Type 2 diabetes mellitus - An autoimmune disease?

Abstract

Inflammation-induced inhibition of the insulin signalling pathway can lead to insulin resistance and contribute to the development of type 2 diabetes mellitus (T2DM). Obesity and insulin resistance are associated with a chronic but subclinical inflammatory process that impairs insulin action in most tissues and could also hamper pancreatic β-cell function. The involvement of monocytic cells and the profiles of the chemokines and cytokines induced by this inflammation suggest an innate immune response. However, emerging data indicate that elements of the adaptive immune system could also be involved. As activation of an adaptive response requires antigen specificity, some researchers have hypothesized that T2DM evolves from an innate immune response to an autoimmune condition. In this Perspectives article, we present the arguments for and against this hypothesis and discuss which mechanisms could be involved in a putative switch from innate immunity to autoimmunity.

Full text

NATURE REVIEWS
|
ENDOCRINOLOGY ADVANCE ONLINE PUBLICATION
|
1
Introduction
Type2 diabetes mellitus (T2DM) is among
the most prevalent diseases in modern soci-
eties, affecting over 340 million people in
the world.
1
Obesity is by far the most impor-
tant predisposing factor for T2DM, and the
rapid increase in the prevalence of obesity
is expected to lead to a similar rise in the
prevalence of T2DM.
The consumption of energy-dense and
fat-rich diets is the main environmen-
tal factor leading to obesity.
2
Clinical and
molecular studies from the past 20years
have provided undisputed evidence for the
role of diet-induced subclinical inflamma-
tion as an important link between obesity
and T2DM.
3
Dietary fats activate signal
transduction through Toll-like receptors
(TLRs) 2 and 4, and induce endoplasmic
reticulum stress (ERS) in adipose tissue,
liver and the hypothalamus.
4–7
TLR and
ERS signalling induce inflammatory activ-
ity, which activates intracellular serine–
threonine kinases that inhibit insulin
signal transduction.
8,9
In addition, TLR and
ERS signalling induce inflammatory gene
transcription, resulting in the production
and secretion of cytokines such as tumour
necrosis factor (TNF) and IL-1β that,
through an extracellular feed-forward loop,
boost intracellular inflammatory signalling
and insulin resistance.
8
Induction of ERS by
saturated fats contributes to activation of
inflammatory mediators
10
and triggers the
dysfunction and death of pancreatic βcells
in patients with T2DM.
9,10
Macrophages and TNF are important
for both induction and perpetuation of the
inflammatory activity linked to obesity,
which led to the concept that insulin resist-
ance results from an anomalous activation
of the innate immune system.
3,8,9,11
However,
other studies suggest that components of the
adaptive immune response also participate
in progression of the inflammatory response
associated with obesity.
12–14
Activation of an
adaptive immune response requires pro-
cessing and presentation of speci fic anti-
gens through interaction between the major
histo compatibility com plex (MHC) and the
T-cell receptor (TCR). No infectious agents
are known to trigger inflammation in either
obesity or T2DM. Thus, endogenous anti-
gens or exogenous molecules (originat ing,
for example, from dietary components)
might be presented, act as antigens or
anoma lously modulate the immune response
and thereby activate subsets of lympho cytes
that target host cells and tissues. If evidence
of such a mechanism could be obtained,
obesity (and perhaps T2DM) might have an
autoimmune component.
Our objective in this Perspectives article
is to discuss the nature of the immune
response in T2DM. We also compare the
inflammatory processes occurring in T2DM
with those of the well-defined autoimmune
disease type1 diabetes mellitus (T1DM).
Autoimmunity in T1DM and T2DM
The autoimmune diseases are a diverse
group of chronic illnesses characterized
by an immune response directed against
specific antigens in the body. Well-known
autoimmune diseases include rheumatoid
arthritis, multiple sclerosis, systemic lupus
erythematosus and T1DM. Our understand-
ing of autoimmunity has evolved consider-
ably since the first attempt was made to
define this term in 1957,
15
and continues
to be debated. In our view, a disease is con-
sidered to have an autoimmune aetiology
if most of the following criteria are met:
pathology is associated with loss of tolerance
to specific antigens; transfer of pathogenic
immune cells and/or anti bodies replicates
the disease in previously unaffected indivi-
duals; immunosuppression or immuno-
modulation modifies the natural history
of the disease; animal models point to an
autoimmune aetiology (including disease
transfer by immune cells); and the disease is
associated with genes that regulate the
immune system, such as HLA.
The evidence suggesting an autoimmune
mechanism of disease in T1DM includes
loss of tolerance to several β-cell antigens,
such as insulin, glutamate decarb oxylase2
(also known as GAD-65), receptor-type
tyrosine-protein phosphatase N2 (also
known as phogrin) and zinc transporter8.
16
More over, animal models of T1DM, such
as nonobese diabetic (NOD) mice and bio-
breeding (BB) rats, have a clear autoimmune
cause of diabetes mellitus,
17
and T1DM can
be ‘transferred’ into immune-deficient
animals by T-cell transfer. Furthermore,
in humans, bone marrow transplantation
between HLA-identical siblings can transfer
OPINION
Type2 diabetes mellitus—an autoimmune
disease?
Lício A.Velloso, Decio L.Eizirik and Miriam Cnop
Abstract | Inflammation-induced inhibition of the insulin signalling pathway can
lead to insulin resistance and contribute to the development of type2 diabetes
mellitus (T2DM). Obesity and insulin resistance are associated with a chronic but
subclinical inflammatory process that impairs insulin action in most tissues and
could also hamper pancreatic β-cell function. The involvement of monocytic cells
and the profiles of the chemokines and cytokines induced by this inflammation
suggest an innate immune response. However, emerging data indicate that
elements of the adaptive immune system could also be involved. As activation of an
adaptive response requires antigen specificity, some researchers have hypothesized
that T2DM evolves from an innate immune response to an autoimmune condition. In
this Perspectives article, we present the arguments for and against this hypothesis
and discuss which mechanisms could be involved in a putative switch from innate
immunity to autoimmunity.
Velloso, L. A. etal. Nat. Rev. Endocrinol. advance online publication 9 July 2013;
doi:10.1038/nrendo.2013.131
Competing interests
The authors declare no competing interests.
PERSPECTIVES
© 2013 Macmillan Publishers Limited. All rights reserved

2
|
ADVANCE ONLINE PUBLICATION www.nature.com/nrendo
T1DM, and specific HLA variants can either
protect against or predispose to T1DM.
18
Finally, therapies that inhibit or modulate
T-lymphocyte activity delay the progres-
sive loss of β-cell mass, although they do
not prevent it.
19
T1DM also has a strong
inflammatory component, and inflamma-
tion could contribute to early stages of the
induction and amplification of the immune
reaction against βcells. Furthermore, in
later stages of this immune response, cross-
talk between invading immune cells and the
target β cells could lead to the progressive
loss of β cells.
20
Whether the pathogenesis of T2DM has
an autoimmune component is less clear.
So far, no potential autoantigenic target
for T lymphocytes has been identified in
this setting, but several reports describe
potential targets for IgG antibodies asso-
ciated with insulin resistance
21
and auto-
antibodies against pancreatic islet antigens
in patients with T2DM.
22
The insulin
resistance phenotype is not reported to be
transferred following bone marrow trans-
plantation, although glucose intolerance
and insulin resistance develop in recipients
of IgG antibody transfers in rodents.
21
In
humans, T-lymphocyte inhibition has no
proven therapeutic or preventive role in
T2DM; however, in mouse models, inhibi-
tion of either B cells
21
or Tlymphocytes
14
can
attenuate the progression of obesity-related
insulin resistance. Although no clear evi-
dence of autoimmunity has been described
in animal models of T2DM, targeting com-
ponents of the adaptive immune system,
such as IFN-γ-expressing type 1 Thelper
cells and B lymphocytes, can improve
insulin resistance.
14,21
Of note, the overlap
between candidate genes for T1DM and
T2DM is very limited; only GLIS3 (which
encodes a zinc finger protein) is associated
with both diseases, among over 50 candidate
genes identified.
23,24
Moreover, none of the
known candidate genes for other inflam-
matory or autoimmune diseases, such as,
IL23R, IL2RA, PTPN2 and a number of HLA
alleles, are associated with T2DM.
25,26
Most
genes linked to obesity are postulated to act
on neuroendocrine circuits that regulate
energy balance and have not been impli-
cated in either innate or adaptive immu-
nity.
27
While these genetic variants explain
only a small fraction of the heritability of
obesity and T2DM, they do not provide evi-
dence for a role of autoimmunity in these
two metabolic conditions.
T1DM (in common with several other
autoimmune conditions) is associated with
an increased risk of other autoimmune dis-
eases, but the same cannot be said of T2DM.
How ever, antibodies against G-protein-
coupled receptors have been detected in
sera from a subgroup of T2DM patients
with an increased risk of hypertension and
cardio vascular complications.
28
In addition,
Rho-kinase-activating auto antibodies are
present in sera from T2DM patients with
maculopathy and macro albuminuria,
29
and autoantibodies against IL-6 have
been detected in sera from 2.5% of Danish
patients with T2DM.
30
Adiposity and inflammation
The stromal vascular fraction of the adipose
tissue is populated by macrophages and
other cells of the immune system, includ-
ing T and B lymphocytes.
31,32
In lean
indivi duals, immune cells residing in the
perivascular space of the adipose tissue
provide a first line of defence against poten-
tial pathogens. However, as adipose tissue
mass increases in individuals with obesity,
these macrophages and adaptive immune
cells become activated even in the absence
of a pathogenic threat.
31,33
Increased adipo-
sity is accompanied by increased numbers
of both CD4
+
and CD8
+
T lymphocytes
in the adipose tissue,
31,33
recruited partly
by C-C motif chemokine 5 (also known
as T-cell-specific protein RANTES).
31
In
obese mice, the increase in numbers of
CD8
+
cells precedes the increase in CD4
+
cells and macrophages.
34
In addition,
depletion of CD8
+
lymphocytes protects
mice from diet-induced insulin resistance,
but not from obesity, whereas adoptive
transfer of CD8
+
lymphocytes can induce
adipose tissue inflammation and insulin
resistance in lean, CD8
+
-depleted mice.
34
Upon stimulation, CD8
+
lymphocytes
from obese mice produce more IFN-γ than
do CD8
+
lymphocytes from lean mice, and
adipocytes from obese mice express more
MHC class I molecules than do those of
lean mice.
33
These data suggest that, during
the early development of obesity, antigens
in the adipose tissue could be presented in
an anomalous way and trigger an adaptive
immune response, preferentially through
CD8
+
lymphocytes.
In line with this apparent key role for
CD8
+
cells in the inflammatory adaptive
immune response in hypertrophic adipose
tissue, one study has shown a reduction in
the FOXP3
+
CD4
+
subpopulation in the vis-
ceral fat depots of obese mice and humans.
13
If a similar subset of cells is experimentally
activated in leptin-deficient Lep
ob/ob
mice,
inflammatory infiltration of the adipose
tissue is reduced and hyper glycaemia is
partially corrected.
35
Regulatory Tcells have
an important role in the control of inflam-
matory responses, and down regulation of
these cells might drive the unrestrained
inflammatory response that is driven by
CD8
+
Tcells in obesity.
13
In lean mice, the
proportion of regulatory T lymphocytes
is high; however, the diversity in their
TCR speci ficity is restricted.
36
During the
develop ment of obesity, the number of regu-
latory Tcells in adipose tissue is reduced,
and their TCR diversity is further restricted,
suggesting the presence of potential adipose
tissue autoantigens.
13,36
Anomalous activation of B lympho-
cytes is also involved in the patho genesis of
obesity-linked insulin resistance.
14
Increased
numbers of B lymphocytes infiltrate the
adipose tissue of obese mice, a change
accompanied by an increased blood concen-
tration of IgG
2c
,
14
and depletion of Blympho-
cytes protects mice from obesity-linked
insulin resistance without affecting body
fat mass. Moreover, IgG anti bodies puri-
fied from obese, insulin-resistant mice can
tr ansfer the glucose-intolerance phenotype.
14
Taken together, these findings indicate
that cells and molecules of the adaptive
immune system have a role in the inflam-
matory activity linking obesity and insulin
resistance. The characterization of this phe-
nomenon has potential clinical relevance,
as disorders related to innate immunity
might respond best to cytokine antago-
nism, whereas disorders related to auto-
immunity might preferentially respond to
therapies that target T and Bcells.
37
Indeed,
treatment with anti-inflammatory drugs
(including TNF blockers, IL-1 receptor
antagonists and salicylate) is beneficial in
animal models of T2DM,
3,38
but resulted in
no or modest reductions in HbA
1c
(≤0.5%)
in patients with T2DM.
39,40
Metabolic inflammation in T2DM
Obesity-associated inflammation is not
restricted to adipose tissue; inflammatory
pro cesses also affect other tissues involved in
whole-body energy homeo stasis, and could
promote the development of glucose intoler-
ance and insulin resistance.
8
For example,
insulin-resistant individuals with obesity
can have inflammatory activity in the liver
41
and skeletal muscle.
42
In the hypothalamus,
signs of inflammation are detected as early
as 1day after introducing a high-fat diet
to rodents.
43
Genetic or pharmaco logical
targeting of this inflammatory process
PERSPECTIVES
© 2013 Macmillan Publishers Limited. All rights reserved

NATURE REVIEWS
|
ENDOCRINOLOGY ADVANCE ONLINE PUBLICATION
|
3
ameliorates not only the obese phenotype
(reviewed elsewhere
44
) but also the defec-
tive regulation of hepatic glucose produc-
tion,
45
the impairment of pancreatic β-cell
f unction
46
and hypertension.
47
The presence of low levels of cytokines
and mild tissue infiltration by immune cells
indicates that inflammation is chronic but
low-level, characteristics that are impor-
tant aspects of the inflammatory process in
metabolic tissues. In contrast to the classic
inflammatory process occurring in most
autoimmune conditions, metabolic inflam-
mation is not associated with the typical
signs of inflammatory activity, such as heat,
oedema, redness and pain, in individuals
with obesity and/or T2DM.
48
More over,
instead of increased energy expenditure,
metabolic inflammation results in reduced
whole-body energy expenditure.
48
Despite
these major differences from classic inflam-
matory processes, however, cryptic anti-
gens could still be exposed as a result of
chronic low-level inflammation and lead
to an adaptive autoimmune response, as
proposedelsewhere.
49
β cells, inflammation and T2DM
The inflammatory process in adipose
tissue seems to be similar in all insulin-
resistant individuals with obesity, regard-
less of whether or not they go on to develop
T2DM.
50
Individuals with obesity will only
develop diabetes mellitus when their pan-
creatic β cells fail to compensate for insulin
resistance. Evidence for a role of innate
and adaptive immunity in β-cell failure in
patients who develop T2DM remains con-
troversial. The presence of macrophages
in pancreatic islets (determined by count-
ing the number of CD68
+
cells per islet)
increased from an average of 0.5 cells
in nondiabetic individuals to 1.5 cells in
people with T2DM.
51,52
The proportion
of islets containing at least five immune
cells increased from 0.6% in nondiabetic
control individuals to 5.6% in patients with
T2DM.
52
The presence of these macro-
phages has been linked to viral infection
of islets
52
and expression of human islet
amyloid polypeptide;
53
however, their role
remains unclear. Microarray analysis of
laser-capture microdissected β cells showed
a minor increase (twofold to threefold) in
levels of the chemokines CCL2, CCL11,
CCL13, and CXCL1 and in the cytokines
IL-1β and IL-8 (but not IFN-γ) in cells
from patients with T2DM, compared with
those from nondiabetic controls.
54
These
cytokine and chemokine profiles indicate
mild inflammation, which is consistent
with an innate but not an adaptive immune
response. A similar response was elicited
invitro by exposing human islets to the
saturated fatty acid palmitate or synthetic
ERS inducers, and this process was depend-
ent on IL-1β production.
54
Blocking IL-1β
with an IL-1 receptor antagonist prevented
palmitate-induced inflammation but not
β-cell apoptosis.
54
Consistent with these
findings, the results of microarray analy-
ses of islet cells from patients with T2DM
linked a group of coexpressed genes (among
which many were IL-1-related genes) to
reduced insulin secretion.
55
An autoimmune response to pan creatic
β cells is present in adults with latent
autoimmune diabetes mellitus, a condi-
tion that shares many features with T1DM
and is often misdiagnosed as T2DM.
56
In
some autoantibody-negative patients with
T2DM, responses to islet-reactive Tcells
were observed by measuring mono nuclear
cell responses to human islet proteins
CD1
+
fatty acid
T lymphocytes
Nonautoimmune adaptive response
against fatty acids
Dietary lipids
Lipid raft
TLR4
a
Dietary
lipids
Dead
adipocyte
Cryptic
antigens
Loss of tolerance to self-antigens
and autoimmunity
Regulatory T lymphocytes
Effector T lymphocytes
Loss of tolerance to self-antigens
and autoimmunity
b
d
Changes in
gut microbiota
LPS and fatty acids
Innate response followed by
adaptive response against
fatty acids leads to
no autoimmunity
T
H
17
autoimmunity
c
TCR
Dietary
lipids
Figure 1 | Putative pathways of progression from innate to adaptive immune responses in obesity
and type2 diabetes mellitus. a | Dietary lipids could act as immunomodulators through various
mechanisms: changing the composition of lipid rafts or activating signal transduction through
TLR2 and TLR4. As a result of immunomodulation, inhibition of regulatory Tlymphocytes
and stimulation of effector T lymphocytes could result in loss of tolerance to self-antigens and
autoimmunity. b | Cell death, induced by innate inflammation in response to fatty acids, could
lead to release of cryptic antigens followed by loss of tolerance to self-antigens and
autoimmunity. c | Dietary lipids could be presented to T lymphocytes in the context of CD1–TCR
signal transduction, leading to a nonautoimmune adaptive response. d | Dietary lipids might
induce changes in gut microbiota, resulting in increased blood levels of LPS and fatty acids. This
mechanism could either modulate the immune response through an anomalous T
H
17 response,
contributing in some individuals to autoimmunity, or induce an innate response followed by a
nonautoimmune adaptive response against fatty acids. Abbreviations: TLR, Toll-like receptor;
TCR, Tcell receptor; LPS, lipopolysaccharide; T
H
17, type 17 Thelper cells.
PERSPECTIVES
© 2013 Macmillan Publishers Limited. All rights reserved

4
|
ADVANCE ONLINE PUBLICATION www.nature.com/nrendo
blotted onto nitrocellulose.
57
This method
of evaluating T-cell activation, however, is
under debate, as it might reflect unspeci-
fic inflammation and not autoimmunity,
owing to lack of control tissues or specific
identification of target antigens. The find-
ings suggesting T-cell responses against islet
proteins in patients with T2DM remain to
be confirmed by other research groups
using conventional assays based on speci-
fic β-cell antigens. In contrast to T1DM,
in which nearly all β cells are eventually
destroyed by the auto immune response,
the decrease in β-cell mass in patients with
T2DM is modest (around 30–60%). This
crucial finding suggests that the mecha-
nisms of β-cell loss are different in T1DM
and T2DM.
58
Possible pathways to T2DM
Accumulating evidence suggests that innate
(and, perhaps, adaptive) immunity have a
role in the development of obesity-related
insulin resistance. For example, several
tissues involved in the control of metabolic
functions are affected by complex sub-
clinical inflammatory activity. Moreover,
the insulin-resistance phenotype can be
partly or completely rescued (depending
on the stage of obesity) by targeting either
the innate or the adaptive components of
inflammation.
8,13,14,21,36,44,48,59
In this Perspectives article we raise the
questions of whether autoimmunity has
a pathogenic role in the onset or progres-
sion of the inflammation associated with
obesity and insulin resistance and whether
autoimmunity influences the development
of T2DM. Although current available data
are not sufficient to give definitive answers
to these questions, we propose four possible
pathways that could lead to the onset and/or
progression of the inflammatory response
in both obesity and T2DM (Figure1).
Dietary fatty acids could modulate
immune activity and lead to activation
of an anomalous adaptive immune res-
ponse against self-antigens by, for example,
chang ing the composition of lipid rafts
60
or
by activating signal transduction through
TLR2 and TLR4.
4,6
In this pathway, fatty
acids could negatively modulate regulatory
T lymphocytes and/or positively modulate
effec tor Tlymphocytes in metabolically rele-
vant tissues, resulting in loss of tolerance to
self-antigens, thus leading to au to immunity
(Figure1a).
Alternatively, fatty acids could trigger an
innate immune response and lead to cell
death in relevant tissues, such as adipose
tissue, exposing cryptic antigens and activat-
ing an anomalous adaptive immune response
against self-antigens (Figure1b). Dietary
lipids activate the innate immune response in
most metabolically relevant tissues,
8–10,44
and
chronic inflammation can result in the death
of key cell types, including adipocytes and
neurons, in at least some of these tissues.
8,44
In this pathway, the cryptic antigens released
as a consequence of i nflammation-induced
cell death could activate an adaptive immune
response against self-antigens and thereby
lead to autoimmunity.
Another possibility is that dietary fatty
acids activate an anomalous adaptive
immune response without the development
of autoimmunity (Figure1c). Endogenous
and exogenous lipids can be presented to
T lymphocytes, and the CD1–TCR signal-
ling system would have a central role in this
process.
61,62
In this pathway, dietary fatty
acids could trigger T-lymphocyte activation
through CD1-mediated antigen presen-
tation, and the ensuing adaptive immune
response would target dietary fats without
leading to autoimmune activity.
Finally, changes in gut microbiota
could result in increased blood levels of
lipo polysaccharide and/or modulation
of immune activity, which might lead
to activation of either nonauto immune,
adap tive res ponses against fatty acids or
of auto immune responses against cryptic
antigens (Figure1d). Changes in the gut
microbiota influence both obesity-linked
systemic inflammation and insulin resist-
ance.
63
The inflammation associated with
diet-induced changes in microbiota can
result from increased gut transposition of
lipo polysaccharides as well as changes in
the absorption of nutrients, particularly
lipids.
63–65
More over, changes in micro-More over, changes in micro-
biota can modify IL-17 production in the
gut, which modulates systemic immune
activity.
66
This pathway can have two pos-
sible outcomes: in the first, changes in lipo-
polysaccharide and lipid absorption in the
gut could result in increased innate immune
activity, followed by the activation of an
adaptive immune response against dietary
lipids, with no autoimmune activity; in
the second, changes in immune regulation
caused by upregulation of type17 Thelper
cells (which increases IL-17 production)
favours the development of autoimmunity.
67
Another interesting possibility is that
multiple pathways might be involved: for
example, the direct activation of an innate
immune response by fatty acids (as pro-
posed in Figure1b), could occur in parallel
with lipid presentation through CD1–TCR,
resulting in an innate immune response as
well as nonautoimmune adaptive immune
activity. Other pathways could possibly also
explain the immune response in obesity
and T2DM. The results of well-controlled
experiments in the future will indicate
whether autoimmunity is indeed activated
in the pathogenesis of T2DM.
Conclusions
The presence of autoimmune activity in the
inflammation linked to obesity and insulin
resistance remains elusive. Even if adaptive
immunity is indeed present, it might not
necessarily have a pathogenic role in the
development of T2DM. Failure of the βcell
is central to the pathogenesis of T2DM;
however, although considerable evidence
supports a pathogenic role for islet inflam-
mation and autoimmunity in T1DM, no
conclusive data suggest that auto immunity
has any role in β-cell death in T2DM. The
autoimmune response in T1DM targets
mainly the pancreatic β cells. By contrast,
the chronic, systemic low-grade inflam-
mation in T2DM affects a variety of tissues,
including adipose tissue, liver, the hypo-
thalamus, β cells and the cardio vascular
system. This broad spectrum of activity
suggests the presence of a metabolically
driven innate immune response rather than
an auto immune disease. Although an innate
immune response evolve into an adap-
tive immune response in adipose tissue, it
has not been shown to do so in β cells in
models of obesity and T2DM.
54,58
Thus, cur-
rently, T2DM cannot be characterized as
an autoimmune condition. Further studies
are required to clarify this important and
therapeutical ly relevant question.
Laboratory of Cell Signalling, Obesity and
Comorbidities Research Centre, University of
Campinas, DCM-FCM UNICAMP, 13,084-970
Campinas, São Paulo, Brazil (L. A. Velloso).
Laboratory of Experimental Medicine, Medical
Faculty, Campus Erasme, 808 Route de Lennik
(D. L. Eizirik), Division of Endocrinology, Erasmus
Hospital, Campus Erasme, 806 Route de Lennik
(M.Cnop), Université Libre de Bruxelles, B-1070
Brussels, Belgium.
Correspondence to: L. A. Velloso
lavelloso.unicamp@gmail.com
1. Danaei, G. etal. National, regional, and global
trends in fasting plasma glucose and diabetes
prevalence since 1980: systematic analysis of
health examination surveys and
epidemiological studies with 370 country-years
and 2.7 million participants. Lancet 378,
31–40 (2011).
2. Schulz, M. etal. Food groups as predictors for
short-term weight changes in men and women
PERSPECTIVES
© 2013 Macmillan Publishers Limited. All rights reserved

NATURE REVIEWS
|
ENDOCRINOLOGY ADVANCE ONLINE PUBLICATION
|
5
of the EPIC-Potsdam cohort. J. Nutr. 132,
1335–1340 (2002).
3. Shoelson, S.E., Lee, J. & Goldfine, A.B.
Inflammation and insulin resistance. J. Clin.
Invest. 116, 1793–1801 (2006).
4. Tsukumo, D.M. etal. Loss-of-function mutation
in Toll-like receptor 4 prevents diet-induced
obesity and insulin resistance. Diabetes 56,
1986–1998 (2007).
5. Ozcan, U. etal. Endoplasmic reticulum stress
links obesity, insulin action, and type2
diabetes. Science 306, 457–461 (2004).
6. Milanski, M. etal. Saturated fatty acids
produce an inflammatory response
predominantly through the activation of TLR4
signaling in hypothalamus: implications for the
pathogenesis of obesity. J. Neurosci. 29,
359–370 (2009).
7. Pal, D. etal. Fetuin-A acts as an endogenous
ligand of TLR4 to promote lipid-induced insulin
resistance. Nat. Med. 18, 1279–1285 (2012).
8. Gregor, M.F. & Hotamisligil, G.S. Inflammatory
mechanisms in obesity. Annu. Rev. Immunol.
29, 415–445 (2011).
9. Cnop, M., Foufelle, F. & Velloso, L.A.
Endoplasmic reticulum stress, obesity and
diabetes. Trends Mol. Med. 18, 59–68 (2012).
10. Eizirik, D.L., Miani, M. & Cardozo, A.K.
Signalling danger: endoplasmic reticulum
stress and the unfolded protein response in
pancreatic islet inflammation. Diabetologia 56,
234–241 (2013).
11. Osborn, O. & Olefsky, J.M. The cellular and
signaling networks linking the immune system
and metabolism in disease. Nat. Med. 18,
363–374 (2012).
12. Guarda, G. etal. Tcells dampen innate immune
responses through inhibition of NLRP1 and
NLRP3 inflammasomes. Nature 460, 269–273
(2009).
13. Feuerer, M. etal. Lean, but not obese, fat is
enriched for a unique population of regulatory
Tcells that affect metabolic parameters. Nat.
Med. 15, 930–939 (2009).
14. Winer, S. etal. Normalization of obesity-
associated insulin resistance through
immunotherapy. Nat. Med. 15, 921–929
(2009).
15. Witebsky, E., Rose, N.R., Terplan, K.,
Paine,J.R. & Egan, R.W. Chronic thyroiditis
and autoimmunization. J. Am. Med. Assoc. 164,
1439–1447 (1957).
16. Ziegler, A.G. & Nepom, G.T. Prediction and
pathogenesis in type1 diabetes. Immunity 32,
468–478 (2010).
17. Yang, Y. & Santamaria, P. Lessons on
autoimmune diabetes from animal models.
Clin. Sci. (Lond.) 110, 627–639 (2006).
18. Roep, B.O. & Peakman, M. Diabetogenic
Tlymphocytes in human type1 diabetes. Curr.
Opin. Immunol. 23, 746–753 (2011).
19. von Herrath, M., Peakman, M. & Roep, B.
Progress in immune-based therapies for type1
diabetes. Clin. Exp. Immunol. 172, 186–202
(2013).
20. Eizirik, D.L., Colli, M.L. & Ortis, F. The role of
inflammation in insulitis and β-cell loss in
type1 diabetes. Nat. Rev. Endocrinol. 5,
219–226 (2009).
21. Winer, D.A. etal. Bcells promote insulin
resistance through modulation of Tcells and
production of pathogenic IgG antibodies. Nat.
Med. 17, 610–617 (2011).
22. Brooks-Worrell, B.M., Juneja, R.,
Minokadeh,A., Greenbaum, C.J. & Palmer, J.P.
Cellular immune responses to human islet
proteins in antibody-positive type2 diabetic
patients. Diabetes 48, 983–988 (1999).
23. Barrett, J.C. etal. Genome-wide association
study and meta-analysis find that over 40 loci
affect risk of type1 diabetes. Nat. Genet. 41,
703–707 (2009).
24. Dupuis, J. etal. New genetic loci implicated in
fasting glucose homeostasis and their impact
on type2 diabetes risk. Nat. Genet. 42,
105–116 (2010).
25. Rafiq, S. etal. Gene variants influencing
measures of inflammation or predisposing to
autoimmune and inflammatory diseases are
not associated with the risk of type2 diabetes.
Diabetologia 51, 2205–2213 (2008).
26. Winkler, C., Raab, J., Grallert, H. & Ziegler, A.G.
Lack of association of type2 diabetes
susceptibility genotypes and body weight on
the development of islet autoimmunity and
type1 diabetes. PLoS ONE 7, e35410 (2012).
27. Speliotes, E.K. etal. Association analyses of
249,796 individuals reveal 18 new loci
associated with body mass index. Nat. Genet.
42, 937–948 (2010).
28. Hempel, P. etal. Sera from patients with type2
diabetes contain agonistic autoantibodies
against G protein-coupled receptors. Scand. J.
Immunol. 70, 159–160 (2009).
29. Zimering, M.B. & Pan, Z. Autoantibodies
in type2 diabetes induce stress fiber
formation and apoptosis in endothelial cells.
J.Clin. Endocrinol. Metab. 94, 2171–2177
(2009).
30. Fosgerau, K. etal. Interleukin-6 autoantibodies
are involved in the pathogenesis of a subset of
type2 diabetes. J. Endocrinol. 204, 265–273
(2010).
31. Wu, H. etal. T-cell accumulation and regulated
on activation, normal Tcell expressed and
secreted upregulation in adipose tissue in
obesity. Circulation 115, 1029–1038 (2007).
32. Corre, J. etal. Human subcutaneous adipose
cells support complete differentiation but not
self-renewal of hematopoietic progenitors.
J.Cell Physiol. 208, 282–288 (2006).
33. Rocha, V.Z. etal. Interferon-gamma, a Th1
cytokine, regulates fat inflammation: a role for
adaptive immunity in obesity. Circ. Res. 103,
467–476 (2008).
34. Nishimura, S. etal. CD8+ effector Tcells
contribute to macrophage recruitment and
adipose tissue inflammation in obesity. Nat.
Med. 15, 914–920 (2009).
35. Ilan, Y. etal. Induction of regulatory Tcells
decreases adipose inflammation and alleviates
insulin resistance in ob/ob mice. Proc. Natl
Acad. Sci. USA 107, 9765–9770 (2010).
36. Yang, H. etal. Obesity increases the production
of proinflammatory mediators from adipose
tissue Tcells and compromises TCR repertoire
diversity: implications for systemic
inflammation and insulin resistance.
J.Immunol. 185, 1836–1845 (2010).
37. Savic, S., Dickie, L.J., Wittmann, M.
& McDermott, M.F. Autoinflammatory
syndromes and cellular responses to stress:
pathophysiology, diagnosis and new treatment
perspectives. Best Pract. Res. Clin. Rheumatol.
26, 505–533 (2012).
38. Araujo, E.P. etal. Infliximab restores glucose
homeostasis in an animal model of diet-
induced obesity and diabetes. Endocrinology
148, 5991–5997 (2007).
39. Larsen, C.M. etal. Interleukin-1-receptor
antagonist in type2 diabetes mellitus. N. Engl.
J. Med. 356, 1517–1526 (2007).
40. Goldfine, A.B. etal. The effects of salsalate on
glycemic control in patients with type2
diabetes: a randomized trial. Ann. Intern. Med.
152, 346–357 (2010).
41. Tuncman, G. etal. Functional invivo
interactions between JNK1 and JNK2
isoforms in obesity and insulin resistance.
Proc. Natl Acad. Sci. USA 103, 10741–10746
(2006).
42. Perreault, M. & Marette, A. Targeted
disruption of inducible nitric oxide synthase
protects against obesity-linked insulin
resistance in muscle. Nat. Med. 7,
1138–1143 (2001).
43. Thaler, J.P. etal. Obesity is associated with
hypothalamic injury in rodents and humans.
J.Clin. Invest. 122, 153–162 (2012).
44. Velloso, L.A. & Schwartz, M.W. Altered
hypothalamic function in diet-induced obesity.
Int. J. Obes (Lond.) 35, 1455–1465 (2011).
45. Milanski, M. etal. Inhibition of hypothalamic
inflammation reverses diet-induced insulin
resistance in the liver. Diabetes 61,
1455–1462 (2012).
46. Calegari, V.C. etal. Inflammation of the
hypothalamus leads to defective pancreatic
islet function. J. Biol. Chem. 286,
12870–12880 (2011).
47. Purkayastha, S., Zhang, G. & Cai, D.
Uncoupling the mechanisms of obesity
and hypertension by targeting hypothalamic
IKK-β and NF-κB. Nat. Med. 17, 883–887
(2011).
48. Calay, E.S. & Hotamisligil, G.S. Turning off the
inflammatory, but not the metabolic, flames.
Nat. Med. 19, 265–267 (2013).
49. Brooks-Worrell, B., Narla, R. & Palmer, J.P.
Biomarkers and immune-modulating therapies
for type2 diabetes. Trends Immunol. 33,
546–553 (2012).
50. Barbarroja, N. etal. Progression from high
insulin resistance to type2 diabetes does not
entail additional visceral adipose tissue
inflammation. PLoS ONE 7, e48155 (2012).
51. Ehses, J.A. etal. Increased number of islet-
associated macrophages in type2 diabetes.
Diabetes 56, 2356–2370 (2007).
52. Richardson, S.J., Willcox, A., Bone, A.J.,
Foulis,A.K. & Morgan, N.G. Islet-associated
macrophages in type2 diabetes. Diabetologia
52, 1686–1688 (2009).
53. Masters, S.L. etal. Activation of the NLRP3
inflammasome by islet amyloid polypeptide
provides a mechanism for enhanced IL-1β in
type2 diabetes. Nat. Immunol. 11, 897–904
(2010).
54. Igoillo-Esteve, M. etal. Palmitate induces a pro-
inflammatory response in human pancreatic
islets that mimics CCL2 expression by β cells
in type2 diabetes. Diabetologia 53,
1395–1405 (2010).
55. Mahdi, T. etal. Secreted frizzled-related
protein4 reduces insulin secretion and is
overexpressed in type2 diabetes. Cell Metab.
16, 625–633 (2012).
56. Tuomi, T. etal. Antibodies to glutamic acid
decarboxylase reveal latent autoimmune
diabetes mellitus in adults with a
non-insulin-dependent onset of disease.
Diabetes 42, 359–362 (1993).
57. Brooks-Worrell, B. & Palmer, J.P. Immunology
in the Clinic Review Series; focus on metabolic
diseases: development of islet autoimmune
disease in type2 diabetes patients: potential
sequelae of chronic inflammation. Clin. Exp.
Immunol. 167, 40–46 (2012).
58. Cnop, M. etal. Mechanisms of pancreatic β-cell
death in type1 and type2 diabetes:
manydifferences, few similarities. Diabetes 54
(Suppl. 2), S97–S107 (2005).
59. Reilly, S.M. etal. An inhibitor of the protein
kinases TBK1 and IKK-ε improves obesity-related
PERSPECTIVES
© 2013 Macmillan Publishers Limited. All rights reserved

6
|
ADVANCE ONLINE PUBLICATION www.nature.com/nrendo
metabolic dysfunctions in mice. Nat. Med. 19,
313–321 (2013).
60. Brassard, P. etal. Modulation of T-cell signalling
by non-esterified fatty acids. Prostaglandins
Leukot. Essent. Fatty Acids 77, 337–343 (2007).
61. Beckman, E.M. etal. Recognition of a lipid
antigen by CD1-restricted αβ
+
Tcells. Nature
372, 691–694 (1994).
62. Cohen, N.R., Garg, S. & Brenner, M.B. Antigen
presentation by CD1 lipids, T cells, and
NKTcells in microbial immunity. Adv. Immunol.
102, 1–94 (2009).
63. Cani, P.D. etal. Metabolic endotoxemia
initiates obesity and insulin resistance.
Diabetes 56, 1761–1772 (2007).
64. Turnbaugh, P.J. etal. An obesity-associated
gut microbiome with increased capacity for
energy harvest. Nature 444, 1027–1031
(2006).
65. Caricilli, A.M. etal. Gut microbiota is a key
modulator of insulin resistance in TLR 2
knockout mice. PLoS Biol. 9, e1001212
(2011).
66. Geddes, K. etal. Identification of an innate
Thelper type17 response to intestinal
bacterial pathogens. Nat. Med. 17, 837–844
(2011).
67. Atarashi, K. & Honda, K. Microbiota in
autoimmunity and tolerance. Curr. Opin.
Immunol. 23, 761–768 (2011).
Acknowledgements
The authors’ research is supported by Fundação de
Amparo à Pesquisa do Estado de São Paulo
(Brazil), the Communauté Française de Belgique
—Actions de Recherche Concertées, and the
European Union projects Naimit and BetaBat,
in the Framework Programme 7 of the European
Community. The authors are also grateful to
MarkPeakman of King’s College London, UK, and
Bart Roep of Leiden University Medical Centre,
TheNetherlands, for helpful discussions.
Author contributions
The authors contributed equally to all aspects of
thisarticle.
PERSPECTIVES
© 2013 Macmillan Publishers Limited. All rights reserved