Macrophages, fat, and the emergence of immunometabolism

Anthony W. Ferrante Jr.

J Clin Invest. 2013;123(12):4992-4993. https://doi.org/10.1172/JCI73658.

Hindsight

Over the last century, our modern concepts and understanding of and have evolved largely in parallel. Notably, during the last decade, there has been a sharpened focus on the convergence of metabolism and immune function. In part motivated by studies originally published in the JCI, we now recognize that the immune system monitors the metabolic state of tissues and and responds in kind by modulating metabolic function. The complexity of these interactions, both adaptive and pathologic, continues to be studied and revealed, with the hope that harnessing the reins that control immune function may one day be used for metabolic benefit.

Find the latest version: https://jci.me/73658/pdf Hindsight Macrophages, fat, and the emergence of immunometabolism Anthony W. Ferrante Jr.

Department of , Naomi Berrie Diabetes Center, Columbia University, New York, New York, USA.

Over the last century, our modern concepts and understanding of metabo- list, it became apparent that many of the lism and immunology have evolved largely in parallel. Notably, during the genes that positively correlated with adi- last decade, there has been a sharpened focus on the convergence of metab- posity in adipose tissue were familiar to me olism and immune function. In part motivated by studies originally pub- from my time in the Stanley laboratory. I lished in the JCI, we now recognize that the immune system monitors the called Richard and asked whether he could metabolic state of tissues and organisms and responds in kind by modulat- provide me with antibodies and some ing metabolic function. The complexity of these interactions, both adaptive macrophage-deficient mice to test whether and pathologic, continues to be studied and revealed, with the hope that the gene signatures were indeed reflective harnessing the reins that control immune function may one day be used for of the macrophage content of the adipose metabolic benefit. tissue. The immunohistochemistry was quite striking. Though we were originally Inflammatory pathways implicated initial observations to demonstrate that doubtful that such a simple explanation in metabolic disease activates the expression of more could explain our expression data, we were For more than 50 years, physicians and than 50 inflammatory molecules in fat (4). quickly convinced that typical rodent and scientists have observed a tight associa- human adipose tissues contain macro- tion between metabolic disorders and sys- The correlation between phages at levels proportionate to adipos- temic inflammation. Studies dating from macrophages and adiposity ity. We also demonstrated that most of the the 1960s found that individuals with Ten years ago Hong Chen, then leading a TNF-α and several other classical inflam- what we now recognize as group at Millenium Pharmaceuticals, and matory molecules were derived from adi- mellitus have higher circulating concen- our laboratory separately studied expres- pose tissue macrophages. trations of active complement and acute sion profiles in adipose tissue of mice. We phase reactants (1, 2), both of which are both surprisingly found transcriptional The intersection of metabolic classic markers of an inflammatory state. evidence for the presence of macrophages dysfunction and inflammation Inflammation is a broad term applied (5, 6). In our study, we profiled adipose These initial observations provided a new to a variety of processes, but most often tissue from mice that varied widely in adi- context for understanding obesity-induced indicates stress to an underlying system. posity due to differences in sex, diet, and inflammation (7). What had previously In the context of metabolic disease, the segregation of alleles that cause genetic been thought to be the isolated release of site of inflammatory stress and the physi- forms of obesity. Because adiposity is a inflammatory molecules by adipocytes in ologic and pathologic implications of the quantitative trait that correlates with med- response to hypertrophy now clearly rep- inflammatory source remained obscure ically important quantitative phenotypes, resented a complex conversation between until about 20 years ago. we reasoned that transcriptional profiling stressed adipocytes and the immune sys- At that time, Gokhan Hotamisligil, might reveal adipose tissue alterations tem. In the intervening years, work by working in Bruce Spiegelman’s laboratory, that varied with mass. In what seems to many laboratories has revealed that obe- found that obesity induced expression of be a relatively simplistic approach by cur- sity induces a complex and broad immune the prototypical inflammatory protein, rent standards, we asked whether we could response involving nearly all innate and TNF-α, in adipose tissue of both rodents identify genes whose expression correlated adaptive immune cell types. The initiating and humans. Additionally, Hotamisligil with measures of adiposity. We identified immune event or events remain unclear; and Spiegelman determined that measur- more than 1,300 such transcripts and I, some studies suggest infiltration of neu- able increases in circulating TNF-α were along with MD-PhD student Stuart Weis- trophils (8), others implicate reductions attributable to adipose tissue and postulat- berg, began to pore over them. in regulator T cells (9), and others propose ed that TNF-α mediates obesity-associated As a graduate student, most of my efforts rapid changes in macrophages (10). None- insulin resistance (3). While the precise role were directed at identifying genes involved theless, it is clear that metabolic-induced of TNF-α in mediating insulin resistance in Drosophila development. However, these changes in immune cell number and func- and type 2 diabetes remains incompletely studies were done with E. Richard Stanley, tion are coordinated and profound (11). defined, other groups have built on these whose laboratory is devoted to studying Although the initial immunometabo- macrophage development. As a conse- lism studies focused on adipose tissue, quence, I sat through many journal clubs it is now evident that metabolic activa- Conflict of interest: The author has declared that no conflict of interest exists. and lab meetings during which mouse mac- tion of the immune system is not limited Citation for this article: J Clin Invest. rophage–specific genes and proteins were to fat of obese individuals. Obesity has 2013;123(12):4992–4993. doi:10.1172/JCI73658. discussed. As Stuart and I looked down the now been shown to change immune cell

4992 The Journal of Clinical Investigation http://www.jci.org Volume 123 Number 12 December 2013 hindsight populations in the liver, bone marrow, key immune responses; very little is known Address correspondence to: Anthony W. Fer- regulator areas of the hypothalamus, and about adaptive homeostatic functions in rante Jr., Department of Medicine, Naomi the circulation. Furthermore, metabolic metabolism. A compelling argument can Berrie Diabetes Center, Columbia Universi- regulation of the immune system is not be made that obesity is an evolutionary ty, New York, New York 10032, USA. Phone: limited to obesity. Fasting, cold challenge, aberration that was not previously expe- 212.851.5322; Fax: 212.851.5335; E-mail: weight loss, and acute lipolysis all increase rienced by a wide swath of any popula- [email protected]. the macrophage number and function in tion; therefore, the immune response to adipose tissue. Perhaps this should not this recent stress reflects uncharacteristic 1. Ganrot PO, Gydell K, Ekelund H. Serum concen- tration of α-2-macroglobulin, haptoglobin and have been surprising given the inhibitory biological clumsiness, resulting in insulin α-1-antitrypsin in diabetes mellitus. Acta Endocrinol effects of starvation and reduced leptin resistance, hepatic steatosis, and β cell (Copenh). 1967;55(3):537–544. concentrations on immune function and failure. However, I suspect underlying 2. Powell ED, Field RA. Studies on salicylates and com- plement in diabetes. Diabetes. 1966;15(10):730–733. effector T cells. The emerging narrative adaptive functions of the immune sys- 3. Hotamisligil GS, Shargill NS, Spiegelman BM. Adi- argues that the immune system monitors tem’s response to obesity will be found. pose expression of tumor necrosis factor-α: direct and responds to the metabolic state of tis- The mammalian immune system has role in obesity-linked insulin resistance. Science. sues and the whole . evolved to restore local homeostasis to 1993;259(5091):87–91. 4. Gregor MF, Hotamisligil GS. Inflammatory Building on earlier work that indicates tissues under endogenous and exogenous mechanisms in obesity. Annu Rev Immunol. 2011; TNF-α mediates obesity-associated insu- stress, and recent work on cold challenge 29:415–445. lin resistance and identifies a role for in mice and weight loss argues that there 5. Weisberg SP, McCann D, Desai M, Rosenbaum M, Leibel RL, Ferrante AW Jr. Obesity is associated immune cells in atheromas as well as a are evolutionarily ancient metabolic states with macrophage accumulation in adipose tissue. great deal of evidence provided by rodent that the immune system recognizes and J Clin Invest. 2003;112(12):1796–1808. studies, a model in which the immune responds to adaptively. 6. Xu H, et al. Chronic inflammation in fat plays a cru- cial role in the development of obesity-related insu- system contributes directly to the devel- lin resistance. J Clin Invest. 2003;112(12):1821–1830. opment of obesity-induced pathologies, The future of immunometabolism 7. Wellen KE, Hotamisligil GS. Obesity-induced including type 2 diabetes, nonalcoholic Although its roots run long and deep in the inflammatory changes in adipose tissue. J Clin Invest. 2003;112(12):1785–1788. fatty liver disease, and, of course, athero- scientific literature, immunometabolism 8. Talukdar S, et al. Neutrophils mediate insulin resis- sclerosis has emerged (12, 13). Although is a young field with many questions to tance in mice fed a high-fat diet through secreted at least one broad classical antiinflamma- be answered. In the past decade, the meta- elastase. Nat Med. 2012;18(9):1407–1412. tory agent (salsalate) improved glycemic bolic and, increasingly, the immunology 9. Feuerer M, et al. Lean, but not obese, fat is enriched for a unique population of regulatory T cells control in a randomize trial in patients communities have succeeded in identify- that affect metabolic parameters. Nat Med. 2009; with type 2 diabetes mellitus, it appears ing key players and providing fundamental 15(8):930–939. to have done so without reducing insu- descriptions of basic immunometabolic 10. Weisberg SP, et al. CCR2 modulates inflammatory and metabolic effects of high-fat feeding. J Clin lin resistance as predicted by the current processes. In the coming decade, I imagine Invest. 2006;116(1):115–124. models (14). Furthermore, despite some that we will uncover and further flesh out 11. Odegaard JI, Chawla A. The immune system as encouraging results from small trials (15), adaptive roles of the immune system in a sensor of the metabolic state. Immunity. 2013; 38(4):644–654. the lack of consistent therapeutic success metabolism. A more detailed understand- 12. Donath MY, Shoelson SE. Type 2 diabetes as an in targeting specific immune molecules ing of immunometabolism will permit inflammatory disease. Nat Rev Immunol. 2011; to treat metabolic disease has provided more focused immune therapies to target 11(2):98–107. 13. Osborn O, Olefsky JM. The cellular and signaling some frustration (16). The relative pauci- metabolic diseases, and a greater apprecia- networks linking the immune system and metabo- ty of immune-related genes uncovered in tion for the ability of metabolic systems lism in disease. Nat Med. 2012;18(3):363–374. genome-wide association studies of obe- to modulate classic immune function will 14. Goldfine AB, et al. Salicylate (salsalate) in patients sity-related phenotypes in humans (17) enhance our understanding of autoim- with type 2 diabetes: a randomized trial. Ann Intern Med. 2013;159(1):1–12. suggests the simple elegance of immune mune and inflammatory disorders. 15. Larsen CM, et al. Interleukin-1-receptor antago- dysfunction leading to metabolic disease nist in type 2 diabetes mellitus. N Engl J Med. indicated by rodent models may be mis- Acknowledgments 2007;356(15):1517–1526. 16. Bernstein LE, Berry J, Kim S, Canavan B, Grinspoon leading. We are awaiting the results of sev- This work was supported by grants from SK. Effects of etanercept in patients with the metabol- eral large-scale clinical trials. the National Institute of Diabetes and ic syndrome. Arch Intern Med. 2006;166(8):902–908. Much of the research characterizing Digestive and Kidney Diseases (NIDDK) 17. Manning AK, et al. A genome-wide approach accounting for body mass index identifies genetic the interaction of the immune and meta- (DK066525 and DK063608) and the Lewis variants influencing fasting glycemic traits and bolic systems has focused on maladaptive Katz Foundation. insulin resistance. Nat Genet. 2012;44(6):659–669.

The Journal of Clinical Investigation http://www.jci.org Volume 123 Number 12 December 2013 4993