What Have We Really Learned About Macrophage Recruitment to Adipose Tissue?

As the rates of obesity continue to rise around the world, many in the scientific community are seeking to understand mechanisms by which obesity increases the risk of different diseases. In particular, the contribution of the immune system to metabolic processes has come to the forefront in the exciting new field of immunometabolism (1). Although this field encompasses interactions of innate and adaptive immune cells in multiple organs, at its core lays a unique interaction between macrophages and adipocytes. Although the interface of resident macrophages and adipocytes in lean mice is necessary for adipose tissue development (2), expansion (3), and homeostasis, in obesity, the recruitment of inflammatory macrophages is thought to impair adipose tissue function (4, 5). Thus, much research has focused on molecules responsible for macrophage recruitment to adipose tissue. In this issue of Endocrinology, Dib et al (6) report their findings that the adipokine leptin may be a key mediator of this recruitment. They show that mice transplanted with bone marrow from leptin receptor-deficient (db/db) mice, when placed on a high-fat diet, have reduced weight gain and adiposity, decreased macrophage infiltration, and subsequently diminished adipose tissue inflammation.

Adipocyte-released adipokines, chemokines, and fatty acids are known to modulate macrophage phenotype and function. For example, the adipokine adiponectin has been shown to promote a more M2-like phenotype of adipose tissue macrophages (ATMs), and thus, it helps to maintain adipose tissue homeostasis (7). More studies, however, have focused on the pathological mechanisms by which adipocytes increase macrophage recruitment to adipose tissue in obesity. An emphasis on various chemokines and their receptors has prevailed. It has been shown that chemokines, such as CC chemokine ligand 2 (CCL2), CCL5, and CCL3, as well as their receptors, CC chemokine receptor 2 (CCR2), and CCR5, are important for recruitment of macrophages to adipose tissue (8–10). In addition, mechanisms by which adipocyte-secreted fatty acids and adipokines promote a proinflammatory M1-like phenotype have been extensively studied. For example, it has been shown that activation of inflammatory signaling pathways, such as Toll-like receptor 4, contribute to the inflammatory status of ATMs (11). However, despite the convincing nature of these studies from many different groups, there are equally convincing data showing that these same chemokines, receptors, and inflammatory mediators do not influence macrophage infiltration or inflammation in adipose tissue (12–16). With respect to leptin and the current report by Dib et al (6), we had previously used an identical bone marrow transplantation technique and shown that recipients of db/db bone marrow were not protected from weight gain or macrophage infiltration into adipose tissue (17). Moreover, an additional study by Fantuzzi and coworkers (18) had shown no effects of hematopoietic leptin receptor deficiency on body weight or adipose tissue inflammation in lean mice. Taken together, these cumulative data raise the question of why similar experimental designs from different laboratories yield contrasting results. This is a critical question, because deciphering the answer may yield important biological information regarding macrophage recruitment and function in adipose tissue.

We suggest that there are key elements that should be considered when designing, analyzing, and reviewing imjmunometabolic studies. First, the background strain is very important, because even slight genetic drift could impact the results obtained. This is highlighted by elegant studies of Dr Attie's group mapping out the genes involved in susceptibility and protection from obesity and metabolic defects in mice (19). Thus, the use of littermate controls is ideal, even if not always possible. Second, an accumulating body of literature suggests that gut microbiota may influence systemic metabolic responses far more than is currently appreciated (20). To account for this, cohousing of experimental and control mice should be performed. Third, matching the experimental and control groups for body weight at the beginning of the study is critical. Even small differences at baseline can become exaggerated over time for reasons related to inherent susceptibility to weight gain, rather than because the manipulation has had an important physiological effect. Fourth, even seemingly unimportant details may be critical. Some examples include: 1) sometimes certain cages of mice do not gain weight as well as other cages; 2) a single short event of cage flooding could stunt the growth of mice and influence their metabolic phenotype for their entire lives; 3) mice housed individually may not thrive as well as those that are group housed; and 4) mice from first litters of a dam may have different metabolic characteristics than those born to multiparous dams (21). Certainly, all investigators do their best to control for these variables. However, as a field, it is important to acknowledge that these factors can impact our results in ways that make the underlying biology more difficult to decipher.

The question still remains, “Does leptin influence macrophage recruitment and function in adipose tissue?” It is well established that leptin-deficient, ob/ob, and leptin receptor-deficient, db/db, mice have altered immune system function (22). A first line of evidence of the involvement of leptin in macrophage recruitment to adipose tissue is that ob/ob and db/db mice have lower numbers of ATMs than might be expected based upon their body weight (4). With regards to the influence of leptin on macrophages, it has been shown that high concentrations of leptin increase endothelial cell adhesion molecule expression and promote macrophage adherence (23). Furthermore, we had previously reported that at lower concentrations, leptin acts as a monocyte and macrophage chemoattractant (24). Finally, plasma leptin levels are positively correlated with the number of macrophages in adipose tissue (17). Thus, testing the hypothesis that leptin could mediate macrophage recruitment to adipose tissue was logical for Dib et al (6), Gove et al (18), and our laboratory (17) to perform. The dissimilar results obtained could have been due to the age of the mice at transplantation (5 wk by Dib et al vs 8 wk by our group). However, Gove et al (18) used the same transplantation age as Dib et al (6), and they found no differences in body weight and adiposity. Second, the 45% percent fat in the diet used by Dib et al vs the 60% diet used in our study, as well as the length of diet feeding (12 vs 16 wk), could account for the difference. Finally, it is possible that the slight bias toward a reduced baseline body weight in the recipients of db/db bone marrow in Dib's study, although not statistically significant, could have had biological significance for the ultimate differences in weight and, thus, adipose tissue inflammation (6). Taken together, the conflicting data available document the complexity of leptin effects on macrophage recruitment and function in the adipose tissue, which might be influenced (even in a subtle manner) by a number of genetic and environmental factors and is probably more complex than initially anticipated.

In the past decade, the field of immunometabolism has made great strides in discovering the complex interactions of the immune system with metabolic processes. In the future, to continue the advancement of this field, publication of both “positive” and “negative” results from well-designed experiments, is critical if we hope to fully disseminate new discoveries. Perhaps, with the whole picture available, we will all be able to advance the current knowledge regarding adipose tissue inflammation more effectively.

Response to Hasty and Gutierrez by L.H.Dib et al:

We thank Hasty and Gutierrez for their interest and comments about our work (6), and we agree with the authors that experimental variables between their study and ours (6, 17) may explain the differences between our findings. For example, although Hasty and Gutierrez report no difference in either body weight or glucose homeostasis parameters after 12 weeks of a high-fat diet (HFD), we, on the other hand, observed significant differences in body weight, serum insulin, and homeostatic model assessment-estimated insulin resistance at the 16-week post-HFD, a time point beyond the scope of observation of Hasty and Gutierrez's study. Our data are consistent in that we both observed no difference in the number of recruited adipose tissue macrophages (ATMs) on a HFD in the absence of leptin signaling. However, our results showed a prevalence of antiinflammatory ATMs as a result of bone marrow leptin receptor deficiency, a point that was not investigated by Hasty and Gutierrez.

Our group agrees that macrophage infiltration into adipose tissue and the macrophage response to obesity is multifactorial. Leptin signaling may be only one of several factors that influence macrophage responses. Because of this, consistency among protocols as discussed by Hasty and Gutierrez is important. Unfortunately, when we started our study in 2009 (25), we were unaware of their work, and it was not possible to foresee their protocol. Nevertheless, closer examination of the datasets and careful control of variables in future experiments will reveal how leptin affects macrophages.