Compelling epidemiological evidence reveals a strong association between being overweight or obese and the risk of developing autoimmune diseases (1). From an immunological standpoint, the cellular and molecular mechanisms linked to this association include the overstimulation of T lymphocytes by nutrient- and energy-sensing pathways. The immunometabolic state of an individual is central to the modulation of immunological self-tolerance that suppresses self-reactivity to avoid autoimmunity. Adipose tissue is an immunologically active organ that influences systemic immune responses through the production of adipocytokines, and, in turn, immune cells affect adipocyte homeostasis and metabolism through the production of pro- and anti-inflammatory cytokines (2). This implies that metabolic overload from obesity can affect immunometabolism, which can alter susceptibility to autoimmune diseases.
Immunological adaptations occur in response to nutritional status: Undernutrition impairs immunity, causing inefficient responses to infections and vaccinations. Conversely, overnutrition favors chronic activation of both innate and adaptive immune cells, with subsequent (low-grade) systemic inflammation. These phenomena occur through the engagement of intracellular nutrient- and energy-sensing pathways and the NACHT, LRR and PYD domains–containing protein 3 (NLRP3) inflammasome, which is a sensor of metabolic stress that is induced by an excess of glucose and lipids, especially in macrophages (
2,
3).
Obesity is a risk factor for autoimmune conditions such as type 1 diabetes (T1D) and multiple sclerosis (MS) (
4,
5). Environmental and lifestyle factors that increase MS risk include smoking, sun exposure, low vitamin D, Epstein-Barr virus infection, and high body mass index (BMI). Prospective longitudinal studies in young obese individuals found a 1.6- to 1.9-fold increase in the risk of developing MS during adolescence and young adulthood (but not at the time of MS onset); this association with obesity was also confirmed in carriers of the human leukocyte antigen (HLA)–DRB1*15:01 susceptibility allele that is responsible for the presentation of myelin self-antigens to autoreactive T cells (
5). Similarly, higher BMI at birth is associated with higher T1D susceptibility in children. Indeed, the incidence of T1D increased almost linearly with a higher birth weight (1.7% increase in incidence per 100-g increase in birth weight) (
4).
Mechanistically, it has been suggested that increased body adiposity promotes the hyperactivation of intracellular nutrient- and energy-sensing pathways [such as mechanistic target of rapamycin (mTOR)] with subsequent metabolic overload in peripheral tissues, including resident immune cells that are involved in both effector and regulatory immune responses (
6). For example, in obese naïve-to-treatment MS patients, the adipocytokine leptin (secreted in proportion to BMI to inhibit food intake), together with elevated amounts of circulating nutrients, was found to boost inflammatory immune responses. High levels of leptin and nutrients cause constitutive overactivation of mTOR in T cells, with subsequent dysregulated T cell receptor (TCR)–mediated signaling. Overactive mTOR in T cells mimics a strong, supra-physiological TCR stimulation that is not permissive for transcription of the forkhead-box P3 (
FOXP3) gene, the expression of which is pivotal for the induction and maintenance of anti-inflammatory CD4
+CD25
+FOXP3
+ regulatory T cells (T
regs) (
2,
6). Through leptin overproduction, obesity impairs the proliferation of anti-inflammatory thymic T
regs and their peripheral differentiation from CD4
+CD25
− conventional T (T
conv) cell precursors (
7). Obesity also promotes conversion of T
conv cells into pathogenic inflammatory T helper 1 (T
H1) and T
H17 cells, thus increasing the risk of altered immunological self-tolerance (see the figure).
Overall, nutrient- and leptin-induced mTOR overactivation inhibits peripheral T
reg proliferation and suppressive function and enhances obesity-associated T
H1 and T
H17 cell differentiation, with a higher risk of MS-associated myelin damage (
2,
7). Further, a recent report demonstrated that obese mice converted the classical T
H2-predominant immune responses of atopic dermatitis into a severe disease predominantly characterized by T
H17-driven inflammation that was caused by reduced activity of the peroxisome proliferator-activated receptor-γ (PPAR-γ) transcription factor (
8). The expression of PPAR-− in adipose tissue was also necessary for the development and function of adipose-tissue resident T
regs, suggesting a further bidirectional link between adipose tissue biology and immune tolerance that involves T
regs (
2,
7).
Physiological nutrients and leptin fluctuations due to daily cycles of fasting and feeding determine oscillations in mTOR activity that are lost in obesity because of excessive food intake. Therefore, in individuals with a normal BMI and physiological cycles of feeding and fasting, the maintenance and perpetuation of self-tolerance are associated with oscillations of mTOR activity in T
regs. This appears to be necessary for T
reg expansion and function in sufficient numbers to suppress pathogenic T
H1 and T
H17 cells and thus autoimmunity (
2,
6,
7).
Of note, mTOR represents a key intracellular node at the crossroad of amino acid, glucose, and lipid metabolism. Furthermore, growth factors linked to nutrition and metabolism, such as leptin, insulin, and insulin-like growth factor 1 (IGF-1), activate mTOR signaling in immune cells, which affects systemic and intracellular immunometabolism and thus inflammation and autoimmunity (
2,
6). Adipose tissue also secretes inflammatory cytokines such as interleukin-1 (IL-1), tumor necrosis factor–α (TNF-α), IL-6, IL-17, and interferon-γ (IFN-γ), as well as leptin, which leads to a higher susceptibility to peripheral tissue damage and autoimmunity. Therefore, I propose that metabolic workload—induced by nutrients, adipocyte-derived growth factors, and adipocytokines—may represent an accelerator of autoimmune disorders in people who typically consume an obesogenic Western diet.
It has been demonstrated in mice and in humans that adaptive and innate immune cells can directly influence the pathophysiological events that lead to obesity and obesity-associated metabolic abnormalities (
2). This could also contribute to the reduction in T
reg numbers observed in obese people. There is an anatomical and functional cross-talk between adipose tissue and the immune system. Indeed, both primary lymphoid organs (bone marrow and thymus) and secondary lymphoid organs (lymph nodes) are generally embedded in and surrounded by adipose tissue. This contiguity allows T cells, T
regs, B cells, dendritic cells, and macrophages to home to adipose tissue. Additionally, adipocytes can express immune-like behaviors (
2). For example, adipocytes can clear intracellular bacteria using the same nuclear-binding oligomerization domain 1 (NOD1) pathogen-sensing system of innate immune cells (
9). Changes in T
reg numbers and function observed in obesity may also affect susceptibility to infections and cancer (
2). Indeed, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection is associated with the production of autoantibodies and is more severe in obese individuals (
10). Additionally, cancer immunotherapy responses are better in obese people than in patients with a lower BMI (
2).
Polygenic obesity (predisposition to obesity caused by multiple genetic variants and environmental factors) has also been proposed to be an autoimmune-like disease, whereby T cells respond to unknown adipocyte antigens and trigger subsequent uncontrolled food intake, although the mechanism remains to be fully elucidated (
11). It is notable that CD4
+ T cells isolated from obese mice could transfer an “obesity memory” by promoting weight gain when injected into normal-weight, immune-deficient recipients (
11). Thus, it appears that obesity associates with a higher susceptibility to develop autoimmunity not only because adipose tissue boosts autoinflammatory responses but also because obesity itself has autoimmune-like features.
A promising possibility is the manipulation of immune tolerance and autoimmunity through immunometabolic interventions: reduced food and/or calorie intake. Although the idea that fasting could modulate immune responses and alleviate symptoms of autoimmune diseases had mostly been dismissed, studies over the past 20 years have provided evidence that supports the therapeutic potential of behavioral changes and nutritional strategies such as diet, caloric restriction (CR), and different fasting regimens (
2,
12). Mild CR, intermittent fasting, and a ketogenic diet have each shown beneficial effects in mouse models of autoimmunity, including experimental autoimmune encephalomyelitis (EAE), experimental rheumatoid arthritis, and experimental colitis (
2,
12). I suggest that “starving” pathogenic inflammatory T
H1 and T
H17 cells could lead to better control of local and systemic inflammation. Similarly, CR allows expansion of T
regs in mice and humans by promoting their generation, proliferation, and function, thereby controlling autoimmunity (
2,
7,
12).
Because adherence to dietary changes is not always possible, a proposed alternative approach is “pseudo-starvation,” whereby drugs that regulate immunometabolism mimic fasting (
13). A prototypical example is the mTOR inhibitor rapamycin. Additionally, metformin, an activator of AMP-activated protein kinase (AMPK) that is used to treat type 2 diabetes and overweight individuals, not only controls glucose tolerance but also has anti-inflammatory actions through AMPK-mediated mTOR inhibition (
13). Metformin attenuated EAE induction by restricting the infiltration of mononuclear cells into the central nervous system (CNS) and down-regulating the expression of inflammatory cytokines, inducible NO synthase, cell-adhesion molecules, matrix metalloproteinase-9, and chemokines in T
H17 cells (
14). These effects were also observed in a study of MS patients with metabolic syndrome (
15).
First-line drug treatments for MS (either IFN-β or glatiramer acetate) in combination with metformin provided a statistically significant improvement of disease and reduced CNS lesions. These effects were associated with lowered circulating leptin and T
H1 and T
H17 inflammatory cytokines and increased numbers of peripheral T
regs (
15). Similarly, pioglitazone, an activator of PPAR-γ with antidiabetic effects, also provided a metabolic signal of pseudo-starvation to immune cells from treated MS patients by increasing insulin sensitivity and reducing circulating glucose and leptin levels (
15). In EAE, pioglitazone treatment controlled the disease course with reduced CNS infiltrates and decreased inflammatory cytokine production and T
H1 and T
H17 differentiation (
15). Also, it is interesting to note that classical anti-inflammatory and immunosuppressive drugs such as salicylate and methotrexate can convey metabolic signals of pseudo-starvation to immune cells through the activation of AMPK (
13), along with their classical mechanisms of action. Overall, conferring metabolic signals of pseudo-starvation could be valuable in down-regulating autoinflammatory responses.
It is remarkable that during CR, T cells reprogram their transcriptional signature toward anti-inflammatory properties that limit tissue damage and prolong life span in mice and humans (
7,
12). Also, CR induces extensive adaptations in the gut microbiota toward the production of anti-inflammatory metabolites that affect local and systemic immunometabolism (
12). Molecules that interact with adipocyte-derived leptin can modulate immune function in various ways depending on metabolic status. For example, neuroendocrine mediators with appetite-stimulating activity such as ghrelin and neuropeptide Y have opposite effects from those of leptin, not only on satiety but also on the peripheral immune responses because they block the secretion of T
H1 and T
H17 cytokines and suppress EAE (
2). Remaining areas for study include the molecular dissection of how single nutrients (i.e., lipids, carbohydrates, and proteins) affect immunological self-tolerance and the temporal window in which CR is an effective therapeutic regimen for obesity-associated autoimmunity.
Acknowledgments
G.M. thanks P. de Candia and A. La Cava for critically reading the manuscript and S. Bruzzaniti for assistance with the figure. G.M. is funded by Fondazione Italiana Sclerosi Multipla (FISM grant 2018/S/5), Progetti di Rilevante Interesse Nazionale (PRIN grant 2017 K55HLC 001), Ministero della Salute (grant RF-2019-12371111), and Ministero dell’Università e Ricerca (INF-ACT grant PE00000007). This work is dedicated to the memory of S. Zappacosta and E. Papa.
