www.medscape.

com

The Role of Inflammation in Insulitis and -cell Loss in Type 1 Diabetes

Dcio L. Eizirik, Maikel L. Colli, Fernanda Ortis

Abstract and Introduction

Abstract

Type 1 diabetes mellitus (T1DM) is a chronic autoimmune disease with a strong inflammatory component. The latest studies indicate that innate immunity and inflammatory mediators have a much broader role in T1DM than initially assumed. Inflammation might contribute to early induction and amplification of the immune assault against pancreatic cells and, at later stages, to the stabilization and maintenance of insulitis. Inflammatory mediators probably contribute to the suppression of -cell function and subsequent apoptosis; they may also inhibit or stimulate -cell regeneration and might cause peripheral insulin resistance. The different effects of inflammation take place in different phases of the course of T1DM, and should be considered in the context of a 'dialog' between invading immune cells and the target cells. This dialog is mediated both by cytokines and chemokines that are released by cells and immune cells, and by putative, immunogenic signals that are delivered by dying cells. In this Review, we divided the role of inflammation in T1DM into three arbitrary stages: induction, amplification and maintenance or resolution of insulitis. These stages, and their progression or resolution, might depend on a patient's genetic background, which contributes to disease heterogeneity.
Introduction

Inflammation is a biological response that is triggered by infection, tissue injury and tissue stress or malfunction. [1] Adequate control of inflammation is essential for the preservation of tissue integrity. Defective resolution of inflammation and 'misreading' of inflammatory signals increases the risk of developing autoimmune and chronic inflammatory diseases, such as type 1 diabetes mellitus (T1DM). In T1DM, local, pancreatic-islet inflammation (insulitis) contributes to the progressive loss of insulin-producing cells, which eventually renders the patientsmany of them children or adolescentsinsulin-dependent for life. [2,3] The latest advances in this field suggest that inflammatory mediators have a broader role in T1DM than initially assumed: they contribute to the induction and amplification of the immune reaction against the cells and, at later stages, to the stabilization and maintenance of insulitis. Inflammation might contribute to -cell destruction, prolonged suppression of -cell function, inhibition or stimulation of -cell regeneration and peripheral insulin resistance. These different roles of inflammation take place during different phases of the course of T1DM, and might be influenced by patients' genetic background, which contributes to disease heterogeneity. In this Review, we divided the role of inflammation in T1DM into three arbitrary stages: induction, amplification and maintenance or resolution of insulitis (Figure 1). We mostly focus on the latest evidence that supports the different roles of inflammation in T1DM. Special attention is given to unanswered questions and future areas of research, bearing in mind the Buddhist saying that "the ignorant who can see his own ignorance is wise at least so far".

Figure 1.

Inflammatory processes that might contribute to T1DM. Inflammation and its mediators might have a key role in the natural history of T1DM; their contribution is arbitrarily divided into three phases. Of note, these 'disease stages' may either progress (arrow) to the next stage, or may be arrested (double bars) by endogenous control mechanisms that remain to be clarified. The genetic background of the affected individual, in interaction with environmental cues, probably modulates induction, amplification and mantainance or resolution of the disease, which contributes to disease heterogeneity at the different levels. The question marks indicate points in the pathogenesis of T1DM that remain to be proven. Abbreviations: dsRNA = double-stranded RNA; ER = endoplasmic reticulum; T1DM = type 1 diabetes mellitus; T H1 = T helper cell type 1; T H17 = T helper cell type 17; T REG = regulatory T-cells; PRRs = pattern-recognition receptors.

Induction of Insulitis
Innate Immunity and PPRs Recognition by the mammalian immune system of invading micro-organisms depends on innate and adaptive components. The adaptive immune system recognizes antigens derived from micro-organisms by highly diverse T-cell surface receptors that are formed via somatic gene recombination. The innate immune system, on the other hand, recognizes microorganisms by pattern-recognition receptors (PRRs), including Toll-like receptors (TLRs), RIG-I, MDA-5 and receptors of the nucleotide-binding oligomerization domain-like receptor (NLR) family ( ). Activation of these signaling molecules contributes to the development of various autoimmune diseases, such as systemic lupus erythematosis, rheumatoid arthritis and autoimmune hepatitis. [4] Some components of the innate immunity response, including TLRs, contribute to the development of insulitis and T1DM in animal models. [5]
Box 1. The role of inflammation in insulitis and -cell loss in type 1 diabetes

The innate immune system recognizes microorganisms by pattern-recognition receptors (PRRs), which are germlineencoded, limited in number and have a broad specificity against conserved molecular patterns that are present in microbes. Activation of PRRs induces expression of several chemokines and proinflammatory cytokines, and upregulates expression of MHC class II and costimulatory molecules on both dendritic cells and macrophages involved in the inflammatory response. [88] This process enhances antigen presentation by dendritic cells and contributes to the adaptive immune response. Three complementary systems recognize pathogen-associated molecular patterns. The first system includes the Tolllike receptors (TLRs), a family of 13 molecules expressed on the cell surface, endoplasmic reticulum and endosomal compartments of mammalian cells. TLR1, TLR2, TLR4, TLR5, TLR6 and TLR11 are located on the cell surface and detect peptidoglycans, lipoproteins and lipids from bacteria, protozoa and fungi, whereas TLR3, TLR7, TLR8 and TLR9 are mostly intracellular and detect viral and bacterial nucleic acids. [89] The second system includes RIG-I and MDA-5, which detect viral nucleic acids. [90] The third system includes receptors of the nucleotide-binding oligomerization domain-like receptor (NLR) family, which mainly recognizes bacterial products in the intracellular compartment, but can also be activated by endogenous stimuli, such as uric-acid crystals, reactive-oxygen species and changes in intracellular ion concentration. [30] A cytosolic, antiviral pattern, known as the IFN-stimulatory DNA response also senses the presence of foreign nucleic acids within the cell. [91] Discrimination between viral and endogenous nucleic acids is, however, imperfect, which might cause autoimmunity. [4] Mammalian cells have evolved mechanisms to decrease this riskpartly by increasing the discriminatory power of endogenous PRRs pathways [92] and also by negative regulation of these pathways. [93] Ligands for PPRs in Insulitis Mouse and human pancreatic islets express TLR2, TLR3, TLR4 and TLR9; of which TLR3 and TLR4 are expressed at high levels. [6,7,8] The expression of TLR3 is upregulated in cells by double-stranded RNA (dsRNA), [7,9] an intermediate nucleic acid which is generated during the life cycle of most viruses. In human islets that are infected with coxsackievirus B5 or exposed to interferon (IFN)- or IFN- and interleukin (IL)-1, increased expression of TLR3, RIG-I and MDA-5 have been observed. [10,11] Intracellular dsRNA and extracellular dsRNA, which is derived from damaged or dying cells, can both bind to TLR3 and trigger -cell apoptosis and cytokine and chemokine production, at least in part, through activation of the transcription factors NFB and IRF-3. [9-14] However, whereas the activation of these transcription factors by extracellular dsRNA is entirely dependent on TLR3, intracellular dsRNA uses alternative pathways, which might include RIG-I/MDA-5 activation [14] (ML Colli, unpublished data). [9]

Internal dsRNA triggers a massive production of type I interferons. When prolonged or excessive, such interferon release can lead to dsRNA-induced -cell apoptosis, which is partly caused by endoplasmic-reticulum (ER) stress. [14] Importantly, high levels of interferons are present in pancreatic tissue of patients with T1DM, [15] and IFN- contributes to viral-induced experimental diabetes. [16,17] These observations suggest that activation of TLR3 and/or RIG-I and MDA-5 in cells leads to a complex molecular response that starts by activation of the key transcription factors NFB and IRF-3. This activation is followed by production of IFN- and IFN-, which leads to paracrine activation of the transcription factor STAT-1, overexpression of MHC class-I antigens, further production of IFN- and IFN- and the release of several chemokines (see below). [9,10,12,14,18] The result of these combined factors is attraction of immune cells that release proinflammatory cytokines, such as IL-1, tumor necrosis factor (TNF) and IFN- (Figure 2). Local inflammation, coupled with triggering of intracellular and extracellular antiviral defenses, should in most cases eradicate the viral infection. In some genetically susceptible individuals, however, these cellular attempts to eradicate or neutralize the invading virus might go wrong: for example, an exaggerated inflammatory response and/or a defective triggering of intracellular antiinflammatory or antiapoptotic responses could occur, such as expression of JunB, [19] which induces progressive inflammation and prolonged -cell loss. How and why these pathological processes transpire is a critical question that remains to be answered.

Figure 2.

Interaction of cells and immune cells leads to induction and amplification of insulitis, and the transition from innate to adaptative immune response. Recognition of endogenous and exogenous ligands by PRRs (TRL3/4, RIG-I, MDA5) leads to activation of key transcription factors, such as NFB, IRF3 and STATs. Activation of these transcription factors induces the release of chemokines and cytokines that recruit and activate immune cells; increase the cells' expression of MHC class-I antigens (that, when associated with ER stress, might lead to presentation of modified antigens to the immune cells), and activate proapoptotic signals that lead to -cell death. Signals from dying cells are presented by professional antigen-presenting cells that contribute to the activation of autoreactive T cells. During this local inflammation process, pro-inflammatory cytokines (for example, IL-1, TNF and IFN-//) are released by the immune cells and induce transcription factors (e.g STAT-1, NF-B and IRF-3) in cells that contribute to the maintenance and amplification of the network described above. This vicious circle can result in a progressive and selective destruction of pancreatic cells. Some transcription factors might function as negative regulators of the above signaling pathways, for instance, overexpression of JunB prevents cytokine-induced ER stress and -cell apoptosis. Abbreviations: ER = endoplasmic reticulum; IFN = interferon; IL = interleukin; PRR = pattern-recognition receptor; TLR = Toll-like receptor; TNF = tumor necrosis factor; + = stimulation; - = inhibition. The model described above assumes that binding of exogenous ligands (in this case, a virus) to PRRs triggers the innate

immune response. This theory is supported by accumulating evidence that viral infections, especially enteroviruses, have a role in the etiology of T1DM. [20] Some enterovirus strains have specific -cell tropism, and, in a study where pancreatic tissue samples of organ donors were examined, such viruses were detected in the cells from three of six patients with recent-onset T1DM. [21] Viruses might also be associated with T1DM by molecular mimicry: if viral antigens have epitopes which are highly similar to those of self-antigens, T-cell receptors sometimes fail to differentiate properly between these molecules. In these cases, T cells are initially activated by both viral antigens and self-antigens, which triggers an acute inflammatory response. After the resolution of the viral infection, the persistence of self antigens might lead to a chronic autoimmune response. This phenomenon has been elegantly demonstrated in the rat insulin promoter lymphocytic choriomeningitis virus model of diabetes mellitus. [22] For such an effect to occur, however, homology must be complete between the self epitope and the viral epitope, which is unlikely to occur in human disease. In fact, molecular mimicry is probably more relevant to the amplification of an ongoing autoimmune process than to initiate one.
[23]

Another possibility (which does not exclude a pathogenic role of viral infection in some individuals) is that endogenous ligands start the inflammatory process by binding to PRRs. [4] In line with this theory, apoptotic mouse cells that are undergoing secondary necrosis trigger T-cell immunity through a TLR2-initiated signaling pathway. Importantly, autoimmune diabetes in two mouse models was markedly inhibited by TLR2 deletion through impaired activation of T cells by antigen-presenting cells following -cell injury. [24] This finding indicates that -cell death, and its detection by TLR2, is a putative trigger of the development of T1DM. Moreover, pancreatic islets isolated from TLR4-deficient mice (C57BL/10ScNJ background) are protected against allograft rejection when transplanted under the kidney capsule of BALB/c. [25] Although the above studies suggest that TLR2 and TLR4 have a central role in experimental T1DM, this hypothesis was challenged by a study that showed no change in the prevalence of diabetes in nonobese, diabetic (NOD) mice with genetic deletion of TLR2 and TLR4. [26] In this model, contribution of mouse protein MyD88a key signaling molecule in TRL2 and TLR4 pathwaysto the development of autoimmune diabetes depends on the constitution of the intestinal microflora, which suggests a complex interaction between the genetic background, enteric bacteria and the innate immune response. [26] From an immunological perspective, multiple pathways and diverse forms of human T1DM might exist. [27] Inbred NOD mice and BB-rats (a diabetes-prone strain of rat with abnormally low levels of lymphocytes) probably have defects in one of these pathways, which makes it particularly difficult to extrapolate findings on interactions between the environment and innate immune response from these models to the heterogeneous human disease. Additional studies are now required to solve this issue, and to determine the mechanisms by which changes in gut microbiota regulate innate and adaptive immune responses of NOD mice. Polymorphisms of NLRs are associated with other autoimmune diseases and allergies, such as Crohn disease [28] and asthma, [29] but whether they contribute to T1DM remains unclear. These receptors detect intracytoplasmic bacterial molecules and trigger major inflammatory pathways by activation of NFB, MAP-kinases and caspase 1. [30] High amounts of NLRs are expressed in the intestine, and one of them, NOD1, has a key role in communication between intestinal bacteria and the innate and adaptive immune system. [31] NLRs are thus interesting candidates for further study of the interaction between the gut microbiota and the immune response in NOD mice.

Transition to Adaptive Immune Response

A crucial transition, which probably takes place in the early stages of insulitis, determines whether the initial inflammatory response will lead to a full-fledged adaptive immune response, which has the potential to generate a prolonged autoimmune reaction, or will resolve and maintain islet integrity (Figure 1). The role of adaptive immunity and T cells in T1DM is well established. Instead of these factors, we will focus on two other components of insulitis: the dialog between cells and the immune system via local production of chemokines and cytokines, and the putative role of ER stress and other mediators of -cell death on antigen presentation. Dialog Between Cells and the Immune System Chemokines are signal molecules that direct leukocyte migration and activation during the early stages of an innate immune response and contribute to the transition to adaptive immunity. [32] The specificity and complexity of the chemokine system derives from both the release of specific chemokines in various inflammatory reactions, and the regulated expression of their receptors in leukocytes, which varies in different immune responses. [33,34] Expression of chemokines and their receptors is altered in several autoimmune diseases, which seems also to be the case in T1DM. Increased levels of T-helper 1 (T H1) cell-derived chemokines (CCL3, CCL4 and CXCL10) are present in the serum of

patients who have recently been diagnosed as having T1DM, [35-37] and a longitudinal study that included 256 such patients identified a negative correlation between the levels of CCL3 and C-peptide. [38] In this study, however, a major overlap was found between chemokine expression profiles in patients and in controls, which reflects the difficulties in understanding proinflammatory mechanisms in early T1DM from studies that have used traditional approaches to examine serum samples or peripheral cells. Indeed, most of the pathological processes that occur in the early phases of T1DM take place in the islet microenvironment and pancreatic draining lymphnodes. These tissues are difficult to access in humans and, as the islets are just a small part of the pancreas (less than 2%), locally generated inflammatory signals are diluted in the general circulation. To circumvent these difficulties, a recent study has used microarray analysis of healthy, peripheral, mononuclear white blood cells exposed to sera from patients with recent-onset or long-standing T1DM, healthy controls, or siblings of patients with T1DM to identify the presence of proinflammatory factors in the serum. [39] Sera obtained from patients with recent-onset T1DM induced a specific expression pattern, including IL-1 cytokine family members and the chemokines CCL2 and CCL7 (which are both involved in the recruitment of monocytes and macrophages). This proinflammatory 'signature' or expression pattern was apparent years before the clinical onset of T1DM in three patients who were studied. [39] The results obtained with this elegant approach need now to be confirmed in large, prospective studies. Studies in NOD mice have shown increased levels of CXCL10, CCL2, CCL20 and IL-15 mRNAs and/or proteins in pancreatic islets during the prediabetic stage. [40-43] During the course of diabetes, macrophages are the first cells to infiltrate the islets of NOD mice and BB rats, and depletion or inactivation of macrophages prevents the development of the disease. [2] CXCL10 and CCL2 attract macrophages, and their early expression in the islets of NOD mice described above contributes to macrophage recruitment during the early stages of insulitis. Transgenic expression, which leads to high production of CCL2 by cells, causes insulitis and autoimmune diabetes. [43] Of interest, high, basal CCL2 production by human islets correlates with a poor clinical outcome following islet transplantation in patients with T1DM. [44] In the adaptive immune response, diabetogenic T 1 cells in NOD mice express the CCR5 receptor and its ligands H (CCL3, CXC10), as well as XCL1, CCL2, CCL7 and CCL12. [45,46] Deletion of CCL3 in NOD mice ameliorates symptoms of insulitis and prevents autoimmune diabetes, [45] whereas deletion of CCR5 leads to a switch from a T H1 response to a T H2 response, which delays islet-allograft destruction in mice. [47] A relevant role for locally produced cytokines and chemokines was also observed in the mouse model of virus-induced autoimmune diabetes, as mentioned above. In these mice, the blockade of CXCL10, but not that of CCL5, prevented the development of autoimmune diabetes after infection with lymphocytic choriomeningitis virus. [48] Conversely, overexpression of CXCL10 accelerated the onset of T1DM. [48] The absence of CXCR3the receptor for CXCL10, CXCL9 and CXCL11delayed, but not prevented, the onset of insulitis and diabetes, [49] which suggests that these chemokines mainly have a role in the early stages of the disease. An important source of chemokine production during insulitis can be the cells themselves. Isolated rat cells exposed to IL-1 and IFN- or to dsRNA have increased expression of mRNAs that encode several cytokines and chemokines, including CCL2, CXCL10, CCL20, CX3CL1 and IL-15. [13,40,41,50,51] Human islets exposed to IL-1 and IFN- have increased mRNA expression of IL15, CXCL10, CCL2, CCL20 and CX3CL1, and secrete IL15, CCL2, CXCL10, CXCL9, CXCL11 and CCL20 into the culture medium. [40,41] -cell expression of chemokines is mostly regulated by the transcription factors NFB [52,53] and STAT-1. [54,55] These two transcription factors are also key mediators of cytokineinduced -cell death (see below). [56] The findings described above suggest the possibility of a dialog between immune cells and the target cells during the course of insulitis, [34,56] where activated macrophages, natural killer (NK) cells and T cells produce cytokines, such as IFN-, IL-1 and TNF, which induce cells to release chemokines and stimulatory cytokines (Figure 2). These molecules will attract more mononuclear cells that also release multiple chemokines. [34,42] If this vicious circle is not interrupted, it will evolve to progressive accumulation of activated macrophages and T cells around and inside the islets. The nature of the T cells that participate in this amplification phase remains an issue of debate, which is beyond the scope of the present article. Of interest is the recent discovery of T H17 cells: T H17 are potent inducers of tissue inflammation and autoimmunity, [57] and a study in NOD mice suggest that they might have a role in T1DM. [58] If confirmed, these observations should promote a re-evaluation of the role of T H1 and regulatory T cells in T1DM. Taking into account the points discussed above, chemokines and their receptors, or the transcription factors that regulate them, are interesting targets for therapeutic interventions to prevent T1DM. [42,56]

ER Stress in Cells and Antigen Presentation -cell apoptosis is probably the main form of -cell death in patients with T1DM. [2,3] As discussed above, chemokine production and the triggering of -cell apoptosis seem to be regulated by similar intracellular signals, for example by the transcription factors NFB and STAT-1. [56] Thus, -cell death occurs parallel to intense inflammation in the islet microenvironment. Products of dying cells might be perceived as danger signals by the immune system, and antigens that are released from such cellsespecially in the presence of inflammatory factors, such as TNF, interferons and chemokinesmight be taken up by professional antigen-presenting cells in pancreatic lymph nodes and boost the autoimmune response. [59] In support of this scenario, knockout mice that lack caspase 3, the major downstream effector enzyme in the apoptotic pathway, are protected against diabetes that is induced by multiple, low-dose streptozotocin injections, which indicates that -cell apoptosis is a required step for T cell activation. [60] Furthermore, apoptotic cells that undergo secondary necrosis trigger TNF production by macrophages and activate autoreactive T cells (Figure 2). [24] This immune-response-enhancing role of dying cells is probably dependent on the physiological context. Low doses of streptozotocin injected in young NOD mice at the preinsulitis stage decrease the prevalence of diabetes, which suggests that the presence of low numbers of apoptotic cells in a noninflammatory environment might lead to tolerance against cell antigens. [59] These data, however, must be interpreted with caution, as streptozotocin might have effects on the immune system and/or on -cell gene expression, which also affects tolerance induction. Exposure of cells to inflammatory cytokines or to dsRNA induces ER stress, which leads to accumulation of misfolded proteins in the ER and triggers the unfolded-protein response. [14,61,62] The unfolded-protein response aims to alleviate stress on the ER and to restore homeostasis by decreasing the arrival of new proteins; increasing the amount of ER chaperones and increasing the extrusion and subsequent degradation of irreversibly misfolded proteins; when the steps described above fail, apoptosis is triggered. [62] Dying cells can transfer immunologically relevant information to dendritic cells, which signals the nature of cell death and determines the immunological outcome of phagocytosis. [63] Peptides from within the ER of dying cells can be loaded onto MHC class-I molecules in dendritic cells without further cytosolic processing. [64] In this way, dying cells provide antigen-presenting cells with an accurate representation of what happened just before their death. [63] Although this signaling might be beneficial when viral infection has triggered apoptosis, it might have dire consequences for insulin-producing cells. Insulin production represents nearly 50% of the total production of protein by cells, and it accumulates in the ER during periods of increased stress. In this case, insulin accumulates (at least in part) in a misfolded configuration, [65] which might increase its antigenicity. [62] High antigenicity results in increased presentation of proinsulin and insulin to antigen-presenting cells, especially in the presence of inflammation. Although this role of ER stress remains to be proved in T1DM, increased antigenicity due to misfolding of HLA-B27 has been suggested to occur in ankylosing spondylitis. [66] Of note, insulin is a key antigen for autoimmune diabetes in both humans and NOD mice.
[67-70]

Maintenance or Resolution of Insulitis

The late stages of the development of T1DM, characterized by stabilization and maintenance of insulitis or, in some cases, its resolution, are probably dominated by the adaptive immune response. Inflammatory mediators, however, continue to have a relevant role at these stages as well. According to a widely held hypothesis, the clinical onset of T1DM occurs when more than 80% of the cells have been destroyed. Nevertheless, the level of -cell loss might actually be less severe than this value and heterogeneous at disease onset in 4090% of patients with T1DM who are older than 10 years of age. [71] The secretion of insulin in these patients can be also higher than expected: after a mixed meal in individuals with newly diagnosed T1DM, the average insulin level approaches 50% of that observed in nondiabetic controls. [72] How can we explain, then, the severe hyperglycemia that is found in most of these patients at diagnosis? Data from studies in NOD mice suggest that inflammatory mediators, such as cytokines, contribute to both functional impairment of cells and peripheral insulin resistance, as female NOD mice develop a progressive glucose intolerance that parallels the aggravation of insulitis. [73] Islets isolated from such animals in the prediabetic period have impaired glucose metabolism and insulin release, but these functions are recovered in vitro after 1 week in tissue culture, [73] and in vivo if the mice are treated with monoclonal antibodies against effector T cells before isolation of islets. [74] Similar observations were made in islets that were isolated from a 14-year-old patient diagnosed as having T1DM 8 months before her death. [75] These observations suggest that inflammatory molecules released by infiltrating mononuclear cells induce a reversible inhibition of -cell function, which precedes actual -cell destruction. This functional impairment is aggravated, at least in NOD mice, by the presence of severe, inflammation-induced insulin resistance. [76]

This inflammatory response might also modulate -cell proliferation, as is suggested by the observation that -cell proliferation occurs at high rates in prediabetic RIP-IFN- transgenic mice [77] and NOD mice. [78,79] This proliferation is inhibited by immunosuppressive agents, such as anti-CD3 antibodies, [80] which suggests that inflammatory mediators directly induce -cell proliferation. On the other hand, in vitro evidence shows that cytokines, such as IL-1 and IFN-, when used in combination, can induce dedifferentiation of newly generated cells. This effect is mediated by reexpression of the Notch-Delta pathway, [81] and by inhibition of the expression of genes that are particularly relevant for the phenotype and function of mature cells, such as PDX1, ISL1 and receptors of GLP-1 and growth hormone (Ortis F et al., unpublished data). [50,82] Another potentially beneficial role of inflammatory components has been described in a mouse model of geneticallydetermined pancreatitis; in this model, macrophages that express matrix metalloproteinase 9 migrate to the pancreas, where they increase angiogenesis and cell proliferation, which preserves -cell mass in the presence of exocrine degeneration. [83] Patients with type 2 diabetes mellitus might also undergo a mild increase in their number of isletassociated macrophages. [84] Unlike in T1DM, however, this increase occurs in the absence of an autoimmune reaction. [3] This observation raises the intriguing possibility that in type 2 diabetes mellitus, these macrophages might actually support -cell survival (as is suggested to occur in the case of pancreatitis), rather than contribute to -cell death, as observed in T1DM. Further research is required to test this hypothesis, and to determine which components of islet inflammation stimulate or prevent -cell regeneration and survival. In many individuals who develop mild insulitis the inflammation might resolve and normal -cell function be regained. This possibility is supported by observations that some individuals who test positive for islet auto-antibodies and already have impaired -cell function seem to recover this function when followed prospectively, [85] and that most individuals who test positive for islet autoantibodies do not show histological signs of insulitis at postmortem examination. [86] In other tissues, recovery from inflammation depends on an active resolution process, which involves coordinated activation of eicosanoids, resolvins and protectins. [87] Unfortunately, no information is available on the mediators of insulitis resolution, and this area deserves future investigation.

Conclusions
Although T1DM cannot be prevented yet, important advances have been made in our understanding of its pathogenesis. Of particular relevance is our improved understanding of the role of innate immunity and inflammatory mediators in the natural history of T1DM. Early induction and amplification of the immune reaction against pancreatic cells might depend on interactions between endogenous and/or exogenous ligands and PRRs, which are expressed on both cells and immune cells. This interaction can trigger ER stress and apoptosis in cells, and lead to insulitis with local production of chemokines and cytokines. -cell death, which occurs parallel to local release of inflammatory mediators, is perceived as a danger signal by the immune system and might contribute to the transition to a full-fledged autoimmune reaction. The intensity of these early responses probably depends on the genetic background of the affected individual. Even after stabilization of insulitis, inflammatory mediators might continue to have a role in the pathogenesis of T1DM: cytokines contribute to long-term functional suppression and death of cells and modulate -cell regeneration, and might also lead to insulin resistance. Further investigation of the role of innate immunity, PRRs and inflammation in different stages of the natural history of T1DM will improve comprehension of the causes of autoimmunity, and the ultimate mechanisms that lead to -cell loss in T1DM. This knowledge should enable clinicians to design rational and targeted therapies to prevent or revert insulitis and T1DM.

CLICK HERE for subscription information about this journal.

Key Points
Innate immunity and inflammatory mediators have a broad and important role in the pathogenesis of type 1 diabetes mellitus Activation of both endogenous and exogenous ligands of pattern-recognition receptors can induce islet inflammation and death of pancreatic cells Amplification of insulitis might depend on a 'dialog' between immune cells and cells that is mediated by local

production of chemokines and cytokines, and danger signals from dying cells After transition to adaptive immune response, inflammatory mediators, such as cytokines, might contribute to prolonged functional suppression and death of cells, modulation of -cell regeneration and insulin resistance Some inflammatory mediators might promote the survival and proliferation of cells, especially in the absence of autoimmune reaction

Review Criteria
Publications discussed in this Review were identified by searching the PubMed database. Different combinations of the following terms were used: "type 1 diabetes", "inflammation", "innate immunity", "cytokines", "ER stress", "Toll-like receptors", "dsRNA", "pancreatic cells", "pancreatic islets" and "apoptosis mechanisms". A manual search of some references cited in these papers, or in relevant articles related to the role of innate immunity and inflammation in other autoimmune diseases, was also performed. All selected papers were English-language, full-text articles. Many of the relevant references identified could not be included because of space restrictions; when available, recent reviews were preferentially quoted.
References

1. Medzhitov, R. Origin and physiological roles of inflammation. Nature 454, 428-435 (2008). 2. Eizirik, D. L. & Mandrup-Poulsen, T. A choice of deaththe signal-transduction of immune-mediated beta-cell apoptosis. Diabetologia 44, 2115-2133 (2001). 3. Cnop, M. et al. Mechanisms of pancreatic beta-cell death in type 1 and type 2 diabetes: many differences, few similarities. Diabetes 54 (Suppl. 2), S97-S107 (2005). 4. Baccala, R. et al. TLR-dependent and TLR-independent pathways of type I interferon induction in systemic autoimmunity. Nat. Med. 13, 543-551 (2007). 5. Zipris, D. Innate immunity and its role in type 1 diabetes. Curr. Opin. Endocrinol. Diabetes Obes. 15, 326-231 (2008). 6. Vives-Pi, M. et al. Evidence of expression of endotoxin receptors CD14, Toll-like receptors TLR4 and TLR2 and associated molecule MD-2 and of sensitivity to endotoxin (LPS) in islet beta cells. Clin. Exp. Immunol. 133, 208218 (2003). 7. Wen, L. et al. The effect of innate immunity on autoimmune diabetes and the expression of Toll-like receptors on pancreatic islets. J. Immunol. 172, 3173-3180 (2004). 8. Giarratana, N. et al. A vitamin D analog downregulates proinflammatory chemokine production by pancreatic islets inhibiting T-cell recruitment and type 1 diabetes development. J. Immunol. 173, 2280-2287 (2004). 9. Rasschaert, J. et al. Toll-like receptor 3 and STAT-1 contribute to double-stranded RNA + interferon-gammainduced apoptosis in primary pancreatic beta-cells. J. Biol. Chem. 280, 33984-33991 (2005). 10. Ylipaasto, P. et al. Global profiling of coxsackievirus- and cytokine-induced gene expression in human pancreatic islets. Diabetologia 48, 1510-1522 (2005). 11. Hultcrantz, M. et al. Interferons induce an antiviral state in human pancreatic islet cells. Virology 367, 92-101 (2007). 12. Liu, D. et al. Double-stranded ribonucleic acid (RNA) induces beta-cell Fas messenger RNA expression and increases cytokine-induced beta-cell apoptosis. Endocrinology 142, 2593-2599 (2001). 13. Liu, D. et al. Double-stranded RNA cooperates with interferon-gamma and IL-1-beta to induce both chemokine expression and nuclear factor--dependent apoptosis in pancreatic beta-cells: potential mechanisms for viralinduced insulitis and beta-cell death in type 1 diabetes mellitus. Endocrinology 143, 1225-1234 (2002).

14. Dogusan, Z. et al. Double-stranded RNA induces pancreatic beta-cell apoptosis by activation of the TLR3 and IRF3 pathways. Diabetes 57, 1236-1245 (2008). 15. Foulis, A. K. et al. Immunoreactive alpha-interferon in insulin-secreting beta cells in type 1 diabetes mellitus. Lancet 2, 1423-1427 (1987). 16. Devendra, D. et al. Interferon-alpha as a mediator of polyinosinic: polycytidylic acid-induced type 1 diabetes. Diabetes 54, 2549-2556 (2005). 17. Lang, K. S. et al. Toll-like receptor engagement converts T-cell autoreactivity into overt autoimmune disease. Nat. Med. 11, 138-145 (2005). 18. Rasschaert, J. et al. Global profiling of double stranded RNA- and IFN-gamma-induced genes in rat pancreatic beta cells. Diabetologia 46, 1641-1657 (2003). 19. Gurzov, E. N. et al. JunB inhibits ER stress and apoptosis in pancreatic beta-cells. PLoS ONE 3, e3030 (2008). 20. Drescher, K. M. & Tracy, S. M. The CVB and etiology of type 1 diabetes. Curr. Top. Microbiol. Immunol. 323, 259-274 (2008). 21. Dotta, F. et al. Coxsackie B4 virus infection of beta cells and natural killer cell insulitis in recent-onset type 1 diabetic patients. Proc. Natl Acad. Sci. USA 104, 5115-5120 (2007). 22. Christen, U. et al. A viral epitope that mimics a self-antigen can accelerate but not initiate autoimmune diabetes. J. Clin. Invest. 114, 1290-1298 (2004). 23. Filippi, C. M. & von Herrath, M. G. Viral trigger for type 1 diabetes: pros and cons. Diabetes 57, 2863-2871 (2008). 24. Kim, H. S. et al. Toll-like receptor 2 senses beta-cell death and contributes to the initiation of autoimmune diabetes. Immunity 27, 321-333 (2007). 25. Goldberg, A. et al. Toll-like receptor 4 suppression leads to islet allograft survival. FASEB J. 21, 2840-2848 (2007). 26. Wen, L. et al. Innate immunity and intestinal microbiota in the development of type 1 diabetes. Nature 455, 11091113 (2008). 27. Wenzlau, J. M. et al. The cation efflux transporter ZnT8 (Slc30A8) is a major autoantigen in human type 1 diabetes. Proc. Natl Acad. Sci. USA 104, 17040-17045 (2007). 28. Hugot, J. P. et al. Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn's disease. Nature 411, 599-603 (2001). 29. Hysi, P. et al. NOD1 variation, immunoglobulin E and asthma. Hum. Mol. Genet. 14, 935-941 (2005). 30. Kanneganti, T. D. et al. Intracellular NOD-like receptors in host defense and disease. Immunity 27, 549-559 (2007). 31. Bouskra, D. et al. Lymphoid tissue genesis induced by commensals through NOD1 regulates intestinal homeostasis. Nature 456, 507-510 (2008). 32. Charo, I. F. & Ransohoff, R. M. The many roles of chemokines and chemokine receptors in inflammation. N. Engl. J. Med. 354, 610-621 (2006). 33. Arimilli, S. et al. Chemokines in autoimmune diseases. Immunol. Rev. 177, 43-51 (2000). 34. Eizirik, D. L. et al. Use of microarray analysis to unveil transcription factor and gene networks contributing to betacell dysfunction and apoptosis. Ann. NY Acad. Sci. 1005, 55-74 (2003). 35. Shimada, A. et al. Elevated serum IP-10 levels observed in type 1 diabetes. Diabetes Care 24, 510-515 (2001).

36. Nicoletti, F. et al. Serum concentrations of the interferon-gamma-inducible chemokine IP-10/CXCL10 are augmented in both newly diagnosed type I diabetes mellitus patients and subjects at risk of developing the disease. Diabetologia 45, 1107-1110 (2002). 37. Hanifi-Moghaddam, P. et al. Altered chemokine levels in individuals at risk of type 1 diabetes mellitus. Diabet. Med. 23, 156-163 (2006). 38. Pfleger, C. et al. Relation of circulating concentrations of chemokine receptor CCR5 ligands to C-peptide, proinsulin and HbA1c and disease progression in type 1 diabetes. Clin. Immunol. 128, 57-65 (2008). 39. Wang, X. et al. Identification of a molecular signature in human type 1 diabetes mellitus using serum and functional genomics. J. Immunol. 180, 1929-1937 (2008). 40. Chen, M. C. et al. Monocyte chemoattractant protein-1 is expressed in pancreatic islets from prediabetic NOD mice and in interleukin-1beta exposed human and rat islet cells. Diabetologia 44, 325-332 (2001). 41. Cardozo, A. K. et al. IL-1beta and IFN-gamma induce the expression of diverse chemokines and IL-15 in human and rat pancreatic islet cells, and in islets from prediabetic NOD mice. Diabetologia 46, 255-266 (2003). 42. Martin, A. P. et al. Islet expression of M3 uncovers a key role for chemokines in the development and recruitment of diabetogenic cells in NOD mice. Diabetes 57, 387-394 (2008). 43. Martin, A. P. et al. Increased expression of CCL2 in insulin-producing cells of transgenic mice promotes mobilization of myeloid cells from the bone marrow, marked insulitis and diabetes. Diabetes 57, 3025-3033 (2008). 44. Piemonti, L. et al. Human pancreatic islets produce and secrete MCP-1/CCL2: relevance in human islet transplantation. Diabetes 51, 55-65 (2002). 45. Cameron, M. J. et al. Differential expression of CC chemokines and the CCR5 receptor in the pancreas is associated with progression to type 1 diabetes. J. Immunol. 165, 1102-1110 (2000). 46. Bradley, L. M. et al. Islet-specific Th1, but not Th2, cells secrete multiple chemokines and promote rapid induction of autoimmune diabetes. J. Immunol. 162, 2511-2520 (1999). 47. Abdi, R. et al. The role of CC chemokine receptor 5 (CCR5) in islet allograft rejection. Diabetes 51, 2489-2495 (2002). 48. Christen, U. et al. Among CXCR3 chemokines, IFN-gamma-inducible protein of 10 kDa (CXC chemokine ligand (CXCL) 10) but not monokine induced by IFN-gamma (CXCL9) imprints a pattern for the subsequent development of autoimmune disease. J. Immunol. 171, 6838-6845 (2003). 49. Frigerio, S. et al. Beta cells are responsible for CXCR3-mediated T-cell infiltration in insulitis. Nat. Med. 8, 14141420 (2002). 50. Cardozo, A. K. et al. Identification of novel cytokine-induced genes in pancreatic beta-cells by high-density oligonucleotide arrays. Diabetes 50, 909-920 (2001). 51. Gysemans, C. A. et al. Dual role of interferon-[gamma] signaling pathway in sensitivity of pancreatic beta-cells to immune destruction. Diabetologia 44, 567-574 (2001). 52. Cardozo, A. K. et al. A comprehensive analysis of cytokine-induced and nuclear factor B-dependent genes in primary rat pancreatic beta-cells. J. Biol. Chem. 276, 48879-48886 (2001). 53. Eldor, R. et al. Conditional and specific NF-B blockade protects pancreatic beta-cell from diabetogenic agents. Proc. Natl Acad. Sci. USA 103, 5072-5077 (2006). 54. Gysemans, C. A. et al. Disruption of the gamma-interferon signaling pathway at the level of signal transducer and activator of transcription-1 prevents immune destruction of beta-cells. Diabetes 54, 2396-2403 (2005).

55. Callewaert, H. I. et al. Deletion of STAT-1 in pancreatic islets protects against streptozotocin-induced diabetes and early graft failure but not against late rejection. Diabetes 56, 2169-2173 (2007). 56. Eizirik, D. L. et al. Use of a systems biology approach to understand pancreatic beta-cell death in type 1 diabetes. Biochem. Soc. Trans. 36, 321-327 (2008). 57. Bettelli, E. et al. Induction and effector functions of T H17 cells. Nature 453, 1051-1057 (2008). 58. Jain, R. et al. Innocuous IFN-gamma induced by adjuvant-free antigen restores normoglycemia in NOD mice through inhibition of IL-17 production. J. Exp. Med. 205, 207-218 (2008). 59. Filippi, C. M. & von Herrath, M. G. Islet beta-cell deathfuel to sustain autoimmunity? Immunity 27, 183-185 (2007). 60. Liadis, N. et al. Caspase-3-dependent beta-cell apoptosis in the initiation of autoimmune diabetes mellitus. Mol. Cell. Biol. 25, 3620-3629 (2005). 61. Cardozo, A. K. et al. Cytokines downregulate the sarcoendoplasmic reticulum pump Ca 2+ ATPase 2b and deplete endoplasmic reticulum Ca 2+ leading to induction of endoplasmic reticulum stress in pancreatic betacells. Diabetes 54, 452-461 (2005). 62. Eizirik, D. L. et al. The role for endoplasmic reticulum stress in diabetes mellitus. Endocr. Rev. 29, 42-61 (2008). 63. Albert, M. L. Death-defying immunity: do apoptotic cells influence antigen processing and presentation? Nat. Rev. Immunol. 4, 223-231 (2004). 64. Blachere, N. E. et al. Apoptotic cells deliver processed antigen to dendritic cells for crosspresentation. PLoS Biol. 3, e185 (2005). 65. Scheuner, D. & Kaufman, R. J. The unfolded protein response: a pathway that links insulin demand with beta-cell failure and diabetes. Endocr. Rev. 29, 317-333 (2008). 66. Smith, J. A. et al. Pathogenesis of ankylosing spondylitis: current concepts. Best Pract. Res. Clin. Rheumatol. 20, 571-591 (2006). 67. Kent, S. C. et al. Expanded T cells from pancreatic lymph nodes of type 1 diabetic subjects recognize an insulin epitope. Nature 435, 224-228 (2005). 68. Nakayama, M. et al. Prime role for an insulin epitope in the development of type 1 diabetes in NOD mice. Nature 435, 220-223 (2005). 69. Wong, F. S. et al. Identification of an MHC class I-restricted autoantigen in type 1 diabetes by screening an organ-specific cDNA library. Nat. Med. 5, 1026-1031 (1999). 70. Skowera, A. et al. CTLs are targeted to kill beta cells in patients with type 1 diabetes through recognition of a glucose-regulated preproinsulin epitope. J. Clin. Invest. 118, 3390-3402 (2008). 71. Klinke, D. J. Extent of beta cell destruction is important but insufficient to predict the onset of type 1 diabetes mellitus. PLoS ONE 3: e1374 (2008). 72. Steele, C. et al. Insulin secretion in type 1 diabetes. Diabetes 53, 426-433 (2004). 73. Strandell, E. et al. Reversal of beta-cell suppression in vitro in pancreatic islets isolated from nonobese diabetic mice during the phase preceding insulin-dependent diabetes mellitus. J. Clin. Invest. 85, 1944-1950 (1990). 74. Strandell, E. et al. Role of infiltrating T cells for impaired glucose metabolism in pancreatic islets isolated from nonobese diabetic mice. Diabetologia 35, 924-931 (1992). 75. Marchetti, P. et al. Function of pancreatic islets isolated from a type 1 diabetic patient. Diabetes Care 23, 701703 (2000).

76. Koulmanda, M. et al. Modification of adverse inflammation is required to cure new-onset type 1 diabetic hosts. Proc. Natl Acad. Sci. USA 104, 13074-13079 (2007). 77. Sarvetnick, N. E. & Gu, D. Regeneration of pancreatic endocrine cells in interferon-gamma transgenic mice. Adv. Exp. Med. Biol. 321, 85-89 (1992). 78. Sreenan, S. et al. Increased beta-cell proliferation and reduced mass before diabetes onset in the nonobese diabetic mouse. Diabetes 48, 989-996 (1999). 79. Sherry, N. A. et al. Effects of autoimmunity and immune therapy on beta-cell turnover in type 1 diabetes. Diabetes 55, 3238-3245 (2006). 80. Ablamunits, V. et al. Autoimmunity and beta cell regeneration in mouse and human type 1 diabetes: the peace is not enough. Ann. NY Acad. Sci. 1103, 19-32 (2007). 81. Darville, M. I. & Eizirik, D. L. Notch signaling: a mediator of beta-cell dedifferentiation in diabetes? Biochem. Biophys. Res. Commun. 339, 1063-1068 (2006). 82. Kutlu, B. et al. Discovery of gene networks regulating cytokine-induced dysfunction and apoptosis in insulinproducing INS-1 cells. Diabetes 52, 2701-2719 (2003). 83. Tessem, J. S. et al. Critical roles for macrophages in islet angiogenesis and maintenance during pancreatic degeneration. Diabetes 57, 1605-1617 (2008). 84. Ehses, J. A. et al. Increased number of islet-associated macrophages in type 2 diabetes. Diabetes 56, 2356-2370 (2007). 85. Eizirik, D. L. et al. Repair of pancreatic beta-cells. A relevant phenomenon in early IDDM? Diabetes 42, 1383-1391 (1993). 86. In't Veld, P. et al. Screening for insulitis in adult autoantibody-positive organ donors. Diabetes 56, 2400-2404 (2007). 87. Serhan, C. N. et al. Resolving inflammation: dual anti-inflammatory and proresolution lipid mediators. Nat. Rev. Immunol. 8, 349-361 (2008). 88. Medzhitov, R. Recognition of microorganisms and activation of the immune response. Nature 449, 819-826 (2007). 89. Akira, S. et al. Pathogen recognition and innate immunity. Cell 124, 783-801 (2006). 90. Takeuchi, O. & Akira, S. MDA5/RIG-I and virus recognition. Curr. Opin. Immunol. 20, 17-22 (2008). 91. Ishii, K. J. et al. A Toll-like receptor-independent antiviral response induced by double-stranded B-form DNA. Nat. Immunol. 7, 40-48 (2006). 92. Schlee, M. et al. Beyond double-stranded RNA-type I IFN induction by 3pRNA and other viral nucleic acids. Curr. Top. Microbiol. Immunol. 316, 207-230 (2007). 93. Stetson, D. B. et al. Trex1 prevents cell-intrinsic initiation of autoimmunity. Cell 134, 587-598 (2008). Acknowledgments This Review is partially based on a lecture given by DL Eizirik at the 13 th European Association for the Study of Diabetes/Juvenile Diabetes Research Foundation Oxford Workshop, Oxford, UK, 811 August 2008. The concepts presented in this article were developed by the authors during research supported by the European Union (STREP SaveBeta, contract no. 036,903; Framework Program 6 of the European Community), the Fonds National de la Recherche Scientifique, Actions de Recherche Concert de la Communaut Franaise, Belgium and the Belgium Program on Interuniversity Poles of Attraction initiated by the Belgian State (IUAP P5/17 and P6/40). M Colli is the recipient of a scholarship from the Brazilian Coordination for the Improvement of Higher Education Personnel. Reprint Address

D. L. Eizirik, Laboratory of Experimental Medicine, Medical Faculty, University Libre de Bruxelles (ULB), 808 Route de Lennik, B-1070 Brussels, Belgium Email: deizirik@ulb.ac.be 2009 Nature Publishing Group