The Overlooked Macrophages in Adipose Tissue

Abstract

Adipose tissue macrophages (ATMs) are heterogeneous, with distinct roles in metabolism. Prior studies mainly used collagenase-based stromal vascular fraction isolation to study ATM heterogeneity. Hu et al. employed multiple approaches, including a revised collagenase-based protocol, to investigate ‘adhesion-related macrophages’, substantially advancing our understanding of this often-overlooked ATM subpopulation.

Adipose tissue macrophages (ATMs) are central regulators of tissue homeostasis and metabolic disease. Over the past two decades, studies based almost exclusively on stromal vascular fraction (SVF) isolation have uncovered remarkable macrophage heterogeneity in adipose tissue [1, 2]. Yet these approaches rely on an implicit assumption that collagenase digestion and SVF isolation faithfully capture the entire ATM compartment [3–5]. Macrophages retained within the adipocyte fraction have long been observed but routinely discarded [6], leaving their identity and function largely unexplored. This methodological blind spot has likely left an important component of adipose immune biology unexplored.

Hu et al. close this gap by isolating macrophages that remain tightly associated with adipocytes after collagenase digestion [7]. These adhesion-related macrophages (ARMs) account for only a minor fraction of ATMs in lean adipose tissue but expand dramatically during obesity, comprising nearly 70% of the ATM compartment (Figure 1). By identifying CAV1 as a selective ARM marker and integrating adipose single-cell and single-nucleus RNA sequencing datasets, the authors further demonstrate that ARMs are largely absent from conventional SVF-based single-cell RNA sequencing but readily detected by single-nucleus RNA sequencing, highlighting how experimental approaches can fundamentally shape our understanding of immune-cell diversity.

Figure 1. Adhesion-related macrophages (ARMs) and their role in metabolism.

Obesity substantially increases adipose tissue “adhesion-related macrophages” (ARMs), macrophages that adhere to adipocytes (1). Caveolin-1 (CAV-1), an integral membrane protein that forms the structural basis of caveolae, is highly expressed on ARMs and plays critical roles in mediating ARM adhesion to adipocytes and maintaining ARM function (2). Through direct adhesion, ARMs can acquire mRNA from adipocytes (3), a process that enhances ARMs’ capacity for lipid handling within adipose tissue, thereby limiting lipotoxicity and improving systemic glucose tolerance under obesity conditions (4). This figure was created using BioRender (www.biorender.com).

Interestingly, although ARMs progressively accumulate during obesity, Hu et al. unexpectedly show that this population serves a protective rather than pathogenic role [7]. Genetic disruption of Cav1, a key determinant of ARM formation, markedly reduced ARM accumulation and exacerbated adipose tissue dysfunction and systemic insulin resistance during obesity. Conversely, adoptive transfer of ARMs into obese mice improved glucose homeostasis, supporting a beneficial role in metabolic homeostasis. Mechanistically, ARMs exhibit a remarkable capacity for lipid uptake and neutral lipid synthesis, enabling them to buffer excess free fatty acids and limit lipotoxic stress within adipose tissue (Figure 1). These findings suggest that distinct macrophage subsets can play fundamentally different roles in obesity.

Because lipid-associated macrophages (LAMs) are widely considered the principal lipid-processing macrophages in adipose tissue [1, 4], an important question is whether ARMs simply represent another manifestation of LAMs. Hu et al. addressed this issue by directly comparing the two populations and found minimal overlap between ARMs (CAV1⁺ ATMs) and LAMs (CD9⁺TREM2⁺ ATMs) [7]. Moreover, ARMs exhibited substantially greater lipid uptake and lipid synthesis than LAMs, indicating that ARMs represent a previously unrecognized and metabolically specialized ATM subset rather than merely a phenotypic variant of LAMs. Furthermore, Hu et al. demonstrated that genetic deletion of LAMs in Trem2-deficient mice did not substantially alter ARM proportion or number [7], further supporting the conclusion that ARMs and LAMs are distinct ATM subsets maintained through different mechanisms.

One of the most intriguing findings of the study by Hu et al. concerns the origin of the lipid metabolic program that distinguishes ARMs. Unexpectedly, several adipocyte-specific metabolic transcripts, including Adipoq, Lep, and Cidec, were highly expressed in ARMs despite lacking corresponding chromatin accessibility, suggesting that they originate from neighboring adipocytes rather than endogenous transcription. Hu et al. further demonstrate that these mRNAs are transferred through direct ARM–adipocyte contact rather than extracellular vesicles. Electron microscopy, RNAscope, and functional perturbation experiments support a CAV1- and actin-dependent transfer mechanism, whereas inflammatory stimulation with TNF-α further enhances this process. These findings uncover a previously unrecognized mode of macrophage–adipocyte communication. Beyond cytokines and extracellular vesicles, stressed adipocytes appear capable of directly transferring metabolic mRNAs to neighboring macrophages, potentially enabling rapid functional adaptation and thereby reinforcing lipid buffering during obesity.

Previous studies have demonstrated that transferred mRNAs can remain functional in recipient cells [8]. Although Hu et al. establish that ARMs acquire adipocyte-derived mRNAs through direct cell-cell contact, whether these transcripts actively shape the ARM phenotype remains an important question. At present, the evidence primarily explains differences at the transcript level. Future studies should determine whether transferred mRNAs are translated into functional proteins and directly contribute to the metabolic phenotype of ARMs. Another key question concerns the developmental origin of ARMs. Although bone marrow transplantation experiments support a contribution from recruited monocytes, it also remains to be established whether ARMs arise predominantly from tissue-resident macrophages, recruited monocytes, or both.

The significance of this study extends beyond the identification of a new macrophage subset. By uncovering ARMs, Hu et al. reveal that a substantial fraction of ATMs has remained hidden because of the methodological limitations of conventional SVF isolation. The detection of similar ARMs in the liver and tumors further raises the possibility that comparable populations have been overlooked across multiple tissues, suggesting that ARMs may represent a broader principle of tissue immunity rather than an adipose-specific phenomenon.

Equally important, this work introduces a new paradigm for intercellular communication. While trogocytosis has traditionally been viewed as a mechanism for transferring membrane components and proteins between cells [9, 10], Hu et al. suggest that direct cell-cell contact can also mediate the transfer of functional mRNAs, enabling rapid metabolic adaptation of recipient macrophages. Whether these transferred transcripts are translated into functional proteins and whether similar mechanisms operate in other tissues remain important questions for future investigation. Nevertheless, by revealing both a previously hidden macrophage population and an unexpected mode of macrophage–adipocyte communication, this study substantially expands the cellular and mechanistic landscape of adipose tissue biology and provides a new conceptual framework for understanding how tissue-resident macrophages dynamically adapt to metabolic stress through intimate interactions with neighboring parenchymal cells.