The Inflammasome in Atherosclerosis and Type 2 Diabetes

See allHide authors and affiliations

Science Translational Medicine  04 May 2011:
Vol. 3, Issue 81, pp. 81ps17
DOI: 10.1126/scitranslmed.3001902


Atherosclerosis is the cause of morbiditiy for 70% of patients with type 2 diabetes. In both of these diseases, a protein complex known as the inflammasome is stimulated to activate interleukin-1β (IL-1β) and IL-18, which are pathogenic inflammatory cytokines. Triggers for the inflammasome are obesity-related factors, such as cholesterol crystals in atherosclerosis, or hyperglycemia, ceramides, and islet amyloid polypeptide in type 2 diabetes. Therapeutics that target IL-1β in clinical trials for type 2 diabetes might also decrease the incidence of atherosclerosis.



Interleukin-1β (IL-1β) has long been known as a proinflammatory cytokine involved in the pathogenesis of type 2 diabetes (T2D) and atherosclerosis. However, mechanisms that govern IL-1β production in these diseases have been described only recently. Certain features of T2D (oligomers of islet amyloid polypeptide, hyperglycemia and elevated ceramide concentrations) and atherosclerosis (cholesterol crystals) have been shown to stimulate the Nlrp3 [nucleotide-binding domain and leucine rich repeat containing (NLR) family pyrin domain containing 3] inflammasome complex (14). The inflammasome is a large, multiprotein complex that oligomerizes in the cytoplasm of innate immune cells to cleave and activate pro–IL-1β. Other factors, such as minimally oxidized low-density lipoprotein (LDL), free fatty acids (FFAs), and glucose metabolism, also influence pro–IL-1β, thus providing a possible link between obesity and inflammation in both of these diseases. A therapy that targets both IL-1β and IL-1α is safe and has already demonstrated efficacy in T2D (5). The Nlrp3 inflammasome might therefore become a major target against which the next generation of palliative and preventative therapies is designed. This Perspective discusses the pathogenic effect of IL-1β in T2D and atherosclerosis as well as recent advances in inflammasome-related mechanisms in these diseases.


The prevalence of T2D is on the rise worldwide, and atherosclerosis accounts for as much as 70% of the morbidity associated with this disease. Remarkably, individuals with T2D have as high a risk of myocardial infarction as nondiabetic individuals who have had a previous myocardial infarct (6). Although atherosclerosis is typically thought of as an inflammatory disease, the arguments for inflammation in T2D have not been as widely propagated (7). Obesity, which is a predisposing condition for both T2D and atherosclerosis, might be the common link between these two diseases. Inflammatory pathways involved in obesity are now offering unique insights into the relationship between T2D and atherosclerosis and the therapies that might be used against them.

This link between obesity and both T2D and atherosclerosis implicates elevated amounts of glucose, oxidized LDL, and FFAs in disease pathogenesis, potentially as triggers for the production of proinflammatory cytokines by macrophages. Macrophages, which are intricately involved in inflammatory signaling pathways, infiltrate adipose tissue to a greater extent in obese relative to nonobese individuals (8). Adipose tissue is a major source of the cytokines IL-1β, IL-6, tumor necrosis factor–α (TNFα), and IL-18, which all play key roles in systemic inflammation. Chronic systemic inflammation is one factor that contributes to insulin resistance via upregulation of SOCS (suppressor of cytokine signaling) proteins, which attenuate insulin receptor signaling (9, 10), and that affects the vasculature owing to endothelial cell dysfunction (11, 12). Local inflammation is also relevant to atherosclerosis, with oxidative, mitochondrial, and endoplasmic reticulum–associated stress pathways resulting in the recruitment of monocytes and platelets, which contributes to eventual plaque rupture in the vasculature (13, 14). Part of the link between metabolic imbalance and pathogenic inflammation is likely the result of a range of innate immune pathways that are activated in these disease states. For example, in the mouse, loss of the innate immune receptor Toll-like receptor 4 (TLR4) can ameliorate vascular inflammation and insulin resistance caused by a high-fat diet (1517) and reduce aortic plaque area caused by apolipoprotein E (ApoE) deficiency (18). It has been shown in mice that TLR2 is important for the progression of atherosclerosis independent of dietary lipids (19, 20) as well as for insulin resistance initiated by a high-fat diet (21). Stimulation of TLRs leads to the transcription of many cytokines, including pro–IL-1β and pro–IL-18. Nlrp3 expression is also increased, which can then feed back to cleave pro–IL-1β and pro–IL-18 into their active, mature species.

IL-1 AND IL-18

A wealth of new information describing the mechanism of IL-1 activation and its profound effect in inflammatory disease has emerged recently (22). An initial role for IL-1β in diabetes was identified because the cytokine can be toxic to pancreatic β cells (23). Unlike in type 1 diabetes, β cell loss occurs later in the course of T2D; however, arguments that link IL-1β to both diseases are persuasive, especially for T2D (24). For example, a prospective study found that individuals with elevated amounts of IL-1β and IL-6, but not IL-6 alone, had a threefold increased risk of developing T2D (25). Furthering this notion, IL-1 plays an intimate role in glucose metabolism, such as the cytokine’s potent effects on glucose uptake and eventual insulin resistance (26, 27). IL-1β has a more central tenet: fever and inflammation, which are tremendously metabolically demanding, with a 1° per day increase in body temperature being equivalent to running 10 km. Therefore, IL-1β must carefully temper any restriction of gluconeogenesis with a provision for these energy-dependent processes. To highlight this dichotomy, small amounts of IL-1β are actually hypoglycemic, although chronic administration decreases the rate of glucose oxidation in pancreatic islets, contributing to hyperglycemia (28). The basis for this difference could be that at lower concentrations of IL-1β, glucose metabolism is elevated to meet the energetic requirements of the inflammatory process. As IL-1β concentrations rise, however, insulin resistance ensues to limit the potentially dangerous effects of this elevated level of glucose metabolism, such as oxidative stress.

To buffer these delicate IL-1 functions, the body deploys a soluble, decoy type II IL-1 receptor and an IL-1 receptor antagonist (IL-1Ra), which block both IL-1β and IL-1α. IL-1Ra is released from myeloid cells, usually during infection, and from hepatic cells, as an acute-phase response to inflammation (29). An imbalance in the amounts of IL-1Ra compared with IL-1β could potentially be more relevant to disease than the amount of IL-1β alone (24); however, genetic evidence for this hypothesis is still preliminary (30). In mice, disturbing this balance by knocking out IL-1Ra yields mice with a lean phenotype, owing to loss of fat mass and increased energy expenditure (31). However, repeated exposure to increased amounts of IL-1β causes hyperglycemia and insulin resistance in rats (32). It was therefore interesting to test whether administering recombinant IL-1Ra– or IL-1–neutralizing antibodies would affect T2D; and in fact, this approach was very effective in both preventing and treating mouse models of the disease (3336). However, the most compelling findings came from intervention in patients with T2D (5). Key measures of disease, such as elevations in glycated hemoglobin and fasting plasma glucose concentrations, were significantly decreased after a 13-week treatment with IL-1Ra. Even 12 months later, treated individuals—whose inflammatory parameters remained decreased—displayed a reduced proinsulin/insulin ratio and an improved production of C-peptide (a measure of insulin processing) after glucose challenge (37). Notably, there was no correlation between the reduction in systemic inflammatory parameters (C-reactive protein and IL-6) and improved insulin secretion, suggesting that IL-1Ra might function locally near the pancreatic β cell itself (5). However, it was just announced, as this article goes to press, that the results of a larger T2D trial in which IL-1β was specifically neutralized did not show a favorable impact on glucose control. Thus the IL-1Ra effect suggests that IL-1α may also have an important role in the disease.

For T2D, IL-1β was known to be toxic to β cells before elevated amounts of the cytokine were observed. Conversely, in atherosclerosis increased IL-1β was detected before a deleterious role for the cytokine was proven. Multiple cell types in the atherosclerotic plaque express IL-1β, particularly foam cells and endothelial cells (38, 39). It is thought that IL-1β is deleterious because of its direct effects on cell function and because it enhances recruitment of effector cell types into the plaque. Specifically, IL-1β has a negative impact on the health of vascular smooth muscle cells (40, 41) and leads to the recruitment in the vasculature of monocytes and activated platelets, which secrete IL-1β to form an inflammatory feed-forward loop (42, 43). Defeating this loop by deleting IL-1β protects mice from diet-induced atherosclerotic plaque formation in ApoE-deficient mice (44), whereas deleting the antagonist, IL-1Ra, leads to spontaneous, lethal, arterial inflammation in mouse atheroslerosis models (45).

Alongside IL-1β, IL-18 is also processed into its mature, secreted form by a protein complex called the inflammasome (Fig. 1). Plasma IL-18 concentrations are elevated in both T2D and atherosclerosis, and increased IL-18 is predictive of outcome, thus suggesting it has a pathogenic role (46, 47). Despite these links, a genetic association of IL-18 with either disease has been less clear, with both positive and negative findings (48). Similar to IL-1β, IL-18 is predominantly secreted by macrophages in the atherosclerotic plaque (49) and is also expressed in and functions on pancreatic islet cells (50). Increasing IL-18 concentrations enhances the instability of atherosclerotic plaques in ApoE-deficient mice (51) and causes vascular inflammation and remodeling, alongside worsening insulin resistance in fructose-fed rats (52). Considerable effort has now clearly established roles for IL-1β and IL-18 in the manifestations of T2D and atherosclerosis, with the animal models mentioned above strongly suggesting a pathogenic role for these cytokines that will be relevant in humans. Thus, the inflammasome complex that creates their active, secreted forms is of primary interest when researching the cause of these diseases and in the development of more specific therapeutics.

Fig. 1. Priming and activation of the Nlrp3 inflammasome in T2D and atherosclerosis.

Priming (signal 1) of the inflammasome proceeds via a MyD88- and Toll/IL-1 receptor (TIR)–domain–containing adaptor inducing IFN-β (TRIF)–dependent TLR receptor—for example, with TLR4 engaged by mmLDL. Nuclear factor κB (NF-κB)–dependent pathways then generate sufficient Nlrp3 and pro–IL-1β levels that are required to process mature IL-1β via the inflammasome. Activation (signal 2) can be provided by cholesterol crystals in atherosclerosis and IAPP oligomers in T2D. These perturb phagolysosomal processes to augment cathepsin B and/or ROS pathways, which leads to the assembly of the Nlrp3 inflammasome complex: Nlrp3, ASC, and caspase-1 proteins. This complex causes homodimerization of caspase-1, which can then cleave pro–IL-1β and pro–IL-18 into their active, secreted forms. Hyperglycemia can induce mitochondrial ROS and Nlrp3 activation in β cells. Glyburide, a current T2D therapy, inhibits Nlrp3 and subsequent IL-1β processing by an unknown mechanism.



The inflammasome complex is composed of a nucleating receptor protein (Nlrp3), the adaptor protein ASC (apoptosis-associated speck-like protein containing a caspase recruitment domain), and the inflammation-associated protease caspase-1. When Nlrp3 is triggered, ASC is oligomerized to autocatalytically activate caspase-1, which cleaves pro–IL-1β and pro–IL-18 into their mature forms (Fig. 1). Activation of the Nlrp3 inflammasome can be initiated by pathogenic stimuli, such as microbial DNA, RNA, cell-wall components, and toxins, as well as endogenous factors and environmental substances, such as uric acid crystals and asbestos, respectively (53, 54). These substances lack structural similarity that would suggest a conserved method of ligand recognition. Instead, it has been postulated that these different agents might all be able to activate the same inflammasome by perturbing a common intracellular homeostatic parameter (55).

Two main themes have emerged in the discussion of potential mechanisms linking inflammasome triggers to Nlrp3 complex formation. The first is reactive oxygen species (ROS) generation, which is common among many of the Nlrp3 inflammasome triggers. Further evidence comes from studies that show that ROS inhibitors block IL-1β and IL-18 production (5658). The second mechanism of inflammasome activation that has been proposed is lysosomal perturbation. Several inflammasome activators are large particles (for example, the adjuvants alum and polystyrene beads) that are taken up by professional phagocytic cells to trigger IL-1β processing. During this phagocytosis, lysosome acidification is altered and lysosomal enzymes, such as cathepsins B and L, are released, causing inflammasome activation (59).

Of course, it could be that neither of these mechanisms is directly involved in inflammasome activation; or, alternatively, it is conceivable that both pathways influence a single mechanism of activation. This single mechanism might be closely linked to cell death, because all known triggers of the Nlrp3 inflammasome eventually lead to death of the cell after IL-1β production (60). Also, many of these inflammasome activators that require ASC appear to rely on an ionic change in which potassium ions (K+) are shuttled out of the cell, which could represent a common biochemical mechanism of activation (61, 62). Our understanding of how a single mechanism of immune activation can be initiated from multiple stimuli is known as the guard hypothesis (54). In plants, an intracellular protein changes its conformation in response to conditions that counter homeostasis; this protein of altered conformation is then the ligand for an inflammasome-like complex. Thus, the inflammasome might detect the conformation of a protein altered by ROS, lysosome dysfunction, or K+ efflux and then bind to and activate the Nlrp3 protein. The first such candidate is thioredoxin-interacting protein (Txnip), a glucose-inducible, proapoptotic factor for pancreatic β cells. Researchers recently showed that Txnip alters its conformation in response to ROS and is potentially a direct ligand for Nlrp3 (63).


Txnip is the single transcript most highly up-regulated by hyperglycemia in human pancreatic islet cells (64). Alongside reports that elevated glucose can stimulate IL-1β production in β cells (65, 66), the identification of Txnip as a direct ligand for Nlrp3 has provided an elegant explanation for how elevated glucose might activate the inflammasome and lead to increased amounts of IL-1β in T2D. What is particularly attractive is the observation that increased ROS causes a conformational change in Txnip, which leads to dissociation from thioredoxin and, in turn, association with Nlrp3 (63). If the source of ROS is damaged mitochondria, this could tie into mitochondrial dysfunction in β cells during T2D (67). Although Txnip is reported to be a ligand for Nlrp3, experiments that show decreased IL-1β production from Txnip-deficient macrophages have not been reproducible (1). It also appears that IL-1β production by β cells ex vivo in response to increased glucose might be technically challenging because different labs have varied results (65, 66, 68). The physiological importance of this is therefore less clear, and some suggest that this would occur only under conditions of cell stress, such as with serum deprivation in vitro (69). It is also interesting to note that glyburide, a current T2D therapy, might prevent Txnip induction from hyperglycemia in the islet and subsequent IL-1β production (Fig. 1) (63). Unfortunately, the concentrations of glyburide required to block the inflammasome are unlikely to be achieved clinically, and glyburide might not influence constitutively high amounts of Txnip in the macrophage (1, 70). Thus, we currently know that under some conditions hyperglycemia can potentiate the production of IL-1β; however, Txnip might only be a cell-type–specific or a nonunique regulator of this process.

In light of the extensive literature that documents particles, fibers, and amyloid as activators of the inflammasome, we sought a similar species that might be present in T2D. Indeed, there is an amyloid that builds up in the pancreatic islet during disease, primarily composed of the islet amyloid polypeptide (IAPP) (71, 72). Up to 90% of individuals with T2D exhibit amyloid deposits in the pancreas at autopsy; however, many have considered this a result of disease and not a contributing factor. In favor of a pathogenic role of IAPP, studies in primates closely match amyloid deposition with the severity of disease (73). Actually, the cytotoxic species of IAPP might be soluble oligomers, which cannot be observed by staining for insoluble amyloid fibrils (74). We examined a transgenic mouse model of T2D in which human IAPP is overexpressed by β cells (hIAPP transgenic mice). Mouse IAPP does not form amyloid, which might be a reason why mice do not develop spontaneous T2D in the same way that humans do. IL-1β stains strongly in the islets of hIAPP transgenic mice, and on the basis of colocalization, the macrophage is likely to be the cell type responsible (1). We have also shown that when macrophages phagocytose IAPP, the phagolysosomal process goes awry and lysosomal enzymes, such as cathepsins B and L, and lysosomal acidification contribute to the process driving inflammasome activation (1). IL-1α is also produced in response to IAPP after inflammasome activation. Others have shown colocalization of IL-1β and macrophage cell markers in preparations of human islets (75). The number of islet macrophages might also be increased in T2D (76, 77). Unlike studies that found that increased glucose potentiates IL-1β release from β cells ex vivo, a difference in macrophages has not been observed in solutions containing up to 25 mM d-glucose with activation from IAPP (1). Although it is difficult to compare cell types with such different abilities to sense glucose, macrophages do require glucose metabolism for IL-1β production to occur. Therefore, further investigation concerning the role of hyperglycemia in vivo will be informative to determine exactly how glucose concentrations regulate the inflammasome in β cells and macrophages. IL-1 blockade has not yet been tested in the hIAPP transgenic mouse; however, together with hyperglycemia, IAPP could be an important signal for IL-1β production in the islet and β cell death, thus increasing the severity of T2D.

Prolonged administration of IL-1β can initiate insulin resistance and hyperglycemia in the absence of IAPP amyloid formation. This observation suggests a role for IL-1β production before an extensive loss of β cell mass has occurred; indeed, Nlrp3 appears to regulate IL-1β at this early stage of developing insulin resistance in response to high-fat diet–induced T2D in mice (2, 78). This regulatory role for Nlrp3 was detected in adipocytes isolated from caspase-1– or Nlrp3-deficient mice; these cells differentiate more efficiently and are more insulin sensitive than those of control mice (78). Also, adipose tissue macrophages from mice fed a high-fat diet express active inflammasome components. Deletion of Nlrp3 in these mice prevents IL-1β expression and consequent activation of interferon-γ (IFN-γ)–producing T cells that normally promote insulin resistance (2). Because hyperlipidemia is known to influence adipose tissue inflammation, the authors then discovered a role for ceramides as possible inflammatory agents that activate the Nlrp3 inflammasome (2). Together, these findings highlight the complex role of IL-1β activation in T2D: responsiveness to multiple and varied metabolic cues, excess of certain lipid moieties, and inappropriately oligomerized proteins, including IAPP. It is therefore possible that obesity initiates IL-1β production in adipose tissue via Nlrp3 activation; this change then leads to insulin resistance, which in turn promotes insulin and IAPP production. The IAPP oligomers can then activate Nlrp3 to produce IL-1β in the pancreas, leading to β cell death and further accelerating disease progression (Fig. 2).

Fig. 2. Nlrp3 and IL-1β in disease pathogenesis.

Key metabolic imbalances attributable to increased adiposity—hyperlipidemia and hyperglycemia—activate Nlrp3 and, thus, IL-1β. In atherosclerosis, this accelerates lesion formation in which cholesterol crystallizes in macrophages and activates the inflammasome, which can promote plaque rupture. In T2D, insulin resistance further elevates blood glucose concentrations and signals for increased insulin and IAPP production. Excess IAPP can form oligomers that activate the inflammasome in islet macrophages, probably leading to β cell death and disease progression.



Similar to IAPP and ceramides in T2D, crystals of cholesterol activate the Nlrp3 inflammasome in atherosclerosis, which results in IL-1β production (4). Cholesterol-crystal deposition in atherosclerotic vessel walls has long been known to occur but has not been thought of widely as pathogenic in the disease, perhaps owing to their dissolution during routine immunohistochemical procedures. Investigations have since shown that these crystals are present even in the very early stages of disease (Fig. 2) (4). The crucial test of inflammasome involvement in atherosclerosis was to show that mice in which the LDL receptor has been knocked out—which are prone to developing atherosclerotic plaques with cholesterol crystal deposition—have significantly reduced aortic lesion sizes and serum IL-18 concentrations when reconstituted with Nlrp3-, ASC-, and IL-1–deficient bone marrow (4). Similar to IAPP, a perturbed phagolysosomal pathway leads to inflammasome activation by cholesterol crystals in mouse macrophages. Importantly, these findings have been replicated using human macrophages stimulated with cholesterol crystals in vitro, in which treatment with siRNAs that target Nlrp3 decreases IL-1β production (79). It is also noteworthy that there is a large body of work highlighting increased oxidative stress in the atherosclerotic lesion, which suggests a potential for Nlrp3 inflammasome activation by mitochondrial ROS, with a subsequent increase in IL-1β.

Contrary to our discussion so far, some of the current blockbuster therapies against atherosclerosis actually promote the production of IL-1β. This was noted by the effect of simvastatin—an inhibitor of 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase, the rate-limiting enzyme in the cholesterol biosynthesis pathway—which potentiates IL-1β release from monocytes (80). It is thought that this might be the result of a block in an isoprenoid (such as steroids, cholesterols, or vitamins A and E) biosynthesis pathway that is normally driven by mevalonate kinase (MK), the enzyme that catalyzes the step in the cholesterol biosynthesis pathway subsequent to that of HMG-CoA reductase. Patients who lack MK function develop an autoinflammatory IL-1β–mediated disease known as hyperimmunoglobulinemia D with recurrent fever syndrome (HIDS) (81). Therefore, although statins might limit cholesterol-crystal formation by inhibiting biosynthesis, decreasing IL-1β production, there might also be an increase in IL-1β owing to inhibition of the MK pathway. The concentration of simvastatin used to increase IL-1β production in vitro might not be relevant for MK function in vivo; however, it is interesting to speculate that statins might synergize with therapies that target IL-1β in the prevention of atherosclerosis. Treatment of HIDS patients with IL-1Ra can be maintained on a prophylactic basis, so that the initiation of a fever attack is abated by the recombinant antagonist (82). Selective biomarkers and more advanced imaging techniques combined with IL-1β–related therapies might facilitate a more personalized and sophisticated approach to the prevention of atherosclerosis.

LDL is intimately linked with cholesterol accumulation in the atherosclerotic plaque, specifically in macrophages that are termed “foam cells.” It is therefore not surprising that LDL can facilitate cholesterol crystal formation, and in turn inflammasome activation (4). However, what is surprising is that LDL can also participate as “signal 1” (Fig. 1) for activation of the inflammasome by increasing pro–IL-1β concentrations (1, 4). This activation proceeds via stimulation of a complex involving TLR4 (83) and is likely to be mediated specifically by minimally oxidized LDL (mmLDL) (84). Serum mmLDL concentration was associated with disease in a very small group of patients with T2D (85); perhaps serum mmLDL could be a shared risk indicator for both atherosclerosis and T2D. Elevated amounts of FFAs are another metabolic insult that might activate TLRs and thus prime the inflammasome in these disease settings (86, 87). Table 1 summarizes these and other clinical and biochemical similarities between the diseases with respect to inflammasome activation.

Table 1. Possible clinical and biochemical features related to inflammasome activation in atherosclerosis and T2D.


View this table:


Recent research has uncovered common inflammatory pathways that underlie T2D and atherosclerosis. The most important question now is how to translate this knowledge to the clinic, by developing therapeutics for T2D and atherosclerosis that intervene during inflammasome activation. Recombinant IL-1Ra requires daily injections, so is unlikely to become widely used. However, newer biologics with longer half-lives might become popular; examples include IL-1β neutralizing antibodies from Novartis (canakinumab), Lilly (Ly2189102), and Xoma (Xoma 052) that are currently in clinical trials for T2D. Preventing the progression of disease by inhibiting the Nlrp3 inflammasome could also become an important therapeutic approach, especially if biomarkers are discovered that track with treatment. Certainly, small-molecule inhibitors of caspase-1 or Nlrp3 that are cheap and orally available would be desirable, either for prophylactic or palliative use.

In fact, the precise clinical problems with which we began this Perspective—cardiac-related complications in T2D—are preventable with caspase-1 inhibitors, when modeled in rats (88), and normalize insulin responsiveness in diabetes-prone ob/ob (leptin-deficient and obese) mice (78). Western populations have not yet shown a move toward reconciling life-style choice with favorable health outcomes, and it is conceivable that even if appropriate biomarkers of pathogenic inflammation are found and made widely accessible, appropriate behavior changes would remain difficult to implement. In 2008, the American Academy of Pediatrics issued a recommendation to treat obese children with statins starting at the age of 8, to prevent cardiovascular complications (89). This policy demonstrates a symptomatic approach that is taken toward a frequently preventable disease. Further dissection of the molecular mechanisms underlying inflammatory manifestations of these diseases is imperative so that there are more efficient, better-targeted therapies available for the increasing number of patients, both young and old, with established T2D and atherosclerosis.


    References and Notes

    1. Competing interests: The authors declare no competing interests.
    • Citation: S. L. Masters, E. Latz, L. A. J. O’Neill, The Inflammasome in Atherosclerosis and Type 2 Diabetes, Sci. Transl. Med. 3, 81ps17 (2011).

    View Abstract

    Stay Connected to Science Translational Medicine

    Navigate This Article