REVIEWS

Atypical MHC class II‑expressing

antigen-presenting cells: can anything
replace a dendritic cell?
Taku Kambayashi1 and Terri M. Laufer2
Abstract | Dendritic cells, macrophages and B cells are regarded as the classical
antigen-presenting cells of the immune system. However, in recent years, there has been a rapid
increase in the number of cell types that are suggested to present antigens on MHC class II
molecules to CD4+ T cells. In this Review, we describe the key characteristics that define an
antigen-presenting cell by examining the functions of dendritic cells. We then examine the
functions of the haematopoietic cells and non-haematopoietic cells that can express MHC
class II molecules and that have been suggested to represent ‘atypical’ antigen-presenting cells.
We consider whether any of these cell populations can prime naive CD4+ T cells and, if not,
question the effects that they do have on the development of immune responses.

1
Department of Pathology
and Laboratory Medicine and
Division of Rheumatology,
Department of Medicine,
Perelman School of Medicine,
University of Pennsylvania,
Philadelphia, Pennsylvania
19104, USA.
2
Philadelphia Veterans Affairs
Medical Center, Philadelphia,
Pennsylvania 19104, USA.
Correspondence to T.M.L.
e-mail:
tlaufer@mail.med.upenn.edu
doi:10.1038/nri3754
Published online
17 October 2014
Dendritic cells (DCs), B cells and macrophages constitu-
tively express MHC class II molecules and are regarded
as the ‘professional’ antigen-presenting cells (APCs) of
the immune system (FIG. 1). However, the universe of
APCs, seemingly so well established, has recently been
changing as more and more cell types are proposed
to express MHC class II molecules and have antigen-
presenting functions (FIG.  1; TABLE 1). A number of
haemato­poietic cell types have been suggested to pre-
sent antigens on MHC class II molecules to CD4+ T cells,
including mast cells, basophils, eosinophils, neutrophils,
innate lymphoid cells (ILCs) and CD4+ T cells themselves.
In addition, MHC class II expression has been detected on
non-haematopoietic cell types in the periphery, such as
endothelial cells, epithelial cells and lymph node stromal
cells (LNSCs). MHC class II‑expressing cells also support
thymocyte development and tolerance induction in the
thymus. However, for the purposes of this article, we focus
on how atypical MHC class II+ cells interact with mature
CD4+ T cells that have exited the thymus. In this Review,
we delineate the qualities that define a true APC and then
examine whether any of the atypical MHC class II+ cell
populations have these functional capabilities.
Classical APC characteristics
A consideration of the properties of DCs that facilitate
the activation of naive CD4+ T cells provides a com-
prehensive list of characteristics that a bona fide APC
should have (BOX 1). Specifically, in addition to their
antigen processing and presentation capabilities, and
their expression of co-stimulatory molecules, DCs
express pattern recognition receptors (PRRs) that medi-
ate activation in response to pathogens. Following PRR-
mediated maturation, DCs can migrate to the lymph
nodes to promote the activation of naive T cells. Indeed,
DCs have been shown to be both necessary and suffi-
cient for the activation of naive T cells1,2. The APC that
initiates an immune response must make multiple deci-
sions — for example, whether to actively respond to or
tolerate the antigenic challenge, and what type of active
immune response to induce. DCs clearly predominate in
the induction of T cell proliferation. Interestingly, how-
ever, DCs are not required for the development of CD4+
T cell responses under certain conditions3,4, suggesting
that other APCs may be able to replace the function of
DCs. Thus, we consider whether any non-classical APC
can prime naive CD4+ T cells, and discuss the roles that
these cells may have in modulating DC‑dependent and
DC‑independent immune responses.
Mast cells as APCs
Mast cells are tissue-resident innate immune cells that are
strategically localized at mucosal and submucosal sites
in close proximity to the external environment. They
have established roles in allergic disease. Following the
activation of high-affinity Fc receptor for IgE (FcεRI),
mast cells degranulate and release pre-stored immuno­
modulatory molecules — such as tumour necrosis factor

NATURE REVIEWS | IMMUNOLOGY VOLUME 14 | NOVEMBER 2014 | 719

© 2014 Macmillan Publishers Limited. All rights reserved

REVIEWS

Professional APCs Atypical APCs

DCs and macrophages B cells Mast cells Basophils Eosinophils ILC3s

Key features
• Phagocytic
• Express receptors for apoptotic
cells, DAMPs and PAMPs
• Localize to tissues
• Localize to T cell zone of lymph
nodes following activation (DCs)
• Constitutively express high levels
of MHC class II molecules and
antigen processing machinery
• Express co-stimulatory molecules
following activation
Key features
• Internalize antigens via
BCRs
• Constitutively express
MHC class II molecules
and antigen processing
machinery
• Express co-stimulatory
molecules following
activation
Key features
• Inducible expression of MHC class II molecules
• Antigen-presenting functions limited to
specific immune environments (especially
type 2 immune settings)
• Lack of compelling evidence that they can
activate naive CD4+ T cells in an antigen-
specific manner

Figure 1 | Key antigen-presenting functions of professional and atypical antigen-presenting cells.  The figure compares
and contrasts the features of ‘professional’ and ‘atypical’ antigen-presenting cells (APCs). Professional
Nature APCs
Reviewsinclude dendritic
| Immunology
cells (DCs) and, with a lesser role, macrophages and B cells. These cells constitutively express MHC class II structural proteins
and associated proteins, and antigen-processing machinery. By contrast, atypical APCs — such as mast cells, basophils,
eosinophils and innate lymphoid cells (ILCs) — do not constitutively express MHC class II molecules, but can upregulate
the expression of MHC class II under certain conditions. Basophils, eosinophils and mast cells that express MHC class II may
contribute to T helper 2 cell differentiation, but there is less evidence that they can prime naive CD4+ T cells. B cell receptor;
DAMP, damage-associated molecular pattern; PAMP, pathogen-associated molecular pattern.

Experimental autoimmune
encephalomyelitis
An experimental model for
the human disease multiple
sclerosis. Autoimmune disease
is induced in experimental
animals by immunization with
myelin or peptides derived
from myelin. The animals
develop a paralytic disease
with inflammation and
demyelination in the brain
and spinal cord.
(TNF) and histamine — which can promote T cell activa-
tion both directly and indirectly, through stimulation of
APCs. Furthermore, mast cell-deficient mice have defec-
tive CD4+ T cell responses in experimental auto­immune
encephalomyelitis5 and in Leishmania major infection6,
suggesting that CD4+ T cell responses could be altered
in the absence of mast cells.
Expression of MHC class II molecules by mast cells. Mast
cells can directly present antigens to T cells. Both rodent7,8
and human9,10 mast cells have been reported to constitu-
tively express MHC class II molecules. The expression of
MHC class II by mast cells correlated with their ability
to present antigen in vitro to naive CD4+ T cells and to
T cell hybridomas. Cytokines such as interleukin‑4 (IL‑4),
interferon‑γ (IFNγ) and granulocyte–macrophage
colony-stimulating factor (GM‑CSF; also known as
CSF2) further enhanced the APC function of mast
cells11,12. However, these reports were contradicted in
follow‑up studies by the same group showing that the
activation of antigen-specific T cells still occurred when
the T cells were cultured with MHC-mismatched mast
cells13. Moreover, co‑culture of mast cells with spleno­
cytes resulted in antigen-independent activation of
T cells14, which was probably mediated by immuno­
logically active exosomes released by mast cells15 (BOX 2).
These results were consistent with subsequent studies
showing that MHC class II molecules mainly reside
in intracellular lysosomal compartments of mast cells,
rather than at the cell surface16.
More recently, several groups have reported that
resting or FcεRI-activated mast cells do not express
MHC class II molecules, either on the cell surface
or intracellularly 17–20. The apparent discrepancies
between the earlier and more recent studies may
reflect the different protocols that have been used
to generate mast cells. The earlier studies generated
mast cells from short-term bone marrow cultures
(~3 weeks) that were supplemented with IL‑3 alone21.
By contrast, later studies used long-term bone
marrow cultures that were supplemented with both IL‑3
and stem cell factor (SCF; also known as KIT ligand).
Many of the cells derived from the short-term cultures
expressed FcεRI but lacked the mast/stem cell growth
factor receptor KIT22, suggesting that these cultures
also contained non-mast cells, including basophils3,23
and FcεRI+ DCs24,25, both of which expressed MHC
class II but not KIT. By contrast, bone marrow cells
grown in cultures supplemented with IL‑3 and SCF
for more than 6 weeks yielded mast cells that express
high, homogeneous levels of both KIT and FcεRI18,20.
Alternatively, the maturation status of mast cells may
have affected their APC function, as only mast cells
from bone marrow cultures less than 3 weeks old were
able to present antigen to T cells26.
Although resting bone marrow-derived mast
cells from long-term cultures do not constitutively
express MHC class II, its expression can be induced on
these cells. After activation of bone marrow-derived
mast cells with Toll-like receptor 4 (TLR4) agonists
and, to a lesser extent, TLR2 agonists in the presence
of IFNγ, a large proportion of cells expressed MHC
class II and other molecules that are associated with
antigen presentation — namely, MHC class II transacti­
vator (CIITA), the invariant chain (Ii) and H2‑DM20.
Similarly, although freshly isolated mast cells from the
peritoneal cavity lacked MHC class II expression, these
mast cells expressed MHC class II after in vitro treat-
ment with IL‑4 and IFNγ27. MHC class II+ mast cells
were also found in lymph node sinuses; treatment of

720 | NOVEMBER 2014 | VOLUME 14 www.nature.com/reviews/immunol

© 2014 Macmillan Publishers Limited. All rights reserved

 REVIEWS

Table 1 | Features of ‘atypical’ antigen-presenting cells

Atypical APC Features of atypical APC
Ability to promote T cell activation Expression of Lymph node Factors that induce Expression of other
co-stimulatory migration by expression of MHC MHC class II‑related
Naive T cells Activated T cells
molecules APC class II on APC genes
Mast cells +/– + +/– (CD80 and + TLR agonists, IFNγ, Ii and HLA‑DM
CD86) GM‑CSF, IL‑4 and Notch
Basophils +/– +/– + (mouse) + FcεRI, IL‑3 and papain Ii and HLA‑DM
– (human)
Eosinophils + + CD80 and CD86 + IL‑3, IFNγ and GM‑CSF Unknown
Neutrophils +/– +/– CD80 and CD86 + Co‑culture with T cells, Unknown
GM‑CSF and IFNγ
ILC2s + Unknown Unknown Unknown Unknown Unknown
ILC3s – Tolerance induction – + Unknown Ii and HLA‑DM
CD4 T cells
+
– Tolerance induction – + TCR stimulation Unknown
LNSCs Tolerance Unknown ICOSL (subset) NA TLR agonists and IFNγ Ii and HLA‑DM
induction
Endothelial cells – + CD137L, OX40L – IFNγ Unknown
and ICOSL
Epithelial cells Tolerance + – – IFNγ and microbial Ii and HLA‑DM
induction stimuli
APC, antigen-presenting cell; CD137L, CD137 ligand (also known as TNFSF9); FcεRI, high-affinity Fc receptor for IgE; GM‑CSF, granulocyte–macrophage
colony-stimulatory factor; ICOSL, ICOS ligand; IFN, interferon; Ii, invariant chain; IL, interleukin; ILC, innate lymphoid cell; LNSC, lymph node stromal cell;
NA, not applicable; OX40L, OX40 ligand (also known as TNFSF4); TCR, T cell receptor; TLR, Toll-like receptor.

mice with lipopolysaccharide or infection with L. major
increased the number of mast cells in the draining
lymph nodes and further upregulated MHC class II
expression20. The apparent constitutive expression of
MHC class II by lymph node mast cells may be owing
to the fact that these cells have matured and migrated
following their activation in the tissue. Alternatively,
endogenous ligands that induce the expression of MHC
class II on mast cells may exist at specific anatomical
sites. The latter hypothesis is supported by data identi­
fying Notch–Delta-like protein 1 interactions as an
inducer of mast cell MHC class II expression19. The
requirement for IFNγ and either TLR-mediated or
Notch-mediated signalling for MHC class II expression
by mast cells may be related to the roles of these path-
ways in inducing the transcription factor PU.1, which
regulates the CIITA promoter 28.
Functions of MHC class II on mast cells. After MHC
class II was expressed by activated bone marrow-derived
mast cells, they could process and present soluble protein
antigens to previously activated CD4+ T cells in vitro, but
they could not prime naive T cells20, perhaps because they
lacked co‑stimulatory molecule expression. However,
given that lymph node-resident and peri­toneal mast
cells that have been activated by IL‑4 and IFNγ express
both CD80 and CD86 (REFS 20,27), it is possible that some
mast cells can present antigens to naive T cells in vivo. In
addition to activated T cells, mast cells appear to prefer-
entially expand antigen-specific regulatory T (TReg) cells
rather than naive T cells20. The activation of TReg cells by
MHC  class II+ mast cells may contribute to the protec-
tive effects of mast cells in mediating tolerance to skin
allografts and in preventing nephrotoxic serum neph­
ritis, which are processes that were proposed to involve
IL‑9 production by TReg cells to recruit mast cells to the
graft or injury site29,30. It has been shown that endogenous
proteins are effectively presented on MHC  class II mol-
ecules of mast cells20 and as such, many of the peptides
presented by mast cell MHC class II may be self antigens
that stimulate TReg cells.
Does IgE support mast cell antigen-presenting functions?
Antigen-specific IgE molecules enhance APC function
in a subset of human DCs that express FcεRI by shut-
tling incorporated antigens to the MHC class II path-
way31. FcεRI could also potentially facilitate the uptake
of antigens into mast cells for more efficient antigen
presentation. Indeed, large particulate antigens were
taken up by mast cells through an IgE-dependent
mechanism both in vitro and in vivo32. Moreover, ear-
lier studies using cultured mast cells demonstrated
that the uptake of antigens through antigen-specific
IgE molecules enhances the T cell-stimulating capac-
ity of mast cells18,22,33. However, in the latter scenario,
the activation of T cells was not mediated through
direct antigen presentation by mast cells, as the mast
cells used in these experiments failed to express MHC
class II 17,18,20. Rather, mast cells transferred FcεRI-
incorporated antigens to DCs that secondarily pre-
sented the antigens to T cells18. Antigens incorporated
by mast cells through FcεRI were found to colocal-
ize with secretory compartments, and were released
by the mast cells upon reactivation32. Thus, although
FcεRI may transport antigens to the MHC class II
pathway under specific conditions22,33, mast cells that

NATURE REVIEWS | IMMUNOLOGY VOLUME 14 | NOVEMBER 2014 | 721

© 2014 Macmillan Publishers Limited. All rights reserved

REVIEWS

Box 1 | What are the key properties of an antigen-presenting cell?

Basophils modulate T cell responses
Basophils are a rare population of granulocytes that
Burnet’s ‘clonal selection’ theory presupposes that the T cell repertoire contains comprise ~1% of circulating peripheral leukocytes, and
a vast number of T cell clones expressing unique antigen receptors. Thus, an effective they are structurally and phenotypically related to mast
adaptive immune response requires a second cell to select and expand those few T cell cells. Like mast cells, basophils contain abundant gran-
clones that express ‘useful’ antigen receptors. Ralph Steinman termed this quality
ules with pre-formed inflammatory mediators that can
‘immunogenicity’ (REF. 145), and crucial work by McDevitt, Sela and Humphrey146–148
showed that immunogenicity required ‘immune response’ genes, which mapped to
be immediately released upon crosslinking of FcεRI34–37.
the MHC locus. Zinkernagel and Doherty149 first demonstrated that the presentation Important new reagents have recently been used to show
of viral antigens was MHC restricted and multiple laboratories subsequently showed that basophils and mast cells have functionally unique
that MHC class I and class II proteins present peptide fragments that are derived from and non-overlapping roles in IgE-mediated and IgG-
antigens to the T cell receptor (TCR). Thus, immunogenicity requires that proteins are mediated allergic inflammation in mice38–43. In addition,
processed and presented on MHC molecules to be recognized by T cells. it has been suggested that basophils are central to the dif-
However, T cell activation requires more than the simple expression of MHC molecules ferentiation of T helper 2 (TH2) cells. Indeed, a number of
by antigen-presenting cells (APCs). The activation of a naive T cell requires interaction studies have suggested that basophil-derived IL‑4 is cru-
with an APC that provides multiple signals (see the figure): ‘signal 1’ is delivered through cial for the development of TH2 cell responses to cysteine
interaction of the TCR with peptide–MHC complexes; ‘signal 2’ involves co-stimulatory
proteases, allergens and extracellular parasites4,23,42,44–50.
molecules; and ‘signal 3’ is mediated by instructive cytokines. The ability to deliver these
three signals is the defining characteristic of a professional APC150.
The search to identify APCs focused on cells that could induce B cell and T cell Basophils as APCs. Although basophil-derived IL‑4
responses in vitro and in vivo. This antigen-presenting function was initially ascribed to can skew T cells to a TH2 cell phenotype, it had been
macrophages, because the crucial ‘accessory cell’ found in these assays was adherent, presumed that antigen-specific T cell activation by
as are macrophages151. However, in 1973, Steinman and Cohn152 identified a different professional APCs, such as DCs, is still necessary for
population of adherent cells with large pseudopods in the mouse spleen; they TH2 cell induction. To examine the requirement for
subsequently demonstrated that these ‘dendritic cells’ (DCs) were the most potent DCs in antigen presentation during TH2 cell differen-
inducers of T cell proliferation in primary mixed lymphocyte responses153. Additionally, tiation, a number of groups have used mice expressing
DCs are particularly well adapted to process proteins into the peptides that are the human diphtheria toxin receptor (DTR) under the
presented by MHC class II molecules. They express the MHC class II structural α- and
Cd11c promoter (Cd11c–DTR mice); in these mice,
β-chains; however, they also express the associated chaperone invariant chains,
HLA‑DM and HLA-DO, which regulate peptide loading, and the complement of
diphtheria toxin treatment results in the ablation of
lysosomal proteases and cathepsins that can operate in the acidic phagolysosomal CD11c‑expressing cells. When bone marrow from
pathway (recently reviewed in REF. 154). Although many analyses probe for expression Cd11c–DTR mice was used to reconstitute irradiated
of the invariant chain (Ii), HLA-DM and HLA-DO, there have been very few stringent wild-type mice, diphtheria toxin treatment of these chi-
evaluations of the phagolysosomal pathways of atypical APCs. meric mice did not alter TH2 cell responses to papain
or Trichuris muris 3,4. Furthermore, papain injection or
DAMPs T.  muris infection of mice in which MHC class  II
and PAMPs
expression was restricted to DCs3,4,51 did not result in a
MHC TH2 cell response. Together, these results suggested
PRR
class II CD4 that MHC class II expression by DCs alone was neither
Peptide necessary nor sufficient for TH2 cell induction.
Signal 1: antigen- This suggested that another MHC class II‑expressing
specific interactions cell type might be required for TH2 cell induction in
TCR these models. Indeed, basophils were found to consti-
Signal 2: co-stimulatory tutively express MHC class II, CD80 and CD86, and the
molecules expression of these molecules was further upregulated
CD40L
CD40 by activation with IL‑3 (REF. 23) or papain3. Moreover,
Signal 3: instructive co‑culture of ovalbumin (OVA)-specific naive CD4+
cytokines
T cells with antigen-pulsed basophils alone was suffi-
IL-12 IL-12R cient to induce T cell proliferation and TH2 cell differen-
Activated DC CD4+ T cell
tiation in vitro3,4,23. Basophils could also present antigens
to T cells in vivo, as antigen-specific naive CD4+ T cells
displayed a TH2 cell phenotype in CIITA-deficient mice
CD40L, CD40 ligand; DAMP, damage-associated molecular pattern; IL‑12, interleukin‑12;
Nature
IL‑12R, IL‑12 receptor; PAMP, pathogen-associated molecular pattern; Reviews
PRR, | Immunology (which lack MHC class II expression) injected with
pattern recognition
receptor. peptide-pulsed basophils23. Basophils were found to
effectively incorporate soluble but not particulate anti-
gens by macropinocytosis3, which was further enhanced
have captured antigens may also act as reservoirs by antigen-specific IgE molecules bound to FcεRI
of antigen that can be presented to T cells by other on basophils. IgE-coated basophils efficiently incor-
APCs. Therefore, evidence from in vitro studies sup- porated specific antigens in vivo 52, and the addition
ports both direct and indirect roles for mast cells in of antigen-specific IgE molecules to CD4+ T cell and
antigen presentation. However, although some mast basophil co‑cultures augmented the induction of TH2
cells certainly express MHC class II when analysed cells23. However, it is unclear whether the augmented
directly ex vivo, whether they truly present antigen T H2 cell response mediated by IgE was owing to
in vivo remains to be determined. enhanced antigen presentation or increased IL‑4

722 | NOVEMBER 2014 | VOLUME 14 www.nature.com/reviews/immunol

© 2014 Macmillan Publishers Limited. All rights reserved


Box 2 | A confounding role for exosomes?
Naive T cells are usually activated by T cell receptor (TCR)-dependent contact with
peptide–MHC complexes on intact antigen-presenting cells (APCs). However, APCs
may also release antigen-presenting vesicles, or exosomes. Exosomes are secreted
membrane vesicles that form within late multivesicular endosomal compartments and
are released into the environment following fusion of the multivesicular bodies with
the plasma membrane.
Exosomes can be secreted by numerous cell types including dendritic cells (DCs),
B cells, mast cells and microglial cells. They are also secreted by several different
tumour cell types. Intestinal epithelial cells (IECs) may also secrete MHC class II‑bearing
exosomes; putative exosomes carrying IEC-specific proteins have been identified by
immunogold analysis of intestinal tissue sections155.
Immunologically active exosomes were first described by Thery and Amigorena156,157,
who noted that exosomes derived from DCs contain immunologically relevant
proteins such as MHC class I, MHC class II and CD86, as well as the heat shock protein
HSC73, which is involved in peptide delivery. Exosomes bearing peptide–MHC
complexes can stimulate naive T cells in vitro; however, Amigorena and his colleagues
have suggested that exosomes must be recaptured by endogenous DCs for antigen
presentation to naive T cells.
We discuss in the main text how mast cells may contribute to the immune response
via the transfer of exosomes to DCs. Similarly, MHC class II‑loaded exosomes may be
secreted by IECs and act as sources of antigen for tissue-resident or migratory DCs158,159.
Thus, the interpretation of many in vivo studies on the immunological functions of
different cell types — including some types of DCs — is confounded by the possibility
that peptide–MHC complexes from non-conventional cells may simply be transferred
to DCs via exosomes for presentation to T cells.
production by FcεRI-activated basophils. Together, these
results suggested that basophils are a source of IL‑4 and
may act as the primary APC in the activation of T cells
in a variety of TH2‑type immune settings.
Different findings in different systems. More recently, the
view that basophils might be the primary APC for TH2 cell
responses in vivo has been challenged by studies suggest-
ing that previous results may have been confounded by
the depletion methods that were used. First, the ablation
of DCs in Cd11c–DTR mice — as opposed to in wild-type
chimeric mice reconstituted with Cd11c–DTR bone mar-
row — markedly reduced the papain-induced TH2 cell
response53. As a proportion of skin-resident migratory
DCs are radioresistant 54–57, diphtheria toxin treatment
Mixed lymphocyte of wild-type mice reconstituted with bone marrow from
responses Cd11c–DTR mice could have spared migratory DCs in
A tissue-culture technique
the skin that were sufficient to induce TH2 cell responses.
for testing T cell reactivity.
The proliferation of one Similarly, Cd11c–Aβb mice lack MHC class II expression
population of T cells — induced on migratory skin Langerhans cells and dermal DCs, and
by exposure to inactivated this may confound the findings from studies using these
MHC-mismatched stimulator mice51. Indeed, surgical excision of the injection site a few
cells — is determined by
measuring the incorporation
hours after antigen challenge also attenuated the TH2 cell
of 3H-thymidine into the response, suggesting that radioresistant skin DC popula-
DNA of dividing cells. tions — either Langerhans cells or dermal DCs — might
be important51. As Langerhans cell depletion by express-
Macropinocytosis
ing DTR under the control of the promoter of the gene
A type of endocytosis (or
phagocytosis) that occurs encoding langerin (also known as CD207) had no effect
during the engulfment of on TH2 cell induction, it was concluded that dermal DCs
apoptotic cells. During were responsible for antigen presentation in this set-
macropinocytosis, large ting51. Dermal DCs incorporated the largest amount of
droplets of fluid are trapped
within the membrane
injected antigen and were the most potent at inducing
protrusions (ruffles) or T cell proliferation in ex vivo co‑culture experiments51.
phagocytic arms. However, DCs alone were unable to skew T cells to a
NATURE REVIEWS | IMMUNOLOGY VOLUME 14 | NOVEMBER 2014 | 723
© 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS
TH2 cell phenotype ex vivo and required the addition of
basophils to the co‑cultures51. Thus, it was concluded that
although DCs present antigen to T cells, basophils were
required for TH2 cell polarization.
A similar caveat to these models lies in the basophil
ablation methods that were used. In a house dust mite
(HDM) allergen model, which induces a TH2 cell response
in the lungs, different results were obtained when baso-
phils were depleted using different antibodies — namely,
MAR‑1 and Ba103 (REF. 24). Treatment with MAR‑1,
which is specific for FcεRI, had a stronger effect than
Ba103 (which recognizes CD200R3) in attenuating
TH2 cell polarization in HDM-challenged mice, although
both antibodies reduced basophil numbers to an
equivalent extent 24. The authors showed that in addi-
tion to basophils, MAR‑1 also ablated a population of
FcεRI+ DCs, whereas Ba103 only depleted basophils.
Furthermore, basophils from HDM-challenged mice were
poor at presenting antigen ex vivo compared to FcεRI+
DCs isolated from the same lymph nodes, suggesting
that DCs mediated antigen presentation. This argument
was supported by experiments showing that diphtheria
toxin treatment of Cd11c–DTR mice abrogated TH2 cell
responses upon HDM challenge24. Two other groups also
questioned the role of basophils in TH2 cell responses by
using novel mouse strains that constitutively lack baso-
phils owing to mast cell protease 8‑driven expression of
Cre recombinase58,59. Papain-induced TH2 cell responses
were intact in these basophil-deficient mice but not in
DC‑deficient mice, suggesting that DCs but not baso-
phils had an important role in papain-induced TH2 cell
polarization58,59. These results are in contrast to the afore-
mentioned studies demonstrating an important role for
basophils in papain-induced TH2 cell polarization3,42,53.
However, a more recent study targeted DTR expression
to basophils under the control of regulatory elements in
the gene encoding IL‑4 and found that basophils were
necessary for TH2 cell differentiation to peptide antigens,
but dispensable for priming to protein antigens60. It is
clear that the immunological effects of acute or consti-
tutive depletion of basophils are quite dependent on the
method of depletion and the biological setting.
Do human basophils show APC functions? Given
the contradictory data generated in mouse systems,
researchers soon began exploring the antigen-presenting
capabilities of human basophils. In humans, basophils
comprise <1% of circulating granulocytes and they
express FcεRI, CD203c (also known as ENPP3) and
CD123 (also known as IL‑3RA). Resting blood baso-
phils are HLA‑DR negative and early studies showed that
short-term activation with allergens, FcεRI engagement
or TLR2 ligands did not induce MHC class II expres-
sion61–63. However, subsequent studies suggested that
human basophils could express MHC class II in certain
disease states (for example, in patients with lupus neph­
ritis), in inflamed tissues that had high levels TH2‑type
cytokines64, and in response to culture with IL‑3 (REF. 65).
However, even these cultured human basophils could
not present allergens61 or exogenous peptides to induce
the activation of human T cells, perhaps because they
REVIEWS
Superantigens
Proteins that bind to and
activate all T cells that express
a particular set of Vβ T cell
receptor genes.
724 | NOVEMBER 2014 | VOLUME 14 www.nature.com/reviews/immunol
lacked expression of relevant co-stimulatory molecules65.
It remains possible that the bone marrow-derived and
tissue basophils that are purified from mice are not well
represented among the blood-derived human basophils
that were used in these studies. Nonetheless, there is not
yet any compelling evidence that human basophils can
present antigens to CD4+ T cells.
Eosinophils as APCs
Eosinophils are a population of circulating granulocytes
that have long been associated with TH2 cell responses.
They are elicited in response to IL‑5 production dur-
ing allergic inflammation and parasitic infections66–69,
and they are equipped with an arsenal of cationic pro-
teins and inflammatory mediators with anti­helminth
effector functions70. In addition to their effector role,
eosinophils may also be involved in the modula-
tion of TH2 cell immune responses. Similar to baso-
phils, eosinophils have constitutive activity at the IL4
locus38,40,71 and they are the major source of IL‑4 upon
challenge with Schistosoma mansoni eggs69. Moreover,
eosinophils release chemokines that are crucial for the
recruitment of T cells to the lungs during allergic airway
hyperresponsiveness72–74.
Like mast cells and basophils, eosinophils can also
express MHC class II. Expression of MHC class II by
eosinophils was first reported in sputum and broncho­
alveolar lavage samples from patients with asthma in the
early 1990s75,76. Although freshly isolated eosinophils
from blood were devoid of MHC class II expression, its
expression could be induced following in vitro culture
with activated T cell-conditioned media, GM‑CSF, IL‑3,
or a combination of IL‑3 and IFNγ77–80. Similar observa-
tions have been made in mice, where airway or lymph
node eosinophils constitutively express MHC class II81–84.
Although mouse eosinophils in the peritoneal cavity
are MHC class II negative, expression of MHC class II
is induced following the culture of these cells with
GM‑CSF85–87. HLA‑DR in human eosinophils localizes to
detergent-resistant lipid rafts that have been suggested
to enhance antigen presentation by professional APCs88
and, in both humans and mice, MHC class II‑expressing
eosinophils are capable of presenting superantigens,
peptides and protein antigens to T cell hybridomas and
T cell lines78,79,85,86,89. These data suggest that eosinophils
have the potential to function as APCs.
Evidence supporting a role of eosinophils in antigen
presentation in vivo came from studies demonstrating
that eosinophils elicited by repeated airway antigen
challenge migrate from endobronchial areas to the
draining mediastinal lymph nodes81. These eosino-
phils were predominantly found in the T cell zones of
the draining lymph nodes and they formed clusters
with antigen-specific T cells87. Lymph node eosinophils
expressed MHC class II, CD80 and CD86 and could
restimulate memory T cells from antigen-challenged
mice81,90. Moreover, intratracheal transfer of antigen-
pulsed eosinophils resulted in the enhanced prolifera-
tion of antigen-experienced CD4+ T cells in the draining
lymph nodes, suggesting that eosinophils could induce
antigen-specific T cell proliferation. Similar results were
© 2014 Macmillan Publishers Limited. All rights reserved
obtained using mice adoptively transferred with antigen-
specific naive T cell receptor (TCR)-transgenic T cells,
suggesting that eosinophils could also activate naive
T cells87. The activation of T cells by eosinophils seemed
to be MHC class II dependent, as the intraperitoneal
injection of Strongyloides stercoralis antigen-loaded
wild-type eosinophils, but not MHC class II-deficient
eosinophils, resulted in the increased production of
TH2 cell-associated cytokines91.
However, some studies have argued that eosinophils
might be incapable of efficiently processing protein
antigens. Although human eosinophils expressed MHC
class II after IL‑3 stimulation in vitro and could stimulate
antigen-specific primed T cells when pulsed with pep-
tides, they were unable to do so when pulsed with whole
antigen80. Similarly, whole protein-pulsed mouse eosino-
phils could not stimulate antigen-specific naive T cells,
although some proliferation of T cells was observed
when the eosinophils were pulsed with peptides84.
A recent study contradicted these findings and showed
that mouse eosinophils were in fact efficient APCs for
naive antigen-specific T cells both in vitro and in vivo91.
They attributed the previously reported lack of antigen-
processing ability of eosinophils to the use of NH4Cl
to lyse red blood cells. In their experiments, eosino-
phils were unable to activate antigen-specific T cells if
they were exposed to NH4Cl, potentially owing to its
lyso­s omotropic activity. Although this method was
used by one of the previous studies84, red blood cells
were removed by density gradients and centrifugation
in the other 80. Alternatively, differences in the human
and mouse systems may contribute to the discrepancies
that were seen in the latter study.
Although some controversy still remains as to their
antigen-processing ability, many studies have convinc-
ingly demonstrated that MHC class II and co‑stimulatory
molecules are expressed by airway-associated and lymph
node eosinophils, and by cytokine-activated circulating
eosinophils77–84. Moreover, eosinophils localize to the
same areas as classical APCs, as they migrate to T cell
zones in the lymph nodes during airway hyper­sensitivity
reactions81,87. Recently, mice that express DTR driven
by eosinophilic peroxidase have been generated, but
T cell-dependent responses in such mice have only been
analysed in an asthma model. At least in an allergic air-
way disease model, eosinophil depletion had no effect
during the antigen sensitization phase92, suggesting that
T cell responses are intact in the absence of eosinophils.
The generation of new methods to selectively ablate
MHC class II expression on eosinophils will be helpful
to further investigate the in vivo role of eosinophils as
APCs in other disease settings.
APC functions of neutrophils
So far, we have examined the potential antigen-
presenting functions of mast cells, basophils and eosino-
phils. However, neutrophils are the most abundant type
of granulocyte and are worth considering. Neutrophils
are rapidly recruited to sites of tissue damage where they
extrude neutrophil extracellular traps (NETs), produce
antimicrobial peptides, phagocytose microorganisms

Neutrophil extracellular
traps
(NETs). A set of extracellular
fibres produced by activated
neutrophils to ensnare
invading microorganisms.
NETs enhance neutrophil
killing of extracellular
pathogens, while minimizing
damage to host cells.
NATURE REVIEWS | IMMUNOLOGY VOLUME 14 | NOVEMBER 2014 | 725
and recruit other immune cells to clear pathogens.
Thus, neutrophils are generally regarded as professional
phagocytes that are involved early in the response to
tissue injury and infection. However, there is increas-
ing evidence that neutrophils may also modulate the
adaptive immune response through the production
of chemokines and cytokines that recruit DCs to sites of
inflammation93. Thioglycollate-elicited peritoneal neutro­
phils in mice express CD80, but neither CD86 nor MHC
class II94. However, co-culture with CD4+ T cells may lead
to a minimal level of cell-surface MHC class II expres-
sion and the ability to process OVA protein and activate
OVA-specific CD4+ T cells94. Similarly, neutrophils puri-
fied from the colons of colitic mice were MHC class II+
and, when pulsed with OVA peptide, could also activate
CD4+ T cells. Human neutrophils could express HLA‑DR
and could stimulate super­antigen-dependent T cell
activation; however, they could not re‑activate tetanus
toxoid-specific T cells95.
Despite the suggestion that neutrophils may be
professional APCs, the interpretation of these data is
quite complicated. First, although neutrophil prepara-
tions in these studies are 95–98% pure, the presence
of only a few contaminating DCs in either the neutro-
phil or T cell preparation could be sufficient to medi-
ate DC‑driven T cell activation. Second, neutrophils
induced by Pseudomonas aeruginosa airway infection lack
MHC class II expression, despite expressing CD80 and
CD86 (REF. 96). Finally, it is notable that in many of these
reports, MHC class II expression occurs in GM‑CSF-rich
environments. In this setting, these ‘differentiated’ neutro­
phils could easily be confused with inflammatory TNF
and iNOS producing (TIP)-DCs97 or a recently described
‘neutrophil–DC hybrid’ cell that shows characteristics of
both cell types98. Thus, there is not compelling evidence
that neutrophils can function as true APCs.
ILCs and CD4+ T cells as APCs
In recent years, there has been growing interest in ILCs,
which function as rapid sources of cytokines early dur-
ing immune responses. ILCs lack rearranged antigen
receptors but otherwise resemble CD4+ T cells in their
developmental pathways, transcription factor profiles
and cytokine-producing patterns. Type 1 ILCs (ILC1s),
ILC2s and ILC3s produce TH1-, TH2- and TH17‑type
cytokines, respectively. Recent evidence implicates ILCs
as crucial regulators of innate immunity and inflamma-
tion at barrier surfaces, including the skin, airways and
gastrointestinal tract (reviewed in REF. 99).
In addition to their early effects on innate immune
responses, there is increasing evidence that ILCs may
directly interact with CD4+ T cells. One group reported
that ILC3s can express CXC-chemokine receptor 5
(CXCR5) and CC-chemokine receptor 7 (CCR7), and
localize to secondary lymphoid organs where they may
control the maintenance of memory CD4+ T cells100,101.
More directly, genome-wide transcriptional profiling of
ILC3s showed that they express genes that encode MHC
class II structural components, as well as proteins that
are necessary for antigen presentation, including Ii and
H2‑DM102. In this report, a subset of ILC3s had surface
© 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS
expression of MHC class II and could acquire, process
and present exogenous antigens in vitro but did not
induce proliferation of CD4+ T cells. It was suggested
that ILCs probably do not activate naive CD4+ T cells;
nevertheless, cell-specific deletion of MHC class II
expression in retinoic acid receptor-related orphan
receptor-γt (RORγt)-expressing ILC3s led to the dysreg-
ulation of CD4+ T cells specific for commensal bac­teria
and the loss of intestinal epithelial integrity 102. Thus,
these authors proposed that MHC class II expression
on ILCs contributed to intestinal homeostasis by coun-
tering professional APCs to prevent the inappropriate
activation of commensal-specific CD4+ T cells. In con-
trast with these initial observations, a second laboratory
examined the function of splenic ILC3s from the same
strain of mice lacking expression of MHC class II in
ILC3s103. Interestingly, in a different animal facility, this
strain of mice did not develop any intestinal pathology,
suggesting a crucial role for microbial exposure103. They
also reported that, in contrast to ‘tolerizing’ intestinal
ILC3s, splenic ILC3s respond to IL‑1β by upregulating
both MHC class II and co-stimulatory molecules, and
thus contribute to CD4+ T cell proliferation. Perhaps
ILC3s localized to different environments have different
effects on CD4+ T cell activation and differentiation.
In agreement with these observations, we recently
showed that CD4+ T cells regulate the number and func-
tion of ILC3s in the mouse intestinal lamina propria in
an MHC class II‑dependent manner104. Although we did
not show that this was directly regulated by MHC class II
expression on the ILCs, we did find that the level of
MHC class II expressed by ILC3s in the intestinal lamina
propria was determined by the presence of functional
CD4+ T cells104. Thus, MHC class II‑expressing ILC3s
may regulate and also be regulated by CD4+ T cells.
There may also be such cross-regulation between
ILC2s and TH2 cells. ILC2s were first characterized as
cells that expressed MHC class II and inducible T cell co-
stimulator (ICOS)105. They are resident in the lungs and
other mucosal tissues, and IL‑2 produced by TH2 cells
may enhance cytokine production by ILC2s106. More
importantly, purified lung ILC2s were shown to present
peptide to TCR-transgenic CD4+ T cells in vitro106; how-
ever, ILC2s could not present OVA protein in this study.
A more recent examination found that peptide-loaded
MHC class II+ ILC2s induced TH2 cell differentiation
in vitro; ILC2s could also internalize and process pro-
tein antigens, but in vitro CD4+ T cell proliferation could
not be detected107. Nonetheless, MHC class II+ ILCs con-
tributed to the clearance of intestinal helminths107. The
study of MHC class II‑dependent functions of ILC2s
and ILC3s is a rapidly evolving area and the data remain
contradictory. Importantly, the systems studied to date
involve ablation of cell-specific MHC class II expression
rather than directly assaying the APC function of ILC3s
in the absence of antigen presentation by DCs.
Among other non‑B cell lymphocytes, CD4+ T cells
also have the ability to express MHC class II and pre-
sent antigens to other CD4+ T cells. Although CD4+
T cells from mice do not express MHC class II, human
CD4+ T cells express HLA‑DR molecules upon TCR
REVIEWS

Lymph node
stromal cell
Endothelial cell
Intestinal epithelial cell

MHC
class II
Constitutive MHC class II Inducible MHC class II Inducible MHC class II
expression expression expression

Function of MHC class II expression unclear

but may have a role in tolerance induction

Figure 2 | Epithelial cells and stromal cells express MHC class II molecules.  Vascular endothelial cells, multiple
types of epithelial cells (including intestinal and pulmonary epithelial cells) and lymph nodeNature
stromalReviews
cells can| Immunology
express
MHC class II molecules. Although these cells have been shown to activate CD4 T cells in vitro, their expression
+

of MHC class II molecules in vivo is thought to be more relevant for the induction and maintenance of immune tolerance.

activation108,109. Antigen presentation by CD4+ T cells
seems to have a dominant tolerance-inducing effect on
other CD4+ T cells. Peptide-loaded T cell clones can
activate each other to induce proliferation. However,
these T cells are defective in response to restimula-
tion110, suggesting that antigen-presenting CD4+ T cells
induce anergy. As MHC class II molecules on CD4+
T cells are loaded with self peptides111, it has been pos-
tulated that activation by HLA‑DR molecules expressed
by CD4 + T  cells may serve to limit self peptide-
reactive CD4+ T cells. The lack of MHC class II expres-
sion by mouse CD4+ T cells has hampered progress in
investigating the antigen-presenting role of CD4+ T cells.
Non-haematopoietic cells as APCs
Up to this point, we have focused on haematopoietic
cells — predominantly myeloid cells — that might act
as APCs. However, there is an extensive literature on the
antigen-presenting abilities of radioresistant endothelial
cells and epithelial cells in particular clinical settings,
especially autoimmunity and transplant tolerance (FIG. 2).
Lymph node stromal cells. The T cell–APC interactions
that mediate T cell activation do not occur in free space
but in tightly regulated anatomic regions of lymph nodes,
which are defined by the presence of non-haematopoietic
stromal cells of mesenchymal and endothelial origin.
LNSCs can be divided into subclasses on the basis of
their surface expression of the glycoproteins CD31 (also
known as PECAM1) and podoplanin (also known as
GP38), and their localization within lymph nodes. These
subclasses include fibroblastic reticular cells (FRCs), folli­
cular DCs (FDCs), lymphatic endothelial cells (LECs),
blood endothelial cells (BECs), pericytes and a small
proportion (<5%) of otherwise undefined stromal cells.
For many years, it was presumed that the sole func-
tion on non-FDC LNSCs was to provide the structural
‘backbone’ for secondary lymphoid organs. However,
the stroma clearly contributes to adaptive responses
(as reviewed in REF. 112) by concentrating antigens in the
lymph nodes and directing DC trafficking. Additionally,
multiple groups have demonstrated that LNSCs ectopi-
cally express tissue-specific antigens, including transgene-
directed neoantigens and endogenous tyrosinase,
and that they can induce the deletion of self-reactive
CD8+ T cells113–117. The mechanisms for the expression of
tissue-specific antigens are not completely clear. Thymic
medullary epithelial cells express autoimmune regulator
(AIRE), which is a transcriptional activator that promotes
the expression of tissue-specific antigens to regulate cen-
tral deletional tolerance. The Anderson laboratory used
an Aire-driven transgenic reporter to identify a popula-
tion of lymph node cells that expressed AIRE and tissue-
specific antigens117. Interestingly, these AIRE+ cells
expressed low levels of CD45 and CD11c, and were radio-
sensitive, suggesting that they were a haematopoietic pop-
ulation that was distinct from other LNSCs117. It is not clear
if AIRE is expressed at significant levels in LNSCs, although
a related transcriptional regulator, deformed epidermal
autoregulatory factor 1 (DEAF1), may be114,115.
LNSCs could either be intrinsically tolerogenic cells
or may simply express MHC class I in the absence of
co-stimulatory molecules and, as such, induce the dele-
tion of self-reactive CD8+ T cells113,114. However, the
ImmGen consortium has shown that FRCs, LECs and
BECs can upregulate surface expression of MHC class II
molecules under inflammatory or infectious condi-
tions118. In this setting, it has been demonstrated that
peripheral AIRE+ cells, but not radioresistant LNSCs,
can also mediate deletional tolerance of autoreactive
CD4+ T cells117. Importantly, there are no data suggest-
ing that CD4+ T cell deletion induced by either LNSCs or
AIRE+ cells is mediated independently of DCs. Indeed,
a recent report suggested that LNSC-mediated deletion
of CD4+ T cells is dependent on peptide–MHC class II
complexes that are acquired from DCs119. Thus, it is
possible that these cells are similar to basophils in
modulating antigen presentation by DCs.

726 | NOVEMBER 2014 | VOLUME 14 www.nature.com/reviews/immunol

© 2014 Macmillan Publishers Limited. All rights reserved


Endothelial cells and epithelial cells. Both human and
mouse vascular endothelial cells and tissue-resident
epithelial cells can express MHC class I and class II
molecules, and there is some evidence that interactions
with CD4+ T cells may be involved in some autoimmune
diseases and in graft rejection. Vascular endothelial cells in
human allografts can express both MHC class I and class II
molecules, and can present processed protein antigens to
T cell clones120,123,124. However, there is no evidence either
in vitro or in vivo that human endothelial cells can express
CD80 or CD86, which are necessary to stimulate naive
alloreactive T cells120. Both rat and mouse allograft models
suggest that DCs in the graft (‘passenger leukocytes’) are
necessary to initiate graft rejection121,122. However, memory
cells have decreased requirements for co-stimulation and
one group has shown in xenograft models that human
memory T cells — specifically the effector memory sub-
set — can directly respond to graft vascular endothelium
to mediate rejection123,124.
Epithelial cells — including intestinal epithelial cells
(IECs), airway epithelial cells and keratinocytes — are
uniquely positioned at the interface between the host
immune system and an environment teeming with anti-
gens, including pathogenic microorganisms and food
antigens, as well as commensal microorganisms. Thus,
decisions about whether to generate pro-inflammatory
or tolerizing responses must continuously be made at
mucosal surfaces. Similar to vascular endothelial cells,
epithelial cells can express MHC class II and may be
uniquely poised to regulate T cell responses to mucosal
antigens. Most work has focused on IECs and lung
alveolar epithelial cells; however, we have previously
suggested that MHC class II expression restricted to
keratinocytes could mediate autoimmune skin disease125.
Multiple reports suggest that both mouse and human
ileal IECs constitutively express MHC class II molecules.
Do these cells have the machinery to present antigen?
Much of the work suggesting that IECs can process and
present antigen has been limited to in vitro studies using
cell lines or primary cells treated with IFNγ. These stud-
ies do show that in the presence of IFNγ, the appropriate
machinery — including Ii, H2‑DM and proteases —
can be expressed by IECs in the small intestine and by
oesophageal epithelial cells126. Interestingly, IECs isolated
from patients with inflammatory bowel disease (IBD)
or from animal models of IBD express higher levels of
MHC class II127–129. Work using IEC-like cell lines has
shown that these cells can stimulate T cell hybridomas
or clones in vitro130–135. However, the evidence that naive
T cells can be stimulated by IECs in vivo is less compel-
ling. Indeed, there is no clear evidence that IECs express
sufficient levels of the appropriate co-stimulatory
molecules to activate naive CD4+ T cells127–129.
There are accumulating data suggesting that IECs may
contribute to the differentiation of TReg cells and the main-
tenance of tolerance. Neoantigen targeted to IECs induced
the expansion of antigen-specific TReg cell populations
in a manner that was independent of DC depletion136.
A more recent in vivo study examining transgenic mice
in which MHC class II expression was restricted to either
IECs or DCs showed that antigen presentation by DCs,
NATURE REVIEWS | IMMUNOLOGY VOLUME 14 | NOVEMBER 2014 | 727
© 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS
but not by IECs, could drive CD4+ T cell-dependent intes-
tinal inflammation137. The requirement for MHC class II
was examined by selectively deleting expression of CIITA
in non-haematopoietic cells, including IECs138. In these
mice, colitis induced by IL‑10 blockade was much more
severe than in co‑housed wild-type littermates; disease
was attributed to increased numbers of local TH1 cells
and an imbalance in the relative numbers of effector
T cells and TReg cells. Similar to investigations of IECs in
the small intestine (the focus of most published work),
it was also found that MHC class II+ colonic epithelial
cells could not stimulate naive CD4+ T cells in vitro;
rather, they suppressed the activation of CD4+ T cells
by professional APCs138,139. The mechanism for this was
not clear, although it was independent of transforming
growth factor-β.
In both mice and humans, as many as half of the MHC
class II+ cells in the lungs are of non-haematopoietic
origin140,141. Non-haematopoietic cells include vascular
and alveolar epithelial cells. Most data suggest that adap-
tive immunity to respiratory pathogens is predominantly
mediated by haematopoietic APCs, including migra-
tory DCs and alveolar macrophages; however, there
are data suggesting that epithelial cells might modulate
these responses. For example, type II alveolar epithelial
cells comprise only 4% of the epithelium140 but they are
strategically located to respond to airborne pathogens,
and they produce antimicrobial peptides and pro-
inflammatory mediators such as complement compo-
nents141,142. Type II alveolar epithelial cells in both mice
and humans express MHC class II that can be upregulated
by IFNγ and microbial stimulation143. Both endogenous
neo-self antigens and exogenous antigens — for example,
from Bacille Calmette–Guérin (BCG) — can be processed
and presented by type II alveolar epithelial cells143. MHC
class II expression on pulmonary non-haematopoietic
cells may contribute to decreased inflammation and
acceptance of orthotopic lung transplants144. In trans-
genic systems, type II alveolar epithelial cells can prime
antigen-specific CD4+ T cells. Interestingly, the T cell
response is skewed towards differentiation of forkhead
box P3 (FOXP3)-expressing CD4+ T cells143. In this regard,
type II alveolar epithelial cells may be similar to vascular
endothelial cells. Thus, endothelial cells and epithelial cells
may be additional cell types that express MHC class II and
modulate the outcome of DC–CD4+ T cell interactions.
Concluding remarks
There are a large number of haematopoietic and
non-haematopoietic cell types that can express MHC
class II molecules and present antigens to CD4+ T cells.
However, MHC class II expression alone is not sufficient
for full APC function, as APCs need to be able to pro-
cess antigens, migrate to secondary lymphoid organs and
express co‑stimulatory molecules. In this regard, pro-
fessional APCs, such as macrophages and B cells, could
potentially replace the function of DCs in certain situ-
ations. However, the non-professional APCs that have
been discussed in this Review are unlikely to replace
DCs, because there is little compelling data (if any) that
these cell types are able to activate naive CD4+ T cells.
REVIEWS
Rather, we propose that these non-conventional APCs
modulate immune responses that have been initiated
by DCs. This is especially true for non-haematopoietic
cells that may mediate the deletion of autoreactive T cells
or may stimulate TReg cells. The requirement for addi-
tional non‑DC APCs is quite context dependent — for
example, basophils, mast cells and eosinophils have
1. Lemos, M. P., Fan, L., Lo, D. & Laufer, T. M.
CD8α+ and CD11b+ dendritic cell-restricted
MHC class II controls Th1 CD4+ T cell immunity.
J. Immunol. 171, 5077–5084 (2003).
2. Lemos, M. P., Esquivel, F., Scott, P. & Laufer, T. M.
MHC class II expression restricted to CD8α+ and
CD11b+ dendritic cells is sufficient for control of
Leishmania major. J. Exp. Med. 199, 725–730
(2004).
3. Sokol, C. L. et al. Basophils function as antigen-
presenting cells for an allergen-induced T helper
type 2 response. Nature Immunol. 10, 713–720
(2009).
4. Perrigoue, J. G. et al. MHC class II‑dependent
basophil‑CD4+ T cell interactions promote TH2
cytokine-dependent immunity. Nature Immunol.
10, 697–705 (2009).
5. Gregory, G. D., Robbie-Ryan, M., Secor, V. H.,
Sabatino, J. J. Jr & Brown, M. A. Mast cells are
required for optimal autoreactive T cell responses
in a murine model of multiple sclerosis.
Eur. J. Immunol. 35, 3478–3486 (2005).
6. Maurer, M. et al. Skin mast cells control T cell-
dependent host defense in Leishmania major
infections. FASEB J. 20, 2460–2467 (2006).
7. Frandji, P. et al. Antigen-dependent stimulation by
bone marrow-derived mast cells of MHC
class II‑restricted T cell hybridoma. J. Immunol.
151, 6318–6328 (1993).
8. Fox, C. C., Jewell, S. D. & Whitacre, C. C. Rat
peritoneal mast cells present antigen to a PPD-specific
T cell line. Cell. Immunol. 158, 253–264 (1994).
9. Dimitriadou, V. et al. Expression of functional major
histocompatibility complex class II molecules on
HMC‑1 human mast cells. J. Leukoc. Biol. 64,
791–799 (1998).
10. Poncet, P., Arock, M. & David, B. MHC
class II‑dependent activation of CD4+ T cell
hybridomas by human mast cells through
superantigen presentation. J. Leukoc. Biol.
66, 105–112 (1999).
11. Frandji, P. et al. Presentation of soluble antigens
by mast cells: upregulation by interleukin‑4 and
granulocyte/macrophage colony-stimulating factor
and downregulation by interferon-γ. Cell. Immunol.
163, 37–46 (1995).
12. Frandji, P. et al. Exogenous and endogenous antigens
are differentially presented by mast cells to CD4+
T lymphocytes. Eur. J. Immunol. 26, 2517–2528
(1996).
13. Skokos, D. et al. Mast cell-derived exosomes induce
phenotypic and functional maturation of dendritic
cells and elicit specific immune responses in vivo.
J. Immunol. 170, 3037–3045 (2003).
14. Tkaczyk, C. et al. In vitro and in vivo
immunostimulatory potential of bone marrow-derived
mast cells on B- and T‑lymphocyte activation.
J. Allergy Clin. Immunol. 105, 134–142 (2000).
15. Skokos, D. et al. Mast cell-dependent B and T
lymphocyte activation is mediated by the secretion
of immunologically active exosomes. J. Immunol.
166, 868–876 (2001).
16. Raposo, G. et al. Accumulation of major
histocompatibility complex class II molecules in
mast cell secretory granules and their release upon
degranulation. Mol. Biol. Cell 8, 2631–2645
(1997).
17. Nakae, S. et al. Mast cells enhance T cell activation:
importance of mast cell costimulatory molecules and
secreted TNF. J. Immunol. 176, 2238–2248
(2006).
18. Kambayashi, T. et al. Indirect involvement of allergen-
captured mast cells in antigen presentation.
Blood 111, 1489–1496 (2008).
19. Nakano, N. et al. Notch signaling confers antigen-
presenting cell functions on mast cells. J. Allergy Clin.
Immunol. 123, 74–81.e1 (2009).
728 | NOVEMBER 2014 | VOLUME 14 www.nature.com/reviews/immunol
been implicated in the induction of TH2 cell responses,
whereas CD4+ T cells, and stromal, endothelial and epi-
thelial cells may contribute to tolerance. Studies in mice
clearly require careful analyses of the cell types that are
affected by manipulation. Similarly, increasing study
of human tissues other than blood will be necessary to
validate findings from mouse systems.
20. Kambayashi, T. et al. Inducible MHC class II
expression by mast cells supports effector and
regulatory T cell activation. J. Immunol. 182,
4686–4695 (2009).
This study shows that MHC class II expression on
mast cells is induced by TLR agonists and IFNγ,
and can support the activation of effector
T cells and TReg cells but not that of naive T cells.
21. Razin, E. et al. Interleukin 3: A differentiation and
growth factor for the mouse mast cell that contains
chondroitin sulfate E proteoglycan. J. Immunol. 132,
1479–1486 (1984).
22. Tkaczyk, C., Villa, I., Peronet, R., David, B. &
Mecheri, S. FcεRI-mediated antigen endocytosis turns
interferon-gamma-treated mouse mast cells from
inefficient into potent antigen-presenting cells.
Immunology 97, 333–340 (1999).
23. Yoshimoto, T. et al. Basophils contribute to TH2‑IgE
responses in vivo via IL‑4 production and presentation
of peptide-MHC class II complexes to CD4+ T cells.
Nature Immunol. 10, 706–712 (2009).
References 3, 4 and 23 show that basophils
express MHC class II and present antigens to
T cells to promote TH2‑type responses.
24. Hammad, H. et al. Inflammatory dendritic cells—not
basophils—are necessary and sufficient for induction
of Th2 immunity to inhaled house dust mite allergen.
J. Exp. Med. 207, 2097–2111 (2010).
This study argues that FcεRI-expressing DCs
and not basophils are responsible for antigen
presentation in response to HDM allergen.
25. Phythian-Adams, A. T. et al. CD11c depletion severely
disrupts Th2 induction and development in vivo.
J. Exp. Med. 207, 2089–2096 (2010).
26. Gong, J. et al. The antigen presentation function of
bone marrow-derived mast cells is spatiotemporally
restricted to a subset expressing high levels of cell
surface FcεRI and MHC II. BMC Immunol. 11, 34
(2010).
27. Gaudenzio, N. et al. Cell-cell cooperation at the
T helper cell/mast cell immunological synapse.
Blood 114, 4979–4988 (2009).
28. Ito, T. et al. Roles of PU.1 in monocyte- and mast cell-
specific gene regulation: PU.1 transactivates CIITA pIV
in cooperation with IFN-γ. Int. Immunol. 21, 803–816
(2009).
29. Lu, L. F. et al. Mast cells are essential intermediaries
in regulatory T‑cell tolerance. Nature 442, 997–1002
(2006).
30. Eller, K. et al. IL‑9 production by regulatory T cells
recruits mast cells that are essential for regulatory
T cell-induced immune suppression. J Immunol. 186,
83–91 (2011).
31. Maurer, D. et al. Fcε receptor I on dendritic cells
delivers IgE-bound multivalent antigens into a
cathepsin S‑dependent pathway of MHC class II
presentation. J. Immunol. 161, 2731–2739 (1998).
32. Shin, J. S., Shelburne, C. P., Jin, C., LeFurgey, E. A. &
Abraham, S. N. Harboring of particulate allergens
within secretory compartments by mast cells following
IgE/FcεRI-lipid raft-mediated phagocytosis.
J. Immunol. 177, 5791–5800 (2006).
33. Tkaczyk, C. et al. Specific antigen targeting to surface
IgE and IgG on mouse bone marrow-derived mast cells
enhances efficiency of antigen presentation.
Immunology 94, 318–324 (1998).
34. Dvorak, A. M., Newball, H. H., Dvorak, H. F. &
Lichtenstein, L. M. Antigen-induced IgE-mediated
degranulation of human basophils. Lab Invest. 43,
126–139 (1980).
35. Ishizaka, T. & Ishizaka, K. Immunological events at the
surface of basophil granulocytes and mast cells which
induce degranulation. Scand. J. Respir. Dis. Suppl.
98, 13–22 (1977).
36. Arinobu, Y., Iwasaki, H. & Akashi, K. Origin of
basophils and mast cells. Allergol Int. 58, 21–28
(2009).
© 2014 Macmillan Publishers Limited. All rights reserved
37. Arinobu, Y. et al. Developmental checkpoints of
the basophil/mast cell lineages in adult murine
hematopoiesis. Proc. Natl Acad. Sci. USA 102,
18105–18110 (2005).
38. Voehringer, D., Shinkai, K. & Locksley, R. M.
Type 2 immunity reflects orchestrated recruitment
of cells committed to IL‑4 production. Immunity 20,
267–277 (2004).
39. Min, B. et al. Basophils produce IL‑4 and accumulate
in tissues after infection with a Th2‑inducing parasite.
J. Exp. Med. 200, 507–517 (2004).
40. Gessner, A., Mohrs, K. & Mohrs, M. Mast cells,
basophils, and eosinophils acquire constitutive IL‑4
and IL‑13 transcripts during lineage differentiation
that are sufficient for rapid cytokine production.
J. Immunol. 174, 1063–1072 (2005).
41. Karasuyama, H., Mukai, K., Tsujimura, Y. & Obata, K.
Newly discovered roles for basophils: a neglected
minority gains new respect. Nature Rev. Immunol.
9, 9–13 (2009).
42. Sokol, C. L., Barton, G. M., Farr, A. G. & Medzhitov, R.
A mechanism for the initiation of allergen-induced
T helper type 2 responses. Nature Immunol.
9, 310–318 (2007).
43. Tsujimura, Y. et al. Basophils play a pivotal role in
immunoglobulin-G‑mediated but not immunoglobulin-
E‑mediated systemic anaphylaxis. Immunity 28,
581–589 (2008).
44. Le Gros, G., Ben-Sasson, S. Z., Seder, R.,
Finkelman, F. D. & Paul, W. E. Generation of
interleukin 4 (IL‑4)-producing cells in vivo and in vitro:
IL‑2 and IL‑4 are required for in vitro generation of
IL‑4‑producing cells. J. Exp. Med. 172, 921–929
(1990).
45. Swain, S. L., Weinberg, A. D., English, M. & Huston, G.
IL‑4 directs the development of Th2‑like helper
effectors. J. Immunol. 145, 3796–3806 (1990).
46. Else, K. J., Finkelman, F. D., Maliszewski, C. R. &
Grencis, R. K. Cytokine-mediated regulation of
chronic intestinal helminth infection. J. Exp. Med.
179, 347–351 (1994).
47. Cohn, L., Homer, R. J., Marinov, A., Rankin, J. &
Bottomly, K. Induction of airway mucus production
By T helper 2 (Th2) cells: a critical role for interleukin 4
in cell recruitment but not mucus production.
J. Exp. Med. 186, 1737–1747 (1997).
48. Min, B. Th2 immunity: a step closer to completion.
Immunol. Cell Biol. 88, 235 (2010).
49. Hida, S., Tadachi, M., Saito, T. & Taki, S.
Negative control of basophil expansion by IRF‑2
critical for the regulation of Th1/Th2 balance.
Blood 106, 2011–2017 (2005).
50. Oh, K., Shen, T., Le Gros, G. & Min, B. Induction of
Th2 type immunity in a mouse system reveals a novel
immunoregulatory role of basophils. Blood 109,
2921–2927 (2007).
51. Allenspach, E. J., Lemos, M. P., Porrett, P. M.,
Turka, L. A. & Laufer, T. M. Migratory and lymphoid-
resident dendritic cells cooperate to efficiently
prime naive CD4 T cells. Immunity 29, 795–806
(2008).
52. Mack, M. et al. Identification of antigen-capturing
cells as basophils. J. Immunol. 174, 735–741
(2005).
53. Tang, H. et al. The T helper type 2 response to
cysteine proteases requires dendritic cell-basophil
cooperation via ROS-mediated signaling.
Nature Immunol. 11, 608–617 (2010).
This study argues that basophils are important
for papain-induced TH2 cell responses not for
their APC function but for their effects on DCs.
54. Merad, M. et al. Langerhans cells renew in the skin
throughout life under steady-state conditions.
Nature Immunol. 3, 1135–1141 (2002).
55. Bursch, L. S. et al. Identification of a novel population
of Langerin+ dendritic cells. J. Exp. Med. 204,
3147–3156 (2007).

56. Ginhoux, F. et al. Blood-derived dermal langerin+
dendritic cells survey the skin in the steady state.
J. Exp. Med. 204, 3133–3146 (2007).
57. Poulin, L. F. et al. The dermis contains langerin+
dendritic cells that develop and function
independently of epidermal Langerhans cells.
J. Exp. Med. 204, 3119–3131 (2007).
58. Ohnmacht, C. et al. Basophils orchestrate chronic
allergic dermatitis and protective immunity against
helminths. Immunity 33, 364–374 (2010).
59. Sullivan, B. M. et al. Genetic analysis of basophil
function in vivo. Nature Immunol. 12, 527–535 (2011).
60. Otsuka, A. et al. Basophils are required for the
induction of Th2 immunity to haptens and peptide
antigens. Nature Commun. 4, 1739 (2013).
61. Eckl-Dorna, J. et al. Basophils are not the key
antigen-presenting cells in allergic patients.
Allergy 67, 601–608 (2012).
62. Kitzmuller, C. et al. Human blood basophils do not act
as antigen-presenting cells for the major birch pollen
allergen Bet v 1. Allergy 67, 593–600 (2012).
References 61 and 62 were the first studies to
propose that basophils might not act as APCs in
the setting of human allergy.
63. Sharma, M. et al. Circulating human basophils lack
the features of professional antigen presenting cells.
Sci. Rep. 3, 1188 (2013).
64. Charles, N., Hardwick, D., Daugas, E., Illei, G. G. &
Rivera, J. Basophils and the T helper 2 environment
can promote the development of lupus nephritis.
Nature Med. 16, 701–707 (2010).
65. Voskamp, A. L., Prickett, S. R., Mackay, F.,
Rolland, J. M. & O’Hehir, R. E. MHC class II
expression in human basophils: induction and lack of
functional significance. PLoS ONE 8, e81777 (2013).
66. Foster, P. S., Hogan, S. P., Ramsay, A. J.,
Matthaei, K. I. & Young, I. G. Interleukin 5 deficiency
abolishes eosinophilia, airways hyperreactivity, and
lung damage in a mouse asthma model. J. Exp. Med.
183, 195–201 (1996).
67. Collins, P. D., Marleau, S., Griffiths-Johnson, D. A.,
Jose, P. J. & Williams, T. J. Cooperation between
interleukin‑5 and the chemokine eotaxin to induce
eosinophil accumulation in vivo. J. Exp. Med. 182,
1169–1174 (1995).
68. Hogan, S. P., Koskinen, A. & Foster, P. S. Interleukin‑5
and eosinophils induce airway damage and bronchial
hyperreactivity during allergic airway inflammation in
BALB/c mice. Immunol. Cell Biol. 75, 284–288 (1997).
69. Sabin, E. A., Kopf, M. A. & Pearce, E. J.
Schistosoma mansoni egg-induced early IL‑4
production is dependent upon IL‑5 and eosinophils.
J. Exp. Med. 184, 1871–1878 (1996).
70. Hogan, S. P. et al. Eosinophils: biological properties
and role in health and disease. Clin. Exp. Allergy 38,
709–750 (2008).
71. Shinkai, K., Mohrs, M. & Locksley, R. M. Helper T cells
regulate type‑2 innate immunity in vivo. Nature 420,
825–829 (2002).
72. Jacobsen, E. A. et al. Allergic pulmonary inflammation
in mice is dependent on eosinophil-induced recruitment
of effector T cells. J. Exp. Med. 205, 699–710 (2008).
73. Walsh, E. R. et al. Strain-specific requirement for
eosinophils in the recruitment of T cells to the lung
during the development of allergic asthma.
J. Exp. Med. 205, 1285–1292 (2008).
74. Fulkerson, P. C. et al. A central regulatory role for
eosinophils and the eotaxin/CCR3 axis in chronic
experimental allergic airway inflammation. Proc. Natl
Acad. Sci. USA 103, 16418–16423 (2006).
75. Hansel, T. T. et al. Sputum eosinophils from asthmatics
express ICAM‑1 and HLA‑DR. Clin. Exp. Immunol. 86,
271–277 (1991).
This is the first study to show that sputum but
not blood eosinophils from a large proportion
of patients with asthma express HLA‑DR.
76. Mengelers, H. J. et al. Immunophenotyping of
eosinophils recovered from blood and BAL of allergic
asthmatics. Am. J. Respir. Crit. Care Med. 149,
345–351 (1994).
77. Beninati, W. et al. Pulmonary eosinophils express
HLA‑DR in chronic eosinophilic pneumonia.
J. Allergy Clin. Immunol. 92, 442–449 (1993).
78. Hansel, T. T. et al. Induction and function of eosinophil
intercellular adhesion molecule‑1 and HLA‑DR.
J. Immunol. 149, 2130–2136 (1992).
79. Mawhorter, S. D., Kazura, J. W. & Boom, W. H.
Human eosinophils as antigen-presenting cells:
relative efficiency for superantigen- and antigen-
induced CD4+ T‑cell proliferation. Immunology 81,
584–591 (1994).
NATURE REVIEWS | IMMUNOLOGY VOLUME 14 | NOVEMBER 2014 | 729
80. Celestin, J. et al. IL‑3 induces B7.2 (CD86) expression
and costimulatory activity in human eosinophils.
J. Immunol. 167, 6097–6104 (2001).
81. Shi, H. Z., Humbles, A., Gerard, C., Jin, Z. &
Weller, P. F. Lymph node trafficking and antigen
presentation by endobronchial eosinophils.
J. Clin. Invest. 105, 945–953 (2000).
This is the first study to show that mouse
eosinophils from antigen-sensitized airways express
MHC class II and can support T cell activation.
82. Shi, H. Z. et al. Endobronchial eosinophils
preferentially stimulate T helper cell type 2 responses.
Allergy 59, 428–435 (2004).
83. Duez, C. et al. Migration and accumulation of
eosinophils toward regional lymph nodes after airway
allergen challenge. J. Allergy Clin. Immunol. 114,
820–825 (2004).
84. van Rijt, L. S. et al. Airway eosinophils accumulate in
the mediastinal lymph nodes but lack antigen-
presenting potential for naive T cells. J. Immunol.
171, 3372–3378 (2003).
85. Tamura, N. et al. Requirement of CD80 and CD86
molecules for antigen presentation by eosinophils.
Scand. J. Immunol. 44, 229–238 (1996).
86. Del Pozo, V. et al. Eosinophil as antigen-presenting
cell: activation of T cell clones and T cell hybridoma by
eosinophils after antigen processing. Eur. J. Immunol.
22, 1919–1925 (1992).
87. Wang, H. B., Ghiran, I., Matthaei, K. & Weller, P. F.
Airway eosinophils: allergic inflammation recruited
professional antigen-presenting cells. J. Immunol.
179, 7585–7592 (2007).
88. Akuthota, P., Melo, R. C., Spencer, L. A. &
Weller, P. F. M. H. C. Class II and CD9 in human
eosinophils localize to detergent-resistant membrane
microdomains. Am. J. Respir. Cell. Mol. Biol. 46,
188–195 (2012).
89. Handzel, Z. T. et al. Eosinophils bind rhinovirus and
activate virus-specific T cells. J. Immunol. 160,
1279–1284 (1998).
90. MacKenzie, J. R., Mattes, J., Dent, L. A. & Foster, P. S.
Eosinophils promote allergic disease of the lung by
regulating CD4+ Th2 lymphocyte function.
J. Immunol. 167, 3146–3155 (2001).
91. Padigel, U. M. et al. Eosinophils act as antigen-
presenting cells to induce immunity to Strongyloides
stercoralis in mice. J. Infect. Dis. 196, 1844–1851
(2007).
92. Jacobsen, E. A. et al. Eosinophil activities modulate the
immune/inflammatory character of allergic respiratory
responses in mice. Allergy 69, 315–327 (2014).
93. Denkers, E. Y., Butcher, B. A., Del Rio, L. &
Bennouna, S. Neutrophils, dendritic cells and
Toxoplasma. Int. J. Parasitol. 34, 411–421 (2004).
94. Abi Abdallah, D. S., Egan, C. E., Butcher, B. A. &
Denkers, E. Y. Mouse neutrophils are professional
antigen-presenting cells programmed to instruct Th1
and Th17 T‑cell differentiation. Int. Immunol. 23,
317–326 (2011).
95. Fanger, N. A. et al. Activation of human T cells by
major histocompatability complex class II expressing
neutrophils: proliferation in the presence of
superantigen, but not tetanus toxoid. Blood 89,
4128–4135 (1997).
This study shows that human neutrophils can
express MHC class II after stimulation with GM‑CSF
and IFNγ, and can support superantigen-mediated
but not peptide-mediated T cell activation.
96. Yamamoto, S. et al. Cutting edge: Pseudomonas
aeruginosa abolishes established lung transplant
tolerance by stimulating B7 expression on neutrophils.
J. Immunol. 189, 4221–4225 (2012).
97. Serbina, N. V., Salazar-Mather, T. P., Biron, C. A.,
Kuziel, W. A. & Pamer, E. G. TNF/iNOS-producing
dendritic cells mediate innate immune defense against
bacterial infection. Immunity 19, 59–70 (2003).
98. Matsushima, H. et al. Neutrophil differentiation into a
unique hybrid population exhibiting dual phenotype
and functionality of neutrophils and dendritic cells.
Blood 121, 1677–1689 (2013).
99. Hepworth, M. R. & Sonnenberg, G. F. Regulation of
the adaptive immune system by innate lymphoid cells.
Curr. Opin. Immunol. 27C, 75–82 (2014).
100. Withers, D. R. et al. Cutting edge: lymphoid tissue
inducer cells maintain memory CD4 T cells within
secondary lymphoid tissue. J. Immunol. 189,
2094–2098 (2012).
101. Lane, P. J., Gaspal, F. M., McConnell, F. M.,
Withers, D. R. & Anderson, G. Lymphoid tissue inducer
cells: pivotal cells in the evolution of CD4 immunity and
tolerance? Front. Immunol. 3, 24 (2012).
© 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS
102. Hepworth, M. R. et al. Innate lymphoid cells regulate
CD4+ T‑cell responses to intestinal commensal
bacteria. Nature 498, 113–117 (2013).
This is the first study to suggest that ILC3s in the
gut regulate CD4+ T cell responses to commensal
bacteria through MHC class II expression.
103. von Burg, N. et al. Activated group 3 innate lymphoid
cells promote T‑cell-mediated immune responses.
Proc. Natl Acad. Sci. USA 111, 12835–12840 (2014).
This study shows that ILC3s stimulated with IL‑1β
express MHC class II and co-stimulatory molecules,
and can stimulate naive CD4+ T cell activation.
104. Korn, L. L. et al. Conventional CD4+ T cells regulate
IL‑22‑producing intestinal innate lymphoid cells.
Mucosal Immunol. 7, 1045–1057 (2014).
105. Neill, D. R. et al. Nuocytes represent a new innate
effector leukocyte that mediates type‑2 immunity.
Nature 464, 1367–1370 (2010).
106. Mirchandani, A. S. et al. Type 2 innate lymphoid cells
drive CD4+ Th2 cell responses. J. Immunol. 192,
2442–2448 (2014).
107. Oliphant, C. J. et al. MHCII-mediated dialog between
group 2 innate lymphoid cells and CD4+ T cells
potentiates type 2 immunity and promotes parasitic
helminth expulsion. Immunity 41, 283–295 (2014).
This is the first study to show that ILC2s express
MHC class II and co‑stimulatory molecules, and can
support peptide-mediated naive CD4+ T cell
activation.
108. Ko, H. S., Fu, S. M., Winchester, R. J., Yu, D. T. &
Kunkel, H. G. Ia determinants on stimulated human
T lymphocytes. Occurrence on mitogen- and antigen-
activated T cells. J. Exp. Med. 150, 246–255 (1979).
109. Evans, R. L. et al. Peripheral human T cells sensitized
in mixed leukocyte culture synthesize and express
Ia‑like antigens. J. Exp. Med. 148, 1440–1445 (1978).
110. LaSalle, J. M., Tolentino, P. J., Freeman, G. J.,
Nadler, L. M. & Hafler, D. A. Early signaling defects
in human T cells anergized by T cell presentation of
autoantigen. J. Exp. Med. 176, 177–186 (1992).
This is the first study to fully characterize the
potential of human CD4+ T cells to anergize other
CD4+ T cells through antigen presentation.
111. Costantino, C. M., Spooner, E., Ploegh, H. L. &
Hafler, D. A. Class, I. I. MHC self-antigen presentation
in human B and T lymphocytes. PLoS ONE 7, e29805
(2012).
112. Card, C. M., Yu, S. S. & Swartz, M. A. Emerging roles
of lymphatic endothelium in regulating adaptive
immunity. J. Clin. Invest. 124, 943–952 (2014).
113. Cohen, J. N. et al. Lymph node-resident lymphatic
endothelial cells mediate peripheral tolerance via
Aire-independent direct antigen presentation.
J. Exp. Med. 207, 681–688 (2010).
114. Fletcher, A. L. et al. Lymph node fibroblastic
reticular cells directly present peripheral tissue
antigen under steady-state and inflammatory
conditions. J. Exp. Med. 207, 689–697 (2010).
115. Lee, J. W. et al. Peripheral antigen display by lymph
node stroma promotes T cell tolerance to intestinal
self. Nature Immunol. 8, 181–190 (2007).
116. Tewalt, E. F. et al. Lymphatic endothelial cells induce
tolerance via PD‑L1 and lack of costimulation leading
to high-level PD‑1 expression on CD8 T cells.
Blood 120, 4772–4782 (2012).
117. Gardner, J. M. et al. Extrathymic Aire-expressing cells
are a distinct bone marrow-derived population that
induce functional inactivation of CD4+ T cells.
Immunity 39, 560–572 (2013).
This study suggests that some APCs termed LNSCs
are actually a bone marrow-derived population of
cells that induce T cell anergy.
118. Malhotra, D. et al. Transcriptional profiling of stroma
from inflamed and resting lymph nodes defines
immunological hallmarks. Nature Immunol. 13,
499–510 (2012).
This transcriptional analysis dissects which LNSCs
can and do express MHC class II that is inducible
by inflammatory signals.
119. Dubrot, J. et al. Lymph node stromal cells acquire
peptide-MHCII complexes from dendritic cells and
induce antigen-specific CD4+ T cell tolerance.
J. Exp. Med. 211, 1153–1166 (2014).
This paper suggests that LNSCs express
endogenous MHC class II molecules and can also
acquire peptide–MHC class II complexes from DCs.
120. Hughes, C. C., Savage, C. O. & Pober, J. S.
Endothelial cells augment T cell interleukin 2
production by a contact-dependent mechanism
involving CD2/LFA‑3 interaction. J. Exp. Med. 171,
1453–1467 (1990).
REVIEWS
121. Lechler, R. I. & Batchelor, J. R. Restoration of
immunogenicity to passenger cell-depleted kidney
allografts by the addition of donor strain dendritic
cells. J. Exp. Med. 155, 31–41 (1982).
122. Lakkis, F. G., Arakelov, A., Konieczny, B. T. & Inoue, Y.
Immunologic ‘ignorance’ of vascularized organ
transplants in the absence of secondary lymphoid
tissue. Nature Med. 6, 686–688 (2000).
123. Shiao, S. L., McNiff, J. M. & Pober, J. S.
Memory T cells and their costimulators in human
allograft injury. J. Immunol. 175, 4886–4896 (2005).
124. Shiao, S. L. et al. Human effector memory CD4+ T cells
directly recognize allogeneic endothelial cells in vitro
and in vivo. J. Immunol. 179, 4397–4404 (2007).
References 123 and 124 are two of many articles
from the Pober laboratory examining the function
of MHC class II+ endothelial cells as APCs during
graft acceptance or rejection. Many of the articles
from this laboratory examine the mechanisms for
stimulating memory but not naive CD4+ T cells.
125. Fan, L. et al. Antigen presentation by keratinocytes
directs autoimmune skin disease. Proc. Natl Acad. Sci.
USA 100, 3386–3391 (2003).
126. Mulder, D. J. et al. Antigen presentation and MHC
class II expression by human esophageal epithelial
cells: role in eosinophilic esophagitis. Am. J. Pathol.
178, 744–753 (2011).
127. Sanderson, I. R., Ouellette, A. J., Carter, E. A.,
Walker, W. A. & Harmatz, P. R. Differential
regulation of B7 mRNA in enterocytes and
lymphoid cells. Immunology 79, 434–438 (1993).
128. Framson, P. E., Cho, D. H., Lee, L. Y. &
Hershberg, R. M. Polarized expression and function of
the costimulatory molecule CD58 on human intestinal
epithelial cells. Gastroenterology 116, 1054–1062
(1999).
129. Nakazawa, A. et al. Functional expression of
costimulatory molecule CD86 on epithelial cells in
the inflamed colonic mucosa. Gastroenterology 117,
536–545 (1999).
130. Kaiserlian, D., Vidal, K. & Revillard, J. P. Murine
enterocytes can present soluble antigen to specific
class II‑restricted CD4+ T cells. Eur. J. Immunol. 19,
1513–1516 (1989).
131. Buning, J. et al. Antigen targeting to MHC
class II‑enriched late endosomes in colonic epithelial
cells: trafficking of luminal antigens studied in vivo in
Crohn’s colitis patients. FASEB J. 20, 359–361 (2006).
132. Bland, P. W. & Warren, L. G. Antigen presentation by
epithelial cells of the rat small intestine. I. Kinetics,
antigen specificity and blocking by anti‑Ia antisera.
Immunology 58, 1–7 (1986).
133. Mayer, L. & Shlien, R. Evidence for function of Ia
molecules on gut epithelial cells in man. J. Exp. Med.
166, 1471–1483 (1987).
References 130, 132 and 133 are among the first
manuscripts to demonstrate that IECs express
MHC class II and present protein antigen to CD4+
T cell hybridomas.
730 | NOVEMBER 2014 | VOLUME 14 www.nature.com/reviews/immunol
134. Hershberg, R. M. et al. Intestinal epithelial cells use
two distinct pathways for HLA class II antigen
processing. J. Clin. Invest. 100, 204–215 (1997).
135. Hershberg, R. M. et al. Highly polarized HLA class II
antigen processing and presentation by human
intestinal epithelial cells. J. Clin. Invest. 102,
792–803 (1998).
136. Westendorf, A. M. et al. CD4+Foxp3+ regulatory T cell
expansion induced by antigen-driven interaction with
intestinal epithelial cells independent of local dendritic
cells. Gut 58, 211–219 (2009).
137. Maggio-Price, L. et al. Lineage targeted MHC‑II
transgenic mice demonstrate the role of dendritic cells
in bacterial-driven colitis. Inflamm. Bowel Dis. 19,
174–184 (2013).
138. Thelemann, C. et al. Interferon-γ induces expression
of MHC class II on intestinal epithelial cells and
protects mice from colitis. PLoS ONE 9, e86844
(2014).
139. Cruickshank, S. M., McVay, L. D., Baumgart, D. C.,
Felsburg, P. J. & Carding, S. R. Colonic epithelial cell
mediated suppression of CD4 T cell activation.
Gut 53, 678–684 (2004).
140. Ward, H. E. & Nicholas, T. E. Alveolar type I and
type II cells. Aust. N. Z. J. Med. 14, 731–734
(1984).
141. Fehrenbach, H. Alveolar epithelial type II cell:
defender of the alveolus revisited. Respir. Res.
2, 33–46 (2001).
142. Strunk, R. C., Eidlen, D. M. & Mason, R. J.
Pulmonary alveolar type II epithelial cells synthesize
and secrete proteins of the classical and alternative
complement pathways. J. Clin. Invest. 81,
1419–1426 (1988).
143. Gereke, M., Jung, S., Buer, J. & Bruder, D.
Alveolar type II epithelial cells present antigen to
CD4+ T cells and induce Foxp3+ regulatory T cells.
Am. J. Respir. Crit. Care Med. 179, 344–355
(2009).
144. Kreisel, D. et al. Cutting edge: MHC class II
expression by pulmonary nonhematopoietic cells
plays a critical role in controlling local inflammatory
responses. J. Immunol. 185, 3809–3813 (2010).
145. Steinman, R. M. Dendritic cells: understanding
immunogenicity. Eur. J. Immunol. 37, S53–S60
(2007).
146. McDevitt, H. O. & Sela, M. Genetic control of the
antibody response. I. Demonstration of determinant-
specific differences in response to synthetic
polypeptide antigens in two strains of inbred mice.
J. Exp. Med. 122, 517–531 (1965).
147. McDevitt, H. O. et al. Genetic control of the immune
response. Mapping of the IR‑1 locus. J. Exp. Med.
135, 1259–1278 (1972).
148. Grumet, F. C. & McDevitt, H. O. Genetic control of
the immune response. Relationship between the
immune response‑1 gene(s) and individual H-2
antigenic specificities. Transplantation 13, 171–173
(1972).
© 2014 Macmillan Publishers Limited. All rights reserved
149. Zinkernagel, R. M. & Doherty, P. C. Restriction of
in vitro T cell-mediated cytotoxicity in lymphocytic
choriomeningitis within a syngeneic or semiallogeneic
system. Nature 248, 701–702 (1974).
150. Banchereau, J. & Steinman, R. M. Dendritic cells and
the control of immunity. Nature 392, 245–252
(1998).
151. Mosier, D. E. A requirement for two cell types
for antibody formation in vitro. Science 158,
1573–1575 (1967).
152. Steinman, R. M. & Cohn, Z. A. Identification of a
novel cell type in peripheral lymphoid organs of mice.
I. Morphology, quantitation, tissue distribution.
J. Exp. Med. 137, 1142–1162 (1973).
153. Steinman, R. M. & Witmer, M. D. Lymphoid dendritic
cells are potent stimulators of the primary mixed
leukocyte reaction in mice. Proc. Natl Acad. Sci. USA
75, 5132–5136 (1978).
154. Blum, J. S., Wearsch, P. A. & Cresswell, P.
Pathways of antigen processing. Annu. Rev. Immunol.
31, 443–473 (2013).
155. Van Niel, G. et al. Intestinal epithelial exosomes
carry MHC class II/peptides able to inform the
immune system in mice. Gut 52, 1690–1697
(2003).
This study suggests that IECs may generate
immunologically active exosomes.
156. Thery, C. et al. Indirect activation of naive CD4+ T cells
by dendritic cell-derived exosomes. Nature Immunol.
3, 1156–1162 (2002).
This is one of the first articles demonstrating
that exosomes may contribute to CD4+ T cell
activation.
157. Thery, C. et al. Molecular characterization of dendritic
cell-derived exosomes. Selective accumulation of the
heat shock protein Hsc73. J. Cell Biol. 147, 599–610
(1999).
158. Mallegol, J., van Niel, G. & Heyman, M. Phenotypic
and functional characterization of intestinal epithelial
exosomes. Blood Cells Mol. Dis. 35, 11–16 (2005).
159. Mallegol, J. et al. T84‑intestinal epithelial exosomes
bear MHC class II/peptide complexes potentiating
antigen presentation by dendritic cells.
Gastroenterology 132, 1866–1876 (2007).
Acknowledgements
Work in the laboratory of T.M.L. is supported by a VA Merit
Award. Research in the laboratory of T.K. is supported by fund‑
ing from the US National Institutes of Health and the American
Asthma Foundation.
Competing interests statement
The authors declare no competing interests.
DATABASES
Immgen Consortium: http://www.immgen.org/
ALL LINKS ARE ACTIVE IN THE ONLINE PDF