitogenic
CD4+ T cells into a naive animal (6, 8, 9). EAE cannot be transferred by antibod-
ies, and so far it has been transferred in only two instances by CD8+ T cells (10,
11), emphasizing the importance of CD4+ T cells. The role of CD4+ T cells in MS
is supported by many parallels with EAE, but it is also supported indirectly by the
fact that certain HLA class II molecules represent the strongest genetic risk factor
for MS, presumably via their role as antigen-presenting molecules to pathogenic
CD4+ T cells.
The above considerations still apply, but research during the past decade has
not only substantially increased our knowledge of the involvement of CD4+
T cells in MS but also shown that the previous concepts were too simplistic and
did not appropriately consider immune factors other than CD4+ T cells. Another
aspect might turn out to be even more important. We have for a long time almost
completely ignored the contribution of the affected organ, the CNS. Pathologic
and imaging studies (3, 5), as well as research of the molecular aspects of the
disease in EAE and MS, now provide ample evidence that CNS-specific factors
are important (12, 13). In this context, it is interesting, although historically not
too surprising, that reviews on organ-specific autoimmune diseases, such as type
1 diabetes or rheumatoid arthritis, focus entirely on alterations of tolerance, spe-
cific immune cells, and other immune aspects, but rarely on the involvement of
factors intrinsic to the target tissue. For MS, such an “immune-centered” view
can not be upheld, and consequently in this chapter we deviate from our previous
review 12 years ago (6) and consider the role of the CNS in targeting the dis-
ease process, in interactions with the immune system, and in the long-term course
of MS.
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 685


Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org
by University of Missouri - Columbia on 03/18/13. For personal use only.

Figure 1 Top: Schematic depiction of the clinical evolution of MS by a clinical scale


(EDSS, red line); the frequency of inflammatory events when studied by MRI (T1
lesions with contrast showing blood-brain barrier opening, blue arrows); T2 lesion
load documenting all tissue damage (blue line); brain atrophy (green line). Pathology:
Main pathological characteristics of MS. On the left, perivascular inflammation with
mononuclear cells and open blood-brain barrier (courtesy of H.F. McFarland, NIB,
NINDS, NIH); on the right, demyelinated areas shown in light blue and white, and,
on the far right, axonal transactions (blue onion bulb-like structure) and segmental
demyelination (from Reference 339, with kind permission of N. Engl. J. Med.). MRI:
Typical MRI characteristics. On the left, T1-weighted image with Gadolinium contrast
enhancement. White lesions indicate areas of fresh inflammation and open blood-brain
barrier. T2-weighted image shows the CSF-filled ventricles in white and MS lesions in
the brain parenchyma. On the right, brain atrophy with widened lateral ventricles and
cortical sulci.
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

686 SOSPEDRA  MARTIN

ETIOLOGY OF MS: IMMUNOGENETIC BACKGROUND


The etiology of MS remains unclear, but according to current data the disease
develops in genetically susceptible individuals and may require additional envi-
ronmental triggers. Virtually hundreds of studies in the past decades have addressed
the genetic contribution, and for details the reader is referred to excellent reviews
and the original articles (14, 15). Here, we try to distill from the existing literature
the most relevant aspects in the context of the immunopathogenesis of MS.
The general population prevalence of MS varies between 60–200/100,000 in
Northern Europe and North America, and 6–20/100,000 in low risk areas such as
Japan. Population, family, and twin studies all show that the prevalence is sub-
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

stantially increased in family members of MS patients (14). First-degree relatives


by University of Missouri - Columbia on 03/18/13. For personal use only.

of affected individuals have an approximately 20- to 50-fold (2%–5%) higher risk


to develop MS, and concordance rates in monozygotic twins vary between 20%
and 35% in different studies, with the most recent studies placing it at 25% (14).
Although the modest concordance rate has been viewed as a sign of environmental
influences, studies of adoptees in MS families (16) and other data indicate that the
genetic risk is probably higher. The search for individual susceptibility genes has
so far been frustrating, despite tremendous advances. More than 20 whole genome
screens have been performed in different MS populations and different geographic
areas, with up to 6000 microsatellite markers and different methodologies (17).
The data are strongest for one or more susceptibility genes on chromosome 6p21
in the area of the major histocompatibility complex [MHC; histocompatibility
leukocyte antigen (HLA) in humans], which is thought to account for 10%–60%
of the genetic risk of MS (18, 19).

THE ROLE OF THE HLA GENE COMPLEX


Similar to other T cell–mediated autoimmune diseases, in MS the specific genes
that confer risk are the HLA-DR and -DQ genes, the HLA-DR15 haplotype in
Caucasians (DRB1∗ 1501, DRB5∗ 0101, DQA1∗ 0102, DQB1∗ 0602) (18), but also
other DRs in ethnically more distant populations. Most of the risk stems from the
two DR alleles that are in very tight linkage disequilibrium, and there is also a dose
effect in DR15 homozygotic MS patients (20). The contribution of DQA1∗ 0102/-
B1∗ 0602 varies, and both additive and independent effects have been described,
particularly in populations with lower overall MS prevalence (21, 22). Additional
MS “risk/protective alleles” are listed in Table 1. Less information exists regarding
genetic risk conferred by HLA class I alleles. Their association with MS appears
to be much lower. HLA-A3 and -B7 are overrepresented in MS patients, and HLA-
A201 has shown protective effects (18, 21, 23) (Table 1). With respect to associa-
tions of HLA-DR/DQ alleles with other genes, or clinical, MRI, or immunological
characteristics, only limited data are available (Table 2). Genes associated with the
DR15 haplotype include transforming growth factor (TGF)-β family members,
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 687

TABLE 1 Association of HLA class I and class II alleles with multiple sclerosis
Ethnic background/population MS subtype
MHC class I/II allele and geographic location/country association Remark

DRB1∗ 1501a Caucasians, many countries and All subtypes Independent and joint with DQ;
backgrounds including Japanese, dose effect
Tasmanians, and many others
DRB1∗ 1503 Martinique None —
DRB1∗ 1506, -1508 India None Joint with ∗ 1501
DRB1∗ 15/DR3 Mexican Mestizos None —
DRB1∗ 0301 Sardinia None —
DRB1∗ 03/A30/B18 Central Sardinia None —
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

DRB1∗ 0405 Sardinia None —


DRB1∗ 04 Turkey/Canary Island None —
by University of Missouri - Columbia on 03/18/13. For personal use only.

DRB1∗ 04 Sweden Progressive MS —


DRB1∗ 04 Russia Higher T2 MRI —
load
DRB1∗ 0405 Japan OCB negativity —
DRB1∗ 04/DQB1∗ 0302 Finland None Weak association
DRB1∗ 0801 Ashkenazi Jews PP-MS —
DRB1∗ 12 Russia Higher T1 MRI —
load, higher
MRI atrophy
DRB1∗ 13 Northern Italy Benign MS —
DRB1∗ 1303 Non-Ashkenazi Jews None —
DRB1∗ 17 Germany/Sweden None —
DR in general Canada None In DR15 negative families,
independent contribution
DRB1∗ 01/07/11 — None Protective
DRB1∗ 01/DRw53 Finland None Protective
DRB1∗ 01/DQB1∗ 0501 Finland None Protective
DRB1∗ 13/DQB1∗ 0603 Finland None Protective
DRB1∗ 15021 Iran None Protective
DQA1∗ 0101 Colombia None —
DQA1∗ 0102 Colombia None —
DQA1∗ 0103 Colombia None Protective
DQB1∗ 0602 Many MS populations All subtypes Independent and joint with
DRB1∗ 1501
DPB1∗ 0301 Japan Classical MS —
DPB1∗ 0501 Japan Opticospinal —
(Asian) MS
A∗ 0301 Caucasians, Russia, Sweden Poor outcome, Partly independent of DR15
none
B∗ 07/B∗ 12 Caucasians, Russia Poor outcome, —
none
A∗ 02 Russia More benign —
outcome
A∗ 0201 Sweden More benign Protective
outcome
a
Note that wherever DRB1∗ 1501 is mentioned in the Table, DRB5∗ 0101 is co-expressed in this haplotype, and therefore
these findings apply to DRB1∗ 1501 and DRB5∗ 0101.
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

688 SOSPEDRA  MARTIN

cytotoxic T lymphocyte–associated antigen (CTLA)-4, the tumor necrosis factor


(TNF) cluster, IL-1 receptor antagonist, IL-1, and estrogen receptor. Clinical fac-
tors include earlier disease onset, more often RR-MS, female gender, optic neuritis,
or spinal involvement as initial event. Immunologically, higher CSF immunoglob-
ulins, oligoclonal bands (OCB), and matrix metalloproteinase 9 (MMP-9) levels
have been reported (24). DR4+ patients often have a worse clinical outcome or
progressive course than patients expressing DR15+ (Tables 1 and 2). The above-
mentioned studies on HLA associations with MS are heterogeneous with respect
to sample size, methodology, ethnic background, and clinical findings. In older
studies, the exact HLA class II gene has not been determined by molecular typing
techniques. However, there is no doubt that HLA-DR and -DQ molecules are by
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

far the strongest genetic risk factors in MS.


by University of Missouri - Columbia on 03/18/13. For personal use only.

Our knowledge of how certain HLA class II genes confer risk for MS or au-
toimmune diseases at the molecular level is very sketchy. Several mechanisms
have been considered: (a) Disease-associated HLA-DR and -DQ molecules have
binding characteristics that lead to preferential presentation of specific sets of self
peptides, e.g., myelin peptides in MS. Currently, little data support this hypothe-
sis, and comparisons of polymorphic residues in the HLA-DR and -DQ binding
pockets have not been conclusive. (b) As a variation of the first possibility, in-
vestigators have speculated that disease-associated HLA molecules could have
binding characteristics that allow only limited sets of peptides to bind, accounting
for less “complete” thymic negative selection of self-reactive T cells. Diabetes-
prone NOD mice and their MHC class II (I-Ag7 ) have been viewed as an example
for this situation (25). Given the high frequency of most autoimmune disease–
associated HLA-DR and -DQ alleles in the population and the normal cellular
immune function in the vast majority, we consider this mechanism unlikely in MS.
(c) Either polymorphic residues of the T cell receptor (TCR)-exposed surfaces of
the α-helical regions of DR/DQ-α and -β chains, such as the “shared motif” in
rheumatoid arthritis–associated class II molecules (26) or TCR-contacting amino
acids of the antigenic peptide, or both, could select an autoimmune-prone T cell
repertoire. Gross abnormalities in T cell repertoires do not exist in MS patients
according to current data (see below). However, we recently observed that clonally
expanded T cells from the CSF of MS patients are capable of utilizing all MS-
associated HLA-DR/DQ molecules in the DR15 haplotype for recognition of large
sets of peptides (M. Sospedra, unpublished observation). (d) Gene and protein ex-
pression of one or several disease-associated DR and DQ alleles could be elevated
in the CNS, enhancing antigen presentation. Comparisons of the expression of
the two MS-associated DR molecules in the DR15 haplotype, DR2a (DRA1∗ 0101
and DRB5∗ 0101) and DR2b (DRA1∗ 0101 and DRB1∗ 1501), in MS patients and
controls did not reveal general or tissue-specific upregulation of one DR allele (E.
Prat, unpublished observation), but differential expression on B cells and mono-
cytes. (e) Antigen presentation in the context of certain DR molecules could be
shaped by proteases involved in antigen processing or by nonpolymorphic class II
molecules such as HLA-DO and -DM that are tightly linked on chromosome 6p21.3
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 689

TABLE 2 Association of HLA-DR15 or other DR haplotypes with additional genes/chromo-


somal regions (part 1), or clinical/MRI/immunological criteria (part 2)
Associated gene or
HLA-DR gene clinical/MRI/immunological factor Remark

Part 1
DRB1∗ 1501a 12p12 Gene not known
DRB1∗ 1501 Microsatellite close to TGFB1 Gene not known
DRB1∗ 1501 TGFB3 —
DRB1∗ 1501 CTLA-4 —
DRB1∗ 1501 Allele in TNF cluster —
DRB1∗ 1501 Area extending to DRA1∗ promoter
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

Gene not known


DRB1∗ 1501 Allele 2 of IL-1 receptor antagonist Association with
by University of Missouri - Columbia on 03/18/13. For personal use only.

RR-MS
DRB1∗ 1501 Association with estrogen receptor —
polymorphism
DRB1∗ 04/05 Association with MBP gene polymorphism in —
Italian and Russian MS patients
Part 2
DRB1∗ 1501 Association with relapse onset MS —
DRB1∗ 1501 Female gender, younger age at onset —
DRB1∗ 1501 Optic neuritis first sign, spinal involvement, —
early onset (all in a non-Japanese population)
DRB1∗ 1501 Optic neuritis in children —
DRB1∗ 1501 Higher CSF OCB and IgG, and MMP-9 —
DRB1∗ 1501 Higher IL-4 and TGF-β levels, RR-MS —
DRB1∗ 1501 Anti-MOG IgA higher in asymptomatic —
relatives
DRB1∗ 15-negative Worse clinical outcome —
status
DRB1∗ 04 Anti-MOG IgM elevated in patients —
DRB1∗ 04 Worse prognosis —
a
Note that wherever DRB1*1501 is mentioned in the Table, DRB5*0101 is co-expressed in this haplotype, and therefore
these findings apply to DRB1*1501 and DRB5*0101.

and fulfill peptide-sorting and -loading functions. DM has been examined, but so
far no association has been found in MS (27). (f) Engagement of HLA class II
molecules leads to intracellular signaling events, e.g., anergy (28), which could be
perturbed in patients with autoimmune diseases. There is currently no information
on this aspect in MS.
HLA class I may act independently of class II in some patients, either via similar
mechanisms or by modulation of NK cell activity. The reduced number of peptide-
occupied HLA class I molecules in MS patients (29), the CD8+ T cell infiltrations
in the CSF and MS plaque tissue (30, 31), and the higher expression of HLA class I
in the brain (32) suggest that the roles of HLA class I, CD8+ T cells, and NK cells
merit further study.
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

690 SOSPEDRA  MARTIN

OTHER RISK-CONFERRING GENES


A recent review on the genetics of MS (14) pointedly remarked that the search for
candidate genes has been plagued by initial positive results on specific genes in
one study and subsequent negative or inconclusive data in several other reports.
Polymorphisms of TCR genes, immunoglobulin loci, CCR5, and CD45 are just a
few examples. Without summarizing the existing literature in depth, a few chromo-
somal loci have been identified in several but not all studies, and these studies used
different methodologies, including the TCRβ chain locus, CTLA-4, TNF-α and -β
alleles, and ICAM-1. Polymorphisms of CCR2, IL-10 receptor α, and Fas-L may
confer protective effects; CCR5, IL-10, IL-4 receptor α, IL-2 receptor β, IFN-γ ,
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

vitamin D, and estrogen receptor confer risk. With respect to CNS-related genes,
by University of Missouri - Columbia on 03/18/13. For personal use only.

Notch4, a signaling molecule that is involved in both myelin development and


immune function, neutral sphingomyelinase activating factor, ciliary neurotrophic
factor, and the myelin basic protein (MBP) gene have been implicated. Other
oligodendrocyte/CNS growth factors have been studied, but no association has
been found. The role of an allele of apolipoprotein E (APOE4), which is involved
in lipid metabolism and associated with the severity of Alzheimer’s disease, re-
mains controversial in MS, although an association between the APOE4 allele and
higher severity/faster progression has been shown.
The above list is far from complete, and there are several reasons for the am-
biguity of candidate gene searches. Methodologies and sample sizes vary; patient
populations are often not stratified with respect to HLA, clinical, MRI-defined, or
pathological phenotype; the ethnic background of subjects differs among studies;
and the search often focuses on a few members of a gene family of interest, e.g., cy-
tokine or TCR genes. Considering the genetic heterogeneity of the outbread human
population and that almost every aspect of immune and nervous system function
occurs and is regulated via highly complex interactions between multiple cell types
and their soluble factors, surface receptors, signaling components, growth charac-
teristics, and many other molecular pathways, our limited understanding is not too
surprising.

GENOMICS STUDIES IN MS
The quantitative genetic trait has been difficult to dissect in MS and other com-
plex diseases. In recent years, numerous groups have examined gene expression
rather than the presence or absence of genetic polymorphisms. The development of
microarray-based methods that allow the interrogation of thousands of genes in one
experiment offers great advantages (33). Several investigators have employed mi-
croarrays to study gene expression patterns in MS brain tissue or peripheral blood
samples. Whitney et al. (34, 35) examined plaque tissue and normal-appearing
white matter in MS patients and EAE and identified four genes consistently overex-
pressed: the transcription factor jun-D, thrombin receptor protease-activated recep-
tor 3, a putative ligand for IL-1 receptor-related molecule T1/ST2, and arachidonic
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 691

acid 5-lipoxygenase, a molecule involved in leukotriene biosynthesis (35). Lock


et al. (36) found increased transcription of MHC class II molecules; complement;
T cell and B cell genes; some cytokine genes (IL-17), as well as their recep-
tors (IL-1R, TNF p75 receptor); and glial fibrillary acidic protein (GFAP) and
transcription factors. However, myelin proteins and neuronal genes were mostly
underexpressed. EAE studies pointed at the relevance of G-CSF and FcRγ (36).
Further studies found differential abundance of CD4, MAPKK1, nerve growth fac-
tors, HLA-DRα, and proinflammatory cytokines including osteopontin, but also
some Th2 genes, α-B crystallin, and others (37–40). These studies have not been
formally compared; however, immune system–related genes are prominently ex-
pressed, particularly at the acute stage of MS, and there are quantitative rather than
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

qualitative differences between early and later stages of the disease. Examination
by University of Missouri - Columbia on 03/18/13. For personal use only.

of normal-appearing white matter, i.e., areas of the brain that appear macroscop-
ically normal but are microscopically abnormal, demonstrated upregulation of
genes involved in homeostasis and neural protection (41).
Expression studies in peripheral blood mononuclear cells (PBMC) have yielded
similarly large numbers (42) of differentially expressed genes, and many are re-
lated to immune function, including MHC class II molecules; cytokines (TNF-α,
IFN-γ , LTB, TNF-α receptor-associated factor 5); adhesion molecules (CD11a,
CD18, CD49, integrin β7); costimulatory molecules (SLAM); T cell transcripts
(TCRα, MAL); B cell or NK cell transcripts; signaling molecules (ZAP70); pro-
teases involved in antigen processing; and many others with unknown relation to
MS (42). The differential expression of only two genes from chromosome 6p21.3,
i.e., heat shock protein 70 and histone family member 2, allowed investigators to
separate patients and controls with 80% accuracy (43), and with the entire set of
differentially expressed genes, one can accurately distinguish the two groups (43;
G. Blevins, unpublished observation). Dissection of the mechanism of action of
MS therapies by gene expression profiling has shown that IFN-β has not only im-
munomodulatory effects (e.g., increase of IL-10) but also proinflammatory effects
(e.g., upregulation of CCR5 and the IL-12 receptor β2 chain) (44). Gene expres-
sion profiling also identified genes associated with partial responsiveness (e.g.,
IL-8 or TRAIL) to IFN-β therapy (45, 46). Although widely perceived as “fishing
expeditions” and not hypothesis-driven experiments, gene expression profiling is
likely to complement genetic studies and also to be instrumental in other aspects
of MS research, such as identifying important functional pathways and treatment
mechanisms.

ETIOLOGY OF MS: NONGENETIC FACTORS AND


INFECTIOUS TRIGGERS

Nongenetic Factors in the Etiology of MS


The relatively low concordance rate of identical twins indicates a contribution of
nongenetic factors to MS etiology (14, 47). This argument has to be considered
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

692 SOSPEDRA  MARTIN

with some caution because studies in congenic mice with gradually increasing
numbers of lupus-associated genes have shown that the rate of disease expression
can be “titrated,” i.e., the fraction of animals that developed lupus was determined
by the number of disease-linked genes under identical environmental influences
(48). Among putative environmental factors, both infectious agents and behav-
ioral or lifestyle influences have been proposed to induce or contribute to disease
expression (49). The fact that women with the disease outnumber men with MS
by 1.6–2.0:1 suggests hormonal variables as risk factors. This is supported by
(a) lower relapse rates during, and disease rebound after, pregnancy (50); (b) the
worsening of MS during menstruation; (c) the correlation of high estradiol and low
progesterone with increased MRI disease activity; (d) gender differences in EAE
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

susceptibility related to the protective effect of testosterone; and finally (e) the
by University of Missouri - Columbia on 03/18/13. For personal use only.

therapeutic effects of estriol in RR-MS (51). The precise mechanisms by which


sex hormones may influence MS susceptibility are not known, but the stimula-
tory effects of estrogens on proinflammatory cytokine secretion and the reverse by
androgens probably represent one mechanism.
Environmental contributions to the etiology for MS are supported by a number
of factors. (a) The north to south gradient in disease prevalence on the north-
ern hemisphere and the opposite on the southern. (b) MS distribution cannot be
explained by population genetics alone. Although regions to which Northern Eu-
ropean descendents migrated show high prevalence rates, these rates among Cau-
casians outside Europe are only half those in many parts of Northern Europe.
(c) Migration studies show that if one migrated from an area of high prevalence
of MS to an area of low prevalence before age 15–16, the low risk was acquired,
whereas migration after 15–16 did not change the risk (52). One proposed causative
factor is the decrease in sunlight exposure depending on the latitude. UV radia-
tion may exert its effects either by influencing immunoregulatory cells or by the
biosynthesis of vitamin D (53). The latter notion is supported by EAE data and the
association of a vitamin D receptor polymorphism with MS in Japan. Melatonin
secretion also depends on sunlight exposure. The lack of sunlight could induce an
excess of melatonin, which enhances Th1 responses.
The geographical distribution also reflects the economic level of the country.
The incidence of MS in Asia is overall low, with the highest prevalence in Japan, the
most developed country in the area. Furthermore, prevalence rates have increased
with the socioeconomic development in previous decades, which has been related
to industrialization, urban living, pollution, occupational exposures to solvents,
changes in diet and breastfeeding, smoking habits, and reduced UV light expo-
sure. Finally, the delayed exposure to or overall reduction in childhood infections
in developed countries is another factor and has led to the “hygiene hypothesis.”
According to this hypothesis, which is supported by findings in type 1 diabetes and
EAE, there is a skewed immune responsiveness and increased propensity to de-
velop autoimmune reactions/diseases (Th1-mediated) and allergy (Th2-mediated)
in populations with delayed exposure to or overall reduction in childhood infec-
tions. However, this hypothesis is difficult to prove.
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 693

INFECTIOUS AGENTS AS TRIGGERS OF MS


Viral and bacterial infections are logical candidates as environmental triggers of
MS. However, what has been stated for genetic studies also applies to research
on infectious agents. Numerous reports have claimed to identify MS triggers, and
almost universally these observations have later not withstood scrutiny (54, 55).
Prospective studies have shown that MS relapses often follow viral infections (56).
The temporal patterns and the occurrence of “MS epidemics,” i.e., sudden increases
in MS incidence in small, previously isolated communities, such as the one on
the Faroe Islands, also point toward an infectious agent (52), although these are
not uncontested. Further evidence for viral or bacterial triggers stems from EAE
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

studies. Almost 100% of transgenic mice expressing a TCR that is specific for
by University of Missouri - Columbia on 03/18/13. For personal use only.

an encephalitogenic peptide of MBP develop EAE when the transgenic mice are
housed under nonpathogen-free conditions, whereas the same animals housed in
a specific-pathogen-free facility remained disease free (57).
The viral etiology of a number of human demyelinating diseases [progres-
sive multifocal leukoencephalopathy caused by papovavirus JC; postinfectious en-
cephalitis and subacute sclerosing panencephalitis (SSPE), both caused by measles
virus; herpes simplex virus (HSV); HIV encephalopathy] explains the continued
interest in viruses as triggers for MS (54, 55). Animal models of virus-induced
demyelinating diseases, such as encephalitis or encephalomyelitis by Theiler’s
murine encephalomyelitis virus (TMEV), canine distemper virus, neurotropic
strains of mouse hepatitis virus, Semliki Forest virus, Visna virus, and rat-adapted
measles virus (54, 55), also support the possible involvement of a virus in MS.
Among viruses that are pathogenic in humans, those that induce persistent in-
fection, such as herpes- or retroviruses, are suitable candidates and have been
studied widely in MS. Herpesviruses are of particular interest owing to their neu-
rotropism, ubiquitous nature, and tendency to produce latent, recurrent infections.
Human herpesvirus 6 (HHV-6) and Epstein-Barr virus (EBV) are the leading can-
didates. The seroprevalence for both is high, i.e., >80% for HHV-6, a lymphotropic
and neurotropic β-herpesvirus, and 90% for EBV, a lymphotropic γ -herpesvirus.
HHV-6 can lead to meningoencephalitis, and several additional observations sug-
gest a role in MS, including its detection in oligodendrocytes in MS plaque tissue
(58) (but also in normal brains), the infection of astrocytes, and the presence of
HHV-6 DNA and anti-HHV-6 IgG and IgM antibodies in serum and CSF of MS
patients. However, the DNA and serological data are controversial (reviewed in
59). The existence of two different HHV-6 variants may account for some of the
discrepancies. The role of HHV-6 variant A in MS is supported by its higher
neurotropism, increased lymphoproliferative responses against variant A in MS
patients (60), and its DNA presence in CSF from MS patients.
EBV has also been linked with MS. Anti-EBV antibodies are elevated in patients
with MS, i.e., the seropositivity rate of MS patients is 100% versus approximately
90% in the general population, and MS patients reactivate latent EBV infections
more often, correlating with relapses (61). Serum anti-EBV IgG levels prior to
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

694 SOSPEDRA  MARTIN

onset of MS have been reported as a strong disease predictor, a history of infectious


mononucleosis is more common in MS patients, and the risk of developing MS
is higher for individuals who suffered from infectious mononucleosis at a young
age (62). Furthermore, some patients with neurological sequelae of primary EBV
infection develop MS.
Human herpesvirus 1 (HSV-1) and varicella zoster virus (VZV or HSV-3) have
also been considered as MS-triggering agents on the basis either of CSF antibody
studies, casuistic observations, or the finding that VZV encephalitis is character-
ized by demyelination. Finally, MS-associated retroviruses have been linked with
disease on the basis of the detection of extracellular virions in plasma and CSF of
MS patients; however, their role is currently not clear.
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

Among bacteria, Chlamydia pneumoniae (Cpn) has been implicated in MS. Cpn
by University of Missouri - Columbia on 03/18/13. For personal use only.

is a Gram-negative intracellular bacterium and common pathogen of the respiratory


system. Following an initial report (63), many studies examined an association
between Cpn and MS. Current data are contradictory. Whereas one study reported
the presence of Cpn in the CSF of a large percentage of MS patients compared
with controls (64), other studies failed to observe an association between Cpn and
MS (65).
The difficulty in identifying a single microorganism as the cause of MS proba-
bly indicates that Koch’s paradigm “one organism, one disease” does not apply to
this complex disease. Current data suggest that MS could be induced and/or exac-
erbated by many different microbial infections, and the responsible agents are most
likely ubiquitous pathogens that are highly prevalent in the general population.

MECHANISMS: HOW INFECTIOUS AGENTS MAY


INDUCE MS
Two main mechanisms have been proposed to explain how infections could in-
duce MS: (a) molecular mimicry, i.e., the activation of autoreactive cells by cross-
reactivity between self-antigens and foreign agents; and (b) bystander activation,
which assumes that autoreactive cells are activated because of nonspecific inflam-
matory events that occur during infections. A third proposal is that infections
induce MS through a combination of these two mechanisms.

Molecular Mimicry
Molecular mimicry involves reactivity of T and B cells with either peptides or
antigenic determinants shared by infectious and self-antigens. The recognition of
self-antigens at intermediate levels of affinity by T cells during thymic selection
leads to positive selection and export of these T cells to the periphery. Cross-
reactivity of these potentially self-reactive T cells with foreign antigens can lead
to activation during infection, migration across the blood-brain barrier (BBB), CNS
infiltration, and, if they recognize antigens expressed in the brain, tissue damage
and potentially an autoimmune disease like MS (Figure 2).
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 695


Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org
by University of Missouri - Columbia on 03/18/13. For personal use only.

Figure 2 The evolution of the molecular mimicry concept and activation of


autoreactive T cells via cross-reactivity with foreign antigens (for details, see
text).

For example, MBP is a candidate autoantigen in MS on the basis of numerous


pieces of evidence (6). MBP-specific T cells can be isolated from MS patients
and controls (66–72). However, their activation state in MS patients, proinflam-
matory phenotype, higher antigen avidity, and preferential memory origin suggest
that they had been activated in vivo, e.g., by cross-reactive infectious antigens
during infections. Many studies have looked for cross-reactive antigens between
MBP and foreign agents. Initially, the search was guided by the concept that
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

696 SOSPEDRA  MARTIN

humoral and cellular immune reactivity is exquisitely specific and that complete
homology between foreign proteins and MBP is required for molecular mimicry.
Although examples of such stringent homology have been reported for MBP and
viruses (73, 74) (Figure 2), complete sequence matching is a rare event. Subse-
quent research of the molecular requirements for T cell recognition found that
certain amino acid positions in a peptide are more critical than others for the
interactions within the trimolecular complex, and most residues, except for the
primary TCR contact, allowed for some degree of variation (75). On the basis of
these observations, a search algorithm assumed that molecular mimicry can oc-
cur as long as a MHC and TCR contact motif is preserved (76) (Figure 2). The
activation of MBP-specific T cell clones (TCC) derived from MS patients by vi-
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

ral and bacterial peptides sharing this motif with MBP confirmed the prediction
by University of Missouri - Columbia on 03/18/13. For personal use only.

that sequence homology was not required for cross-recognition (76) (Figure 2).
Subsequently, the recognition by a MBP(83−99) -specific TCC was systematically
dissected using single amino acid substitutions in each position of the peptide
sequence (77). These data demonstrate that cross-reactivity can occur with pep-
tides that share no amino acid in their sequence and that each amino acid in the
peptide contributes independently to TCR recognition (78) (Figure 2). Recently,
the concept evolved even further; Lang et al. (79) showed that different peptides
bound to different class II molecules can lead to cross-reactivity by the same TCR
as long as the complexes share similarity in charge distribution and overall shape
(Figure 2). Together, these observations offer new perspectives on the concept
of molecular mimicry and indicate that cross-reactivity occurs frequently. Addi-
tional evidence for molecular mimicry stems from animal experiments showing
that mice expressing viral proteins as tissue-specific transgenes develop autoim-
mune diseases after viral infection (80, 81). Recently, a model for virus infection
that leads to molecular mimicry has been developed, in which an encephalito-
genic virus (TEMV) encodes a mimic peptide for an encephalitogenic myelin
proteolipid protein (PLP) that is naturally expressed by Haemophilus influenzae.
The infection with this recombinant virus induces early onset of disease, which
indicates that CNS infection with a pathogen containing a mimic epitope for a
self-myelin antigen can induce a cross-reactive T cell response, resulting in au-
toimmune demyelinating disease (82). Although all these findings demonstrate that
molecular mimicry is a viable hypothesis that can explain the link between infec-
tion and MS, evidence for this phenomenon in human autoimmune diseases is still
scarce.

Bystander Activation
Bystander activation mechanisms can be classified into two categories. The first
category encompasses TCR-independent bystander activation of autoreactive T
cells by inflammatory cytokines, superantigens, and molecular pattern recog-
nition, e.g., Toll-like receptor (TLR) activation. The second category involves
the unveiling of host antigens and the adjuvant effect of infectious agents on
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 697

antigen-presenting cells (APCs). Several proinflammatory cytokines and chemo-


kines are produced during infection, and these molecules have long been con-
sidered the main activators of virus-specific CD8+ T cells and inducers of the
autoimmune process (83). Most of the activated CD8+ T cells are specific for viral
antigens, however, and cytokines alone are unlikely to cause the activation and dif-
ferentiation of T cells in the absence of a specific antigen (84), which suggests that
bystander activation requires the cooperation of several mechanisms to induce au-
toimmunity. Although the administration of cytokines can induce disease relapses
in EAE, there are few examples in which the local overexpression of inflammatory
cytokines/chemokines alone can break tolerance in healthy animals. The local ex-
pression of IL-2, IL-12, and IFN-γ -inducible protein (IP)-10 in diabetes can lead
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

to inflammation but not to clinical disease, and only the overexpression of IFN-γ in
by University of Missouri - Columbia on 03/18/13. For personal use only.

pancreatic β cells disrupts tolerance to autoantigens, probably owing to enhanced


presentation of self-antigens.
Superantigen exposure has also been proposed as a bystander activation mech-
anism. These toxins can induce relapses in the EAE model via interactions with
MBP-specific TCC that express certain TCR Vβ chains (85).
Bystander activation via another group of infectious agent–derived and proin-
flammatory factors, such as TLRs, has also been described (86). One of them,
lipopolysaccharide (LPS), binds to TLR4 and initiates innate immune responses
to common Gram-negative bacteria such as Cpn. TLR4 activation by LPS increases
the expression of cytokines as well as of reactive oxygen species. TLR4 in the CNS
is mainly expressed on microglia but not on astrocytes or oligodendrocytes. LPS-
TLR4 interactions may occur in MS during an infection with bacteria such as
Cpn and induce the activation of monocytes and microglia, i.e., the adjuvant effect
on APCs. Alternatively, LPS-TLR4 interactions may activate autoreactive T cells
in the periphery. Bacteria injected into the brain parenchyma are able to induce
inflammatory responses only after peripheral sensitization.
Another mechanism of bystander activation that depends on specific TCR recog-
nition is the unveiling of host antigens as a consequence of viral tissue damage.
Activated virus-specific T cells traffic to the infected tissue, where they recognize
viral epitopes and kill infected cells, resulting in the destruction of self-tissue and
the release of autoantigens. The presentation of autoantigens together with the
adjuvant effect of infectious agents can then result in the de novo activation of au-
toreactive T cells and later to epitope spreading. This process occurs in the TMEV
mouse model for MS, in which an initial virus-specific T cell response broadens or
spreads to myelin proteins during persistent infection of the CNS (87). Spreading of
the T cell response can include the presentation of cryptic epitopes that are usually
not processed or presented as immunodominant epitopes but that can be presented
during specific conditions associated with viral infection (88). During viral in-
fection, the expression of self-proteins in the infected tissue is often upregulated
(89), tissue-specific APCs are activated, and the expression pattern of proteases in
these APCs can be altered, which leads to processing of cryptic epitopes that are
not generated during “normal” processing (90). The recognition of such cryptic
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

698 SOSPEDRA  MARTIN

epitopes is probably more important during progression and perpetuation of the


autoimmune response.
Recently, Pender (91) suggested that MS, like other chronic autoimmune dis-
eases, could be based on infection of autoreactive B lymphocytes by EBV. In this
scenario, autoreactive B cells are infected by EBV, proliferate, and turn into latently
infected B cells that are resistant to apoptosis because they express virus-encoded
antiapoptotic molecules. The presence of infected B cells in the target tissue can
result in costimulation of autoreactive T cells, which prevents these cells from
undergoing activation-induced apoptosis.
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org
by University of Missouri - Columbia on 03/18/13. For personal use only.

THE “MAJOR PLAYERS”


CD4+ T Cells
EVIDENCE FOR INVOLVEMENT OF CD4+ T CELLS IN MS Following the description
of MS by Charcot in 1868 (92), and the observation by Pasteur at the turn to the
twentieth century (93) of acute postvaccinal encephalomyelitis in rabies vaccinees,
Rivers showed in 1933 (94) that the injection of spinal cord or brain homogenates
into healthy primates caused a disease similar to MS, leading to the hypothe-
sis that MS is an autoimmune disease. Several decades later, investigators began
to study systematically the experimental disease in rodents and made the semi-
nal observations that still dominate our thinking about the pathogenesis of MS
(8, 9, 95). They showed that the injection of defined protein components of the
myelin sheath together with an adjuvant into naive susceptible animals caused
either an acute, chronic, or relapsing-remitting encephalomyelitis, which is now
referred to as EAE. The observation that EAE could be transferred by in vitro
reactivated myelin-specific CD4+ T cells (passive or adoptive transfer EAE) (8,
9) convincingly documented that EAE can be directly induced with autoreactive
T cells in naive animals. Unlike myasthenia gravis, which is also an autoimmune
disease affecting striated muscle, EAE cannot be transferred by antibodies. This
fact led investigators to conclude that MS is likely a T cell–mediated autoim-
mune disease. As we discuss below, this view is too simplistic. However, current
evidence on the induction and perpetuation of MS still favors CD4+ autoreac-
tive T cells as a central factor for the autoimmune pathogenesis of MS by the
following arguments: (a) CD4+ T cells contribute to the CNS- and CSF-infiltrating
inflammatory cells in MS; (b) genetic risk is to a substantial degree conferred by
HLA-DR and -DQ molecules; (c) humanized transgenic mice expressing either
HLA-DR or -DQ molecules are susceptible to EAE (96–98), and mice
expressing both MS-associated HLA-DR molecules and MS patient–derived MBP-
specific TCR develop spontaneous or induced EAE (99, 100); (d) a therapeutic trial
with an altered peptide ligand (APL) of MBP(83−99) induced cross-reactive CD4+
T cells with Th1 phenotype that led to disease exacerbations of MS patients (101);
(e) antibody production, CD8+ maturation, and many other steps of adaptive and
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 699

innate immune function are at least in part controlled by CD4+ helper T cells.
Unlike our previous review of this subject (6), we do not describe EAE data in
detail here, but rather refer the reader to the original literature and reviews (8, 102,
103). We focus instead on MS.
One of the first striking observations in the early investigations of the involve-
ment of CD4+ T cells in MS was that MBP-specific T cells were readily found in
both MS patients and healthy controls (66), indicating that previous concepts about
the efficiency of central tolerance mechanisms and the elimination of autoreactive
T cells were probably not correct (68–71). The fact that such autoreactive T cells
from the normal T cell repertoire of Lewis rats can induce EAE (104) suggested
to investigators that their equivalent in humans might also be relevant for MS.
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

During the subsequent two decades, every aspect of CD4+ T cells in MS has been
by University of Missouri - Columbia on 03/18/13. For personal use only.

the subject of exhaustive research. We are not able to cover all these data in detail,
but we summarize the main findings.

FREQUENCY OF CD4+ AUTOREACTIVE T CELLS Frequencies of autoreactive T cells


in MS patients and healthy controls vary greatly depending on the methodology
(71, 72, 105–109). Whereas tissue culture–based techniques have shown frequen-
cies of about 1 MBP-specific cell per 106 −107 PBMC (71, 105), approximately
1–2 orders of magnitude higher numbers were observed with enzyme-linked im-
munospot (ELISPOT) assays, which detect IFN-γ -secreting cells (72). Newer
methods, which employ quantitative polymerase chain reaction to follow individ-
ual TCC via their specific TCR CDR3 regions, observe frequencies of 1/104 or
even higher (107, 109). Flow cytometry–based techniques that follow the prolifer-
ating cell fraction upon stimulation with a myelin antigen observe frequencies in a
similar range (110). Tetramer-based assays currently do not work for autoreactive
HLA class II–restricted T cells, probably owing to low-affinity TCR recognition
of autoantigens. Most studies comparing MBP- or PLP-specific T cells in MS
patients and controls observe elevations in precursor frequencies in MS patients
(111). Furthermore, the number of myelin-specific T cells with mutations of the
hypoxanthine phosphoribosyl transferase gene, which occur in the proliferating
T cell pool, is elevated in MS (112, 113). Up to 2000-fold expansions of MBP(83−99) -
specific T cells have been described during exacerbation of patients in a treatment
trial with an APL based on the immunodominant MBP(83−99) peptide (101). Most
APL-specific T cells cross-reacted with MBP(83−99) , and these cells were also
found in the CSF and exhibited a Th1 phenotype, all supporting their involvement
in disease exacerbation (101). Finally, TCR CDR3-based molecular tracking of
individual clones showed that MBP- and APL-specific T cells had preexisted in
the patient’s peripheral blood long before APL therapy but were markedly ex-
panded during disease exacerbation (107). Most studies of autoreactive T cells
in MS used relatively high concentrations (10–50 µg/ml) of either whole native
or recombinant proteins or peptides. Under normal conditions, and even under
disease conditions such as stroke, such high concentrations of myelin antigens are
probably rarely reached, and T cell activation by autoantigens will only result if
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

700 SOSPEDRA  MARTIN

other factors, such as strong activation of innate immune mechanisms, upregula-


tion of MHC and costimulatory molecules, and proinflammatory cytokines, occur.
However, high-avidity T cells that respond at low antigen concentrations to myelin
proteins/peptides and that are likely more relevant to disease are clearly also in-
creased in MS patients and mostly express a proinflammatory phenotype (108).

Antigen Specificity of Myelin-Specific CD4+ T Cells


MYELIN BASIC PROTEIN (MBP) MBP is the best-studied myelin protein in MS. It is
the second most abundant myelin protein (approximately 30%–40%) after PLP, is
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

relatively easy to isolate owing to its physicochemical characteristics, and was the
first that was used extensively in EAE. There are five MBP isoforms with 14.0–21.5
by University of Missouri - Columbia on 03/18/13. For personal use only.

kDa molecular weights in mammals that result from differential splicing of eleven
axons within the Golli-MBP locus (114). The highly basic MBP is positioned at
the intracellular surface of myelin membranes, and via interactions with acidic
lipid moieties it is involved in maintaining the structure of compact myelin. The
most abundant 18.5 kDa isoform (170 amino acid length) has been used in most
immunological studies. Unlike myelin oligodendrocyte glycoprotein (MOG) and
PLP, MBP is found in significant quantities in both central and peripheral myelin,
and MBP transcripts have also been demonstrated in peripheral lymphoid organs
(115). EAE can be induced with MBP in several mouse and rat strains, guinea
pigs, and nonhuman primates (103). The most important encephalitogenic areas
are depicted in Figure 3. An important parallel between rodent and primate EAE
models and MBP-specific immune responses in humans is the striking overlap be-
tween epitopes that are encephalitogenic in the context of EAE-associated MHC
class II alleles and MBP regions that are immunodominant in the context of MS-
associated HLA-DR alleles, i.e., HLA-DR2a (DRB5∗ 0101), -DRb (DRB1∗ 1501),
and -DRB1∗ 0401/0404/0405 (69–71, 116, 117) (Figure 3). This applies to the im-
munodominant MBP(83−99) or MBP(84−102) epitope, a promiscuous binder to all
the above MS-associated HLA-DR molecules (118–120), as well as to the immun-
odominant MBP(111−129) epitope in the context of DRB1∗ 0401 (121) and the region
of MBP that is immunodominant with DR2a (71, 122) and other DR alleles (123).
For high-avidity myelin-specific T cells, MBP(83−99) is not immunodominant, but
MBP(13−32) , MBP(111−129) , and MBP(146−170) are (108). Most of these peptides are
promiscuous HLA-DR binders; however, the predicted affinity to HLA-DR2a, -
DR2b, -DR4, and other DR alleles is low, indicating that deletion of T cells with
high functional avidity for these MBP epitopes in the thymus is incomplete. This
situation is similar to MBP Ac1-11 epitope in PL/J mice (114, 124). The poor
binding affinity of the latter MBP epitope to IAu supports the view that complexes
of MBP Ac1-11 with IAu are unstable and therefore inefficient in negative selec-
tion (125). MBP Ac1-11-specific TCR transgenic mice develop EAE depending
on the level of microbial exposure or after induction with pertussis (57, 126). The
encephalitogenic potential of a MS patient–derived T cell was demonstrated in a
transgenic mouse expressing a MBP(84−104) -specific TCR and HLA-DR15 (99).
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 701


Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org
by University of Missouri - Columbia on 03/18/13. For personal use only.

Figure 3 Immunodominant regions of MBP in humans and various EAE models in


different species. MHC class II–restricted epitopes are shown on top (green), those
that are recognized in the context of MHC class I and/or encephalitogenic in EAE at
the bottom (blue).

EAE could readily be induced, and about 4% of these animals developed spon-
taneous disease. Furthermore, the same TCR cross-reacts with an EBV-derived
peptide in the context of DR2a, supporting molecular mimicry (79). The complex
of HLA-DR2b and MBP(84−102) was also detected in the brains of MS patients via
staining with a monoclonal antibody, which supports the notion that the immun-
odominant autoantigenic peptide is presented locally (127). A recent humanized
transgenic mouse model that combines another MS patient–derived MBP(83−99) -
specific TCR and DR2a also readily develops active and passive EAE, although
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

702 SOSPEDRA  MARTIN

we do not know yet whether spontaneous disease will occur (J. Shukaliak-Quandt,
unpublished observation). MBP(83−99) has received the most attention; however,
EAE can also be induced in humanized transgenic mice expressing a MBP(111−129) -
specific MS patient–derived TCR together with the restriction element DRB1∗ 0401
(100). Interestingly, only adoptive transfer EAE was inducible in this model, and
some animals not only develop signs of conventional EAE, i.e., limp tail, flac-
cid hind limb paresis, or paralysis, but also show signs of involvement of caudal
cranial nerves with swallowing difficulties and ataxia, which indicate that clini-
cal/phenotypic heterogeneity is related to the inducing myelin peptide (102).
Additional evidence supporting a role for MBP in MS includes cross-reactivity
between MBP(84−102) -specific Th1 cells and an identical sequence in the U24
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

antigen of HHV-6 (74), broader responses to MBP, and intra- and interindividual
by University of Missouri - Columbia on 03/18/13. For personal use only.

fluctuations of the specificities over time, without clear relation to inflammatory


MRI activity.

PROTEOLIPID PROTEIN (PLP) PLP is the most abundant CNS myelin protein (about
50%), highly hydrophobic and evolutionarily conserved across species. In mice,
there are two main transcripts, the full-length 276 amino acid isoform; and DM-
20, an isoform that lacks 35 amino acids and is mainly expressed in brain and
spinal cord prior to myelination but also in peripheral lymphoid organs, where
full-length PLP is barely found (114, 115, 128). The differential peripheral expres-
sion is relevant for one major encephalitogenic and immunodominant PLP(139−154)
peptide that is contained in full-length PLP, but is not contained in DM-20 (115,
128) and therefore is not available for thymic negative selection. Consequently,
high frequencies of PLP(139−154) -specific T cells have been observed even in
naive unprimed animals (128, 129). PLP is a stronger encephalitogen compared
with MBP, at least in some EAE models, particularly in SJL/J mice, in which
PLP(139−151) is dominant (130, 131). PLP TCR transgenic mice on the SJL/J
background develop spontaneous EAE with very high frequency (129). Upon
EAE induction with whole spinal cord homogenate in SJL/J mice, the dominant
T cell response is directed against PLP(139−151) , and during disease relapses pre-
dictable epitope spreading occurs to PLP(178−191) and later to MBP(89−101) (130,
131). If EAE in SJL/J mice is induced with either the secondary PLP(178−191) epi-
tope or with MBP(89−101) , further waves of the disease always involve reactivity
to PLP(139−151) (130, 131). Numerous other PLP peptides are encephalitogenic,
including PLP(178−191) , PLP(43−64) , PLP(56−70) , and PLP(104−117) in SJL/J mice,
PLP(217−233) in Lewis rats, and PLP(56−70) in Biozzi mice. Although examined less
extensively for PLP, the above parallels between encephalitogenic MBP epitopes
in EAE and immunodominant peptides in humans are also observed. PLP(104−117) ,
PLP(142−153) , PLP(184−199) , and PLP(190−209) peptides are immunodominant in the
context of the MS-associated DR2 alleles, but these peptides also bind to other
HLA-DR alleles (132, 133). Further immunodominant epitopes are PLP(30−49) ,
PLP(40−60) , PLP(89−106) , and PLP(95−116) (97, 134−136). As is the case in the
mouse, the human thymus does not express PLP(139−151) , which at least in part
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 703

explains the immunodominance in humans. The frequencies and skew toward a


Th1 phenotype of PLP-specific T cells are increased in MS, but not in every study.
PLP(139−151) and PLP(178−191) are main targets of high-avidity T cells and are clearly
elevated in MS patients (108).

MYELIN OLIGODENDROCYTE GLYCOPROTEIN (MOG) MOG, a 218 amino acid


transmembrane glycoprotein of the Ig superfamily, is much less abundant than
the major myelin proteins (0.01%–0.05%), and it is not located in compact myelin
but rather on the outer surface of the oligodendrocyte membrane. Owing to this
“strategic” location, it is directly accessible to antibodies and believed to be rel-
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

evant as a target for both cellular and humoral immune responses in MS. MOG
is expressed late in myelination and is only found in the brain/spinal cord and
by University of Missouri - Columbia on 03/18/13. For personal use only.

the retina, not in peripheral nerve. Furthermore, MOG expression is either com-
pletely or almost completely lacking in peripheral lymphoid tissues (114, 115).
MOG-induced EAE is best examined in C57/BL6 mice, in which the MOG(35−55)
peptide induces a chronic, nonrelapsing EAE (137). A recent MOG TCR transgenic
mouse model on the B6 background showed spontaneous EAE with inflammation,
demyelination, and axonal damage in brain and spinal cord in a small fraction of
animals, while 35% developed spontaneous optic neuritis (138). Optic neuritis is
also seen in MOG-induced EAE in DA rats (102), and the relatively higher ex-
pression of MOG in the optic nerve has been proposed as one explanation for the
involvement of the optic nerve (138). Differences in lesion location, as well as in
the involvement of antibodies versus T cells in different EAE models support the
notion that the inducing antigens and immunogenetic background contribute to
disease phenotype (102, 139).
Overall, much less information is available on the fine specificity of human
MOG-reactive T cells when compared with MBP and PLP. Immunodominant
epitopes have been located in the Ig-like extracellular domain of MOG(1−22) ,
MOG(11−30) , MOG(21−40) , MOG(31−50) , MOG(34−56) , MOG(63−87) , MOG(64−96) ,
MOG(71−90) (140−142), which also harbor several encephalitogenic epitopes (143),
but immunodominant areas have also been found in the intracellular parts of
MOG. MOG(146−154) is immunodominant with both DR15 (DRB1∗ 1501) and DR4
(DRB1∗ 0401) (144). Weissert et al. (144) reported stronger responses toward intra-
cellular portions of MOG and to different MOG peptides in MS patients, whereas
the reverse was observed by Lindert et al. (145). MOG(1−20) and MOG(35−55) pep-
tides are among the 6/15 myelin peptides from MBP, PLP, MOG, and CNPase that
account for clearly elevated high-avidity myelin-specific T cell responses in MS
patients, which supports the importance of MOG (108).

Other Myelin and Nonmyelin Antigens as


Targets for CD4+ T Cells
Investigators have examined the role of a few other myelin components and
nonmyelin proteins and glycolipids as antigens for CD4+ T cells. The order in
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

704 SOSPEDRA  MARTIN

which they are mentioned here does not reflect their importance, which is not yet
known.

MYELIN-ASSOCIATED GLYCOPROTEIN (MAG) MAG is a large (approximately


100 kDa) myelin glycoprotein located at the inner surface of the myelin sheath
opposing the axon surface. It accounts for less than 1% of total myelin protein
in the CNS and is even less abundant in the peripheral nervous system (PNS).
The pathogenetic relevance of MAG has been documented for polyneuropathies
by anti-MAG IgM (146). MAG(97−112) is encephalitogenic in ABH (H-2Ag7 ) mice
(147), and elevated MAG-specific T and B cell responses have been observed in
the CSF of MS patients by ELISPOT assays (148). Among the few MAG peptides
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

that have been examined, C-terminal areas, i.e., MAG(596−612) and MAG(609−626) ,
by University of Missouri - Columbia on 03/18/13. For personal use only.

are relatively immunodominant (148). The preferential location of CNS lesions in


cerebellum, centrum semiovale, and forebrain in MAG-induced EAE in Lewis rats
supports the notion that the antigen specificity is related to lesion location (102).

2 ,3 -CYCLIC NUCLEOTIDE 3 PHOSPHODIESTERASE (CNPase) CNPase exists in two


splice variants (CNPase I and II, 46 kDa and 48 kDa) and makes up 3%–4% of
total myelin protein. It is located in oligodendrocytes, mainly around the nucleus
and in the paranodal loops, but it is also expressed in peripheral Schwann cells and,
although much less, in lymphoid tissues. Its exact role is not clear. Encephalito-
genicity could not be demonstrated so far (147); however, immunization of Lewis
rats with a CNPase peptide with homology with mycobacterial HSP65 resulted
in protection from EAE (149). CNPase is immunogenic both in rodents and in
humans, and studies of the reactivity to either recombinant or native CNPase and
to overlapping CNPase peptides have located a number of areas with promiscuous
binding to several HLA-DR alleles, including the MS-associated DR15 molecules
(150, 151). A C-terminal area [CNPase(343−373) ] is one of the immunodominant
epitopes that is recognized preferentially by high-avidity myelin-specific T cells
of MS patients (108).

MYELIN-ASSOCIATED OLIGODENDROCYTIC BASIC PROTEIN (MOBP) MOBP was


discovered recently. Several splice variants exist, and the 81 amino acid isoform
is most abundant in rodent and human myelin. MOBP is exclusively expressed in
oligodendrocytes, appears late in myelination, and is located in the major dense line
of compact myelin. MOBP is encephalitogenic in SJL/J mice, and the encephalito-
genic epitope is located within amino acids 37–60 (152, 153). Preliminary studies
of cellular anti-MOBP responses in MS patients and controls identified one im-
munodominant region, MOBP(21−39) (152), and the reactivity of MOBP-specific
T cells cofluctuated with inflammatory MRI activity (154).

OLIGODENDROCYTE-SPECIFIC GLYCOPROTEIN (OSP) OSP is the third most abun-


dant myelin protein (7%), is expressed in the CNS and testis, and is located in
tight junctions. These characteristics led it to be grouped in the family of tight
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 705

junction proteins and to be renamed as OSP/Claudin-11. Several OSP peptides


induce EAE in SJL/J mice, and OSP-specific antibodies are found in the CSF of
RR-MS patients (155). By testing PBMC from RR-MS and SP-MS patients with
overlapping OSP peptides, investigators identified a number of immunogenic ar-
eas and observed overall strong responses in both healthy controls and RR-MS
patients but decreased reactivity in SP-MS (156).

α-B CRYSTALLIN (αB-C) Unlike the myelin proteins discussed above, αB-C was
identified as a candidate target in MS patients and not in EAE models. Van Noort
and colleagues (157) fractionated MS brain–derived proteins and then tested the
proliferation of PBMC from MS patients and healthy controls against brain protein
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

fractions. They observed prominent reactivity in one of the fractions and identi-
by University of Missouri - Columbia on 03/18/13. For personal use only.

fied the small heat shock protein αB-C as the relevant antigen (157). αB-C is a
major constituent of the eye lens, but it is also expressed in astrocytes and oligo-
dendrocytes in active MS lesions. A cryptic epitope of α-B crystallin, αB-C(1−16) ,
is weakly encephalitogenic in Biozzi ABH mice. In addition to the demonstra-
tion of strong responses to αB-C-containing MS brain–derived protein fractions,
DRB1∗ 1501-restricted CD4+ Th1 T cells in MS patients responded to peptides
αB-C(21−40) and αB-C(41−60) , although less to αB-C(131−150) (158). Other inves-
tigators documented comparable T cell responses to αB-C in MS patients and
healthy controls (159).

S100β PROTEIN Linington and colleagues (160) examined the astrocyte-derived


calcium-binding protein S100 in Lewis rats and observed a strong immune response
against the S100β epitope (amino acids 76–91). Unlike myelin antigens, S100 im-
munization or adoptive transfer of S100-specific T cells led to a panencephalitis
and uveoretinitis. However, disease induction with S100 led to little if any clinical
deficit (160). The lack of clinical disease was related to the decreased macrophage
recruitment, despite massive T cell infiltrates (160). Also, unlike MBP-specific
T cell lines, S100-specific T cells did not show cytotoxic activity. These observa-
tions parallel data from MS patients. Both CD4+ and CD8+ T cells specific for
S100β can be isolated with no differences among the groups (159, 161). S100-
specific CD4+ T cells exhibited cytotoxic activity less often compared with MBP-
specific T cells from the same donors (161).

TRANSALDOLASE-H (Tal-H) Tal-H was discovered on the basis of homologies with


the gag p17 protein of human T lymphotropic virus type I (162). Tal-H is a key
enzyme of the pentose phosphate pathway and is expressed in oligodendrocytes,
Schwann cells, and lymphoid tissues. High-affinity antibodies against Tal-H have
been found in the serum and CSF of MS patients, and Tal-H also stimulates pro-
liferation of MS PBMC (163).

IMMUNOGLOBULINS AS T CELL ANTIGENS Vartdal and colleagues (164) examined


the interesting hypothesis that the intrathecal Ig synthesis is involved in
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

706 SOSPEDRA  MARTIN

perpetuating the CD4+ T cell response. They found proliferative reactivity of


T cells to CSF Ig in 14 out of 21 MS patients and 4 out of 17 other neurological
controls, and preliminary studies indicate that CD4+ T cells responded in a DR-
restricted fashion (164). We have recently identified an IgG peptide as one target
of a CD4+ TCC (MN36) that was clonally expanded in the CSF of a MS patient
during exacerbation (M. Sospedra, unpublished observation). The specificity of
the TCC was identified with an unbiased technique, i.e., positional scanning com-
binatorial peptide libraries, and it supports the above hypothesis (164, 165) that
CSF-derived Ig may serve as an autoantigen that perpetuates the autoreactive T
cell response.
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

LIPID COMPONENTS AS ANTIGENS FOR CD4+ T CELLS IN MS Paralleling the ob-


by University of Missouri - Columbia on 03/18/13. For personal use only.

servation of elevated antiganglioside antibodies, particularly in the chronic pro-


gressive form of MS (166), Pender and colleagues (167) found enhanced T cell
reactivity against gangliosides GM3 and GQ1b in PP-MS patients only. These
findings suggest that ganglioside-specific T cells can contribute to axonal damage
in PP-MS, although further studies are required, particularly with respect to which
cell types and molecular mechanisms are involved in lipid recognition in MS.

ANTIGEN AVIDITY: CROSS-REACTIVITY AND DEGENERACY OF CD4+ T CELL RECOG-


NITION Despite long-held views that cellular immune responses are highly spe-
cific, it is now firmly established that the ability of T cells for cross-reactivity or
degeneracy is a normal phenomenon. Degenerate T cell recognition is not only
required for thymic positive selection on thymic self peptides but also for host
protection against potential antigens that outnumber the available T cells/TCRs by
several orders of magnitude. For self-antigens expressed in the thymus (e.g., MBP),
only T cells recognizing these antigens with low functional avidity are positively se-
lected. In agreement, most MBP-specific T cell lines respond to antigen at relatively
high concentrations in the micromolar range (69–71, 168), and pathogen-derived
peptides have been identified that activate MBP-specific TCC at several orders
of magnitude lower concentrations (169). However, high-avidity myelin-reactive
TCC also exist in the periphery, and they are elevated in MS patients compared
with controls (108). Furthermore, during a trial with an APL peptide derived from
MBP(83−99) , we identified TCC responding to both the APL peptide and MBP(83−99)
at subnanomolar concentrations (101), and one of these TCC was already present in
the peripheral blood seven years before the APL trial (107). It is not clear whether
changes in functional avidity occur during the disease process in the periphery,
either by changes in the requirements for costimulation or the TCR-associated
signaling machinery, or whether high-avidity T cells pass “thymic inspection” be-
cause central tolerance is less stringent in MS. With respect to self-antigens that
are not or are barely expressed in the thymus, such as MOG or the full length PLP
isoform, it is expected that autoreactive T cells are deleted less efficiently in the
thymus. This has been confirmed for PLP(139−154) -specific T cells in SJL/J mice.
Furthermore, the six myelin peptides (three MBP, one PLP, and two MOG peptides)
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 707

that are immunodominant for high-avidity T cells are either derived from proteins
that are not expressed in the thymus (PLP(139−154) and the two MOG peptides)
or that reside in areas that poorly bind to the MS-associated HLA-DR (108).
Recent observations suggest that the extent of cross-reactivity increases ei-
ther during the disease process, upon entering of the CNS/CSF, or during long-
term antigen stimulation within the CNS. CSF-derived and clonally expanded
TCC during disease exacerbation show a considerably higher degree of cross-
reactivity/degeneracy than previously studied peripheral blood-derived TCC (M.
Sospedra, unpublished observation; 169, 170). Some of these TCC further demon-
strate a high degree of promiscuity in HLA restriction, i.e., restriction by both
MS-associated DR and DQ molecules (M. Sospedra, unpublished observation).
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org
by University of Missouri - Columbia on 03/18/13. For personal use only.

HLA-RESTRICTION AND FUNCTIONAL CHARACTERISTICS OF AUTOREACTIVE CD4+


T CELLS Most myelin-specific CD4+ T cells are restricted by HLA-DR molecules
(68, 69, 71, 116–118, 121, 168, 171). MBP-specific T cell lines are mainly restricted
by MS-associated DR alleles (Figure 3), and the immunodominant peptides are
either promiscuous binders to all of them [MBP(83−99) ] or are recognized with one
HLA molecule, e.g., MBP(111−129) with DRB1∗ 0401 (121). Investigators do not
currently know what accounts for these parallels, but the peptide binding character-
istics of the disease-associated class II alleles, as well as the preferential generation
of certain protein fragments by the processing machinery, are probably important.
A few myelin-specific T cell lines have been restricted by DQ or DP molecules
(68, 71), although the reasons are not clear. Nonetheless, the observation that
myelin-specific T cells in Japanese patients with Asian-type (optico-spinal) MS
were in part HLA-DP5-restricted supports the notion that important interactions
occur among specific HLA class II molecules, T cell response, and disease char-
acteristics, because HLA-DP5 is associated with Asian-type MS (172). Another
study suggests that HLA-DP can be relevant for epitope spreading in MS (173).
With respect to function, the aggregate data support the idea that myelin-specific
T cells are skewed toward a Th1 phenotype (72, 108, 111, 132, 174). It is important
to consider the methodology of the respective study and the patients’ disease state
and subtype. Fluctuations of cytokine secretion have, for example, been linked
to the MRI-documented inflammatory activity, and elevated expression of IFN-γ
and TNF-α and reduced IL-10 have correlated with disease activity (175). Even
though inflammation decreases during SP-MS (Figure 1), the secretion of proin-
flammatory cytokines IL-12, IL-18, and IFN-γ are elevated during later stages
(176, 177), the activation of Th1 cells is less strictly controlled, and Th1 markers
CCR5 and TIM-3 are upregulated (178, 179). When the specificity and cytokine
expression of high-avidity T cells has been linked to MRI characteristics in MS pa-
tients, correlations have been observed between MOG-specific Th0/2 cells and less
inflammation, as well as between MBP-specific Th1 cells and a more destructive
disease process (108).
The cytotoxic activity of CD4+ T cells is relatively poorly understood compared
with that of CD8+ T cells. MBP-specific CD4+ T cells mediate both perforin- and
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

708 SOSPEDRA  MARTIN

Fas/Fas-L-mediated cytotoxicity of MBP or MBP peptide–pulsed targets. When


comparing MBP-specific T cells restricted by either DR2a or DR2b, only the
former employ perforin-mediated killing or are noncytotoxic, whereas DR2b-
restricted T cells exclusively exhibit Fas/Fas-L-mediated cytolysis (122, 171, 180,
181). Currently, it is not known how this relates to the pathogenesis of MS. It is
unlikely that direct lysis of oligodendrocytes, and even fewer neurons, involves
CD4+ T cells because neither type of CNS cells express HLA class II. A sub-
type of MBP-specific CD4+ TCRαβ + T cells expresses the neural cell adhesion
molecule family member CD56 (also a marker for NK cells) and is capable of
lysing CD56+ target cells via homotypic CD56-CD56 interactions independent
of HLA restriction (180). A number of CNS cells, including oligodendrocytes,
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

express CD56, and CD4+ CD56+ T cells can indeed lyse oligodendrocytes in an
by University of Missouri - Columbia on 03/18/13. For personal use only.

HLA-unrestricted fashion (182).


Furthermore, the requirement for costimulation, i.e., the interaction of CD80/86
on APCs with CD28 on T cells, as well as the control via the negative costimulator
CTLA-4, is perturbed in CD4+ T cells in MS. CD4+ myelin-specific T cells, as
well as T cells with specificity for other antigens, are less dependent or independent
of costimulation (183–186), and they do not respond, or respond less, to CTLA-4
(185, 187). The latter is due to the absence of CTLA-4 upon activation on CD4+
CD28− T cells. Furthermore, this cell population is characterized by a clear Th1
skew, seemingly increased proliferative capacity, and relative enrichment for au-
toreactive T cells (185). The susceptibility to activation-induced cell death via
Fas/Fas-L interactions is not generally impaired. However, data—including the
increased expression of the antiapoptotic molecules survivin, bcl-2, and inhibitor
of apoptosis (IAP) family members IAP, IAP-2 and X-IAP in MS T cells, their
heightened expression during disease exacerbations, and downregulation by IFN-
β—all suggest that the regulation of apoptosis is perturbed in MS, although some
aspects are not different from controls.
CD4+ immunoregulatory T cells (Tregs) are characterized by CD25high expres-
sion and the transcription factor Fox-P3 (188, 189). CD4+ CD25+ Tregs suppress
T cell proliferation by both cell-cell contact and cytokine-mediated mechanisms.
The number and function of CD4+ CD25+ Tregs appear reduced in MS patients
(190). CD4+ Th2/3 cells and their cytokines IL-4, IL-10, and TGF-β are probably
largely beneficial in MS (191); however, under certain circumstances Th2 cells
can induce EAE (192), and Th2-controlled cell populations, e.g., mast cells, can
contribute to tissue damage in MS.

TCR REPERTOIRE Early EAE studies have demonstrated a restricted TCR reper-
toire in some models. Initial data in MS patients appeared to confirm these data,
and a restricted expression of Vβ17 was described (193); however, this particular
report was heavily influenced by data from one individual. Subsequent research
has described a restricted TCR repertoire either within single MS patients, but
not interindividually (194), or across the entire MS populations (195), or not
(196). Among the most often found Vβ chains are Vβ5.2, Vβ5.3, and Vβ6.2
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 709

(117, 195, 197), or in PLP-specific T cells Vβ2 (136). Subsequent studies have
examined CDR3 spectratypes or oligoclonality by single-strand conformational
polymorphism typing and sequencing (136, 197–201). They described (a) an as-
sociation of oligoclonal TCR CDR3 spectratypes, particularly in Vβ5.2 T cells
(200); (b) prevalence of TCR Vβ13-associated junctional sequences at disease
onset (198); (c) increased MBP reactivity and IFN-γ and IL-2 secretion in CD4+
and CD8+ T cells with altered CDR3 length distributions (199); (d) an oligo-
clonal expansion of T cells with distinct TCRs in the CSF (201; M. Sospedra,
unpublished observation); and (e) the observation of Vβ5.2-associated junctional
sequences from MS brain–derived TCRs similar to an MBP TCC (117, 197), as
well as CDR3 motifs in PLP-specific T cells that showed homologies with TCRs
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

from MS brains (136). Finally, the comparison of the TCR Vα chain usage in
by University of Missouri - Columbia on 03/18/13. For personal use only.

monozygous concordant and discordant twins showed that the overall TCR Vα
chain repertoire in discordant twins is different, and not only in MBP-specific but
also in tetanus-specific T cells (202). A recent study that focused on the CDR3
spectratypes in naive T cells of discordant monozygous twins found similar distor-
tions in both the healthy and diseased twins (203). Such repertoire shifts in naive T
cells may predispose one to MS development, but they are probably not sufficient.

CD8+ T Cells
Much less is known about CD8+ T cells than CD4+ T cells, not only in MS
but in other human autoimmune diseases as well. Technical difficulties in grow-
ing and characterizing CD8+ TCC have probably contributed to this temporary
neglect. In the context of effector functions, however, CD8+ T cells are much
better suited than CD4+ T cells to mediate CNS damage for the following reasons:
(a) Except for microglia, none of the resident CNS cells express MHC class II; it
can be induced on astrocytes by IFN-γ (32), but not on oligodendrocytes or neu-
rons, and therefore the latter can only be recognized by CD8+ T cells (204, 205);
(b) prominent oligoclonal expansions of CD8+ memory T cells have been found
in the CSF (31) and in MS brain tissue (206), and a persistence of CD8+ TCC
in CSF and blood (206); (c) CD8+ T cells are more prevalent in MS brain tissue
than are CD4+ T cells (207); (d) MHC class I can be induced on neurons that
are functionally compromised (208), and CD8+ virus-specific T cells can directly
lyse neurons via Fas/Fas-L-mediated cytolysis (205); (e) a number of HLA class
I–restricted myelin epitopes have been described for MBP, PLP, MAG, and others
(110, 209–211), and the CD8+ cytotoxic T cell response to MBP is increased
in MS patients (211); (f) CD8+ myelin-specific T cells secrete chemoattractants
(IL-16 and IP-10) for CD4+ myelin-specific T cells (212); and (g) the MBP(79−87) -
specific CD8+ TCC from wild-type C3H mice are encephalitogenic and induce a
disease phenotype that resembles MS more closely with respect to the presence of
ataxia and spasticity than some of the CD4+ T cell–mediated EAE models (10)
(Figure 2). Further data supporting a role for CD8+ T cells in MS are the increased
production of lymphotoxin (LT) in SP-MS patients, their increased adhesion to
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

710 SOSPEDRA  MARTIN

brain venules, an increased frequency of CD8+ T cells against EBV epitopes in


MS patients, and a correlation between cytokine production by CD8+ T cells and
MRI-documented tissue destruction (213). Taken together, there is little doubt that
both CD4+ and CD8+ T cell responses contribute to MS pathogenesis, albeit at
different steps and with different roles.

B Cells and Antibodies in MS


The observation that Igs are elevated in the CSF of MS patients (214) has been the
most important and earliest evidence suggesting a role for B cells and antibodies in
the pathology of MS. The correlations between increased CSF Ig with episodes of
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

worsening and the absence of OCB in some patients with benign MS also suggest
by University of Missouri - Columbia on 03/18/13. For personal use only.

the involvement of humoral responses.


B cells do not cross the intact BBB; however, once inflammation has started,
B cells, antibodies, and complement can enter the CNS. The observation of in-
creased Igs in the CSF in MS patients (214), but not in the serum, indicates local pro-
duction. B cell activation can occur because of stimulation with antigen from either
self or foreign proteins, through a random bystander effect during inflammation in
MS lesions, or by superantigen stimulation. Sequence analysis of the Ig variable
regions have shown a high frequency of clonally expanded memory B cells that ex-
press variable heavy chain-4 type in MS CSF (215) and also in lesions (216, 217),
which suggests selection by a specific antigen. In addition, CSF Igs of MS patients
show an oligoclonal distribution, i.e., only a limited number of B cell clones con-
tributes to the increased CSF Igs (218, 219). B cells and antibodies can contribute to
MS disease pathogenesis in various ways. (a) B cells can serve as APCs for autore-
active T cells. Supporting this mechanism is the observation that the epitope speci-
ficity of the antibodies generated during EAE, the encephalitogenic T cell epitopes,
and the immunodominant T and B cell epitopes in humans often overlap (220, 221).
(b) B cells provide costimulation to autoreactive T cells. (c) B cells and tissue-
bound Ig can recruit autoreactive T cells to the CNS (222). (d) Idiotope-specific
T cells may be activated by CSF Igs, and these T cells sustain B cells that pro-
duce such idiotopes (164). (e) The production of myelin-specific antibodies and
the destruction of myelin within plaques appear to be the most important way that
B cells contribute to pathogenesis.
In 1959, investigators demonstrated that humoral factors may have a role in in-
flammatory demyelination by the in vitro demyelinating activity of a serum factor
(223), which was later identified as myelin-specific Igs. Further support came from
histopathological studies of CNS tissue and the analysis of CSF. B cells, plasma
cells, and myelin-specific antibodies are detected in MS plaques and in areas of
active demyelination in MS patients (224–226). Antibodies can cause demyeli-
nation by opsonization of myelin for phagocytosis (227–229). Another antibody-
mediated mechanism of demyelination acts via complement activation, leading to
membrane attack complex (MAC) deposition and complement-mediated cytolysis
(230). Studies of MS lesions found complement in areas of active demyelination
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 711

(5, 231), and neurological disability and terminal complement concentrations in


CSF correlate in MS (232). MAC-enriched vesicles in MS CSF support this mech-
anism (233). The demyelinating potential of antibody in EAE has been corre-
lated with complement fixation, and soluble complement receptor inhibits EAE
severity.
The antigen specificity or specificities of CSF antibodies in MS have yet to
be established. Most of the CSF OCB are not directed against the major myelin
components (234), but sometimes against infectious agents (235). Several com-
prehensive reviews have been published on this subject (236–238), and below
we emphasize recent data. Assuming a pathogenic role of autoantibodies in MS,
the search for autoantigens has focused on myelin proteins and other CNS com-
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

ponents. A pathogenic contribution of MBP-specific antibodies in EAE has not


by University of Missouri - Columbia on 03/18/13. For personal use only.

been established and is controversial in MS. Although some studies emphasize


the relevance of MBP-specific antibodies (239, 240), others fail to confirm these
data. Unbiased screenings of antigen libraries with CSF Igs did not identify MBP
epitopes (241). Technical considerations, as well as the low-affinity interactions
of these antibodies (242), contribute to the controversial results. Increased num-
bers of anti-PLP-secreting B cells have been detected in the CSF of MS patients
(243), and those with a prominent anti-PLP response are distinct from patients
with anti-MBP antibodies (244). Igs against minor myelin components have also
been shown. MOG is the most interesting candidate B cell autoantigen in MS.
Anti-MOG antibodies are able to cause myelin destruction in EAE (245–248), in
contrast to anti-MBP or -PLP antibodies (249). Anti-MOG antibodies have also
been found in human MS lesions (226). The B cell response to MOG is enhanced
in MS (145, 250). Serum anti-MOG antibodies in patients with first CNS symp-
toms of MS and MRI lesions are predictive of subsequent exacerbations and the
diagnosis of definitive MS (251); however, because anti-MOG antibodies are also
frequent in controls, important questions remain.
Antibodies with specificity against minor myelin components, other autoanti-
gens, lipids, and DNA are summarized in Table 3. The observation that autoan-
tibodies against ubiquitous antigens are present in MS suggests that less biased
search approaches should be applied. A recent flow cytometry study compared
antibody binding to human cell lines between patients with MS and patients with
other inflammatory CNS diseases (252). The study observed antibodies to oligo-
dendrocyte precursors without differences between RR-MS and SP-MS. Binding
to a neuronal cell line was increased in SP-MS. Although the antigens targeted
by these antibodies have to be elucidated, this approach to identifying accessible
cell surface autoantigens that could mediate demyelination or neuronal damage
appears promising. Another unbiased method employs array-based or proteomics
technologies to characterize autoantibodies directed against candidate antigens in
cohorts of autoimmune and control patients, as well as in experimental autoim-
mune conditions (253).
Interestingly, antibodies may also play beneficial roles in two ways. First, they
may skew the cytokine pattern toward Th2. Rats treated with an encephalitogenic
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

712 SOSPEDRA  MARTIN

TABLE 3 Antibody specificities against CNS components other than MOG and MBP
Target antigen Remarks

Myelin-associated glycoprotein (MAG) Low titers in MS; possible involvement in


progression
Oligodendrocyte-specific protein (OSP) Minor myelin component
2 3 cyclic nucleotide —
3 -phosphodiesterase (CNPase)
Transaldolase-H Oligodendrocyte component
Glyco-shingolipids Lipid component of myelin
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

Sulfatides Lipid component of myelin


by University of Missouri - Columbia on 03/18/13. For personal use only.

GD1a and GM3 (ganglioside) Lipid component of myelin


Galactocerebroside (Gal-C) Major myelin lipid; anti-Gal-C has demyelinating
activity in vitro; anti-Gal-C antibodies exacerbate
EAE
α-B crystallin (small heat shock protein) Detected in MS sera; isotype prevalence
controversial
Neurofilament-L (NF-L) Elevated in MS CSF, suggested as indicator of
axonal damage; elevated in the CSF in
progressive MS
AN2 (oligodendrocyte surface Expression of AN2 on oligodendrocyte precursor
glycoprotein) cells suggests involvement in suppression of
remyelination
Nogo-A (neurite outgrowth inhibitor) Anti-Nogo antibodies are frequent in serum and
CSF of MS patients, but also in controls
Proteasome (protein complex involved in Anti-proteasome antibodies found in serum and
processing and chaperone function) CSF in MS
DNA High-affinity antibodies found in MS CSF; from
the role of anti-DNA antibodies in CNS lupus, it
is speculated that they might bind to neurons and
oligodendrocytes

peptide of MBP coupled to monoclonal anti-IgD are resistant to induction of


EAE after sequent challenge with MBP in CFA (254). Second, antibodies against
CNS components, e.g., Nogo-A, can foster myelin repair. IgM antibodies against
certain CNS antigens enhance remyelination in different animal models of MS
(255). Further evidence for a beneficial role of antibodies stems from the use of
pooled intravenous Ig in the therapy of MS (256), which acts through a num-
ber of mechanisms, including Fc-receptor blockade, inactivation of cytokines,
complement inhibition, blocking of CD4 and MHC, and modulation of apoptosis
(257).
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 713

INNATE IMMUNE MECHANISMS IN MS


Innate immune responses involve (a) recognition of conserved molecular struc-
tures produced by microbial pathogens via TLRs, cells such as macrophages,
neutrophils, and mast cells, and effector mechanisms such as the production of
lysozyme, lactoferrin, phagocyte oxidase, and nitric oxide; and (b) the recognition
of molecular structures expressed only on normal, uninfected host cells that serve as
indicators of “normality” and inhibit immune activation [NK cells, complement,
and receptors of the C-type lectin family (258)]. Although the innate immune
system’s main role is self-protection and maintenance of homeostasis, innate im-
mune mechanisms can in some circumstances result in destructive autoimmunity.
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

We summarize important findings and include observations that suggest a role for
by University of Missouri - Columbia on 03/18/13. For personal use only.

NKT cells and γ δ T cells in the pathogenesis of MS.

Toll-Like Receptors (TLR)


TLR function as sentinels by recognizing conserved pathogen-associated molecu-
lar patterns and generating proinflammatory signals that initiate adaptive immune
responses. They are expressed by a wide array of immune and nonimmune cells.
Inappropriate TLR signaling may contribute to diseases such as MS. TLR engage-
ment on dendritic cells (DCs) inhibits immunosuppressive effects of CD4+ CD25+
regulatory cells on effector T cells via IL-6 (259), and mice deficient in IL-6 are
more resistant to induction of autoimmune diseases (260). TLRs could further play
a role by breaking peripheral tolerance to self-antigens during chronic infections.
Assuming that autoreactive T cells are part of the normal T cell population and
tightly regulated, it has been hypothesized that tolerance is maintained in a similar
manner to mature T cells after recognition of antigen presented by resting or in-
activated DCs (261). Under normal conditions, APCs remain in their resting state
and induce tolerance in autoreactive T cells, whereas danger signals activate APCs
and may convert tolerized autoreactive T cells into effector cells. The increase of
MS exacerbations around viral infections supports this concept (56), and pretreat-
ment of mice with bacterially derived DNA exacerbates EAE (262). In a recent
EAE study, the stimulation of TLR9 with CpG oligonucleotides breaks tolerance
and renders lymph node cells reactive against a self-antigen to which they were
previously unresponsive (86, 263).

Mast Cells
Mast cells are activated during allergic reactions through crosslinking of surface
IgE receptors, which leads to degranulation of multiple mediators. Mast cells
are ubiquitously distributed among tissues including the brain, but their numbers
in the CNS are low and their role unclear. Investigators have suggested several
effects of mast cells in MS (264). Elevated numbers in MS plaques were originally
shown in 1890 (265) and later confirmed by others (266). They are attracted to MS
lesions via chemokines. RANTES, a potent attractant for mast cells, is elevated in
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

714 SOSPEDRA  MARTIN

MS lesions (39). Interestingly, mast cell–released mediators such as tryptase and


histamine are increased in the CSF of MS patients. Gene expression profiling of
MS plaques also demonstrated elevated expression of mast cell mediators in acute
lesions (36). Mast cells and their mediators can act in MS during BBB opening
and augment CNS infiltration via increased recruitment, adhesion, rolling, and
extravasation of leukocytes through the chemokines/cytokines lymphotactin and
IL-16, through TNF-α and IL-1-mediated induction of ICAM-1 and VCAM-1
expression, and through the effects of histamine and tryptase on leukocyte rolling.
Mast cell proteases such as tryptase and chymase activate matrix metalloproteinase
(MMP) precursors, and mast cells can also synthesize MMP-2 and MMP-9 directly.
It has been suggested that mast cells may act as APCs and influence MS by shaping
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

Th1/Th2 responses, but a clear demonstration of this role is lacking. Finally, mast
by University of Missouri - Columbia on 03/18/13. For personal use only.

cell degranulation in response to MBP can lead to demyelination in vitro via


proteolytic enzymes. Mast cell mediators can participate in the destruction of
oligodendrocytes and neurons. With respect to the gender bias of MS and the
elevated inflammatory activity in women, it is interesting to note that mast cells
express estrogen receptors.

Nitric Oxide Synthase


Phagocytes (granulocytes and macrophages) are equipped with the enzymatic ma-
chinery to generate highly toxic reactive oxygen and nitrogen intermediates, which
exert potent antimicrobial activities. The enzyme inducible nitric oxide synthase
(iNOS) generates large amounts of nitric oxide (NO), a short-lived and bioactive
free radical that is toxic to bacteria. NOS has been found in MS lesions, suggesting
a role in MS pathology (267). Although initial studies have shown that NO can
mediate microglia-induced cytotoxicity (268) and also necrosis of rodent oligo-
dendrocytes (269), the actual role of NOS in CNS injury in MS is not clear. Results
from blocking NOS in EAE are not conclusive, and additional data suggest that
NO may even have an antiapoptotic effect or modulate immune responses in a
beneficial way.

NK Cells
An association between decreased NK cell activity and MS was first reported in
1980 (270), and later studies expanded this knowledge (271), although findings
remain controversial. Potential explanations are disease heterogeneity among pa-
tient groups and fluctuations of NK activity and number during the disease course.
NK lysis is reduced prior to and during acute exacerbations compared with chronic
disease (271, 272) and normal during stable phases. Multi-parameter flow cytome-
try demonstrated that NK cells are significantly reduced in MS (273). Furthermore,
NK deficiencies exist in peripheral blood, placques, and CSF of MS patients (274).
Furthermore, NK cell depletion in two different EAE models exacerbate disease
(275, 276), whereas the transfer of in vitro generated NK cells decrease autoimmu-
nity (277). NK cells could suppress autoimmunity by cytokine production (IL-5,
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 715

IL-13, TGF-β) or by the induction of target lysis via perforin- and/or TRAIL-
dependent mechanisms. Supporting data come from perforin-deficient lpr mice
that developed severe autoimmunity (278) and blockage of TRAIL-exacerbated
EAE (279). In a recent phase II clinical trial with a humanized monoclonal anti-
body against the IL-2 receptor α chain in MS (280), we observed marginal effects
on CD4+ T cells but an expansion of CD56bright immunoregulatory NK cells. The
relative and absolute expansion of the latter NK cell population and their increased
perforin expression correlate highly with the reduction of the inflammatory activ-
ity, and in vitro experiments demonstrated direct lysis of activated CD4+ T cells
via perforin (B. Bielekova, unpublished observation). These observations indicate
that NK cells may exert important immunoregulatory functions in MS.
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org
by University of Missouri - Columbia on 03/18/13. For personal use only.

Complement
Complement serves as an auxiliary system in antimicrobial defenses. The human
brain is considered an immunoprivileged site and separated from the periphery
via the BBB. Nevertheless, all major CNS cells produce most of the complement
proteins. Astrocytes are the main CNS complement source, thus providing immune
defense against pathogens, and also contributing to damage in some diseases.
Demyelination not only results from an autoimmune response against myelin via
the classical pathway, but also from direct complement activation after binding of
complement to myelin. Purified CNS myelin, but not PNS myelin, can activate
the classical pathway (281). Furthermore, mature rat oligodendrocytes are lysed
in vitro by complement in the absence of antimyelin antibodies (282). MOG may
be capable of binding and activating the C1q component of complement (283)
because it harbors a domain similar to the C1q-binding sequence of antibodies.
Complement activation results in oligodendrocyte lysis and chemoattraction of
macrophages. Susceptibility of oligodendrocytes to complement injury could be
facilitated by the lack of the protective and ubiquitously distributed complement
inhibitors. CR1 (CD35), membrane cofactor protein (CD46), and homologous
restriction factor were not expressed on oligodendrocytes, whereas CD59 showed
substantial heterogeneity (283, 284).

NKT Cells
NKT cells share characteristics with T and NK cells and play a regulatory role in
autoimmunity as well as in immune responses to tumors and infections via secretion
of high levels of IL-4 and IFN-γ . Both CD4− and CD4+ cells contain NKT cells,
and in humans CD4− and CD4+ cells express a conserved canonical TCRα chain,
Vα24JαQ, paired with a selected Vβ11 segment. NKT cells recognize glycolipids
presented by the nonclassical class I–like CD1d molecule (285). A considerable
reduction of Vα24JαQ+ cells among Vα24+ cells has been observed in MS blood
(286) and confirmed by another group that also showed reduced Vα24 Vβ11+ NKT
cells (287). A further study failed to detect decreased NKT cells within Vα24+
cells, but did detect reduced production of IL-4 by Vα24JαQ TCC (288). A role
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

716 SOSPEDRA  MARTIN

for NKT cells in MS is supported by EAE data. An analog of α-galactosylceramide


(GalCer), a synthetic glycolipid that binds CD1d and stimulates mouse and human
NKT cells, suppresses EAE by selective IL-4 induction (289). CD4+ NKT cells
are probably the main NKT regulatory population.

γ δ T Cells
γ δ T cells represent another distinct lymphocyte population that mediates host
defense and immunoregulatory functions. The expression of NK cell inhibitory
receptors on human γ δ T cells indicates a role for γ δ T cells in tumor immu-
nity and autoimmunity. Two main fractions of γ δ T cells have been described.
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

One fraction of γ δ T cells expresses Vγ 1 within epithelial tissues, where it may


provide a first line of defense against infections and cancer. The second fraction
by University of Missouri - Columbia on 03/18/13. For personal use only.

that expresses Vγ 2 represents the majority of peripheral blood γ δ T cells. This


fraction infiltrates chronic lesions and is detected in the CSF of MS patients (290,
291). Interestingly, oligodendrocytes selectively stimulate the expansion of the
Vγ 2 subtype of γ δ T cells (292). Limited TCR heterogeneity of CSF-infiltrating
γ δ T cells in MS suggests a common antigen reactivity (293, 294). Human γ δ T
cells can lyse oligodendrocytes via perforin without the need for APCs, possibly
through recognition of heat shock proteins (291, 295), αB crystallin, or even non-
peptide antigens (296). These findings, together with EAE studies in which γ δ T
cells appear to be important early mediators of damage (297), support a role for
γ δ T cells in MS pathogenesis.

CYTOKINES AND CHEMOKINES IN MS


Cytokines
Cytokines orchestrate all phases of immune responses, act in highly complex, dy-
namic networks in paracrine and/or autocrine fashion, and often exert overlapping
and in part redundant functions via multi-component receptor molecules that may
be shared by different cell types. To maintain homeostasis, a dynamic balance
between pro- and anti-inflammatory cytokines is required. Proinflammatory cy-
tokines are thought to play a role in the pathogenesis of MS via immune system
activation in the periphery and/or by directly damaging the oligodendrocyte/myelin
unit. Anti-inflammatory cytokines, e.g., IL-4, have been considered beneficial. We
summarize the main findings about proinflammatory cytokines (IFN-γ , TNF-α,
IL-12, IL-17, and IL-23), anti-inflammatory cytokines (IL-4, IL-10), and others
exerting both effects (IL-6) in MS (298). Proinflammatory cytokines can partici-
pate in the pathogenesis of MS at different points. Elevated numbers of blood cells
expressing TNF-α mRNA (299), serum TNF-α concentrations (300), and PBMC
secreting TNF-α (301) have been reported in MS patients. Nevertheless, therapy
with a soluble TNF-α receptor Ig fusion protein or anti-TNF-α leads to increased
and prolonged MS exacerbations (302). Results about IFN-γ in the blood of MS
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 717

patients are conflicting. Although higher numbers of PBMC expressing IFN-γ


mRNA and serum levels (300) have been found in MS, other studies found no
differences (303). A therapeutic trial with IFN-γ in MS resulted in disease exacer-
bation (304). The role of IFN-γ in EAE is also not clear. The prevailing perception
is that IFN-γ -secreting T cells are encephalitogenic, but IFN-γ -knockout animals
develop much worse or even lethal EAE compared with wild-type littermates. IL-
12, a main stimulator of IFN-γ has been implicated as a proinflammatory cytokine
(305), but recent data indicate that IL-23, a cytokine that shares the p40 chain
with IL-12, is the main mediator of these effects (306). In MS, some studies have
reported higher numbers of PBMC expressing IL-12 p40 mRNA (307), but other
studies found no differences (308).
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

Data about anti-inflammatory cytokines in MS are similarly contradictory. De-


by University of Missouri - Columbia on 03/18/13. For personal use only.

creased numbers of PBMC secreting IL-10 and lower serum levels of IL-10 in
MS have been reported (309). Moreover, investigators have described decreases
in IL-10 expression but elevated numbers of PBMC expressing IL-10 mRNA be-
fore clinical relapses (310). Therefore, the role of IL-10 in MS is currently not
clear. Increased levels of IL-6, a cytokine with pro- and anti-inflammatory ca-
pacities, have been shown in MS patient serum (301). Within the CSF and brain,
proinflammatory cytokines can damage the oligodendrocyte/myelin unit. Higher
numbers of mononuclear CSF cells expressing TNF-α and IFN-γ have been de-
tected in MS patients. TNF-α has proinflammatory functions but is also involved
in tissue repair in the brain. Proinflammatory cytokines have also been found
in active MS lesions (311, 312). The expression of TNF-α is elevated in active
demyelinating lesions compared with inactive/remyelinating lesions (313), and
transgenic mice overexpressing TNF-α and IFN-γ driven by the astrocyte-specific
GFAP promoter induced demyelination (314). Investigators have proposed differ-
ent mechanisms for this demyelination: (a) TNF-α and IFN-γ may be toxic for
oligodendrocytes; (b) cytokines may activate macrophages and microglia, which
then phagocytose myelin; and (c) proinflammatory cytokines may be involved in
apoptosis induction/execution and subsequent demyelination. The addition of IFN-
γ to cultured oligodendrocytes renders them susceptible to Fas ligand–mediated
apoptosis by inducing Fas expression on their surface (315). The proinflammatory
cytokines IL-12 and IL-17 are also elevated in CSF and brain lesion of MS patients
(36).
Unexpectedly high numbers of cells expressing IL-4 mRNA have been observed
in MS CSF lesions (316). Studies on IL-10 and IL-6 have been contradictory.
Assuming a beneficial role of the Th2 cytokines, at least two interpretations can be
proposed: (a) These cytokines, mainly IL-10, could be involved in MS pathogenesis
by augmenting B cell proliferation, differentiation, and antibody production. In
line with this hypothesis, a correlation between IL-10 levels and IgG in the CSF
of MS patients has been reported (317). (b) The presence of IL-4, IL-10, and
TGF-β in CSF or MS brain parenchyma could reflect ongoing immunoregulatory
mechanisms that are initiated after disease exacerbations and are important for
disease resolution/prevention in EAE (318).
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

718 SOSPEDRA  MARTIN

Chemokines
Chemokines and their receptors play a central role in the inflammatory recruit-
ment of leukocytes and other cell types. Trafficking of inflammatory T cells into
the CNS is a crucial step in MS and begins with weak adhesion and rolling on
the endothelium of the BBB, followed by firm arrest on the luminal side of the
endothelium and subsequent diapedesis across the BBB. Chemokines induce and
activate leukocyte adhesion molecules that mediate firm adhesion to the endothe-
lium and establish a chemotactic concentration gradient that results in recruitment
across the endothelial monolayer. The induction of proteolytic enzymes facilitates
BBB opening (319), and subsequently chemokines mediate retention of leukocytes
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

in the CNS. Numerous reports analyze the roles of chemokines and their receptors
by University of Missouri - Columbia on 03/18/13. For personal use only.

in intrathecal accumulation of T cells in MS (320). Among the various chemokine


receptors, CCR5 and CXCR3 have received attention as key receptors on Th1 cells,
as have CCR3 and CCR4 on Th2 cells. Furthermore, CCR7, an important marker
for the capacity of mononuclear cells to migrate to secondary lymphoid organs, is
also of interest. This section summarizes the main findings on chemokines in the
blood, CSF, and lesions in MS.

BLOOD CCR5 expression is increased on circulating T cells in MS patients (321,


322) and during disease relapse, suggesting a pathogenic role of CCR5+ T cells
(323). Increased CXCR3 expression on circulating T cells has been shown in some
but not all studies (322). T cells expressing CCR5 and CXCR3 in MS produce high
quantities of IFN-γ and TNF-α (324), and MBP-specific Th1 cells express high
levels of CXCR3 and CXCR6 (325). The effects of IFN-β treatment on chemokine
and chemokine receptor expression are controversial.

CSF CCL5 (RANTES) and CXCL10 (IP-10) are elevated in MS CSF, whereas
CCL2 (MCP-1) is significantly decreased (326). The increase of CXCL10 (IP-10)
and decrease of CCL2 (MCP-1) has been confirmed to take place during MS exac-
erbations and not to occur during remissions (327). CCL2 (MCP-1) decreases cor-
relate with active MRI, i.e., presence of inflammation and gadolinium-enhancing
lesions in the brain (328), suggesting a Th1 polarization in active MS. CCL3
(MIP-1α) has been found in the CSF of MS patients, as well as in other neuroin-
flammatory diseases. The source of these chemokines in the CSF remains to be
elucidated.
With respect to chemokine receptors, initial studies document a higher propor-
tion of CSF T cells that express CXCR3 and CCR5 (326) compared with PBMC.
Because CSF T cells are enriched for the CD4+ /CD45RO+ subset, corrections for
this bias have shown that only CXCR3, but not other receptors (CCR1-3, CCR5,
and CCR6), is relatively increased on CSF (329). Interestingly, the same has been
observed in controls and interpreted such that the presence of CXCR3+ cells in
the CSF is independent of CNS inflammation (326). CXCR3 expression probably
facilitates the entry of T cells into the CSF, and CXCL10 (IP-10) mediates the
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 719

retention in the inflamed CNS. CCR5+ and CXCR3+ Th1 cells in the CSF also
express CCR7 (330), and CSF-infiltrating monocytes express higher CCR1 and
CCR5 levels (331). But similar results were obtained in controls, suggesting that
the presence of CCR1+ /CCR5+ monocytes in the CSF is independent of CNS
inflammation.

BRAIN LESIONS A number of chemokines and the corresponding receptors have


been detected in MS brain lesions, indicating that they might evolve into interesting
therapeutic targets. CCL3 (MIP-1α), CCL4 (MIP-1β), and CCL5 (RANTES) are
expressed within MS lesions, CCL4 in parenchymal inflammatory cells (macro-
phages and microglia), CCL3 also in parenchymal inflammatory cells and activated
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

neuroglia (332), and CCL5 in perivascular inflammatory cells and (though less so)
by University of Missouri - Columbia on 03/18/13. For personal use only.

in astrocytes (39, 332, 333). Other chemokines in active MS lesions include CCL2
(MCP-1), CCL7 (MCP-3), CCL8 (MCP-2), and CXCL10 (IP-10). CXCR3 is ex-
pressed on the majority of perivascular T cells in MS brain lesions, and CCR5 on a
subset of these cells. CCR1 has been found on newly infiltrating monocytes (331),
CCR2 and CCR3 on macrophages (333), and CCR5 on infiltrating monocytes and
activated microglia cells (324, 326, 333).
A role for chemokines and their receptors in MS is supported by EAE data.
Increased expression of CCL2, CCL3, CCL5, and CXCL10 in EAE is associated
with disease progression, and in vivo depletion improves EAE (334). Mice deficient
in CCR2 (335), and to a lesser extent in CCR1 (336), fail to exhibit EAE symptoms.
In contrast, CCR5-deficient mice showed disease severity similar to controls (337),
which suggests that T cell accumulation in the CNS during EAE does not function
through CCR5.
Polymorphisms in genes for chemokines and their receptors have been proposed
to confer susceptibility or protection in MS, although definitive evidence is still
lacking. The CCR5 32 mutation leads to a nonfunctional receptor that has been
associated with decreased severity of MS. Although homozygous individuals for
CCR5 32 were not protected from MS, heterozygosity for 32 has been linked
to prolonged disease-free intervals and a delay in MS onset. Microsatellite poly-
morphisms in CCL7 (MCP-3) have also been associated with disease resistance
to MS.

PATHOGENETIC STAGES IN THE DISEASE PROCESS


IN MS: LESION PATHOLOGY
Figure 4 summarizes the most important events in MS. Potentially autoreactive
CD4+ T cells are activated in the periphery by recognizing, for example, a vi-
ral peptide in the context of costimulatory and other less-defined signals (step
1). Factors that contribute to a proinflammatory environment include a number
of cytokines from both T cells and APCs (e.g., IL-12, IFN-γ ), the strength of
activation, and the infectious context (“danger”). Activated autoreactive T cells
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

720 SOSPEDRA  MARTIN


Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org
by University of Missouri - Columbia on 03/18/13. For personal use only.

Figure 4 Schematic diagram depicting the pathogenetic steps and contributing factors that
lead to tissue damage in MS (see text for details).
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 721

adhere to the BBB endothelium via adhesion molecules (LFA-1 and VLA-4), and
transmigrate into the brain parenchyma through cerebrovascular endothelial cells
(step 2). Several mechanisms are still unclear, including what guides autoreactive
CD4+ T cells to the CNS; whether antigen presentation is required in deep cervi-
cal lymph nodes, a putative draining site for brain-derived antigen; and whether
a chemokine gradient from inside the brain parenchyma to the blood exists dur-
ing the initial event. However, experiments in EAE have shown that adoptively
transferred encephalitogenic T cells are transiently found in deep cervical lymph
nodes and then locally reactivated in the CNS, as shown by downmodulation of
their TCR (338). Subsequently, proinflammatory cytokines (IFN-γ , IL-23, TNF-
α, LT) and chemokines (RANTES, IP-10, IL-8, and others) (a) activate resident
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

cells, such as microglia and astrocytes; (b) recruit other immune cells, including
by University of Missouri - Columbia on 03/18/13. For personal use only.

monocytes, CD8+ T cells, B cells, and mast cells, from the peripheral blood; and
(c) orchestrate the formation of the inflammatory lesion (step 3) . The formation of
the inflammatory lesion is characterized by an open BBB with tissue edema after
mediator/protease release from mast cells, monocytes, and T cells, as well as by a
host of proinflammatory molecules and oxygen and nitrogen radicals. Damage of
CNS tissue, i.e., the myelin sheath, oligodendrocytes, and axons, occurs already
at this early inflammatory stage (step 4). During the above steps, CD4+ autoreac-
tive T cells are likely driving the process, whereas their role in the effector phase
is probably secondary. Numerous processes may lead to myelin/oligodendrocyte
and axonal damage, including radicals, TNF-α, LT, and direct complement deposi-
tion, as well as antibody-mediated complement activation and antibody-dependent
cellular cytotoxicity via Fc-receptors, myelin phagocytosis, direct lysis of axons
by CD8+ cytotoxic T lymphocytes, the secretion of proteases, and apoptosis of
oligodendrocytes. Furthermore, the increased production and decreased degrada-
tion or reuptake of the excitatory neurotransmitter glutamate by astrocytes leads
to glutamate-mediated excitotoxicity of oligodendrocytes via glutamate receptor-
mediated calcium influx (12). The inflammatory event lasts from a few days to
two weeks. The “aftermath” is characterized by stretches of demyelinated axons,
apoptotic oligodendrocytes and T cells, axon transsections with onion bulb-like
protrusions owing to interrupted axonal transport (339) (Figure 1), macrophages
loaded with phagocytosed myelin lipids, and the activation and beginning pro-
liferation of astrocytes. Besides clearing debris, lesion resolution (step 5) further
includes a relative dominance of Th2/Th3 cytokines, such as IL-10 and TGF-β, and
the secretion of various growth factors (brain-derived neurotrophic factor, platelet-
derived growth factor, ciliary neurotrophy factor, and fibroblast growth factors) by
both resident cells and T cells. Oligodendrocyte precursors that are still present
in the adult CNS are also activated, and surviving oligodendrocytes begin to re-
myelinate denuded internode areas, although the original thickness of the compact
myelin is not reached again and hence nerve conduction velocity is slower in these
“repaired” areas, despite some compensatory redistribution of sodium channels.
Inhibitory signals between axonal and myelin structures, including Nogo, MAG,
and OMgp, all of which interact with Nogo receptors and are physiologically
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

722 SOSPEDRA  MARTIN

relevant during shaping and maintenance of the intricate cytoarchitecture in the


CNS, impede the repair process (340). “Repaired” myelin differs from mature
adult myelin in its relative composition of myelin protein isoforms and/or post-
translational modifications. The relative abundance of citrullinated C8-MBP is
one example. C8-MBP is less basic than mature MBP owing to modification of six
arginines into citrullines (341). C8-MBP increases during the course of MS and
probably results in functionally impaired myelin, as well as in increased vulnera-
bility to further damage (342). C8-specific T cells have been found in MS patients.
During the months and years following an inflammatory event, the cellular compo-
sition of the plaque changes dramatically. Chronic plaques may show smoldering
inflammation, but they are often devoid of inflammatory cells and characterized
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

by loss of myelin and axons, relative increases in astrocytes (but overall lower
by University of Missouri - Columbia on 03/18/13. For personal use only.

cellularity), and deposition of scar tissue.


The cellular composition and involved molecular pathways vary among patients
(5). Besides the recognition of pathological heterogeneity, other “forgotten” as-
pects, such as axonal injury, have received new attention (339). Investigators have
identified four pathologic MS subtypes on the basis of the relative contribution
of different immune cells, antibody and complement deposition, myelin loss, and
oligodendrocyte death (5). Interestingly, patterns differ interindividually, but not
intraindividually. The following pathologic subtypes are described by Brück and
colleagues (343):
Pattern I. This pattern is predominated by T cells and macrophages, and
candidate effector molecules include TNF-α, IFN-γ , and radical species.
Pattern II. In this pattern, antibody and complement deposition predominate,
and both MOG- and MBP-specific antibodies are involved. The mechanism
of tissue destruction shares similarities with those observed in Guillain Barré
syndrome, an acute inflammatory demyelinating disease of the PNS.
Pattern III. Lesions impress by preferential loss of MAG and oligoden-
drogliopathy, and a vasculitic mechanism is suspected on the basis of parallels
with focal cerebral ischemia. Furthermore, the vulnerability of oligodendro-
cytes may be increased by immune responses against heat shock proteins in
this pattern.
Pattern IV. This pattern, marked by nonapoptotic oligodendrocyte degenera-
tion, is the least common and occurs primarily in PP-MS (343). The overall
extent of inflammation is highest in RR-MS and declines over time, with the
evolution into SP-MS (Figure 1).
Further pathologic observations that deserve mention include the axonal loss,
which was originally described by Charcot (92) and others in the mid-nineteenth
century. Substantial axonal loss may occur during the earliest stages of disease.
It is closely related to neurological disability, and, as mentioned above, several
effector mechanisms, including antiganglioside antibodies and CD8+ CTL, are
probably involved. Finally, a number of findings indicate that contributors to
tissue vulnerability and aberrant repair include the vulnerability of CNS tissue,
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 723

the local dysregulation of apoptosis mechanisms such as higher expression of


bcl-2 on oligodendrocytes during RR-MS, glutamate-mediated excitotoxicity (12,
344), and the reexpression of developmentally important recognition molecules
(Jagged1/Notch) (13).

LESSONS FROM THERAPIES


The concept of MS as an autoimmune inflammatory disease is supported by the
response to immunomodulatory and suppressive treatments. Glucocorticoids, ap-
plied intravenously and at high doses during acute clinical exacerbations of MS,
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

act broadly as anti-inflammatory agents by reducing edema and arachidonic acid


by University of Missouri - Columbia on 03/18/13. For personal use only.

metabolites and by decreasing and modulating IL-1, IL-2, IL-4, IL-5, IL-6, IFN-γ ,
TNF-αβ, fibrin deposition, and other mechanisms.
A number of chemotherapeutic agents with similarly broad activities but long-
term immunosuppressive effects are used at more advanced stages of the disease,
i.e., the transition from RR-MS to SP-MS, or in patients with aggressive disease
who do not respond or who incompletely respond to the approved agents. Im-
munosuppressants include mitoxantrone, cyclophosphamide, methotrexate, aza-
thioprine, cladribin, and mycophenolate. Interestingly, their mechanism of action
in autoimmune diseases is relatively poorly understood; however, we do know that
cyclophosphamide not only has apoptosis-inducing activities but also induces Th2
cells in MS.
IFN-β is approved for treatment of RR-MS and is currently the agent that is most
broadly used. It was originally explored as an antiviral agent, but in recent years
it has been shown that it has immunomodulatory activities. These immunomod-
ulatory activities include the upregulation and increased shedding of adhesion
molecules, induction of IL-10 and neurotrophic factors, blocking of BBB opening
via inhibition of MMP-2 and -9, and reduction of cell adhesion to the BBB. IFN-β
reduces disease exacerbations by only about 30% and has a modest impact on dis-
ease progression. IFN-β is a clear step forward in MS therapy, but the frequency of
subcutaneous injections of IFN-β, the flu-like symptoms that occur at the begin-
ning of therapy, the modest activity required of patients, and the treatment failures
are all reasons to search for better agents.
Glatiramer-acetate (GA, copolymer-1, Cop-1) is another approved therapy for
RR-MS, with similar or slightly lower efficacy than IFN-β at high doses. However,
GA has a more favorable side-effect profile than IFN-β (345). GA is a random
copolymer of the four amino acids Ala, Lys, Glu, and Tyr, with various lengths and
fixed molar ratios of 4.5:3.6:1.5:1 (346). It was originally developed as a mimic of
MBP and to induce EAE (346). Fortuitously, GA blocks the experimental disease
(346). Initially, it was assumed that it acts primarily by displacing autoantigenic
peptides from HLA class II binding grooves, i.e., via competition for binding.
Later, a host of other activities were shown, including polyclonal T cell stimulation,
partial agonist effects, Th2 activation and cross-reactivity with myelin peptides,
shift of the antibody response toward IgG4, interference with DC differentiation,
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

724 SOSPEDRA  MARTIN

and induction of brain-derived neurotrophic factors (347, 348). The most important
effect of GA is most likely the relative skew toward Th2 reactivity. When inflamma-
tory activity is monitored via MRI, IFN-β reduces BBB opening almost immedi-
ately, but it takes much longer until an effect of GA is observed. Currently, attempts
are ongoing to develop better defined and more active peptidic compounds.
Other promising therapeutic strategies include humanized monoclonal anti-
bodies against VLA-4 (natalizumab), which blocks BBB migration of T cells and
their activation and reduces brain inflammation (349), and against the IL-2 re-
ceptor α chain (daclizumab), which activates and expands immunoregulatory NK
cells (280). Daclizumab reduces brain inflammation by almost 80% in patients
with high disease activity who have failed IFN-β treatment (280), but daclizumab
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

is also effective as monotherapy (J. Rose, personal communication). Anti-CD52


by University of Missouri - Columbia on 03/18/13. For personal use only.

leads to long-lasting lymphopenia and also reduces inflammatory activity in MS


(350). Numerous other monoclonal antibodies (e.g., against CD20+ B cells and
CD40L) are either in preclinical or clinical testing. Promising results have also
been observed with estriols (51) and the cholesterin-lowering statins (351, 352).
Further strategies include modulators of cAMP levels, e.g., phosphodiesterase
type 4 inhibitor, pentoxifylline (353), and β-adrenergic agents, inhibitors of chem-
okine receptors (CCR2 antagonists), blocking agents of CD4, retinoic acid, vita-
min A and D derivatives, peroxisome proliferator-activated receptoragonists, DNA
vaccination, and numerous others. Many of these agents have appeared promising
in EAE but later not shown activity in MS. For example, the phosphodiesterase
type 4 inhibitor Rolipram had shown prophylactic and therapeutic activity in EAE,
but in a recent clinical trial in MS showed no disease reduction (B. Bielekova, un-
published observation). Paradoxic effects have been observed with therapies aimed
at TNF or TNF-receptors (see above). Furthermore, in rheumatoid arthritis, where
TNF-blocking agents have become standard therapy, a number of cases developed
acute inflammatory demyelination or MS. Caution must therefore be used both in
extrapolating animal data to humans or information from one autoimmune disease
to another.

REESTABLISHING TOLERANCE
Reestablishing tolerance to autoantigens and specific and subtle therapeutic inter-
ventions remain important goals. The question is whether they can be achieved in
complex and heterogeneous diseases such as MS. Currently, investigators are pur-
suing two main lines. The first is modulation of antigen-specific T cell responses via
induction of anergy or activation-induced cell death. The latter is achieved through
intravenous immunization with either autoantigenic peptides, proteins/fusion pro-
teins, or DNAs that code for these proteins with and without covaccination with
DNAs coding for anti-inflammatory cytokines, or by APL peptides. The second
investigative line is methods to induce anti-idiotypic T cells directed either at TCR
CDR2 or CDR3 regions of autoantigen-specific TCR chains, or vaccination with
whole, inactivated, autoreactive T cells.
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 725

Currently, it is not clear whether either of these approaches can be successful. We


have already mentioned the negative experience with a high dose of APL peptide,
which not only led to hypersensitivity reactions but also to disease exacerbations
(101). Lower doses appeared to induce a Th2 bias, and a currently ongoing trial
with the same APL will hopefully solve whether APL therapy is viable. T cell
vaccination (354) or immunization with CDR2 peptides have been well tolerated
(355). With respect to their mechanisms, the induction of CDR2-specific T cells
that secreted IL-10 has been shown (355). T cell vaccination has led to the reduction
of MBP-specific T cells, apparently via CD8+ cytolytic T cells, and treatment with
anti-Vβ5.2/5.3 led to the reduction of Vβ5.2/5.3-expressing T cells; however, the
effect on reducing MRI activity was below expectations (356). In summary, our
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

incomplete understanding of antigen-specific T cell responses in MS, the disease


by University of Missouri - Columbia on 03/18/13. For personal use only.

heterogeneity, and its complex pathogenesis are among the factors that render
specific immune intervention very challenging.
A more drastic approach toward reestablishing tolerance is hematopoietic stem
cell transplantation (HSCT) via abrogation of the hematopoietic or lymphopoi-
etic system by chemotherapy/irradiation, or optionally via lymphocyte-depleting
steps such as antilymphocyte antibodies and subsequent infusion of autologous
hematopoietic (CD34+ ) stem cells. Although still considered a high-risk proce-
dure, HSCT offers the prospect of stopping the autoimmune process and curing at
least the inflammatory aspects. Recent trials revealed that (a) inflammatory disease
activity is completely halted in the majority of patients (357); (b) progression of
clinical disability continues in patients with advanced disease (358, 359) and there-
fore HSCT probably must be applied earlier when neurological deficit is limited but
the patient clearly has aggressive disease; (c) low-risk protocols have to be explored
and improved, and such studies are currently ongoing; and (d) long-term follow-
up is necessary, and we need to understand the mechanism of action. With respect
to the latter point, HSCT indeed leads to rejuvenation of the immune system, with
increased recent thymic emigrants, reactivation of the thymus, a net increase of
naive CD4+ T cells, and the reestablishment of a more diverse TCR repertoire.
Transiently after HSCT, there is also increased apoptosis of T cells and a relative in-
crease of CD4+ CD25+ regulatory T cells (P.A. Muraro, unpublished observation).
New approaches toward specific immune intervention clearly need to be defined,
but recent progress in immunomodulation has been very promising. We believe
that future therapies toward tissue repair and neuroprotection are only meaningful
if the inflammatory components of MS can be contained or completely stopped.

CONCLUDING REMARKS
Exciting progress has been made in understanding MS pathogenesis in the past
decade. Every aspect has become more complicated, and our previous concept
of MS as “simply” a CD4+ Th1 cell–mediated autoimmune disease must be re-
visited. We now recognize that the complex genetic background in concert with
environmental triggers—most likely common viral infections, but mitigated by
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

726 SOSPEDRA  MARTIN

many other factors—is responsible for the heterogeneity of every aspect of the
disease, including pathologic mechanisms, clinical and MRI presentation, and
response to treatments. Many components of the innate and adaptive immune sys-
tems, and in the latter CD4+ T cells, CD8+ T cells, and antibodies, all contribute
to different aspects of the disease process. In addition, factors other than the au-
toimmune response clearly shape the disease. The vulnerability of the CNS to
inflammatory insult and/or its inability to repair tissue is equally heterogeneous
among different patients, and we will only understand the full scope of disease
etiology and pathogenesis if we consider both immune system and nervous system
and their mutual interactions in MS. The latter point is particularly relevant when
we try to block the disease process or even repair already inflicted damage and
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

prevent it in the future through single agent, cell-based, or combination therapies.


by University of Missouri - Columbia on 03/18/13. For personal use only.

These treatments will likely become much more complex than those currently
applied.

APPENDIX
Abbreviations used: αB-C, α-B crystallin; ADCC, antibody-dependent cellular
cytotoxicity; APC, antigen-presenting cell; APL, altered peptide ligand; APOE,
apolipoprotein E; BBB, blood-brain barrier; CNPase, 2 ,3 -cyclic nucleotide 3
phosphodiesterase; CNS, central nervous system; Cpn, Chlamydia pneumoniae;
CTLA, cytotoxic T lymphocyte–associated antigen; DC, dendritic cell; EAE,
experimental allergic encephalomyelitis; EBV, Epstein-Barr virus; ELISPOT,
enzyme-linked immunospot; GA, glatiramer-acetate; G-CSF, granulocyte colony-
stimulating factor; GFAP, glial fibrillary acidic protein; HHV, human herpesvirus;
HLA, histocompatibility leukocyte antigen; HSCT, hematopoietic stem cell trans-
plantation; HSV, herpes simplex virus; IAP, inhibitor of apoptosis; IFN, inter-
feron; IP, IFN-γ -inducible protein; LT, lymphotoxin; MAC, membrane attack
complex; MAG, myelin-associated glycoprotein; MBP, myelin basic protein; MCP,
monocyte chemoattractant protein; MHC, major histocompatibility complex; MRI,
magnetic resonance imaging; MMP, matrix metalloproteinase; MOBP, myelin-
associated oligodendrocytic basic protein; MOG, myelin oligodendrocyte glyco-
protein; MS, multiple sclerosis; NO, nitric oxide; NOS, nitric oxide synthase;
OCB, oligoclonal bands; OSP, oligodendrocyte-specific glycoprotein; PBMC, pe-
ripheral blood mononuclear cells; PLP, proteolipid protein; PNS, peripheral ner-
vous system; PP-MS, primary progressive MS; RR-MS, relapsing-remitting MS;
SP-MS, secondary progressive MS; TCC, T cell clones; TCR, T cell receptor;
TGF, transforming growth factor, TLR, Toll-like receptor; TNF, tumor necrosis
factor; TRAIL, TNF-related apoptosis-inducing ligand; Treg, immunoregulatory
T cell; VZV, varicella zoster virus.

ACKNOWLEDGMENTS
We realize that many studies could not be considered, and we apologize to the
authors of these works. Our summary views try, however, to take them into account.
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 727

I (R.M.) thank Hans-Wolfgang Kreth, MD, Department of Pediatrics, University


of Würzburg, Germany, my first teacher in immunology, and Henry F. McFarland,
MD, Neuroimmunology Branch, NINDS, NIH, Bethesda, my long-term mentor,
for his advice and support. Further, we acknowledge the work of our coworkers
and prior investigators at the Cellular Immunology Section and Neuroimmunology
Branch, NINDS, NIH, whose research during recent years has contributed to this
review.

The Annual Review of Immunology is online at


http://immunol.annualreviews.org
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

LITERATURE CITED
by University of Missouri - Columbia on 03/18/13. For personal use only.

1. McFarlin DE, McFarland HF. 1982. Mul- myelin basic protein: requirement for Lyt
tiple sclerosis (first of two parts). N. Engl. 1+ 2-T lymphocytes. J. Immunol. 127:
J. Med. 307:1183–88 1420–23
2. McFarlin DE, McFarland HF. 1982. Mul- 10. Huseby ES, Liggitt D, Brabb T, Schna-
tiple sclerosis (second of two parts). N. bel B, Ohlen C, Goverman J. 2001. A
Engl. J. Med. 307:1246–51 pathogenic role for myelin-specific CD8+
3. McFarland HF. 1999. Correlation be- T cells in a model for multiple sclerosis.
tween MR and clinical findings of disease J. Exp. Med. 194:669–76
activity in multiple sclerosis. Am. J. Neu- 11. Sun DM, Whitaker JN, Huang ZG, Liu
roradiol. 20:1777–78 D, Coleclough C, et al. 2001. Myelin
4. Raine CS, Scheinberg LC. 1988. On the antigen-specific CD8+ T cells are en-
immunopathology of plaque development cephalitogenic and produce severe disease
and repair in multiple sclerosis. J. Neu- in C57BL/6 mice. J. Immunol. 166:7579–
roimmunol. 20:189–201 87
5. Lucchinetti C, Bruck W, Parisi J, Schei- 12. Pitt D, Werner P, Raine CS. 2000. Gluta-
thauer B, Rodriguez M, Lassmann H. mate excitotoxicity in a model of multiple
2000. Heterogeneity of multiple sclerosis sclerosis. Nat. Med. 6:67–70
lesions: implications for the pathogenesis 13. John GR, Shankar SL, Shafit-Zagardo B,
of demyelination. Ann. Neurol. 47:707– Massimi A, Lee SC, et al. 2002. Multiple
17 sclerosis: re-expression of a developmen-
6. Martin R, McFarland HF, McFarlin DE. tal pathway that restricts oligodendrocyte
1992. Immunological aspects of demyeli- maturation. Nat. Med. 8:1115–21
nating diseases. Annu. Rev. Immunol. 10: 14. Dyment DA, Ebers GC, Sadovnick AD.
153–87 2004. Genetics of multiple sclerosis.
7. Hafler DA. 2004. Multiple sclerosis. J. Lancet Neurol. 3:104–10
Clin. Invest. 113:788–94 15. Haines JL, Bradford Y, Garcia ME, Reed
8. Zamvil SS, Steinman L. 1990. The T AD, Neumeister E, et al. 2002. Multiple
lymphocyte in experimental allergic en- susceptibility loci for multiple sclerosis.
cephalomyelitis. Annu. Rev. Immunol. 8: Hum. Mol. Genet. 11:2251–56
579–621 16. Ebers GC, Sadovnick AD, Risch NJ.
9. Pettinelli CB, McFarlin DE. 1981. Adop- 1995. A genetic basis for familial aggrega-
tive transfer of experimental allergic en- tion in multiple sclerosis. Canadian Col-
cephalomyelitis in SJL/J mice after in laborative Study Group. Nature 377:150–
vitro activation of lymph node cells by 51
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

728 SOSPEDRA  MARTIN

17. GAMES, Transatlantic Multiple Sclero- viduals specify the same inferred amino
sis Genetics Cooperative. 2003. A meta- acid sequence as the DRβ1 and DRβ2
analysis of whole genome linkage screens genes of a DR4 (Dw14) haplotype. Arthri-
in multiple sclerosis. J. Neuroimmunol. tis Rheum. 32:251–58
143:39–46 27. Ristori G, Carcassi C, Lai S, Fiori P,
18. Hillert J, Olerup O. 1993. HLA and MS. Cacciani A, et al. 1997. HLA-DM poly-
Neurology 43:2426–27 morphisms do not associate with multiple
19. Haines JL, Terwedow HA, Burgess K, sclerosis: an association study with anal-
Pericak-Vance MA, Rimmler JB, et al. ysis of myelin basic protein T cell speci-
1998. Linkage of the MHC to familial ficity. J. Neuroimmunol. 77:181–84
multiple sclerosis suggests genetic hetero- 28. Matsuoka T, Tabata H, Matsushita S.
geneity. Hum. Mol. Genet. 7:1229–34 2001. Monocytes are differentially acti-
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

20. Barcellos LF, Oksenberg JR, Begovich vated through HLA-DR, -DQ, and -DP
by University of Missouri - Columbia on 03/18/13. For personal use only.

AB, Martin ER, Schmidt S, et al. 2003. molecules via mitogen-activated protein
HLA-DR2 dose effect on susceptibility to kinases. J. Immunol. 166:2202–8
multiple sclerosis and influence on disease 29. Li F, Linan MJ, Stein MC, Faustman
course. Am. J. Hum. Genet. 72:710–16 DL. 1995. Reduced expression of peptide-
21. Fogdell-Hahn A, Ligers A, Gronning M, loaded HLA class I molecules on multiple
Hillert J, Olerup O. 2000. Multiple scle- sclerosis lymphocytes. Ann. Neurol. 38:
rosis: a modifying influence of HLA class 147–54
I genes in an HLA class II associated 30. Babbe H, Roers A, Waisman A, Lassmann
autoimmune disease. Tissue Antigens 55: H, Goebels N, et al. 2000. Clonal expan-
140–48 sions of CD8+ T cells dominate the T cell
22. Fernandez O, Fernandez V, Alonso A, infiltrate in active multiple sclerosis le-
Caballero A, Luque G, et al. 2004. sions as shown by micromanipulation and
DQB1∗ 0602 allele shows a strong asso- single cell polymerase chain reaction. J.
ciation with multiple sclerosis in patients Exp. Med. 192:393–404
in Malaga, Spain. J. Neurol. 251:440– 31. Jacobsen M, Cepok S, Quak E, Happel
44 M, Gaber R, et al. 2002. Oligoclonal ex-
23. Bertram J, Kuwert E. 1982. HLA antigen pansion of memory CD8+ T cells in cere-
frequencies in multiple sclerosis. Eur. J. brospinal fluid from multiple sclerosis pa-
Neurol. 7:74–79 tients. Brain 125:538–50
24. Sellebjerg F, Jensen J, Madsen HO, Svej- 32. Head JR, Griffin WST. 1985. Functional
gaard A. 2000. HLA DRB1∗ 1501 and in- capacity of solid tissue transplants in
trathecal inflammation in multiple sclero- brain: evidence for immonological priv-
sis. Tissue Antigens 55:312–18 ilege. Proc. R. Soc. London Ser. B 224:
25. Ridgway WM, Ito H, Fasso M, Yu C, 375–87
Fathman CG. 1998. Analysis of the role 33. DeRisi J, Penland L, Brown PO, Bittner
of variation of major histocompatibility ML, Meltzer PS, et al. 1996. Use of a
complex class II expression on nonobese cDNA microarray to analyse gene expres-
diabetic (NOD) peripheral T cell re- sion patterns in human cancer. Nat. Genet.
sponse. J. Exp. Med. 188:2267–75 14:457–60
26. Merryman PF, Crapper RM, Lee S, 34. Whitney LW, Becker KG, Tresser NJ,
Gregersen PK, Winchester RJ. 1989. Caballero-Ramos CI, Munson PJ, et al.
Class II major histocompatibility complex 1999. Analysis of gene expression in mu-
gene sequences in rheumatoid arthritis. tiple sclerosis lesions using cDNA mi-
The third diversity regions of both DRβ1 croarrays. Ann. Neurol. 46:425–28
genes in two DR1, DRw10-positive indi- 35. Whitney LW, Ludwin SK, McFarland HF,
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 729

Biddison WE. 2001. Microarray analy- tients and healthy controls: identifying
sis of gene expression in multiple scle- pathways relevant to disease. Hum. Mol.
rosis and EAE identifies 5-lipoxygenase Genet. 12:2191–99
as a component of inflammatory lesions. 44. Wandinger KP, Sturzebecher CS,
J. Neuroimmunol. 121:40–48 Bielekova B, Detore G, Rosenwald A,
36. Lock C, Hermans G, Pedotti R, Bren- et al. 2001. Complex immunomodulatory
dolan A, Schadt E, et al. 2002. Gene- effects of interferon-β in multiple sclero-
microarray analysis of multiple sclerosis sis include the upregulation of T helper 1-
lesions yields new targets validated in au- associated marker genes. Ann. Neurol. 50:
toimmune encephalomyelitis. Nat. Med. 349–57
8:500–8 45. Sturzebecher S, Wandinger KP, Rosen-
37. Tajouri L, Mellick AS, Ashton KJ, Tan- wald A, Sathyamoorthy M, Tzou A, et al.
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

nenberg AE, Nagra RM, et al. 2003. Quan- 2003. Expression profiling identifies re-
by University of Missouri - Columbia on 03/18/13. For personal use only.

titative and qualitative changes in gene sponder and non-responder phenotypes to


expression patterns characterize the activ- interferon-β in multiple sclerosis. Brain
ity of plaques in multiple sclerosis. Brain 126:1419–29
Res. Mol. Brain Res. 119:170–83 46. Wandinger KP, Lunemann JD, Wengert
38. Mycko MP, Papoian R, Boschert U, Raine O, Bellmann-Strobl J, Aktas O, et al.
CS, Selmaj KW. 2004. Microarray gene 2003. TNF-related apoptosis inducing lig-
expression profiling of chronic active and and (TRAIL) as a potential response
inactive lesions in multiple sclerosis. Clin. marker for interferon-β treatment in mul-
Neurol. Neurosurg. 106:223–29 tiple sclerosis. Lancet 361:2036–43
39. Baranzini SE, Elfstrom C, Chang SY, Bu- 47. McFarland HF. 1992. Twin studies and
tunoi C, Murray R, et al. 2000. Transcrip- multiple sclerosis. Ann. Neurol. 32:722–
tional analysis of multiple sclerosis brain 23
lesions reveals a complex pattern of cy- 48. Wakeland EK, Liu K, Graham RR,
tokine expression. J. Immunol. 165:6576– Behrens TW. 2001. Delineating the ge-
82 netic basis of systemic lupus erythemato-
40. Chabas D, Baranzini SE, Mitchell D, sus. Immunity 15:397–408
Bernard CC, Rittling SR, et al. 2001. 49. Coo H, Aronson KJ. 2004. A systematic
The influence of the proinflammatory cy- review of several potential non-genetic
tokine, osteopontin, on autoimmune de- risk factors for multiple sclerosis. Neu-
myelinating disease. Science 294:1731– roepidemiology 23:1–12
35 50. Runmarker B, Andersen O. 1995. Preg-
41. Graumann U, Reynolds R, Steck AJ, nancy is associated with a lower risk
Schaeren-Wiemers N. 2003. Molecular of onset and a better prognosis in mul-
changes in normal appearing white matter tiple sclerosis. Brain 118(Part 1):253–
in multiple sclerosis are characteristic of 61
neuroprotective mechanisms against hy- 51. Sicotte NL, Liva SM, Klutch R, Pfeiffer P,
poxic insult. Brain Pathol. 13:554–73 Bouvier S, et al. 2002. Treatment of multi-
42. Achiron A, Gurevich M, Friedman N, ple sclerosis with the pregnancy hormone
Kaminski N, Mandel M. 2004. Blood tran- estriol. Ann. Neurol. 52:421–28
scriptional signatures of multiple sclero- 52. Kurtzke JF. 1983. Epidemiology of multi-
sis: unique gene expression of disease ac- ple sclerosis. In Multiple Sclerosis, ed. JF
tivity. Ann. Neurol. 55:410–17 Hallpike, CWM Adams, WW Tourtelotte,
43. Bomprezzi R, Ringner M, Kim S, Bit- pp. 49–95. Baltimore, MD: Williams &
tner ML, Khan J, et al. 2003. Gene ex- Wilkins
pression profile in multiple sclerosis pa- 53. Hayes CE. 2000. Vitamin D: a natural
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

730 SOSPEDRA  MARTIN

inhibitor of multiple sclerosis. Proc. Nutr. tem in multiple sclerosis. Ann. Neurol.
Soc. 59:531–35 46:6–14
54. Johnson RT. 1994. The virology of 65. Pucci E, Taus C, Cartechini E, Morelli M,
demyelinating diseases. Ann. Neurol. Giuliani G, et al. 2000. Lack of Chlamydia
36(Suppl.):S54–60 infection of the central nervous system in
55. Soldan SS, Jacobson S. 2001. Role of multiple sclerosis. Ann. Neurol. 48:399–
viruses in etiology and pathogenesis of 400
multiple sclerosis. Adv. Virus Res. 56: 66. Burns J, Rosenzweig A, Zweiman B,
517–55 Lisak RP. 1983. Isolation of myelin ba-
56. Sibley WA, Bamford CR, Clark K. 1985. sic protein-reactive T-cell lines from nor-
Clinical viral infections and multiple scle- mal human blood. Cell Immunol. 81:435–
rosis. Lancet 1:1313–15 40
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

57. Goverman J, Woods A, Larson L, Weiner 67. Richert JR, McFarlin DE, Rose JW, Mc-
by University of Missouri - Columbia on 03/18/13. For personal use only.

LP, Hood L, Zaller DM. 1993. Trans- Farland HF, Greenstein JI. 1983. Expan-
genic mice that express a myelin basic sion of antigen-specific T cells from cere-
protein-specific T cell receptor develop brospinal fluid of patients with multiple
spontaneous autoimmunity. Cell 72:551– sclerosis. J. Neuroimmunol. 5:317–24
60 68. Chou YK, Vainiene M, Whitham R, Bour-
58. Challoner PB, Smith KT, Parker JD, dette D, Chou CH, et al. 1989. Response of
MacLeod DL, Coulter SN, et al. 1995. human T lymphocyte lines to myelin basic
Plaque-associated expression of human protein: association of dominant epitopes
herpesvirus 6 in multiple sclerosis. Proc. with HLA class II restriction molecules.
Natl. Acad. Sci. USA 92:7440–44 J. Neurosci. Res. 23:207–16
59. Moore FG, Wolfson C. 2002. Human her- 69. Martin R, Jaraquemada D, Flerlage M,
pes virus 6 and multiple sclerosis. Acta Richert J, Whitaker J, et al. 1990. Fine
Neurol. Scand. 106:63–83 specificity and HLA restriction of myelin
60. Soldan SS, Leist TP, Juhng KN, Mc- basic protein-specific cytotoxic T cell
Farland HF, Jacobson S. 2000. Increased lines from multiple sclerosis patients and
lymphoproliferative response to human healthy individuals. J. Immunol. 145:540–
herpesvirus type 6A variant in multiple 48
sclerosis patients. Ann. Neurol. 47:306– 70. Pette M, Fujita K, Kitze B, Whitaker JN,
13 Albert E, et al. 1990. Myelin basic protein-
61. Wandinger KP, Jabs W, Siekhaus A, specific T lymphocyte lines from MS pa-
Bubel S, Trillenberg P, et al. 2000. As- tients and healthy individuals. Neurology
sociation between clinical disease activ- 40:1770–76
ity and Epstein-Barr virus reactivation in 71. Ota K, Matsui M, Milford EL, Mackin
MS. Neurology 55:178–84 GA, Weiner HL, Hafler DA. 1990. T-
62. Martyn CN, Cruddas M, Compston DA. cell recognition of an immunodominant
1993. Symptomatic Epstein-Barr virus in- myelin basic protein epitope in multiple
fection and multiple sclerosis. J. Neurol. sclerosis. Nature 346:183–87
Neurosurg. Psychiatry 56:167–68 72. Olsson T, Zhi WW, Hojeberg B, Kostu-
63. Sriram S, Mitchell W, Stratton C. las V, Jiang YP, et al. 1990. Autoreactive
1998. Multiple sclerosis associated with T lymphocytes in multiple sclerosis de-
Chlamydia pneumoniae infection of the termined by antigen-induced secretion of
CNS. Neurology 50:571–72 interferon-γ . J. Clin. Invest. 86:981–85
64. Sriram S, Stratton CW, Yao S, Tharp A, 73. Fujinami RS, Oldstone MB. 1985. Amino
Ding L, et al. 1999. Chlamydia pneumo- acid homology between the encephalito-
niae infection of the central nervous sys- genic site of myelin basic protein and
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 731

virus: mechanism for autoimmunity. Sci- 83. Tough DF, Sprent J. 1996. Viruses and
ence 230:1043–45 T cell turnover: evidence for bystander
74. Tejada-Simon MV, Zang YCQ, Hong J, proliferation. Immunol. Rev. 150:129–
Rivera VM, Zhang JWZ. 2003. Cross- 42
reactivity with myelin basic protein and 84. Murali-Krishna K, Altman JD, Suresh M,
human herpesvirus-6 in multiple sclero- Sourdive DJ, Zajac AJ, et al. 1998. Count-
sis. Ann. Neurol. 53:189–97 ing antigen-specific CD8 T cells: a reeval-
75. Evavold BD, Sloan-Lancaster J, Hsu BL, uation of bystander activation during viral
Allen PM. 1993. Separation of T helper infection. Immunity 8:177–87
1 clone cytolysis from proliferation and 85. Brocke S, Gaur A, Piercy C, Gautam A,
lymphokine production using analog pep- Gijbels K, et al. 1993. Induction of relaps-
tides. J. Immunol. 150:3131–40 ing paralysis in experimental autoimmune
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

76. Wucherpfennig KW, Strominger JL. encephalomyelitis by bacterial superanti-


by University of Missouri - Columbia on 03/18/13. For personal use only.

1995. Molecular mimicry in T cell- gen. Nature 365:642–44


mediated autoimmunity: viral peptides 86. Waldner H, Collins M, Kuchroo VK.
activate human T cell clones specific for 2004. Activation of antigen-presenting
myelin basic protein. Cell 80:695–705 cells by microbial products breaks self tol-
77. Hemmer B, Vergelli M, Gran B, Ling erance and induces autoimmune disease.
N, Conlon P, et al. 1998. Predictable J. Clin. Invest. 113:990–97
TCR antigen recognition based on pep- 87. Miller SD, Vanderlugt CL, Begolka WS,
tide scans leads to the identification of ag- Pao W, Yauch RL, et al. 1997. Persis-
onist ligands with no sequence homology. tent infection with Theiler’s virus leads to
J. Immunol. 160:3631–36 CNS autoimmunity via epitope spreading.
78. Hemmer B, Pinilla C, Gran B, Vergelli M, Nat. Med. 3:1133–36
Ling N, et al. 2000. Contribution of indi- 88. Horwitz MS, Bradley LM, Harbertson J,
vidual amino acids within MHC molecule Krahl T, Lee J, Sarvetnick N. 1998. Dia-
or antigenic peptide to TCR ligand po- betes induced by Coxsackie virus: initia-
tency. J. Immunol. 164:861–71 tion by bystander damage and not molec-
79. Lang HL, Jacobsen H, Ikemizu S, Ander- ular mimicry. Nat. Med. 4:781–85
sson C, Harlos K, et al. 2002. A func- 89. Barnaba V. 1996. Viruses, hidden self-
tional and structural basis for TCR cross- epitopes and autoimmunity. Immunol.
reactivity in multiple sclerosis. Nat. Im- Rev. 152:47–66
munol. 3:940–43 90. Opdenakker G, Van Damme J. 1994.
80. Ohashi PS, Oehen S, Buerki K, Pircher H, Cytokine-regulated proteases in autoim-
Ohashi CT, et al. 1991. Ablation of “tol- mune diseases. Immunol. Today 15:103–7
erance” and induction of diabetes by virus 91. Pender MP. 2003. Infection of autoreac-
infection in viral antigen transgenic mice. tive B lymphocytes with EBV, causing
Cell 65:305–17 chronic autoimmune diseases. Trends Im-
81. Oldstone MB, Nerenberg M, Southern P, munol. 24:584–88
Price J, Lewicki H. 1991. Virus infection 92. Charcot J. 1868. Histologie de la sclerose
triggers insulin-dependent diabetes melli- en plaque. Gaz. Hopitaux 41:554–66
tus in a transgenic model: role of anti-self 93. Remlinger J. 1905. Accidents paraly-
(virus) immune response. Cell 65:319–31 tiques au cours du traitment antirabique.
82. Olson JK, Croxford JL, Miller SD. 2004. Ann. Inst. Pasteur 19:625–46
Innate and adaptive immune requirements 94. Rivers TM, Sprunt DH, Berry GP. 1933.
for induction of autoimmune demyelinat- Observations on attempts to produce acute
ing disease by molecular mimicry. Mol. disseminated encephalomyelitis in mon-
Immunol. 40:1103–8 keys. J. Exp. Med. 58:39–53
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

732 SOSPEDRA  MARTIN

95. Ben-Nun A, Cohen IR. 1982. Experimen- the central and peripheral nervous system.
tal autoimmune encephalomyelitis (EAE) Lab. Invest. 76:355–64
mediated by T cell lines: process of selec- 103. Wekerle H, Kojima K, Lannes-Vieira J,
tion of lines and characterization of the Lassmann H, Linington C. 1994. Animal
cells. J. Immunol. 129:303–8 models. Ann. Neurol. 36:S47–53
96. Das P, Drescher KM, Geluk A, Bradley 104. Schluesener HJ, Wekerle H. 1985. Au-
DS, Rodriguez M, David CS. 2000. Com- toaggressive T lymphocyte lines recog-
plementation between specific HLA-DR nizing the encephalitogenic region of
and HLA-DQ genes in transgenic mice myelin basic protein: in vitro selection
determines susceptibility to experimen- from unprimed rat T lymphocyte popu-
tal autoimmune encephalomyelitis. Hum. lations. J. Immunol. 135:3128–33
Immunol. 61:279–89 105. Martin R, Voskuhl R, Flerlage M, Mc-
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

97. Kawamura K, Yamamura T, Yokoyama K, Farlin DE, McFarland HF. 1993. Myelin
by University of Missouri - Columbia on 03/18/13. For personal use only.

Chui DH, Fukui Y, et al. 2000. HLA-DR2- basic protein-specific T-cell responses
restricted responses to proteolipid protein in identical twins discordant or concor-
95–116 peptide cause autoimmune en- dant for multiple sclerosis. Ann. Neurol.
cephalitis in transgenic mice. J. Clin. In- 34:524–35
vest. 105:977–84 106. Bieganowska KD, Ausubel LJ, Modabber
98. Forsthuber TG, Shive CL, Wienhold W, Y, Slovik E, Messersmith W, Hafler DA.
deGraaf K, Spack EG, et al. 2001. T cell 1997. Direct ex vivo analysis of activated,
epitopes of human myelin oligodendro- Fas-sensitive autoreactive T cells in hu-
cyte glycoprotein identified in HLA-DR4 man autoimmune disease. J. Exp. Med.
(DRB1∗ 0401) transgenic mice are en- 185:1585–94
cephalitogenic and are presented by hu- 107. Muraro PA, Wandinger KP, Bielekova B,
man B cells. J. Immunol. 167:7119–25 Gran B, Marques A, et al. 2003. Molecular
99. Madsen LS, Andersson EC, Jansson L, tracking of antigen-specific T cell clones
Krogsgaard M, Andersen CB, et al. 1999. in neurological immune-mediated disor-
A humanized model for multiple sclero- ders. Brain 126:20–31
sis using HLA-DR2 and a human T-cell 108. Bielekova B, Sung MH, Kadom N, Si-
receptor. Nat. Genet. 23:343–47 mon R, McFarland H, Martin R. 2004. Ex-
100. Quandt JA, Baig M, Yao K, Kawamura pansion and functional relevance of high-
K, Huh J, et al. 2004. Unique clini- avidity myelin-specific CD4+ T cells in
cal and pathological features in HLA- multiple sclerosis. J. Immunol. 172:3893–
DRB1∗ 0401-restricted MBP 111–129– 904
specific humanized transgenic mice. J. 109. Hong J, Zang YCQ, Li SF, Rivera VM,
Exp. Med. 200:223–34 Zhang JWZ. 2004. Ex vivo detection of
101. Bielekova B, Goodwin B, Richert N, myelin basic protein-reactive T cells in
Cortese I, Kondo T, et al. 2000. Encephal- multiple sclerosis and controls using spe-
itogenic potential of the myelin basic pro- cific TCR oligonucleotide probes. Eur. J.
tein peptide (amino acids 83–99) in multi- Immunol. 34:870–81
ple sclerosis: results of a phase II clinical 110. Crawford MP, Yan SX, Ortega SB,
trial with an altered peptide ligand. Nat. Mehta RS, Hewitt RE, et al. 2004. High
Med. 6:1167–75 prevalence of autoreactive, neuroantigen-
102. Berger T, Weerth S, Kojima K, Linington specific CD8+ T cells in multiple sclerosis
C, Wekerle H, Lassmann H. 1997. Exper- revealed by novel flow cytometric assay.
imental autoimmune encephalomyelitis: Blood 103:4222–31
the antigen specificity of T lymphocytes 111. Olsson T, Sun J, Hillert J, Hojeberg B,
determines the topography of lesions in Ekre HP, et al. 1992. Increased numbers
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 733

of T cells recognizing multiple myelin ba- nition by human T cell clones. J. Exp.
sic protein epitopes in multiple sclerosis. Med. 179:279–90
Eur. J. Immunol. 22:1083–87 120. Vogt AB, Kropshofer H, Kalbacher H,
112. Allegretta M, Nicklas JA, Sriram S, Al- Kalbus M, Rammensee HG, et al. 1994.
bertini RJ. 1990. T cells responsive to Ligand motifs of HLA-DRB5∗ 0101 and
myelin basic protein in patients with mul- DRB1∗ 1501 molecules delineated from
tiple sclerosis. Science 247:718–21 self-peptides. J. Immunol. 153:1665–73
113. Trotter JL, Damico CA, Cross AH, Pel- 121. Muraro PA, Vergelli M, Kalbus M, Banks
frey CM, Karr RW, et al. 1997. HPRT DE, Nagle JW, et al. 1997. Immunodom-
mutant T-cell lines from multiple sclerosis inance of a low-affinity major histocom-
patients recognize myelin proteolipid pro- patibility complex-binding myelin basic
tein peptides. J. Neuroimmunol. 75:95– protein epitope (residues 111–129) in
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

103 HLA-DR4 (B1∗ 0401) subjects is associ-


by University of Missouri - Columbia on 03/18/13. For personal use only.

114. Seamons A, Perchellet A, Goverman J. ated with a restricted T cell receptor reper-
2003. Immune tolerance to myelin pro- toire. J. Clin. Invest. 100:339–49
teins. Immunol. Res. 28:201–21 122. Vergelli M, Kalbus M, Rojo SC, Hem-
115. Bruno R, Sabater L, Sospedra M, Ferrer- mer B, Kalbacher H, et al. 1997. T cell
Francesch X, Escudero D, et al. 2002. response to myelin basic protein in
Multiple sclerosis candidate autoantigens the context of the multiple sclerosis-
except myelin oligodendrocyte glycopro- associated HLA-DR15 haplotype: peptide
tein are transcribed in human thymus. Eur. binding, immunodominance and effector
J. Immunol. 32:2737–47 functions of T cells. J. Neuroimmunol. 77:
116. Pette M, Fujita K, Wilkinson D, Altmann 195–203
DM, Trowsdale J, et al. 1990. Myelin au- 123. Richert JR, Robinson ED, Deibler GE,
toreactivity in multiple sclerosis: recog- Martenson RE, Dragovic LJ, Kies MW.
nition of myelin basic protein in the con- 1989. Human cytotoxic T-cell recognition
text of HLA-DR2 products by T lympho- of a synthetic peptide of myelin basic pro-
cytes of multiple sclerosis patients and tein. Ann. Neurol. 26:342–46
healthy donors. Proc. Natl. Acad. Sci. 124. Wraith DC, Smilek DE, Mitchell DJ,
USA 87:7968–72 Steinman L, McDevitt HO. 1989. Anti-
117. Martin R, Howell MD, Jaraquemada D, gen recognition in autoimmune en-
Flerlage M, Richert J, et al. 1991. A cephalomyelitis and the potential for
myelin basic protein peptide is recognized peptide-mediated immunotherapy. Cell
by cytotoxic T cells in the context of four 59:247–55
HLA-DR types associated with multiple 125. Fairchild PJ, Wraith DC. 1992. Peptide-
sclerosis. J. Exp. Med. 173:19–24 MHC interaction in autoimmunity. Curr.
118. Valli A, Sette A, Kappos L, Oseroff C, Opin. Immunol. 4:748–53
Sidney J, et al. 1993. Binding of myelin 126. Lafaille JJ, Nagashima K, Katsuki M,
basic protein peptides to human histo- Tonegawa S. 1994. High incidence
compatibility leukocyte antigen class II of spontaneous autoimmune encephalo-
molecules and their recognition by T cells myelitis in immunodeficient anti-myelin
from multiple sclerosis patients. J. Clin. basic protein T cell receptor transgenic
Invest. 91:616–28 mice. Cell 78:399–408
119. Wucherpfennig KW, Sette A, Southwood 127. Krogsgaard M, Wucherpfennig KW, Can-
S, Oseroff C, Matsui M, et al. 1994. nella B, Hansen BE, Svejgaard A, et al.
Structural requirements for binding of 2000. Visualization of myelin basic pro-
an immunodominant myelin basic protein tein (MBP) T cell epitopes in multi-
peptide to DR2 isotypes and for its recog- ple sclerosis lesions using a monoclonal
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

734 SOSPEDRA  MARTIN

antibody specific for the human histocom- ond T cell epitope of human proteolipid
patibility leukocyte antigen (HLA)-DR2- protein (residues 89–106) recognized by
MBP 85–99 complex. J. Exp. Med. 191: proliferative and cytolytic CD4+ T cells
1395–412 from multiple sclerosis patients. J. Neu-
128. Klein L, Klugmann M, Nave K-A, Tuohy roimmunol. 53:153–61
VK, Kyewski B. 2000. Shaping of the 136. Kondo T, Yamamura T, Inobe J, Ohashi T,
autoreactive T-cell repertoire by a splice Takahashi K, Tabira T. 1996. TCR reper-
variant of self protein expressed in thymic toire to proteolipid protein (PLP) in mul-
epithelial cells. Nat. Med. 6:56–62 tiple sclerosis (MS): homologies between
129. Waldner H, Whitters MJ, Sobel RA, PLP-specific T cells and MS-associated
Collins M, Kuchroo VK. 2000. Fulmi- T cells in TCR junctional sequences. Int.
nant spontaneous autoimmunity of the Immunol. 8:123–30
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

central nervous system in mice transgenic 137. Mendel I, Kerlero de Rosbo N, Ben-
by University of Missouri - Columbia on 03/18/13. For personal use only.

for the myelin proteolipid protein-specific Nun A. 1995. A myelin oligoden-


T cell receptor. Proc. Natl. Acad. Sci. USA drocyte glycoprotein peptide induces
97:3412–17 typical chronic experimental autoimmune
130. Kennedy MK, Tan LJ, Dal Canto MC, encephalomyelitis in H-2b mice: fine
Tuohy VK, Lu ZJ, et al. 1990. Inhibition specificity and T cell receptor Vβ expres-
of murine relapsing experimental autoim- sion of encephalitogenic T cells. Eur. J.
mune encephalomyelitis by immune tol- Immunol. 25:1951–59
erance to proteolipid protein and its en- 138. Bettelli E, Pagany M, Weiner HL, Lin-
cephalitogenic peptides. J. Immunol. 144: ington C, Sobel RA, Kuchroo VK. 2003.
909–15 Myelin oligodendrocyte glycoprotein-
131. Vanderlugt CL, Miller SD. 2002. Epi- specific T cell receptor transgenic mice
tope spreading in immune-mediated dis- develop spontaneous autoimmune optic
eases: implications for immunotherapy. neuritis. J. Exp. Med. 197:1073–81
Nat. Rev. Immunol. 2:85–95 139. Tsunoda I, Kuang LQ, Theil DJ, Fuji-
132. Correale J, McMillan M, McCarthy nami RS. 2000. Antibody association with
K, Le T, Weiner LP. 1995. Isolation a novel model for primary progressive
and characterization of autoreactive pro- multiple sclerosis: induction of relapsing-
teolipid protein-peptide specific T-cell remitting and progressive forms of EAE in
clones from multiple sclerosis patients. H2s mouse strains. Brain Pathol. 10:402–
Neurology 45:1370–78 18
133. Greer JM, Csurhes PA, Cameron KD, 140. Kerlero de Rosbo N, Hoffman M, Mendel
McCombe PA, Good MF, Pender MP. I, Yust I, Kaye J, et al. 1997. Predominance
1997. Increased immunoreactivity to two of the autoimmune response to myelin
overlapping peptides of myelin prote- oligodendrocyte glycoprotein (MOG) in
olipid protein in multiple sclerosis. Brain multiple sclerosis: reactivity to the ex-
120(Part 8):1447–60 tracellular domain of MOG is directed
134. Pelfrey CM, Trotter JL, Tranquill LR, Mc- against three main regions. Eur. J. Im-
Farland HF. 1993. Identification of a novel munol. 27:3059–69
T cell epitope of human proteolipid pro- 141. Wallstrom E, Khademi M, Andersson M,
tein (residues 40–60) recognized by pro- Weissert R, Linington C, Olsson T. 1998.
liferative and cytolytic CD4+ T cells from Increased reactivity to myelin oligoden-
multiple sclerosis patients. J. Neuroim- drocyte glycoprotein peptides and epitope
munol. 46:33–42 mapping in HLA DR2(15)+ multiple scle-
135. Pelfrey CM, Trotter JL, Tranquill LR, Mc- rosis. Eur. J. Immunol. 28:3329–35
Farland HF. 1994. Identification of a sec- 142. Koehler NK, Genain CP, Giesser B,
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 735

Hauser SL. 2002. The human T cell re- shock protein alters the course of EAE. J.
sponse to myelin oligodendrocyte glyco- Neurosci. Res. 44:381–96
protein: a multiple sclerosis family-based 150. Rosener M, Muraro PA, Riethmuller A,
study. J. Immunol. 168:5920–27 Kalbus M, Sappler G, et al. 1997. 2 ,3 -
143. Iglesias A, Bauer J, Litzenburger T, cyclic nucleotide 3 -phosphodiesterase: a
Schubart A, Linington C. 2001. T- and B- novel candidate autoantigen in demyeli-
cell responses to myelin oligodendrocyte nating diseases. J. Neuroimmunol. 75:28–
glycoprotein in experimental autoimmune 34
encephalomyelitis and multiple sclerosis. 151. Muraro PA, Kalbus M, Afshar G, Mc-
Glia 36:220–34 Farland HF, Martin R. 2002. T cell
144. Weissert R, Kuhle J, de Graaf KL, Wien- response to 2 ,3 -cyclic nucleotide 3 -
hold W, Herrmann MM, et al. 2002. High phosphodiesterase (CNPase) in multiple
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

immunogenicity of intracellular myelin sclerosis patients. J. Neuroimmunol. 130:


by University of Missouri - Columbia on 03/18/13. For personal use only.

oligodendrocyte glycoprotein epitopes. J. 233–42


Immunol. 169:548–56 152. Holz A, Bielekova B, Martin R, Oldstone
145. Lindert RB, Haase CG, Brehm U, Lin- MB. 2000. Myelin-associated oligoden-
ington C, Wekerle H, Hohlfeld R. 1999. drocytic basic protein: identification of an
Multiple sclerosis: B- and T-cell re- encephalitogenic epitope and association
sponses to the extracellular domain of with multiple sclerosis. J. Immunol. 164:
the myelin oligodendrocyte glycoprotein. 1103–9
Brain 122(Part 11):2089–100 153. Kaye JF, Kerlero de Rosbo N, Mendel I,
146. Steck AJ, Murray N, Meier C, Page N, Flechter S, Hoffman M, et al. 2000. The
Perruisseau G. 1983. Demyelinating neu- central nervous sytem-specific myelin
ropathy and monoclonal IgM antibody to oligodendrocytic basic protein (MOBP)
myelin-associated glycoprotein. Neurol- is encephalitogenic and a potential target
ogy 33:19–23 antigen in multiple sclerosis (MS). J. Neu-
147. Morris-Downes MM, McCormack K, roimmunol. 102:189–98
Baker D, Sivaprasad D, Natkunarajah J, 154. Arbour N, Holz A, Sipe JC, Naniche D,
Amor S. 2002. Encephalitogenic and im- Romine JS, et al. 2003. A new approach
munogenic potential of myelin-associated for evaluating antigen-specific T cell re-
glycoprotein (MAG), oligodendrocyte- sponses to myelin antigens during the
specific glycoprotein (OSP) and 2 ,3 - course of multiple sclerosis. J. Neuroim-
cyclic nucleotide 3 -phosphodiesterase munol. 137:197–209
(CNPase) in ABH and SJL mice. J. Neu- 155. Bronstein JM, Lallone RL, Seitz RS, El-
roimmunol. 122:20–33 lison GW, Myers LW. 1999. A humoral
148. Andersson M, Yu M, Soderstrom M, response to oligodendrocyte-specific pro-
Weerth S, Baig S, et al. 2002. Multiple tein in MS: a potential molecular mimic.
MAG peptides are recognized by circulat- Neurology 53:154–61
ing T and B lymphocytes in polyneuropa- 156. Vu T, Myers LW, Ellison GW, Mendoza
thy and multiple sclerosis. Eur. J. Neurol. F, Bronstein JM. 2001. T-cell responses to
9:243–51 oligodendrocyte-specific protein in multi-
149. Birnbaum G, Kotilinek L, Schlievert P, ple sclerosis. J. Neurosci. Res. 66:506–9
Clark HB, Trotter J, et al. 1996. Heat 157. van Noort JM, van Sechel AC, Bajramovic
shock proteins and experimental autoim- JJ, el Ouagmiri M, Polman CH, et al.
mune encephalomyelitis (EAE): I. Im- 1995. The small heat-shock protein αB-
munization with a peptide of the myelin crystallin as candidate autoantigen in mul-
protein 2 ,3 cyclic nucleotide 3 phospho- tiple sclerosis. Nature 375:798–801
diesterase that is cross-reactive with a heat 158. Chou YK, Burrows GG, LaTocha D,
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

736 SOSPEDRA  MARTIN

Wang C, Subramanian S, et al. 2004. CD4 167. Pender MP, Csurhes PA, Wolfe NP,
T-cell epitopes of human αB-crystallin. J. Hooper KD, Good MF, et al. 2003. In-
Neurosci. Res. 75:516–23 creased circulating T cell reactivity to
159. Saez-Torres I, Brieva L, Espejo C, Bar- GM3 and GQ1b gangliosides in primary
rau MA, Montalban X, Martinez-Caceres progressive multiple sclerosis. J. Clin.
EM. 2002. Specific proliferation towards Neurosci. 10:63–66
myelin antigens in patients with multiple 168. Meinl E, Weber F, Drexler K, Morelle C,
sclerosis during a relapse. Autoimmunity Ott M, et al. 1993. Myelin basic protein-
35:45–50 specific T lymphocyte repertoire in multi-
160. Kojima K, Berger T, Lassmann H, Hinze- ple sclerosis. Complexity of the response
Selch D, Zhang Y, et al. 1994. Experi- and dominance of nested epitopes due to
mental autoimmune panencephalitis and recruitment of multiple T cell clones. J.
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

uveoretinitis transferred to the Lewis rat Clin. Invest. 92:2633–43


by University of Missouri - Columbia on 03/18/13. For personal use only.

by T lymphocytes specific for the S100β 169. Hemmer B, Fleckenstein BT, Vergelli M,
molecule, a calcium binding protein of as- Jung G, McFarland H, et al. 1997. Identi-
troglia. J. Exp. Med. 180:817–29 fication of high potency microbial and self
161. Schmidt S, Linington C, Zipp F, Sot- ligands for a human autoreactive class II-
giu S, de Waal Malefyt R, et al. 1997. restricted T cell clone. J. Exp. Med. 185:
Multiple sclerosis: comparison of the hu- 1651–59
man T-cell response to S100β and myelin 170. Zhao Y, Gran B, Pinilla C, Markovic-
basic protein reveals parallels to rat ex- Plese S, Hemmer B, et al. 2001. Combina-
perimental autoimmune panencephalitis. torial peptide libraries and biometric score
Brain 120(Part 8):1437–45 matrices permit the quantitative analysis
162. Banki K, Colombo E, Sia F, Hal- of specific and degenerate interactions be-
laday D, Mattson DH, et al. 1994. tween clonotypic TCR and MHC peptide
Oligodendrocyte-specific expression and ligands. J. Immunol. 167:2130–41
autoantigenicity of transaldolase in mul- 171. Jaraquemada D, Martin R, Rosen-
tiple sclerosis. J. Exp. Med. 180:1649–63 Bronson S, Flerlage M, McFarland HF,
163. Colombo E, Banki K, Tatum AH, Daucher Long EO. 1990. HLA-DR2a is the domi-
J, Ferrante P, et al. 1997. Comparative nant restriction molecule for the cytotoxic
analysis of antibody and cell-mediated au- T cell response to myelin basic protein
toimmunity to transaldolase and myelin in DR2Dw2 individuals. J. Immunol. 145:
basic protein in patients with multiple 2880–85
sclerosis. J. Clin. Invest. 99:1238–50 172. Minohara M, Ochi H, Matsushita S, Irie
164. Holmoy T, Vandvik B, Vartdal F. 2003. A, Nishimura Y, Kira J. 2001. Differ-
T cells from multiple sclerosis patients ences between T-cell reactivities to major
recognize immunoglobulin G from cere- myelin protein-derived peptides in opti-
brospinal fluid. Mult. Scler. 9:228–34 cospinal and conventional forms of multi-
165. Holmoy T, Vartdal F. 2004. Cerebrospinal ple sclerosis and healthy controls. Tissue
fluid T cells from multiple sclerosis pa- Antigens 57:447–56
tients recognize autologous Epstein-Barr 173. Yu M, Kinkel RP, Weinstock-Guttman B,
virus-transformed B cells. J. Neurovirol. Cook DJ, Tuohy VK. 1998. HLA-DP:
10:52–56 a class II restriction molecule involved
166. Sadatipour BT, Greer JM, Pender MP. in epitope spreading during the devel-
1998. Increased circulating antiganglio- opment of multiple sclerosis. Hum. Im-
side antibodies in primary and secondary munol. 59:15–24
progressive multiple sclerosis. Ann. Neu- 174. Hemmer B, Vergelli M, Calabresi P,
rol. 44:980–83 Huang T, McFarland HF, Martin R. 1996.
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 737

Cytokine phenotype of human autoreac- 182. Antel JP, McCrea E, Ladiwala U, Qin
tive T cell clones specific for the immun- YF, Becher B. 1998. Non-MHC-restricted
odominant myelin basic protein peptide cell-mediated lysis of human oligoden-
(83–99). J. Neurosci. Res. 45:852–62 drocytes in vitro: relation with CD56 ex-
175. Correale J, Gilmore W, McMillan M, Li pression. J. Immunol. 160:1606–11
S, McCarthy K, et al. 1995. Patterns of 183. Scholz C, Patton KT, Anderson DE, Free-
cytokine secretion by autoreactive prote- man GJ, Hafler DA. 1998. Expansion of
olipid protein-specific T cell clones dur- autoreactive T cells in multiple sclerosis
ing the course of multiple sclerosis. J. Im- is independent of exogenous B7 costimu-
munol. 154:2959–68 lation. J. Immunol. 160:1532–38
176. Balashov KE, Smith DR, Khoury SJ, 184. Lovett-Racke AE, Trotter JL, Lauber
Hafler DA, Weiner HL. 1997. Increased J, Perrin PJ, June CH, Racke MK.
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

interleukin 12 production in progressive 1998. Decreased dependence of myelin


by University of Missouri - Columbia on 03/18/13. For personal use only.

multiple sclerosis: induction by activated basic protein-reactive T cells on CD28-


CD4+ T cells via CD40 ligand. Proc. Natl. mediated costimulation in multiple
Acad. Sci. USA 94:599–603 sclerosis patients. A marker of acti-
177. Karni A, Koldzic DN, Bharanidharan P, vated/memory T cells. J. Clin. Invest. 101:
Khoury SJ, Weiner HL. 2002. IL-18 is 725–30
linked to raised IFN-γ in multiple scle- 185. Markovic-Plese S, Cortese I, Wandinger
rosis and is induced by activated CD4+ T KP, McFarland HF, Martin R. 2001.
cells via CD40-CD40 ligand interactions. CD4+ CD28− costimulation-independent
J. Neuroimmunol. 125:134–40 T cells in multiple sclerosis. J. Clin. In-
178. Monney L, Sabatos CA, Gaglia JL, vest. 108:1185–94
Ryu A, Waldner H, et al. 2002. Th1- 186. Chitnis T, Khoury SJ. 2003. Role of cos-
specific cell surface protein Tim-3 regu- timulatory pathways in the pathogenesis
lates macrophage activation and severity of multiple sclerosis and experimental au-
of an autoimmune disease. Nature 415: toimmune encephalomyelitis. J. Allergy
536–41 Clin. Immunol. 112:837–50
179. Khademi M, Illes Z, Gielen AW, Marta 187. Oliveira EM, Bar-Or A, Waliszewska AI,
M, Takazawa N, et al. 2004. T Cell Ig- Cai G, Anderson DE, et al. 2003. CTLA-4
and mucin-domain-containing molecule- dysregulation in the activation of myelin
3 (TIM-3) and TIM-1 molecules are dif- basic protein reactive T cells may dis-
ferentially expressed on human Th1 and tinguish patients with multiple sclerosis
Th2 cells and in cerebrospinal fluid- from healthy controls. J. Autoimmun. 20:
derived mononuclear cells in multiple 71–81
sclerosis. J. Immunol. 172:7169–76 188. Hori S, Nomura T, Sakaguchi S. 2003.
180. Vergelli M, Le H, van Noort JM, Dhib- Control of regulatory T cell development
Jalbut S, McFarland H, Martin R. 1996. A by the transcription factor Foxp3. Science
novel population of CD4+ CD56+ myelin- 299:1057–61
reactive T cells lyses target cells express- 189. Zhang X, Koldzic DN, Izikson L, Reddy
ing CD56/neural cell adhesion molecule. J, Nazareno RF, et al. 2004. IL-10 is
J. Immunol. 157:679–88 involved in the suppression of experi-
181. Vergelli M, Hemmer B, Muraro PA, Tran- mental autoimmune encephalomyelitis by
quill L, Biddison WE, et al. 1997. Hu- CD25+ CD4+ regulatory T cells. Int. Im-
man autoreactive CD4+ T cell clones munol. 16:249–56
use perforin- or Fas/Fas ligand-mediated 190. Viglietta V, Baecher-Allan C, Weiner
pathways for target cell lysis. J. Immunol. HL, Hafler DA. 2004. Loss of functional
158:2756–61 suppression by CD4+ CD25+ regulatory
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

738 SOSPEDRA  MARTIN

T cells in patients with multiple sclerosis. quet D, Gachelin G, Dormont D. 2003.


J. Exp. Med. 199:971–79 The same TCR (N)Dβ(N)Jβ junctional
191. Weiner HL, Friedman A, Miller A, region is associated with several different
Khoury SJ, Al-Sabbagh A, et al. 1994. vβ13 subtypes in a multiple sclerosis pa-
Oral tolerance: immunologic mechanisms tient at the onset of the disease. Neurobiol.
and treatment of animal and human organ- Dis. 14:470–82
specific autoimmune diseases by oral ad- 199. Laplaud DA, Ruiz C, Wiertlewski S,
ministration of autoantigens. Annu. Rev. Brouard S, Berthelot L, et al. 2004. Blood
Immunol. 12:809–37 T-cell receptor β chain transcriptome in
192. Lafaille JJ, Keere FV, Hsu AL, Baron JL, multiple sclerosis. Characterization of the
Haas W, et al. 1997. Myelin basic protein- T cells with altered CDR3 length distri-
specific T helper 2 (Th2) cells cause ex- bution. Brain 127:981–95
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

perimental autoimmune encephalomyeli- 200. Matsumoto Y, Yoon WK, Jee Y, Fujihara


by University of Missouri - Columbia on 03/18/13. For personal use only.

tis in immunodeficient hosts rather than K, Misu T, et al. 2003. Complementarity-


protect them from the disease. J. Exp. determining region 3 spectratyping analy-
Med. 186:307–12 sis of the TCR repertoire in multiple scle-
193. Wucherpfennig KW, Ota K, Endo N, rosis. J. Immunol. 170:4846–53
Seidman JG, Rosenzweig A, et al. 1990. 201. Gestri D, Baldacci L, Taiuti R, Galli E,
Shared human T cell receptor Vβ usage to Maggi E, et al. 2001. Oligoclonal T cell
immunodominant regions of myelin basic repertoire in cerebrospinal fluid of pa-
protein. Science 248:1016–19 tients with inflammatory diseases of the
194. Ben-Nun A, Liblau RS, Cohen L, nervous system. J. Neurol. Neurosurg.
Lehmann D, Tournier-Lasserve E, et al. Psychiatry 70:767–72
1991. Restricted T-cell receptor Vβ gene 202. Utz U, Biddison WE, McFarland HF, Mc-
usage by myelin basic protein-specific T- Farlin DE, Flerlage M, Martin R. 1993.
cell clones in multiple sclerosis: predomi- Skewed T-cell receptor repertoire in ge-
nant genes vary in individuals. Proc. Natl. netically identical twins correlates with
Acad. Sci. USA 88:2466–70 multiple sclerosis. Nature 364:243–47
195. Kotzin BL, Karuturi S, Chou YK, Laf- 203. Haegert DG, Galutira D, Murray TJ,
ferty J, Forrester M, et al. 1991. Preferen- O’Connor P, Gadag V. 2003. Identical
tial T-cell receptor β-chain variable gene twins discordant for multiple sclerosis
use in myelin basic protein-reactive T-cell have a shift in their T-cell receptor reper-
clones from patients with multiple sclero- toires. Clin. Exp. Immunol. 134:532–37
sis. Proc. Natl. Acad. Sci. USA 88:9161– 204. Jurewicz A, Biddison WE, Antel JP. 1998.
65 MHC class I-restricted lysis of human
196. Afshar G, Muraro PA, McFarland HF, oligodendrocytes by myelin basic protein
Martin R. 1998. Lack of over-expression peptide-specific CD8 T lymphocytes. J.
of T cell receptor Vβ5.2 in myelin ba- Immunol. 160:3056–59
sic protein-specific T cell lines derived 205. Medana IM, Gallimore A, Oxenius A,
from HLA-DR2 positive multiple scle- Martinic MM, Wekerle H, Neumann H.
rosis patients and controls. J. Neuroim- 2000. MHC class I-restricted killing of
munol. 84:7–13 neurons by virus-specific CD8+ T lym-
197. Oksenberg JR, Stuart S, Begovich AB, phocytes is effected through the Fas/FasL,
Bell RB, Erlich HA, et al. 1990. Limited but not the perforin pathway. Eur. J. Im-
heterogeneity of rearranged T-cell recep- munol. 30:3623–33
tor V α transcripts in brains of multiple 206. Skulina C, Schmidt S, Dornmair K, Babbe
sclerosis patients. Nature 345:344–46 H, Roers A, et al. 2004. Multiple sclero-
198. Demoulins T, Mouthon F, Clayette P, Be- sis: brain-infiltrating CD8+ T cells persist
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 739

as clonal expansions in the cerebrospinal dred cases of multiple sclerosis and in


fluid and blood. Proc. Natl. Acad. Sci. other diseases. Am. J. Med. Sci. 219:55–
USA 101:2428–33 64
207. Cabarrocas J, Bauer J, Piaggio E, Liblau 215. Qin Y, Duquette P, Zhang Y, Talbot P,
R, Lassmann H. 2003. Effective and selec- Poole R, Antel J. 1998. Clonal expansion
tive immune surveillance of the brain by and somatic hypermutation of VH genes of
MHC class I-restricted cytotoxic T lym- B cells from cerebrospinal fluid in multi-
phocytes. Eur. J. Immunol. 33:1174–82 ple sclerosis. J. Clin. Invest. 102:1045–50
208. Neumann H. 2003. Molecular mecha- 216. Owens GP, Kraus H, Burgoon MP, Smith-
nisms of axonal damage in inflammatory Jensen T, Devlin ME, Gilden DH. 1998.
central nervous system diseases. Curr. Restricted use of VH4 germline segments
Opin. Neurol. 16:267–73 in an acute multiple sclerosis brain. Ann.
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

209. Tsuchida T, Parker KC, Turner RV, Neurol. 43:236–43


by University of Missouri - Columbia on 03/18/13. For personal use only.

McFarland HF, Coligan JE, Biddison 217. Baranzini SE, Jeong MC, Butunoi C, Mur-
WE. 1994. Autoreactive CD8+ T-cell re- ray RS, Bernard CC, Oksenberg JR. 1999.
sponses to human myelin protein-derived B cell repertoire diversity and clonal ex-
peptides. Proc. Natl. Acad. Sci. USA 91: pansion in multiple sclerosis brain lesions.
10859–63 J. Immunol. 163:5133–44
210. Honma K, Parker KC, Becker KG, McFar- 218. Walsh MJ, Tourtellotte WW. 1986. Tem-
land HF, Coligan JE, Biddison WE. 1997. poral invariance and clonal uniformity of
Identification of an epitope derived from brain and cerebrospinal IgG, IgA, and
human proteolipid protein that can in- IgM in multiple sclerosis. J. Exp. Med.
duce autoreactive CD8+ cytotoxic T lym- 163:41–53
phocytes restricted by HLA-A3: evidence 219. Sharief MK, Thompson EJ. 1991. In-
for cross-reactivity with an environmental trathecal immunoglobulin M synthesis in
microorganism. J. Neuroimmunol. 73:7– multiple sclerosis. Relationship with clin-
14 ical and cerebrospinal fluid parameters.
211. Zang YCQ, Li SF, Rivera VM, Hong Brain 114(Part 1A):181–95
J, Robinson RR, et al. 2004. Increased 220. Wang LY, Fujinami RS. 1997. Enhance-
CD8+ cytotoxic T cell responses to ment of EAE and induction of autoanti-
myelin basic protein in multiple sclerosis. bodies to T-cell epitopes in mice infected
J. Immunol. 172:5120–27 with a recombinant vaccinia virus encod-
212. Biddison WE, Cruikshank WW, Cen- ing myelin proteolipid protein. J. Neu-
ter DM, Pelfrey CM, Taub DD, Turner roimmunol. 75:75–83
RV. 1998. CD8+ myelin peptide-specific 221. Wucherpfennig KW, Catz I, Hausmann S,
T cells can chemoattract CD4+ myelin Strominger JL, Steinman L, Warren KG.
peptide-specific T cells: importance of 1997. Recognition of the immunodomi-
IFN-inducible protein 10. J. Immunol. nant myelin basic protein peptide by au-
160:444–48 toantibodies and HLA-DR2-restricted T
213. Killestein J, Eikelenboom MJ, Izeboud T, cell clones from multiple sclerosis pa-
Kalkers NF, Ader HJ, et al. 2003. Cy- tients. Identity of key contact residues in
tokine producing CD8+ T cells are corre- the B-cell and T-cell epitopes. J. Clin. In-
lated to MRI features of tissue destruction vest. 100:1114–22
in MS. J. Neuroimmunol. 142:141–48 222. Lou YH, Park KK, Agersborg S, Alard
214. Kabat EA, Freedman DA, Murray JP, P, Tung KS. 2000. Retargeting T cell-
Knaub V. 1950. A study of the cristalline mediated inflammation: a new perspective
albumin, gamma globulin and total pro- on autoantibody action. J. Immunol. 164:
tein in the cerebrospinal fluid of one hun- 5251–57
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

740 SOSPEDRA  MARTIN

223. Bornstein MB, Appel SH. 1959. De- the complement system and disability in
myelination in cultures of rat cerebel- multiple sclerosis. J. Neurol. Sci. 157:
lum produced by experimental allergic 168–74
encephalomyelitic serum. Trans. Am. 233. Scolding NJ, Morgan BP, Houston WA,
Neurol. Assoc. 84:165–66 Linington C, Campbell AK, Compston
224. Esiri MM. 1977. Immunoglobulin- DA. 1989. Vesicular removal by oligoden-
containing cells in multiple-sclerosis drocytes of membrane attack complexes
plaques. Lancet 2:478 formed by activated complement. Nature
225. Mattson DH, Roos RP, Arnason BG. 339:620–22
1980. Isoelectric focusing of IgG eluted 234. Trotter JL, Rust RS. 1989. Human Cere-
from multiple sclerosis and subacute brospinal Fluid Immunology, ed. RM
sclerosing panencephalitis brains. Nature Herndon, RA Brumback, pp. 179–226.
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

287:335–37 Amsterdam: Martinus Nyhoff


by University of Missouri - Columbia on 03/18/13. For personal use only.

226. Genain CP, Cannella B, Hauser SL, Raine 235. Sindic CJ, Monteyne P, Laterre EC. 1994.
CS. 1999. Identification of autoantibodies The intrathecal synthesis of virus-specific
associated with myelin damage in multi- oligoclonal IgG in multiple sclerosis. J.
ple sclerosis. Nat. Med. 5:170–75 Neuroimmunol. 54:75–80
227. Trotter J, DeJong LJ, Smith ME. 1986. 236. Archelos JJ, Storch MK, Hartung HP.
Opsonization with antimyelin antibody 2000. The role of B cells and autoantibod-
increases the uptake and intracellular ies in multiple sclerosis. Ann. Neurol. 47:
metabolism of myelin in inflammatory 694–706
macrophages. J. Neurochem. 47:779–89 237. Cross AH, Trotter JL, Lyons J. 2001.
228. Goldenberg PZ, Kwon EE, Benjamins JA, B cells and antibodies in CNS demyeli-
Whitaker JN, Quarles RH, Prineas JW. nating disease. J. Neuroimmunol. 112:1–
1989. Opsonization of normal myelin by 14
anti-myelin antibodies and normal serum. 238. Dharmasaroja P. 2003. Specificity of au-
J. Neuroimmunol. 23:157–66 toantibodies to epitopes of myelin pro-
229. van der Laan LJ, Ruuls SR, Weber KS, teins in multiple sclerosis. J. Neurol. Sci.
Lodder IJ, Dopp EA, Dijkstra CD. 1996. 206:7–16
Macrophage phagocytosis of myelin in 239. Warren KG, Catz I. 1993. Increased syn-
vitro determined by flow cytometry: thetic peptide specificity of tissue-CSF
Phagocytosis is mediated by CR3 and in- bound anti-MBP in multiple sclerosis. J.
duces production of tumor necrosis factor- Neuroimmunol. 43:87–96
α and nitric oxide. J. Neuroimmunol. 70: 240. Reindl M, Linington C, Brehm U, Egg
145–52 R, Dilitz E, et al. 1999. Antibodies
230. Mead RJ, Singhrao SK, Neal JW, Lass- against the myelin oligodendrocyte gly-
mann H, Morgan BP. 2002. The mem- coprotein and the myelin basic protein
brane attack complex of complement in multiple sclerosis and other neurolog-
causes severe demyelination associated ical diseases: a comparative study. Brain
with acute axonal injury. J. Immunol. 168: 122(Part 11):2047–56
458–65 241. Cortese I, Tafi R, Grimaldi LM, Martino
231. Storch MK, Piddlesden S, Haltia M, G, Nicosia A, Cortese R. 1996. Identifica-
Iivanainen M, Morgan P, Lassmann H. tion of peptides specific for cerebrospinal
1998. Multiple sclerosis: in situ evidence fluid antibodies in multiple sclerosis by
for antibody- and complement-mediated using phage libraries. Proc. Natl. Acad.
demyelination. Ann. Neurol. 43:465–71 Sci. USA 93:11063–67
232. Sellebjerg F, Jaliashvili I, Christiansen M, 242. O’Connor KC, Chitnis T, Griffin DE,
Garred P. 1998. Intrathecal activation of Piyasirisilp S, Bar-Or A, et al. 2003.
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 741

Myelin basic protein-reactive autoanti- in multiple sclerosis. J. Immunol. 146:


bodies in the serum and cerebrospinal 1490–95
fluid of multiple sclerosis patients are 251. Berger T, Rubner P, Schautzer F, Egg R,
characterized by low-affinity interactions. Ulmer H, et al. 2003. Antimyelin antibod-
J. Neuroimmunol. 136:140–48 ies as a predictor of clinically definite mul-
243. Sun JB, Olsson T, Wang WZ, Xiao tiple sclerosis after a first demyelinating
BG, Kostulas V, et al. 1991. Autoreac- event. N. Engl. J. Med. 349:139–45
tive T and B cells responding to myelin 252. Lily O, Palace J, Vincent A. 2004. Serum
proteolipid protein in multiple sclerosis autoantibodies to cell surface determi-
and controls. Eur. J. Immunol. 21:1461– nants in multiple sclerosis: a flow cyto-
68 metric study. Brain 127:269–79
244. Warren KG, Catz I, Johnson E, Mielke B. 253. Robinson WH, Fontoura P, Lee BJ, de
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

1994. Anti-myelin basic protein and anti- Vegvar HE, Tom J, et al. 2003. Pro-
by University of Missouri - Columbia on 03/18/13. For personal use only.

proteolipid protein specific forms of mul- tein microarrays guide tolerizing DNA
tiple sclerosis. Ann. Neurol. 35:280–89 vaccine treatment of autoimmune en-
245. Schluesener HJ, Sobel RA, Linington C, cephalomyelitis. Nat. Biotechnol. 21:
Weiner HL. 1987. A monoclonal antibody 1033–39
against a myelin oligodendrocyte glyco- 254. Saoudi A, Simmonds S, Huitinga I, Ma-
protein induces relapses and demyelina- son D. 1995. Prevention of experimental
tion in central nervous system autoim- allergic encephalomyelitis in rats by tar-
mune disease. J. Immunol. 139:4016–21 geting autoantigen to B cells: evidence
246. Linington C, Bradl M, Lassmann H, that the protective mechanism depends
Brunner C, Vass K. 1988. Augmenta- on changes in the cytokine response and
tion of demyelination in rat acute allergic migratory properties of the autoantigen-
encephalomyelitis by circulating mouse specific T cells. J. Exp. Med. 182:335–
monoclonal antibodies directed against 44
a myelin/oligodendrocyte glycoprotein. 255. Rodriguez M, Lennon VA. 1990. Im-
Am. J. Pathol. 130:443–54 munoglobulins promote remyelination in
247. Storch MK, Stefferl A, Brehm U, Weissert the central nervous system. Ann. Neurol.
R, Wallstrom E, et al. 1998. Autoimmu- 27:12–17
nity to myelin oligodendrocyte glycopro- 256. Achiron A, Gabbay U, Gilad R, Hassin-
tein in rats mimics the spectrum of mul- Baer S, Barak Y, et al. 1998. In-
tiple sclerosis pathology. Brain Pathol. 8: travenous immunoglobulin treatment in
681–94 multiple sclerosis. Effect on relapses.
248. Litzenburger T, Fassler R, Bauer J, Lass- Neurology 50:398–402
mann H, Linington C, et al. 1998. B lym- 257. Sewell WA, Jolles S. 2002. Immunomod-
phocytes producing demyelinating au- ulatory action of intravenous immunog-
toantibodies: development and function lobulin. Immunology 107:387–93
in gene-targeted transgenic mice. J. Exp. 258. van Kooyk Y, Geijtenbeek TB. 2002.
Med. 188:169–80 A novel adhesion pathway that regulates
249. Genain CP, Nguyen MH, Letvin NL, Pearl dendritic cell trafficking and T cell inter-
R, Davis RL, et al. 1995. Antibody facili- actions. Immunol. Rev. 186:47–56
tation of multiple sclerosis-like lesions in 259. Pasare C, Medzhitov R. 2003. Toll
a nonhuman primate. J. Clin. Invest. 96: pathway-dependent blockade of CD4+
2966–74 CD25+ T cell-mediated suppression by
250. Sun J, Link H, Olsson T, Xiao BG, Ander- dendritic cells. Science 299:1033–36
sson G, et al. 1991. T and B cell responses 260. Ohshima S, Saeki Y, Mima T, Sasai
to myelin-oligodendrocyte glycoprotein M, Nishioka K, et al. 1998. Interleukin
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

742 SOSPEDRA  MARTIN

6 plays a key role in the development a possible pathogenetic factor. Clin. Exp.
of antigen-induced arthritis. Proc. Natl. Immunol. 39:657–62
Acad. Sci. USA 95:8222–26 271. Kastrukoff LF, Morgan NG, Zecchini D,
261. Hawiger D, Inaba K, Dorsett Y, Guo M, White R, Petkau AJ, et al. 1998. A role for
Mahnke K, et al. 2001. Dendritic cells in- natural killer cells in the immunopatho-
duce peripheral T cell unresponsiveness genesis of multiple sclerosis. J. Neuroim-
under steady state conditions in vivo. J. munol. 86:123–33
Exp. Med. 194:769–79 272. Kastrukoff LF, Lau A, Wee R, Zecchini
262. Tsunoda I, Tolley ND, Theil DJ, Whitton D, White R, Paty DW. 2003. Clinical re-
JL, Kobayashi H, Fujinami RS. 1999. Ex- lapses of multiple sclerosis are associated
acerbation of viral and autoimmune ani- with ‘novel’ valleys in natural killer cell
mal models for multiple sclerosis by bac- functional activity. J. Neuroimmunol. 145:
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

terial DNA. Brain Pathol. 9:481–93 103–14


by University of Missouri - Columbia on 03/18/13. For personal use only.

263. Ichikawa HT, Williams LP, Segal BM. 273. Munschauer FE, Hartrich LA, Stewart
2002. Activation of APCs through CD40 CC, Jacobs L. 1995. Circulating natural
or Toll-like receptor 9 overcomes toler- killer cells but not cytotoxic T lympho-
ance and precipitates autoimmune dis- cytes are reduced in patients with active
ease. J. Immunol. 169:2781–87 relapsing multiple sclerosis and little clin-
264. Zappulla JP, Arock M, Mars LT, Liblau ical disability as compared to controls. J.
RS. 2002. Mast cells: new targets for mul- Neuroimmunol. 62:177–81
tiple sclerosis therapy? J. Neuroimmunol. 274. Weber WE, Buurman WA, Vandermeeren
131:5–20 MM, Medaer RH, Raus JC. 1987. Fine
265. Neumann J. 1890. Über das Vorkommen analysis of cytolytic and natural killer T
der sogenannten “Mastzellen” bei pathol- lymphocytes in the CSF in multiple scle-
ogischen Veränderungen des Gehirns. Vir- rosis and other neurologic diseases. Neu-
chow’s Arch. Pathol. Anat. 122:378– rology 37:419–25
80 275. Zhang B, Yamamura T, Kondo T, Fuji-
266. Kruger PG. 2001. Mast cells and multi- wara M, Tabira T. 1997. Regulation of ex-
ple sclerosis: a quantitative analysis. Neu- perimental autoimmune encephalomyeli-
ropathol. Appl. Neurobiol. 27:275–80 tis by natural killer (NK) cells. J. Exp.
267. Bo L, Dawson TM, Wesselingh S, Mork S, Med. 186:1677–87
Choi S, et al. 1994. Induction of nitric ox- 276. Matsumoto Y, Kohyama K, Aikawa Y,
ide synthase in demyelinating regions of Shin T, Kawazoe Y, et al. 1998. Role of
multiple sclerosis brains. Ann. Neurol. 36: natural killer cells and TCR γ δ T cells
778–86 in acute autoimmune encephalomyelitis.
268. Merrill JE, Ignarro LJ, Sherman MP, Eur. J. Immunol. 28:1681–88
Melinek J, Lane TE. 1993. Microglial cell 277. Smeltz RB, Wolf NA, Swanborg RH.
cytotoxicity of oligodendrocytes is medi- 1999. Inhibition of autoimmune T cell re-
ated through nitric oxide. J. Immunol. 151: sponses in the DA rat by bone marrow-
2132–41 derived NK cells in vitro: implications for
269. Mitrovic B, Ignarro LJ, Vinters HV, Akers autoimmunity. J. Immunol. 163:1390–97
MA, Schmid I, et al. 1995. Nitric oxide in- 278. Peng SL, Moslehi J, Robert ME, Craft J.
duces necrotic but not apoptotic cell death 1998. Perforin protects against autoimmu-
in oligodendrocytes. Neuroscience 65: nity in lupus-prone mice. J. Immunol. 160:
531–39 652–60
270. Benczur M, Petranyl GG, Palffy G, Varga 279. Hilliard B, Wilmen A, Seidel C, Liu
M, Talas M, et al. 1980. Dysfunction of TS, Goke R, Chen Y. 2001. Roles of
natural killer cells in multiple sclerosis: TNF-related apoptosis-inducing ligand in
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 743

experimental autoimmune encphalomye- 2001. Decreases in interleukin-4 secre-


litis. J. Immunol. 166:1314–19 tion by invariant CD4− CD8− Vα24JαQ
280. Bielekova B, Richert N, Howard T, T cells in peripheral blood of patients
Blevins G, Markovic-Plese S, et al. 2004. with relapsing-remitting multiple sclero-
Humanized anti-CD25 (daclizumab) in- sis. Clin. Immunol. 98:11–17
hibits disease activity in multiple sclerosis 289. Miyamoto K, Miyake S, Yamamura T.
patients failing to respond to interferon β. 2001. A synthetic glycolipid prevents au-
Proc. Natl. Acad. Sci. USA 101:8705–8 toimmune encephalomyelitis by inducing
281. Vanguri P, Shin ML. 1986. Activation TH2 bias of natural killer T cells. Nature
of complement by myelin: identifica- 413:531–34
tion of C1-binding proteins of human 290. Triebel F, Hercend T. 1989. Subpopula-
myelin from central nervous tissue. J. tions of human peripheral T γ δ lympho-
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

Neurochem. 46:1535–41 cytes. Immunol. Today 10:186–88


by University of Missouri - Columbia on 03/18/13. For personal use only.

282. Scolding NJ, Morgan BP, Houston A, 291. Battistini L, Salvetti M, Ristori G, Falcone
Campbell AK, Linington C, Compston M, Raine CS, Brosnan CF. 1995. γ δ T cell
DA. 1989. Normal rat serum cytotoxi- receptor analysis supports a role for HSP
city against syngeneic oligodendrocytes. 70 selection of lymphocytes in multiple
Complement activation and attack in the sclerosis lesions. Mol. Med. 1:554–62
absence of anti-myelin antibodies. J. Neu- 292. Freedman MS, Bitar R, Antel JP. 1997.
rol. Sci. 89:289–300 γ δ T-cell-human glial cell interactions. II.
283. Johns TG, Bernard CC. 1997. Bind- Relationship between heat shock protein
ing of complement component Clq to expression and susceptibility to cytolysis.
myelin oligodendrocyte glycoprotein: a J. Neuroimmunol. 74:143–48
novel mechanism for regulating CNS in- 293. Nick S, Pileri P, Tongiani S, Uematsu Y,
flammation. Mol. Immunol. 34:33–38 Kappos L, De Libero G. 1995. T cell re-
284. Piddlesden SJ, Morgan BP. 1993. Killing ceptor γ δ repertoire is skewed in cere-
of rat glial cells by complement: defi- brospinal fluid of multiple sclerosis pa-
ciency of the rat analogue of CD59 is the tients: molecular and functional analyses
cause of oligodendrocyte susceptibility to of antigen-reactive γ δ clones. Eur. J. Im-
lysis. J. Neuroimmunol. 48:169–75 munol. 25:355–63
285. Park SH, Bendelac A. 2000. CD1- 294. Stinissen P, Vandevyver C, Medaer R,
restricted T-cell responses and microbial Vandegaer L, Nies J, et al. 1995. Increased
infection. Nature 406:788–92 frequency of γ δ T cells in cerebrospinal
286. Illes Z, Kondo T, Newcombe J, Oka N, fluid and peripheral blood of patients with
Tabira T, Yamamura T. 2000. Differen- multiple sclerosis. Reactivity, cytotoxic-
tial expression of NK T cell Vα24JαQ ity, and T cell receptor V gene rearrange-
invariant TCR chain in the lesions of ments. J. Immunol. 154:4883–94
multiple sclerosis and chronic inflamma- 295. Birnbaum G, Kotilinek L, Albrecht L.
tory demyelinating polyneuropathy. J. Im- 1993. Spinal fluid lymphocytes from a
munol. 164:4375–81 subgroup of multiple sclerosis patients
287. van der Vliet HJ, von Blomberg BM, Nishi respond to mycobacterial antigens. Ann.
N, Reijm M, Voskuyl AE, et al. 2001. Cir- Neurol. 34:18–24
culating Vα24+ Vβ11+ NKT cell num- 296. Constant P, Davodeau F, Peyrat MA, Po-
bers are decreased in a wide variety of quet Y, Puzo G, et al. 1994. Stimulation
diseases that are characterized by autore- of human γ δ T cells by nonpeptidic my-
active tissue damage. Clin. Immunol. 100: cobacterial ligands. Science 264:267–70
144–48 297. Rajan AJ, Gao YL, Raine CS, Brosnan
288. Gausling R, Trollmo C, Hafler DA. CF. 1996. A pathogenic role for γ δ T cells
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

744 SOSPEDRA  MARTIN

in relapsing-remitting experimental aller- CA, Joyce B, et al. 2003. Interleukin-23


gic encephalomyelitis in the SJL mouse. rather than interleukin-12 is the critical
J. Immunol. 157:941–49 cytokine for autoimmune inflammation of
298. Bielekova B, Martin R. 2004. Develop- the brain. Nature 421:744–48
ment of biomarkers in multiple sclerosis. 307. Matusevicius D, Kivisakk P, Navikas
Brain 127:1463–78 V, Soderstrom M, Fredrikson S, Link
299. Navikas V, He B, Link J, Haglund M, H. 1998. Interleukin-12 and perforin
Soderstrom M, et al. 1996. Augmented ex- mRNA expression is augmented in blood
pression of tumour necrosis factor-α and mononuclear cells in multiple sclerosis.
lymphotoxin in mononuclear cells in mul- Scand. J. Immunol. 47:582–90
tiple sclerosis and optic neuritis. Brain 308. Heesen C, Sieverding F, Schoser BG,
119(Part 1):213–23 Hadji B, Kunze K. 1999. Interleukin-12
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

300. Hohnoki K, Inoue A, Koh CS. 1998. El- is detectable in sera of patients with mul-
by University of Missouri - Columbia on 03/18/13. For personal use only.

evated serum levels of IFN-γ , IL-4 and tiple sclerosis—association with chronic
TNF-α/unelevated serum levels of IL-10 progressive disease course? Eur. J. Neu-
in patients with demyelinating diseases rol. 6:591–96
during the acute stage. J. Neuroimmunol. 309. Huang WX, Huang P, Link H, Hillert
87:27–32 J. 1999. Cytokine analysis in multiple
301. Ozenci V, Kouwenhoven M, Huang YM, sclerosis by competitive RT—PCR: a de-
Kivisakk P, Link H. 2000. Multiple scle- creased expression of IL-10 and an in-
rosis is associated with an imbalance be- creased expression of TNF-α in chronic
tween tumour necrosis factor-α (TNF-α)- progression. Mult. Scler. 5:342–48
and IL-10-secreting blood cells that is cor- 310. Navikas V, Link J, Palasik W, Soderstrom
rected by interferon-β (IFN-β) treatment. M, Fredrikson S, et al. 1995. Increased
Clin. Exp. Immunol. 120:147–53 mRNA expression of IL-10 in mononu-
302. The Lenercept Multiple Sclerosis Study clear cells in multiple sclerosis and optic
Group and The University of British neuritis. Scand. J. Immunol. 41:171–78
Columbia MS/MRI Analysis Group. 311. Windhagen A, Newcombe J, Dangond F,
1999. TNF neutralization in MS: results Strand C, Woodroofe MN, et al. 1995. Ex-
of a randomized, placebo-controlled mul- pression of costimulatory molecules B7-1
ticenter study. Neurology 53:457–65 (CD80), B7-2 (CD86), and interleukin 12
303. Nguyen LT, Ramanathan M, Munschauer cytokine in multiple sclerosis lesions. J.
F, Brownscheidle C, Krantz S, et al. 1999. Exp. Med. 182:1985–96
Flow cytometric analysis of in vitro proin- 312. Cannella B, Raine CS. 1995. The adhesion
flammatory cytokine secretion in periph- molecule and cytokine profile of multiple
eral blood from multiple sclerosis pa- sclerosis lesions. Ann. Neurol. 37:424–35
tients. J. Clin. Immunol. 19:179–85 313. Bitsch A, Kuhlmann T, Da Costa C,
304. Panitch HS, Hirsch RL, Schindler J, John- Bunkowski S, Polak T, Bruck W. 2000.
son KP. 1987. Treatment of multiple scle- Tumour necrosis factor α mRNA ex-
rosis with γ interferon: exacerbations as- pression in early multiple sclerosis le-
sociated with activation of the immune sions: correlation with demyelinating ac-
system. Neurology 37:1097–102 tivity and oligodendrocyte pathology.
305. Segal BM, Dwyer BK, Shevach EM. Glia 29:366–75
1998. An interleukin (IL)-10/IL-12 im- 314. Owens T, Wekerle H, Antel J. 2001. Ge-
munoregulatory circuit controls suscepti- netic models for CNS inflammation. Nat.
bility to autoimmune disease. J. Exp. Med. Med. 7:161–66
187:537–46 315. Pouly S, Becher B, Blain M, Antel
306. Cua DJ, Sherlock J, Chen Y, Murphy JP. 2000. Interferon-γ modulates human
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 745

oligodendrocyte susceptibility to Fas- chemokine signaling. J. Neuroimmunol.


mediated apoptosis. J. Neuropathol. Exp. 114:207–12
Neurol. 59:280–86 324. Balashov KE, Rottman JB, Weiner
316. Link J, Soderstrom M, Olsson T, Ho- HL, Hancock WW. 1999. CCR5+ and
jeberg B, Ljungdahl A, Link H. 1994. CXCR3+ T cells are increased in mul-
Increased transforming growth factor-β, tiple sclerosis and their ligands MIP-1α
interleukin-4, and interferon-γ in multi- and IP-10 are expressed in demyelinating
ple sclerosis. Ann. Neurol. 36:379–86 brain lesions. Proc. Natl. Acad. Sci. USA
317. Nakashima I, Fujihara K, Misu T, Okita 96:6873–78
N, Takase S, Itoyama Y. 2000. Signifi- 325. Calabresi PA, Yun SH, Allie R,
cant correlation between IL-10 levels and Whartenby KA. 2002. Chemokine recep-
IgG indices in the cerebrospinal fluid of tor expression on MBP-reactive T cells:
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

patients with multiple sclerosis. J. Neu- CXCR6 is a marker of IFNγ -producing


by University of Missouri - Columbia on 03/18/13. For personal use only.

roimmunol. 111:64–67 effector cells. J. Neuroimmunol. 127:96–


318. Issazadeh S, Mustafa M, Ljungdahl Å, 105
Höjeberg B, Dagerlind Å, et al. 1995. 326. Sorensen TL, Tani M, Jensen J, Pierce V,
Interferon-γ , interleukin-4 and trans- Lucchinetti C, et al. 1999. Expression of
forming growth factor β in experimental specific chemokines and chemokine re-
autoimmune encephalomyelitis in Lewis ceptors in the central nervous system of
rats: dynamics of cellular mRNA expres- multiple sclerosis patients. J. Clin. Invest.
sion in the central nervous system and 103:807–15
lymphoid cells. J. Neurosci. Res. 40:579– 327. Sorensen TL, Sellebjerg F, Jensen CV,
90 Strieter RM, Ransohoff RM. 2001.
319. Sindern E. 2004. Role of chemokines and Chemokines CXCL10 and CCL2: differ-
their receptors in the pathogenesis of mul- ential involvement in intrathecal inflam-
tiple sclerosis. Front Biosci. 9:457–63 mation in multiple sclerosis. Eur. J. Neu-
320. Trebst C, Ransohoff RM. 2001. Investi- rol. 8:665–72
gating chemokines and chemokine recep- 328. Sindern E, Niederkinkhaus Y, Henschel
tors in patients with multiple sclerosis: op- M, Ossege LM, Patzold T, Malin JP. 2001.
portunities and challenges. Arch. Neurol. Differential release of β-chemokines in
58:1975–80 serum and CSF of patients with relapsing-
321. Calabresi PA, Martin R, Jacobson S. remitting multiple sclerosis. Acta Neurol.
1999. Chemokines in chronic progres- Scand. 104:88–91
sive neurological diseases: HTLV-1 asso- 329. Kivisakk P, Trebst C, Liu Z, Tucky
ciated myelopathy and multiple sclerosis. BH, Sorensen TL, et al. 2002. T-cells in
J. Neurovirol. 5:102–8 the cerebrospinal fluid express a similar
322. Strunk T, Bubel S, Mascher B, Schlenke repertoire of inflammatory chemokine re-
P, Kirchner H, Wandinger KP. 2000. In- ceptors in the absence or presence of CNS
creased numbers of CCR5+ interferon- inflammation: implications for CNS traf-
γ - and tumor necrosis factor-α-secreting ficking. Clin. Exp. Immunol. 129:510–18
T lymphocytes in multiple sclerosis pa- 330. Giunti D, Borsellino G, Benelli R, March-
tients. Ann. Neurol. 47:269–73 ese M, Capello E, et al. 2003. Phenotypic
323. Misu T, Onodera H, Fujihara K, Mat- and functional analysis of T cells homing
sushima K, Yoshie O, et al. 2001. into the CSF of subjects with inflamma-
Chemokine receptor expression on T cells tory diseases of the CNS. J. Leukoc. Biol.
in blood and cerebrospinal fluid at re- 73:584–90
lapse and remission of multiple scle- 331. Trebst C, Sorensen TL, Kivisakk P, Cath-
rosis: imbalance of Th1/Th2-associated cart MK, Hesselgesser J, et al. 2001.
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

746 SOSPEDRA  MARTIN

CCR1+ /CCR5+ mononuclear phagocytes onal transection in the lesions of multi-


accumulate in the central nervous system ple sclerosis. N. Engl. J. Med. 338:278–
of patients with multiple sclerosis. Am. J. 85
Pathol. 159:1701–10 340. Karnezis T, Mandemakers W, McQual-
332. Boven LA, Montagne L, Nottet HS, De ter JL, Zheng B, Ho PP, et al. 2004. The
Groot CJ. 2000. Macrophage inflamma- neurite outgrowth inhibitor Nogo A is in-
tory protein-1α (MIP-1α), MIP-1β, and volved in autoimmune-mediated demyeli-
RANTES mRNA semiquantification and nation. Nat. Neurosci. 7:736–44
protein expression in active demyelinat- 341. Whitaker JN, Kirk KA, Herman PK, Zhou
ing multiple sclerosis (MS) lesions. Clin. SR, Goodin RR, et al. 1992. An immuno-
Exp. Immunol. 122:257–63 chemical comparison of human myelin
333. Simpson J, Rezaie P, Newcombe J, basic protein and its modified, citrulli-
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

Cuzner ML, Male D, Woodroofe MN. nated form, C8. J. Neuroimmunol. 36:
by University of Missouri - Columbia on 03/18/13. For personal use only.

2000. Expression of the β-chemokine re- 135–46


ceptors CCR2, CCR3 and CCR5 in mul- 342. Wood DD, Bilbao JM, O’Connors P,
tiple sclerosis central nervous system tis- Moscarello MA. 1996. Acute multiple
sue. J. Neuroimmunol. 108:192–200 sclerosis (Marburg type) is associated
334. Elhofy A, Kennedy KJ, Fife BT, Kar- with developmentally immature myelin
pus WJ. 2002. Regulation of experi- basic protein. Ann. Neurol. 40:18–24
mental autoimmune encephalomyelitis by 343. Bruck W, Lucchinetti C, Lassmann H.
chemokines and chemokine receptors. Im- 2002. The pathology of primary progres-
munol. Res. 25:167–75 sive multiple sclerosis. Mult. Scler. 8:93–
335. Fife BT, Huffnagle GB, Kuziel WA, Kar- 97
pus WJ. 2000. CC chemokine receptor 2 344. Werner P, Pitt D, Raine CS. 2001. Multi-
is critical for induction of experimental ple sclerosis: altered glutamate homeosta-
autoimmune encephalomyelitis. J. Exp. sis in lesions correlates with oligodendro-
Med. 192:899–905 cyte and axonal damage. Ann. Neurol. 50:
336. Rottman JB, Slavin AJ, Silva R, Weiner 169–80
HL, Gerard CG, Hancock WW. 2000. 345. Johnson KP, Brooks BR, Cohen JA, Ford
Leukocyte recruitment during onset of ex- CC, Goldstein J, et al. 1995. Copoly-
perimental allergic encephalomyelitis is mer 1 reduces relapse rate and improves
CCR1 dependent. Eur. J. Immunol. 30: disability in relapsing-remitting multiple
2372–77 sclerosis: results of a phase III multicen-
337. Tran EH, Kuziel WA, Owens T. 2000. In- ter, double-blind placebo-controlled trial.
duction of experimental autoimmune en- Neurology 45:1268–76
cephalomyelitis in C57BL/6 mice defi- 346. Arnon R. 1996. The development of Cop1
cient in either the chemokine macrophage (Copaxone), an innovative drug for the
inflammatory protein-1α or its CCR5 treatment of multiple sclerosis: personal
receptor. Eur. J. Immunol. 30:1410– reflections. Immunol. Lett. 50:1–15
15 347. Ziemssen T, Kumpfel T, Klinkert WE,
338. Flugel A, Berkowicz T, Ritter T, Labeur Neuhaus O, Hohlfeld R. 2002. Glatiramer
M, Jenne DE, et al. 2001. Migratory activ- acetate-specific T-helper 1- and 2-type
ity and functional changes of green fluo- cell lines produce BDNF: implications for
rescent effector cells before and during ex- multiple sclerosis therapy. Brain-derived
perimental autoimmune encephalomyeli- neurotrophic factor. Brain 125:2381–
tis. Immunity 14:547–60 91
339. Trapp BD, Peterson J, Ransohoff RM, 348. Chen M, Valenzuela RM, Dhib-Jalbut S.
Rudick R, Mork S, Bo L. 1998. Ax- 2003. Glatiramer acetate-reactive T cells
19 Feb 2005 14:11 AR AR239-IY23-21.tex XMLPublishSM (2004/02/24) P1: JRX

IMMUNOLOGY OF MULTIPLE SCLEROSIS 747

produce brain-derived neurotrophic fac- cination in multiple sclerosis: results of a


tor. J. Neurol. Sci. 215:37–44 preliminary study. J. Neurol. 249:212–18
349. Miller DH, Khan OA, Sheremata WA, 355. Vandenbark AA, Chou YK, Whitham R,
Blumhardt LD, Rice GP, et al. 2003. A Mass M, Buenafe A, et al. 1996. Treat-
controlled trial of natalizumab for relaps- ment of multiple sclerosis with T-cell re-
ing multiple sclerosis. N. Engl. J. Med. ceptor peptides: results of a double-blind
348:15–23 pilot trial. Nat. Med. 2:1109–15
350. Coles AJ, Wing MG, Molyneux P, 356. Killestein J, Olsson T, Wallstrom E,
Paolillo A, Davie CM, et al. 1999. Svenningsson A, Khademi M, et al.
Monoclonal antibody treatment exposes 2002. Antibody-mediated suppression of
three mechanisms underlying the clinical Vβ5.2/5.3+ T cells in multiple sclerosis:
course of multiple sclerosis. Ann. Neurol. results from an MRI-monitored phase II
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

46:296–304 clinical trial. Ann. Neurol. 51:467–74


by University of Missouri - Columbia on 03/18/13. For personal use only.

351. Youssef S, Stuve O, Patarroyo JC, Ruiz 357. Mancardi GL, Saccardi R, Filippi M, Gua-
PJ, Radosevich JL, et al. 2002. The HMG- landi F, Murialdo A, et al. 2001. Au-
CoA reductase inhibitor, atorvastatin, pro- tologous hematopoietic stem cell trans-
motes a Th2 bias and reverses paralysis in plantation suppresses Gd-enhanced MRI
central nervous system autoimmune dis- activity in MS. Neurology 57:62–68
ease. Nature 420:78–84 358. Burt RK, Cohen BA, Russell E, Spero
352. Vollmer T, Key L, Durkalski V, Tyor W, K, Joshi A, et al. 2003. Hematopoi-
Corboy J, et al. 2004. Oral simvastatin etic stem cell transplantation for progres-
treatment in relapsing-remitting multiple sive multiple sclerosis: failure of a total
sclerosis. Lancet 363:1607–8 body irradiation-based conditioning regi-
353. Weber F, Polak T, Gunther A, Kubuschok men to prevent disease progression in pa-
B, Janovskaja J, et al. 1998. Synergistic tients with high disability scores. Blood
immunomodulatory effects of interferon- 102:2373–78
β1b and the phosphodiesterase inhibitor 359. Nash RA, Bowen JD, McSweeney PA,
pentoxifylline in patients with relapsing- Pavletic SZ, Maravilla KR, et al. 2003.
remitting multiple sclerosis. Ann. Neurol. High-dose immunosuppressive therapy
44:27–34 and autologous peripheral blood stem cell
354. Zhang JZ, Rivera VM, Tejada-Simon MV, transplantation for severe multiple sclero-
Yang DY, Hong J, et al. 2002. T cell vac- sis. Blood 102:2364–72
P1: JRX
February 21, 2005 13:52 Annual Reviews AR239-FM

Annual Review of Immunology


Volume 23, 2005

CONTENTS
FRONTISPIECE—Tadamitsu Kishimoto x
INTERLEUKIN-6: FROM BASIC SCIENCE TO MEDICINE–40 YEARS IN
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

IMMUNOLOGY, Tadamitsu Kishimoto 1


by University of Missouri - Columbia on 03/18/13. For personal use only.

TNF/TNFR FAMILY MEMBERS IN COSTIMULATION OF T CELL


RESPONSES, Tania H. Watts 23
DEVELOPMENT AND REGULATION OF CELL-MEDIATED IMMUNE
RESPONSES TO THE BLOOD STAGES OF MALARIA: IMPLICATIONS
FOR VACCINE RESEARCH, Michael F. Good, Huji Xu, Michelle Wykes,
and Christian R. Engwerda 69
THE T CELL RECEPTOR: CRITICAL ROLE OF THE MEMBRANE
ENVIRONMENT IN RECEPTOR ASSEMBLY AND FUNCTION,
Matthew E. Call and Kai W. Wucherpfennig 101
CHEMOKINES, SPHINGOSINE-1-PHOSPHATE, AND CELL MIGRATION IN
SECONDARY LYMPHOID ORGANS, Jason G. Cyster 127
MARGINAL ZONE B CELLS, Shiv Pillai, Annaiah Cariappa,
and Stewart T. Moran 161
HOW NEUTROPHILS KILL MICROBES, Anthony W. Segal 197
NK CELL RECOGNITION, Lewis L. Lanier 225
IPC: PROFESSIONAL TYPE 1 INTERFERON-PRODUCING CELLS AND
PLASMACYTOID DENDRITIC CELL PRECURSORS, Yong-Jun Liu 275
TYPE I INTERFERONS (α/β) IN IMMUNITY AND AUTOIMMUNITY,
Argyrios N. Theofilopoulos, Roberto Baccala, Bruce Beutler,
and Dwight H. Kono 307
PENTRAXINS AT THE CROSSROADS BETWEEN INNATE IMMUNITY,
INFLAMMATION, MATRIX DEPOSITION, AND FEMALE FERTILITY,
Cecilia Garlanda, Barbara Bottazzi, Antonio Bastone,
and Alberto Mantovani 337
MAINTENANCE OF SERUM ANTIBODY LEVELS, Rudolf A. Manz,
Anja E. Hauser, Falk Hiepe, and Andreas Radbruch 367
CATERPILLER: A NOVEL GENE FAMILY IMPORTANT IN IMMUNITY,
CELL DEATH, AND DISEASES, Jenny P-Y. Ting
and Beckley K. Davis 387

v
P1: JRX
February 21, 2005 13:52 Annual Reviews AR239-FM

vi CONTENTS

B CELL SIGNALING AND TUMORIGENESIS, Hassan Jumaa,


Rudolf W. Hendriks, and Michael Reth 415
THE NOD MOUSE: A MODEL OF IMMUNE DYSREGULATION,
Mark S. Anderson and Jeffrey A. Bluestone 447
ANTIGEN-SPECIFIC MEMORY B CELL DEVELOPMENT,
Louise J. McHeyzer-Williams and Michael G. McHeyzer-Williams 487
THE B7 FAMILY REVISITED, Rebecca J. Greenwald, Gordon J. Freeman,
and Arlene H. Sharpe 515
TEC FAMILY KINASES IN T LYMPHOCYTE DEVELOPMENT AND
FUNCTION, Leslie J. Berg, Lisa D. Finkelstein, Julie A. Lucas,
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org

and Pamela L. Schwartzberg 549


by University of Missouri - Columbia on 03/18/13. For personal use only.

MOLECULAR GENETICS OF T CELL DEVELOPMENT, Ellen V. Rothenberg


and Tom Taghon 601
UNDERSTANDING PRESENTATION OF VIRAL ANTIGENS TO CD8+
T CELLS IN VIVO: THE KEY TO RATIONAL VACCINE DESIGN,
Jonathan W. Yewdell and S.M. Mansour Haeryfar 651
IMMUNOLOGY OF MULTIPLE SCLEROSIS, Mireia Sospedra
and Roland Martin 683
MAST CELLS AS “TUNABLE” EFFECTOR AND IMMUNOREGULATORY
CELLS: RECENT ADVANCES, Stephen J. Galli, Janet Kalesnikoff,
Michele A. Grimbaldeston, Adrian M. Piliponsky, Cara M.M. Williams,
and Mindy Tsai 749
NETWORK COMMUNICATIONS: LYMPHOTOXINS, LIGHT, AND TNF,
Carl F. Ware 787
ROLE OF C5A IN INFLAMMATORY RESPONSES, Ren-Feng Guo
and Peter A. Ward 821
DNA DEGRADATION IN DEVELOPMENT AND PROGRAMMED CELL
DEATH, Shigekazu Nagata 853
TOWARD AN UNDERSTANDING OF NKT CELL BIOLOGY: PROGRESS
AND PARADOXES, Mitchell Kronenberg 877
MACROPHAGE RECEPTORS AND IMMUNE RECOGNITION,
P.R. Taylor, L. Martinez-Pomares, M. Stacey, H-H. Lin, G.D. Brown,
and S. Gordon 901
REGULATION OF LYMPHOID DEVELOPMENT, DIFFERENTIATION, AND
FUNCTION BY THE NOTCH PATHWAY, Ivan Maillard, Terry Fang,
and Warren S. Pear 945
CELL BIOLOGY OF ANTIGEN PROCESSING IN VITRO AND IN VIVO,
E. Sergio Trombetta and Ira Mellman 975
P1: JRX
February 21, 2005 13:52 Annual Reviews AR239-FM

CONTENTS vii

INDEXES
Subject Index 1029
Cumulative Index of Contributing Authors, Volumes 13–23 1065
Cumulative Index of Chapter Titles, Volumes 13–23 1072

ERRATA
An online log of corrections to Annual Review of Immunology chapters
may be found at http://immunol.annualreviews.org/errata.shtml
Annu. Rev. Immunol. 2005.23:683-747. Downloaded from www.annualreviews.org
by University of Missouri - Columbia on 03/18/13. For personal use only.

You might also like