REVIEW
Mechanisms of mesenchymal stromal cell
immunomodulation
Karen English
Multipotent mesenchymal stromal cells (MSCs) have generated considerable interest in the fields of regenerative medicine, cell
therapy and immune modulation. Over the past 5 years, the initial observations that MSCs could enhance regeneration and
modulate immune responses have been significantly advanced and we now have a clearer picture of the effects that MSCs have
on the immune system particularly in the context of inflammatory-mediated disorders. A number of mechanisms of action have
been reported in MSC immunomodulation, which encompass the secretion of soluble factors, induction of anergy, apoptosis,
regulatory T cells and tolerogenic dendritic cells. It is clear that MSCs modulate both innate and adaptive responses
and evidence is now emerging that the local microenvironment is key in the activation or licensing of MSCs to become
immunosuppressive. More recently, studies have suggested that MSCs have the capacity to sense their environment and have a
role in pathogen clearance in conjunction with the resolution of insult or injury. This review focuses on the mechanisms of MSC
immunomodulation discussing the multistep process of MSC localisation at sites of inflammation, the cross talk between MSCs
and the local microenvironment as well as the subsequent mechanisms of action used to resolve inflammation.
Immunology and Cell Biology (2013) 91, 19–26; doi:10.1038/icb.2012.56; published online 23 October 2012
Keywords: mesenchymal stromal cells; immune modulation; cell contact; soluble factors; tolerogenic DC; Treg
INTRODUCTION TO MULTIPOTENT MESENCHYMAL STROMAL
CELLS (MSCs)
Although MSCs were first identified in the bone marrow ov er 40 years
ago,
1–3
their intriguing capacity to modulate the immune response was
only identified 30 years later with reports that MSCs could suppress
T-cell proliferation in vitro
4–5
and prolong the survival of allogeneic
skin grafts in vivo.
4
Since then our understanding of how MSCs
mediate their immune-suppressive effects has significantly advanced. A
large number of in vitro studies have provided insight into the effects
of MSCs on both innate and adaptive immune responses. Specifically,
MSCs have the capacity to interfere with many components of the
innate immune system including complement, Toll-lik e r eceptor
(TLR) signalling, macrophages, dendritic cells, neutrophils, mast
cells and natural killer cells.
6–11
In regard to the adaptive immune
response, MSCs can directly inhibit T-cell function, shift the T helper
lymphocyte balance and induce functional regulatory T cells
(Tregs).
12–14
Less is known about the effect of MSCs on B cells but
some studies suggest that MSCs can also modulate B-cell proliferation
and function.
15
Mor eov er, MSCs have been demonstrated to exert
their anti-inflammatory effects in a number of in vivo models
including graft versus host disease (GvHD), experimental
autoimmune encephalomyelitis, inflammatory bowel disease and
allergic airway disease.
10,12,16–18
Based on this wealth of data
supporting an anti-inflammatory and pro-reparative role for MSCs,
these cells have been used for the treatment of various inflammatory
diseases in clinical trials (clinicaltrials.gov).
Although a small number of studies have identified the mechan-
isms involved in MSC protective effects, for the most part we do not
fully understand how MSCs mediate their effect in vivo. Given that
MSCs are already being utilised for the treatment of patients in
clinical trials, it is imperative that the field gains a better under-
standing of exactly how MSCs mediate their effects in these different
inflammatory disorders to ensure that MSC therapy can be utilised
with optimal therapeutic efficacy and safety.
Considerable data support an anti-inflammatory effect of MSCs on
immune cells, however, there are conflicting reports, which suggest
that MSCs enhance immune cell survival and function. It is crucial
that we delineate these disparate findings to ensure that MSC therapy
does not exacerbate inflammatory disease. In the context of patho-
genic insult or excessive sterile inflammation (inflammation in the
absence of micro-organisms), it seems logical that MSCs would
orchestrate the clearance of pathogens or necrotic cells associated with
sterile inflammation through promotion of immune cell survival and
function followed by resolution of inflammation through suppressive
mechanisms. Combined with the idea that MSCs have the capacity to
sense their environment, this suggests that MSCs are receptive to local
biochemical signals and deploy a stepwise strategic approach to
resolving inflammation and encouraging tissue repair.
Cellular Immunology Group, Institute of Immunology, Department of Biology, National University of Ireland Maynooth, Co., Kildare, Ireland
Correspondence: Dr K English, Cellular Immunology Group, Institute of Immunology, Department of Biology, National University of Ireland Maynooth, Co., Kildare, Ireland.
E-mail: karen.english@nuim.ie
Received 13 September 2012; revised 18 September 2012; accepted 19 September 2012; published online 23 October 2012
Immunology and Cell Biology (2013) 91, 19–26
&
2013 Australasian Society for Immunology Inc. All rights reserved 0818-9641/13
www.nature.com/icb
In this review, the major mechanisms of action mediated by MSCs
in modulating inflammation and the capacity of MSCs to sense and
react to their local microenvironment are discussed.
MSC MIGRATION TO SITES OF INFLAMMATION/INSULT/
INJURY
The immune system is adept at recognising and responding to
pathogenic insult
19
through specialised pathogen recognition
receptors. These pathogen recognition receptors not only recognise
pathogen-associated molecular patterns but also molecules associated
with damaged cells and tissues (so called DAMPs) during sterile
inflammation.
20
Following recognition, a signalling cascade triggers
the release of complement components, acute phase proteins, pro-
inflammatory cytokines and chemokines, which lead to the
recruitment of innate immune cells such as neutrophils and
macrophages.
21
It seems reasonable to assume that endogenous or
adoptively transferred MSCs would also respond to these cues and
migrate to the site of inflammation. However, some of the key
limitations for MSC research have been the lack of specific markers
and useful tracking studies to examine the migration and engraftment
of MSCs in vivo. Currently, little is known about the capacity for
MSCs to migrate to sites of tissue inflammation, and indeed in some
cases MSCs have been shown to exert their immunosuppressive effect
from a distance. For example, human bone marrow MSCs trapped in
the lung microvasculature as microemboli secrete tumour necrosis
factor (TNF)-a-stimulated gene/protein 6 (TSG-6), which suppresses
the early immune response in murine models of myocardial
infarction
22
and corneal injury.
23
MSCs have the capacity to
migrate in vitro to a number of complement proteins, growth
factors, cytokines and chemokines including complement
component 1 subcomponent q (C1q),
24
C3a and C5a,
25
stromal cell
derived factor-1 (SDF-1),
26,27
platelet derived growth factor AB,
insulin like growth factor-1, epidermal growth factor and
hepatocyte growth factor,
27
interleukin (IL)-1b,
28
TNF-a,CC
chemokine ligand 5
29
and macrophage derived chemokine.
27
Moreover, short-term exposure to TLR-3 and 4 ligands significantly
enhanced MSC migration in vitro,
29
and pre-stimulation of MSCs
with TNF-a enhanced migration towards chemokines through
upregulation of chemokine receptors.
27
In vivo studies have
demonstrated that MSCs upregulated expression of CXCR4 and
CCR7 and selectively migrated to ischaemic kidney in response to
SDF-1a or to wounded skin in response to secondary lymphoid tissue
chemokine in vivo.
26,30
Thus, it seems that MSCs have the capacity to
migrate in response to signals produced by inflamed tissues and these
signals may have a role in determining the function of MSCs, be that
promotion of pathogen clearance or suppression of inflammation.
MSC LICENSING/ACTIVATION AT SITES OF INFLAMMATION
An emerging concept in the MSC field is that MSCs are not
spontaneously immunosuppressive but require ‘licensing’ or activa-
tion to exert their immunosuppressive effects. In particular, interferon
(IFN)-g,TNF-a or IL-1b have been demonstrated to be required for
the activation of MSCs to modulate immune responses.
18,31,32
Importantly, differential regulation of a number of MSC immuno-
modulatory molecules (including indoleamine 2,3-dioxygenase
(IDO), prostaglandin E-2 (PGE-2), transforming growth factor
(TGF)-b, TSG-6 and nitric oxide (NO)) by pro-inflammatory
cytokines has been observed.
18,33
In addition to pro-inflammatory
cytokines, TLR signalling has also been implicated in the licensing of
MSCs. TLR3 and TLR4 activation of MSCs enhanced MSC
immunosuppression in vitro through IDO induction via IFN-b and
protein kinase R signalling.
34
In addition, TLR2 activation of human
bone marrow MSCs resulted in the upregulation of the immune-
suppressive protein galectin-3.
35
In contrast, Liotta et al.
36
observed
the opposite effect, with TLR3 and TLR4 ligand binding leading to
the downregulation of Jagged-1 and the failure of MSCs to modulate
T-cell responses. The discrepancy in these studies may be resolved by
the findings that TLR3 and TLR4 may differentially licence MSCs;
with TLR4 priming inducing a pro-inflammatory phenotype and
secretion of IL-6, IL-8 and TGF-b
29
a process reportedly enhanced by
co-stimulation with IFN-g.
37
In contrast, TLR3 priming induced anti-
inflammatory MSCs (producing IDO, PGE-2, IL-4 and IL-1RA).
29
Further research is required on the effects of TLR activation on MSCs,
however, these finding indicate that MSCs are receptive to
environmental cues and may have the capacity to promote
pathogen clearance or immune suppression. MSCs have the
capacity to trigger complement activation, a process that would
normally result in lysis of the complement activating cell. Surprisingly,
high levels of C3 activation correlate with enhanced immuno-
suppressive capacity of MSCs.
38
Importantly, MSCs express CD59, a
complement regulatory protein, and also release complement factor
H, which protects them from complement lysis.
38,39
Overall, it seems
clear that the local microenvironment significantly influences MSC
activation and immunoregulatory function (Figure 1).
MECHANISMS OF MSC IMMUNOMODULATION
MSCs possess an arsenal of immunosuppressive mechanisms, which
canbedeployedinthemodulationofinflammation.Twovery
interesting paradigms have recently been proposed which postulate
that (1) MSCs have sentinel functions that allow them to sense their
microen vironment and act accordingly
40
and (2) MSCs become
polarised towards either a pro-inflammatory phenotype or an
immunosuppressive phenotype depending on the TLR signals
received.
29
Together, these concepts help to resolve some of the
conflicting data showing that in some cases MSCs enhance immune
cell survival and function and in others they inhibit inflammation and
encourage repair.
MSC immunomodulation takes place over a multistage process
involving (1) MSC responsiveness to inflammation and possible
migration to the site of tissue injury, (2) licensing or activation of
MSCs, (3) promotion of pathogen clearance if required and
(4) modulation of inflammation (Figure 1). As discussed earlier,
MSCs may exert their immunosuppressive effects at a distance,
23,41
but many studies demonstrate that MSCs require contact with
immune cells to exert their effects. Two very elegant studies have
shown that mouse MSCs can also chemoattract T lymphocytes
through the secretion of CXCL9, CXCL10,
18
and CCL2 (monocyte
chemotactic protein-1).
12
The production of CXCL9 and CXCL10 by
MSCs was induced by IFN-g,TNF-a,IL-1a or IL-1b,
18
whereas CCL2
secretion by MSCs was regulated by the TNF receptor superfamily,
member 6 (FAS).
12
Once MSCs have attracted effector T cells, this
provides a platform for MSC contact with effector T cells and
facilitates the direct immunomodulation of the T cells via production
of NO by MSCs
18
or FAS/FASL (Fas ligand)-induced apoptosis.
12
Although differences have been reported in the mechanism of action
by mouse and human MSCs,
32
the study by Akiyama et al.
12
definitively shows that the data obtained in the mouse model
(demonstrating higher levels of apoptotic effector T cells and
increased numbers of Treg as well as elevated serum levels of
TGF-b) could also be observed in systemic sclerosis patients treated
with MSCs, indicating that basic research findings on the mechanisms
of action of MSCs can be extrapolated to the clinic.
Mechanisms of MSC immune modulation
KEnglish
20
Immunology and Cell Biology
MSCs RE-EDUCATE MONOCYTES/MACROPHAGES IN THE
CONTEXT OF TISSUE REPAIR
In response to activating signals, macrophages become polarised into
either a classical M1 phenotype (pro-inflammatory, stimulated by
TLR engagement or IFN-g) or an alternative M2 phenotype (anti-
inflammatory, stimulated by IL-4/IL-13). M1 macrophages are
characterised by their high level production of pro-inflammatory
cytokines including TNF-a and IL-1b whereas M2 are commonly
associated with secretion of IL-10.
42
Given that macrophages have key
roles at sites of inflammation,
21
it seems plausible that MSCs could
interact with these cells and may even influence their polarisation.
Indeed human MSCs derived from bone marrow, umbilical cord and
cord blood have the capacity to modulate monocyte function
in vitro.
7
Moreover, an array of studies demonstrate that MSCs
alternatively activate macrophages, downregulating the production of
TNF-a,IL-1a, IL-6 and IL-12p70 and increasing the production of
IL-10 and enhancing the phagocytic activity.
6,9,10,43–45
In addition,
two separate studies have shown that MSCs impair microglial
activation and alternatively activate microglia to produce IGF-1,
galectin-3 and to express factors associated with a neuroprotective
phenotype.
46,47
Production of IDO and PGE-2 have been implicated
in MSC modulation of macrophages,
43,44
and MSCs cultured in
three-dimensional spheroids also have the capacity to reprogram
macrophages through the production of PGE-2.
48
Perhaps the most
notable studies are those that build a stepwise picture of how MSCs
orchestrate macrophage polarisation and the influence the local
microenvironment has on that process. First of these is the study
by Nemeth et al.
10
, in which MSCs ameliorate sepsis through
alternative activation of macrophages. The authors elegantly
elucidate the mechanisms of action, showing that LPS and TNF-a
activate TLR4 and TNFR1 on MSCs to activate nuclear factor-kB
signalling. This in turn leads to the expression of cycloxygenase
(COX)-2 and synthesis of PGE-2 by MSCs, which bind EP2 and EP4
receptors on macrophages resulting in increased production of IL-10
and facilitating the resolution of inflammation. The second study
utilised a zymosan-induced peritonitis model in which MSCs exerted
anti-inflammatory effects through the production of TSG-6. In this
case, zymosan, a TLR2 agonist, stimulated the activation of TLR2 and
nuclear factor-kB signalling in macrophages. TNF-a secreted by the
activated macrophages leads to the production of TSG-6 by MSCs.
TSG-6 is then thought to limit TLR2/nuclear factor-kB signalling
through direct interaction with CD44 expressed on the macrophage
to initiate a negative feedback loop inhibiting the inflammatory
response.
6
On the whole, micro-environmental cues present at the site
of MSC activation seem to determine the particular mechanism of
actions deployed by MSCs in modulating the immune response and
resolving inflammation.
MSCs PRODUCE IMMUNOSUPPRESSIVE SOLUBLE FACTORS
MSC modulation of immune responses is mediated through an array
of mechanisms, however, most of these mechanisms involve the
production of immunosuppressive factors. The majority of these
soluble factors are not constitutively produced by MSCs but in fact
are induced through the licensing or activation of MSCs as described
earlier. Herein, the key MSC-derived soluble factors and their modes
of action are discussed.
Figure 1 Activation by inflammatory mediators determines the effector mechanisms utilised by MSCs. MSCs are licensed/activated directly through TLR
stimulation (pathogen-associated molecular patterns; (lipopolysaccharide (LPS), double-stranded RNA (dsRNA) and DAMPS) or indirectly by activated
macrophages producing pro-inflammatory cytokines (IFN-g,TNF-a and IL-1b) during tissue injury. MSCs have the capacity to trigger complement activation
(C3), which in normal circumstances would target MSCs for complement lysis. However, complement activation of MSCs has been correlated with increased
immunosuppressive activity. Importantly, MSCs express CD59 (a complement regulatory protein), and release complement factor H, which protects MSCs
from complement lysis. Depending on the stimulus received, MSCs are thought to have the capacity to promote pathogen clearance or immune modulation
(during the effector phase). MSCs may promote pathogen clearance through secretion of pro-inflammatory cytokines (IL-6 and IL-8), polarisation of pro-
inflammatory M1 macrophages, anti-microbial activity,
100
through the enhanced survival and function of neutrophils or through other as of yet unidentified
mechanisms. MSCs promote immune modulation through the secretion of immunosuppressive soluble factors (IDO, PGE-2, TGF-b, NO, TSG-6 and factor H
among others), promotion of alternatively activated anti-inflammatory M2 macrophages, tolerogenic DC (Tol DC) or Treg. In response to pathogenic stimuli,
MSCs may initially promote the clearance of pathogens followed by suppression of the immune response in the resolution of inflammation. A full colour
version of this figure is available at the Immunology and Cell Biology journal online.
Mechanisms of MSC immune modulation
KEnglish
21
Immunology and Cell Biology
INDOLEAMINE 2,3-DIOXYGENASE
IDO is an enzyme that catabolises tryptophan (an essential amino acid
required for T -cell proliferation) into kynurenine metabolites that
regulate T-cell proliferation.
49
MSC expression of IDO is induced by
stimulation with IFN-g
31,33,50
or through stimulation with TLR3 and
TLR4 ligands, which involve the activation of protein kinase R,
autocrine IFN-b signalling and activation of signal transducer and
activator of transcription-1/interferonregulatoryfactor(STAT-1/IRF).
34
In contrast, Waterman et al.
29
showed that TLR3 but not TLR4
priming induced IDO production, while another study reported that
TLR3 and TLR4 activation of MSCs abrogated the immunosuppressive
effect of MSCs.
36
Differences in experimental set-up , timing, cell types
and concentration of ligands may explain the differences observed,
however, a clear definition of the precise parameters is still awaited.
MSC-derived IDO has been associated with the re-education of
immune cells including macrophage polarisation to the anti-
inflammatory M2 phenotype,
43
induction of tolerogen ic dendritic
cells (DCs) and Treg in vivo
13
as well as promoting a Th1–Th2
switch.
13
In addition, MSC production of IDO directly impacts T -c ell
differentiation
51
and T and natural killer cell proliferation.
11,31,33,50
Blocking studies using the inhibitor 1-methyl- L-tryptophan or IDO
knockout MSCs have demonstrated the important role for IDO in
MSC suppression of these immune responses.
13,31,33,43,50,52
IDO is also
known to have a role in microbial defence and it may be plausible that
MSC-derived IDO could enhance microbial clearance as well as
inhibition of immune responses.
PROSTAGLANDIN E-2
PGE-2 is a rapidly released and short acting small lipid mediator
known to have a role in immune modulation. The pathway for
prostaglandin synthesis inv olves the CO X enzyme (COX-1 and COX-2)
production of prostaglandin H2 from arachidonic acid, followed by
con vers ion of prostaglandin H2 into prostaglandins via prostaglandin
synthases.
53
COX-2isconstitutivelyexpressedbyMSCsandCOX-2
inhibitor studies have shown that CO X-2 is required for the production
of PGE-2 by MSCs.
33,50,54
MSCs constitutively produce PGE-2 that is
enhanced by stimulation with IFN-g and TNF-a
33
as well as by TLR3
but not TRL4 ligands.
29
Furthermore, at least one study has identified
that IL-6 is required for the production of PGE-2 and subsequent
inhibition of inflammation in an experimental arthritis model,
55
however, this may be specific to an arthritic environment. A large
body of data support the role of MSC-derived PGE-2 in the
suppression of T-cell activation and proliferation both in vitro and
in vivo.
33,50,56,57
In addition to T lymphocyte-specific effects, PGE-2
produc ed by MSCs has an important role in MSC reprogramming
of macrophages
10,44,48
and DCs.
58
More recently , MSCs have been
shown to inhibit mast cell function through a COX-2-dependent
mechanism.
59
Importantly, the biochemical events involve d in the
mechanism of action of MSC-derived PGE-2 have been somewhat
clarified by separate studies, which revealed that PGE-2 binds to either
EP2 or EP4 on macrophages altering their phenotype
10
or to EP4 on
CD4
þ
T cells inhibiting Th17 differentiation.
54
Further elucidation of
the interactions and signalling pathways involved will provide a better
insight into how best to utilise this particular constitutively active
property of MSCs in clinical application.
TNF-a STIMULA TED GENE/PROTEIN 6
TSG-6 is an IL-1/TNF -indu cible protein with anti-inflammatory
properties. TNF-a stimulation of MSCs leads to the production of
significant amounts of TSG-6. The observation that the majority of
intravenously injected MSCs were found localised in the lung and not
at the site of inflammation in a mouse model of my ocardial infarction
promoted the hypothesis that MSCs produc ed a potent anti-inflam-
matoryagentthatcouldactsystemically.Themoleculeinvolvedwas
identified as TSG-6 through microarray analysis of mouse lungs after
infusion of human MSCs.
41
MSC-derivedTSG-6wasshownto
mediateprotectiveeffectsinmurinemodelsofmyocardial
infarction,
41
corneal injury ,
23
allogeneic corneal transplant
60
as well as
zymosan-induced peritonitis.
6
In all models, TSG-6 inhibited the early
inflammatory response, and specifically neutrophil infiltration and pro-
inflammatory cytokines. Enhanced allogeneic corneal graft survival was
associated with decreased activation of antigen-pre senting cells in the
graft and also in the draining lymph node.
60
The anti-inflammatory
effect of TSG-6 in the corneal injury model was dose dependent and
the inhibition of the early immune response significantly reduced the
neovascularisation and subsequent development of opacity .
23
MSC-
derived TSG-6 in the myocar dial infarct model reduced infarct size and
improv ed left ventricular function.
41
Clear evidence for the importance
of MSC-derived TSG-6 was provided in systems using small interfering
RNA knockdown or infusion of recombinant TSG-6 in place of MSCs.
Ofthesestudies,perhapsthezymosan-inducedperitonitismodel
provides most insight into the cascade of events involv ed in MSC
production of TSG-6 and subsequent suppression of macrophages as
discussed above. Similar to the formation of microemboli in the lung,
6
MSCs can be cultured in spheroids in vitro and have been
demonstrated to repr ogram macrophages through production of
PGE-2,
48
it may be interesting to examine a collaborative role for
PGE-2 and TSG-6 in vivo.
NITRIC OXIDE
NO is produced as a result of the enzymatic reaction of inducible NO
synthase and has the capacity to inhibit T-cell proliferation and
induce T-cell apoptosis.
18
Stimulation of MSCs with IFN-g and TNF-a-
or IL-1-induced expression of inducible NO synthase in mouse
MSCs.
18
NO is a potent molecule that mediates its effects at close
proximity and two studies have elegantly shown that mouse MSCs
lure T cells through the production of chemokines for subsequent
suppression by local production of NO.
18,61
MSC-derived NO also
induced apoptosis of alloreactive T cells through suppression of
STAT-5 phosphorylation.
62
MSC production of NO enhanced cardiac
allograft survival,
63
attenuated delayed-type hypersensitivity responses
through induction of T-cell apoptosis
18,61
and prevented GvHD.
18
However, important species differences exist with regard to the role of
NO in immune modulation. It may be that the importance of NO is
more obvious in mouse MSCs compared with human.
32
Although
NO may not be useful in MSC therapy in the clinic—the studies
examining the role of NO have identified the very intriguing capacity
of MSCs to chemoattract T cells to facilitate close contact allowing the
provision of short acting molecules like NO.
18
MSCs ALTER THE HELPER T-CELL BALANCE (Th1/Th2/Th17)
CD4
þ
T helper cells become activated in response to pathogen- or
danger-associated signals. Depending on the threat encountered
CD4
þ
T cells (Th0) differentiate into various T-cell subsets with
distinct cytokine and gene expression profiles. The most common of
these are the Th1, Th2 and Th17 subsets.
64,65
CD4
þ
helper T cells
have an essential role in host defence against pathogens, and for the
most part, this component of the immune system is well regulated.
However, in some circumstances, excessive immune responses can
lead to significant tissue damage perhaps culminating in allergic or
autoimmune diseases such as asthma, type-1 diabetes or multiple
sclerosis.
16,66,67
MSCs have the capacity to modulate T-cell
Mechanisms of MSC immune modulation
KEnglish
22
Immunology and Cell Biology
proliferation and function
68–70
and in some cases MSCs mediate their
protective effect through shifting the balance from Th1-driven
responses to a more anti-inflammatory Th2 profile and vice versa
for Th2-driven pathologies. Evidence from in vitro studies as well as
in vivo models clearly demonstrate the ability for MSCs to shift the
balance from a pro-inflammatory Th1 phenotype secreting IFN-g and
TNF-a to a more anti-inflammatory Th2 profile secreting increased
levels of IL-4, IL-5, IL-10, IL-13 and the Th2 chemokine I-309.
66,67,71
The promotion of a switch towards a Th2 phenotype by MSCs was
associated with delayed onset of type-1 diabetes in NOD mice
67
and
amelioration of experimental autoimmune encephalomyelitis.
66
Contrastingly, in the context of allergic disease where Th2 cells
drive allergic pathology, MSCs decrease the production of Th2-
associated cytokines and enhance Th1 cytokine secretion
72,73
to create
a greater balance and provide protection in allergic airway disease
72
and improve the symptoms of patients with sclerodermatous chronic
GvHD.
73
MSCs also modulate Th17 differentiation
51,54
in favour of
IL-4-producing Th2 cells or the generation of Treg.
66,74
Both IDO and
PGE-2 have been implicated in MSC inhibition of Th17
differentiation,
51,68
and Duffy et al.
54
go a step further to elucidate
the steps involved in this mechanism in their system. Specifically,
contact-dependent COX-2 induction in MSCs leads to the production
of PGE-2 and direct inhibition through EP4.
54
Importantly, MSCs
can also mediate this effect through manipulation of the relative
plasticity of T cells in suppressing the Th17 transcription factor
retinoic-acid-receptor-related orphan receptor-g t(RORgt) and
upregulating Foxp3 to induce a Treg phenotype producing IL-10.
74
MSC production of the anti-inflammatory cytokine, TGF-b, has been
shown to have a partial role in shifting the balance of Th1/Th2/Th17
and Treg in a rat model of experimental autoimmune myasthenia
gravis.
14
Furthermore, MSCs re-educated Th1 cells acquired the
capacity to inhibit T-cell proliferation in vitro.
74
Thus, there is now
a considerable body of evidence to suggest that MSCs may provide
protection in autoimmune and allergic diseases through shifting the
balance of Th1/Th2 and Th17/Treg phenotypes.
MSCs INDUCE TOLEROGENIC DCs
The main function of dendritic cells is to act as sentinel cells and as
such to present antigens activating antigen-specific helper T cells.
These cells have a critical role in host defence and therefore in the
generation of immune responses. MSCs can interfere with the
development of both conventional and plasmacytoid DCs
75–77
but
also with the key features of DC function; migration, maturation and
antigen presentation
8
and an array of mechanisms have been
implicated in these effects. Most notably, MSCs downregulate
the expression of DC maturation markers including major
histocompatibility complex (MHC) class II, CD40, CD80 and
CD86
8,76–79
and modulate expression of the lymph node homing
chemokine receptor CCR7 in vitro
8
and in vivo.
80
Moreover, MSC-
mediated preservation of E-cadherin
8
expression by DCs fits with the
concept that MSCs may prevent DC homing to the local lymph node.
Interestingly, both soluble factors and contact-dependent signals have
been identified in MSC modulation of DC maturation markers. MSC
production of IL-6 has been shown to be involved in MSC
downregulation of maturation markers.
8,76,77
Conversely, Li et al.
78
reported that IL-6 was not required and showed that contact-
dependent Notch signalling was necessary for DC modulation. In
support of a role for Notch signalling in this scenario, Zhang et al.
found a partial role for contact-dependent Jagged-2 (a ligand of the
Notch signalling pathway) signalling in the generation of regulatory
DCs.
79
Theimportanceoftheoriginalin vitro data on MSC
modulation of CCR7 (English et al.
8
) and the subsequent hypothesis
that MSCs inhibit DC migration to the lymph node in vivo,
69,70
has
recently been supported.
80
Nevertheless, the exact mechanism utilised
by MSCs to achieve this effect remains to be elucidated.
Akin to the effects of MSCs on macrophage polarisation discussed
earlier, MSCs can also re-programme conventional lymphocyte
stimulatory DCs into anti-inflammatory DCs with a tolerogenic
phenotype.
13,58,78,79,81,82
DCs generated in the presence of MSCs
produce higher levels of anti-inflammatory cytokines including
IL-10 and lower levels of the pro-inflammatory cytokines IL-12 and
TNF-a. The encounter with MSCs also results in enhanced phagocytic
activity
79,82
typical of tolerogenic DCs. On a functional level,
tolerogenic DCs generated by MSCs inhibited delayed-type
hypersensitivity responses in vivo
79,82
and failed to induce activation
of CD4
þ
T cells (in vitro and in vivo).
58,80
Instead DCs that
encountered MSCs promoted the generation of antigen-specific
Treg in v itro.
78
The capacity of MSC educated DCs to induce a
state of tolerance in the context of solid organ transplantation is a
very exciting prospect and at least one in vivo study implicates MSC-
induced tolerogenic DCs in kidney allograft survival in the presence
of low-dose immunosuppression.
81
Although there are several
observations that MSCs induce tolerogenic DCs as well as clear
supporting evidence of the immunosuppressive or regulatory role
played by these DCs, the intricate details of how MSCs induce these
tolerogenic DCs is somewhat ambiguous. Given that IL-6 secreted by
MSCs has been shown to be partially involved in the downregulation
of DC maturation markers,
8,76,77
it initially seemed plausible that IL-6
might also be implicated in the generation of tolerogenic DCs.
However, two independent studies did not find a role for IL-6 but
instead demonstrated that PGE-2 and/or cell contact-dependent
activation of the Notch signalling pathway were required for MSC
induction of tolerogenic DCs.
58,78
A partial role for contact-
dependent activation of Jagged-2 was also suggested by Zhang
et al.
79
In further support of a contact-dependent mechanism,
Chiesa et al.
80
propose that MSCs induce tolerogenic DCs through
activation of AKT, which impaired nuclear factor-kB signalling, but
could not find a role for secreted IL-10. Finally, a recent publication
has shown that mouse embryonic fibroblast-derived MSCs generate
a novel population of IL-10-dependent tolerogenic DCs through an
IL-10-activated suppressor of cytokine signalling-3 (SOCS-3)-
dependent mechanism.
82
Overall, the mechanisms of action
mediated by MSCs in the generation of tolerogenic DCs are
extremely varied and complex and no doubt are influenced by the
context in which MSCs see DCs or DC precursors. Indeed, some of
these variations may be explained by differences between in vitro and
in vivo environments. The capacity for MSCs to induce tolerogenic
DCs is uncontested, however, we must endeavour to ask the right
questions and to fastidiously investigate the mechanisms of action
involved for a greater understanding of how best to utilise MSCs in
the clinic.
MSC INDUCTION OF TREG AND IMMUNE TOLERANCE
Tregs have an important role in the regulation of immune responses
and in the prevention of autoimmune disease. Currently, there is
significant interest in utilising Treg as a prospective therapy, in the
setting of autoimmunity
83
and particularly organ transplantation
84,85
as Treg not only control allo- and autoreactive T-cell responses but
also induce and maintain tolerance to self and non-self antigens. As
previously discussed, MSCs favour the generation of Treg and this
corresponds with a decrease in Th1, Th2 and Th17
lymphocytes.
12,14,74,86
The promotion of Treg by human bone
Mechanisms of MSC immune modulation
KEnglish
23
Immunology and Cell Biology