Nerve-associated macrophages control adipose homeostasis across lifespan and restrain age-related inflammation

Gonzalez-Hurtado, Elsie; Leveau, Claire; Li, Keyi; Mishra, Manish; Qu, Rihao; Goldberg, Emily L.; Sidorov, Sviatoslav; Damani-Yokota, Payal; Yeung, Stephen T.; Khairallah, Camille; Gonzalez, David; Shepard, Taverlyn M.; Camell, Christina; Artyomov, Maxim N.; Kluger, Yuval; Khanna, Kamal M.; Dixit, Vishwa Deep

doi:10.1038/s43587-025-00952-9

Download PDF

Article
Open access
Published: 02 September 2025

Nerve-associated macrophages control adipose homeostasis across lifespan and restrain age-related inflammation

Nature Aging volume 5, pages 1828–1843 (2025)Cite this article

24k Accesses
1 Citations
160 Altmetric
Metrics details

Subjects

Abstract

Age-related inflammation or ‘inflammaging’ increases disease burden and controls lifespan. Adipose tissue macrophages (ATMs) are critical regulators of inflammaging; however, the mechanisms involved are not well understood in part because the molecular identities of niche-specific ATMs are unknown. Using intravascular labeling to exclude circulating myeloid cells followed by single-cell sequencing with orthogonal validation via multiparametric flow cytometry, we define sex-specific changes and diverse populations of resident ATMs through lifespan in mice. Aging led to depletion of vessel-associated macrophages, expansion of lipid-associated macrophages and emergence of a unique subset of CD38⁺ age-associated macrophages in visceral adipose tissue with inflammatory phenotype. Notably, CD169⁺CD11c⁻ ATMs are enriched in a subpopulation of nerve-associated macrophages (NAMs) that declines with age. Depletion of CD169⁺ NAMs in aged mice increases inflammaging and impairs lipolysis suggesting catecholamine resistance in visceral adipose tissue. Our findings reveal NAMs are a specialized ATM subset that control adipose homeostasis and link inflammation to tissue dysfunction during aging.

Macrophage function in adipose tissue homeostasis and metabolic inflammation

Article 03 April 2023

Senescent cells promote tissue NAD⁺ decline during ageing via the activation of CD38⁺ macrophages

Article 16 November 2020

Atherosclerotic cardiovascular disease in aging and the role of advanced cardiovascular imaging

Article Open access 02 August 2024

Main

The sympathetic nervous system along with innate immune system are the first lines of defense in sensing danger and protection of tissues from internal and external threats. The emergence of chronic inflammation is considered a hallmark feature of the aging process that is directly linked to a decline in tissue function^1,2. Adipose tissue macrophages (ATMs) are abundant immune cells in the adipose. ATMs do not display binary M1 or M2-like phenotypes and are known to decline with age^3,4,5. ATMs can reside in specialized niches such as fat-associated lymphoid clusters (FALCs), crown-like structures that contain necrotic adipocytes, and near vasculature and/or sympathetic nerves. Recent studies have revealed the presence of diverse macrophage states such as lipid-associated macrophages (LAMs) and vascular-associated macrophages (VAMs)^6,7; however, the changes that may occur among distinct ATM subsets in aging and the functional role these cells play through lifespan remain largely unknown.

A population of tissue-resident macrophages referred to here as nerve-associated macrophages (NAMs) have also been described in several tissues including: the gut, lung, skin and adipose tissue^{3,8,9,10,11,12,13}. NAMs have been linked to a wide array of functions including the regulation of peristalsis (gut), the immunoregulatory response to influenza (lung), nerve surveillance and repair (skin and sciatic nerve) and catecholamine degradation (adipose)^{3,8,9,10,11,12,13,14}. In the adipose tissue, NAMs seem to regulate obesity and may play a role in fasting-induced lipolysis during aging^3,12; however, given the lack of specific markers or Cre drivers that target NAMs, or distinguish them from other ATMs, the precise functional role of NAMs in adipose tissue homeostasis and aging remains unclear. Moreover, despite several subsequent comprehensive single-cell transcriptomic analyses, the exact transcriptional identity of adipose tissue NAMs continues to remain elusive, as so far, no study has yet identified a NAM-specific subcluster or transcriptomic signature using an unbiased single-cell RNA sequencing (scRNA-seq) approach.

Here, we implemented an unbiased approach to meticulously dissect the resident F4/80⁺CD11b⁺ population in young and aged visceral white adipose tissue (VAT). Using intravascular (iv) labeling in combination with bulk RNA sequencing we show that resident F4/80⁺CD11b⁺ cells are transcriptionally distinct from circulating F4/80⁺CD11b⁺ cells. Furthermore, using scRNA-seq we find that the resident F4/80⁺CD11b⁺ population can be subdivided into 13 subsets that include VAMs, LAMs, interferon-associated macrophages (IAMs), CD169⁺ NAMs and CD38⁺ age-associated macrophages (AAMs) that emerge only in aging. Using transgenic CD169 diphtheria toxin receptor (CD169-DTR) mice, we demonstrate that CD169⁺ NAMs control lipolysis and are required for preventing catecholamine resistance and adipose tissue dysfunction. We further reveal that CD169⁺ NAMs diminish in adipose tissue with age and are crucial for mitigating age-induced inflammation.

Results

Transcriptional landscape of resident adipose tissue macrophages

The VAT expands with aging, is highly vascularized and contains circulating myeloid cells that are likely transcriptionally distinct from tissue-resident macrophages or monocytes that have recently undergone extravasation. To investigate this, we performed iv labeling with anti-CD45.2 antibody to tag circulating (CD45⁺CD45.2iv⁺) and resident (CD45⁺CD45.2iv⁻) cells^15,16. Resident and circulating F4/80⁺CD11b⁺ cells were sorted from the VAT of young (2 months) and aged (22 months) male and female mice (Fig. 1a). Flow cytometry analysis showed that in the VAT, approximately 20–30% of F4/80⁺CD11b⁺ cells were in the vasculature and 70–80% of F4/80⁺CD11b⁺ cells resided in the tissue (Fig. 1b). We next conducted bulk RNA sequencing of resident (iv⁻) and circulating (iv⁺) F4/80⁺CD11b⁺ cells sorted from two functionally distinct adipose depots: the VAT and brown adipose tissue (BAT) (Fig. 1c). Principal-component analysis (PCA) revealed that 72% of the variation in F4/80⁺CD11b⁺ populations was driven by their resident or circulating origin, while adipose depot type accounted for only 10% (Fig. 1c). Analysis of differentially expressed genes (DEGs) revealed that resident F4/80⁺CD11b⁺ cells are enriched in antigen presentation genes (H2-A, H2-Ab1, H2Eb1 and Cd74) as well as Retnla, Fclrs, Mgl2, Lpl, Dab2, C1qa and Lmna (Extended Data Fig. 1a,b). Of note, genes such as Sepp1, Lilra5 and Siglec1 were enriched in BAT resident F4/80⁺CD11b⁺ cells compared to VAT (Extended Data Fig. 1c). Further, pathway analysis suggested resident F4/80⁺CD11b⁺ cells downregulate genes involved in leukocyte transendothelial migration and upregulate extracellular matrix (ECM) genes (Extended Data Fig. 1d,e). Together, these data highlight that approximately 75% of the F4/80⁺CD11b⁺ cells isolated from the stromal vascular fraction (SVF) are in the tissue parenchyma and exhibit a gene expression profile that is distinct from circulating F4/80⁺CD11b⁺ cells in blood vessels.

To examine the age-related changes in the tissue-resident compartment (CD45⁺CD45.2iv⁻), we conducted scRNA-seq of F4/80⁺CD11b⁺ cells sorted from the VAT of young and aged male and female mice (CD45⁺CD45.2iv⁻CD3⁻CD19⁻SiglecF⁻) (Fig. 1a). Unsupervised clustering of 15,709 quality control (QC)-positive cells revealed 16 clusters as visualized by Uniform Manifold Approximation and Projection (UMAP) (Extended Data Fig. 2a,b and Supplementary Table 1). Clusters 9, 12 and 14 did not express significant levels of Ptprc (CD45), Itgam (CD11b) and Adgre1 (F4/80), suggesting that these clusters were likely contaminating cells and were therefore excluded from future analyses (Extended Data Fig. 2c). Approximately 60–70% of cells clustered into five main populations (clusters 0–4) and 30–40% of cells clustered into the remaining eight populations (clusters 5–8, 10–13 and 15) (Fig. 1f). Of note, some clusters were differentially distributed in male and female mice: cluster 0 was the dominant cluster in male mice, whereas cluster 5 was dominant in female mice (Fig. 1f and Extended Data Fig. 3a).

Clusters 0, 1, 4 and 5 represented approximately 20%, 13%, 10% and 6% of all cells profiled, respectively, and exhibited high expression levels of Adgre1 (F4/80) relative to other clusters (Fig. 1f and Extended Data Fig. 2c). Only a few genes such as Cx3cr1 were differentially expressed in cluster 1 compared to other clusters (Fig. 1e and Supplementary Table 2). RNA velocity inference analysis suggested that cluster 1 represented an early differentiation and/or precursor state that could give rise to macrophages from cluster 0 (Extended Data Fig. 3b). Visualization of the top 25 DEGs for each cluster showed that clusters 0, 4 and 5 were enriched for unique as well as a shared subset of genes that included complement genes (C1qa, C1qb, C1qc, C4b, C3ar1 and C5ar1), and genes associated with tissue-resident macrophages such as Cd163, Cd209f, Cd209g, Csf1r, Folr2 and Gas6 (Fig. 1e, Extended Data Fig. 2b and Supplementary Table 2)^8,17,18. To delineate between clusters 0, 4 and 5, we compared DEGs between cluster 0 versus clusters 4 and 5. This analysis showed that cluster 0 was enriched for Lyve1, Cd163, Ednrb, F13a1, Apoe, Cd36 and Ccl24 expression (Fig. 1e, Extended Data Fig. 3c and Supplementary Table 2). The gene signature for cluster 0 corresponded to a population of tissue-resident VAMs previously described in the adipose tissue, lungs and heart^6,7,13,17,19. Cluster 4 was enriched for Cxcl13, C3, Cd55, Ly6e, Ccl8, Colq and Mmp9 genes, and was unique in that it only emerged in the aged condition; hence, we refer to this cluster as AAMs (Figs. 1e and 2a, Extended Data Fig. 3c and Supplementary Table 2). Cluster 5 was enriched for expression of Lilra5, H2Eb1, Cd74, Cd86, Vcam1, Ccl12, Csf1r, St3gal6 and Xist (Fig. 1e, Extended Data Fig. 3d and Supplementary Table 2). Notably, cluster 5 showed enrichment for monoamine oxidase A (Maoa), an enzyme required for the breakdown and inactivation of catecholamines that was previously shown to be expressed by sympathetic nerve-associated macrophages (referred to here as NAMs)^3,12 (Extended Data Fig. 4a). The expression of the catecholamine transporter Slc6a2 was not detected in our dataset (Extended Data Fig. 4b). To assess whether cluster 5 was a bona fide NAM cluster, we calculated a gene set enrichment score (GSES) for each cell using a list of NAM-associated genes from studies characterizing NAMs in other tissues^{3,8,9,10,11,12,13,20,21} (Supplementary Table 3). Our analysis showed cluster 5 had the highest GSES, suggesting this cluster represented NAMs in VAT (Extended Data Fig. 4c). Moreover, relative to clusters 0 and 4, cluster 5 expressed low levels of Lyve1, a gene that has been used to discriminate between macrophages adjacent to vessels (Lyve1^hiMHCII^hi) or nerves (Lyve1^loMHCII^hi)⁸ (Fig. 1e and Extended Data Fig. 4a). Of note, Ccr2 expression was lowest in clusters 0, 4 and 5, with the lowest levels observed in cluster 4 (Extended Data Fig. 4a). Conversely, expression of Timd4, Folr2 and Lyve1 was largely restricted to cells in clusters 0 and 4, and to a lesser extent cluster 5, suggesting as a previous report did, that these macrophage subsets might be maintained via self-renewal with minimal monocyte input (Extended Data Fig. 4d)¹⁸.

**Fig. 2: Aging remodels resident adipose tissue macrophages.**

Cluster 2 represented approximately 11% of all cells profiled (Fig. 1f). Analysis of DEGs for cluster 2 showed it was enriched for Ccr2, Chil3, Plac8, Thbs1, Clec4e, Mpeg1, Il1b and Ly6c2 genes, suggesting that these ATMs may be of monocyte origin (ATM-1) (Fig. 1e and Supplementary Table 2)^17,22. Clusters 3, 7, 8 and 13 represented approximately 10%, 5%, 4% and 1.5% of all cells profiled, respectively, and exhibited intermediate expression levels of Adgre1 (F4/80) (Fig. 1f, Extended Data Fig. 1c and Supplementary Table 2). DEGs for clusters 3, 7, 8 and 13 showed that they were enriched for Fn1, Lyz1, Ear2 and Cd52, with clusters 3 and 7 exhibiting the highest and lowest expression levels of these genes, respectively (Fig. 1e and Supplementary Table 2). Cluster 3 was distinguished by high expression of Retnla, Lpl, Cd226 and Socs6, a gene signature similar to that of small peritoneal macrophages and to a monocyte-derived macrophage cluster previously described in the adipose tissue and heart (ATM-2) (Fig. 1e and Supplementary Table 2)^6,17,22,23. Cluster 7 was enriched for the cell cycle genes Mki67, Top2a, Hmgb2 and Ube2c, and cluster 13 for the ribosomal genes Rpl32, Rpl10, Rpl11 and Rps18 (Fig. 1e and Supplementary Table 2). The gene signature for cluster 7 suggested it represented cycling ATMs (cATMs), which have been reported in the lung and heart^13,19,22. Clusters 10 and 11 represented approximately 3% and 2.5% of all cells profiled, respectively (Fig. 1f). Cluster 10 was characterized by expression of Trem2, Spp1, Lgals3, Lpl, Ctsl, Cstb, Cd9 and Fabp5, a gene signature previously associated with LAMs (Fig. 1e and Supplementary Table 2)^6,17. Cluster 11 expressed Cxcl9, Cxcl10, Rsad2 and Stat1 and interferon-stimulated genes such as Ifit1, Ifit2, Irf7, Ifi47 and Isg15 (Fig. 1e and Supplementary Table 2). The gene signature for cluster 11 has been observed in a subset of macrophages from the heart, hence we refer to them here as IAMs²². Clusters 6 and 15 represented about 6% and 0.7% of all cells profiled and expressed the lowest levels of Adgre1 (F4/80) (Fig. 1e and Extended Data Fig. 1b). While only cluster 6 expressed significant levels of Cd209a and Itgam (CD11b), both clusters 6 and 15 were enriched for the dendritic cell (DC)-associated genes Itgax, Flt3, Napsa, Nr4a3, Zbtb46, Ccl17, Ccl22 and Dpp4 suggesting that they represented CD11b⁺ DCs (DC-1 and DC-2) (Fig. 1e and Supplementary Table 2)¹⁷. Together, the gene signatures and RNA velocity inference analysis suggest that clusters 13, 8 and 1 represent early myeloid precursor (MP) populations (MP-1, MP-2 and MP-3) that may give rise to more terminally differentiated tissue-resident macrophages (clusters 0, 4 and 5) or monocyte-derived-macrophages (clusters 2, 3, 7, 10 and 11) and monocyte-derived DCs (clusters 6 and 15) (Fig. 1d, Extended Data Fig. 3b and Supplementary Table 2).

Next, we calculated a GSES for each cell using the hallmark gene collection from the Molecular Signatures Database (MSigDB)^24,25,26. This analysis confirmed upregulation of complement genes in clusters 0, 4 and 5, cell cycle genes in cluster 7, oxidative phosphorylation and adipogenesis genes in cluster 10, and interferon response genes in cluster 11 (Fig. 1g). Clusters 4 and 5 were enriched for genes involved in the p53 pathway and apoptosis (Fig.1g). In addition, cluster 4 showed enrichment for KRAS signaling and protein-secretion genes and cluster 5 for hypoxia and UV-response genes (Fig. 1g).

Altogether, our analysis unveils unique ATM states with distinct transcriptional signatures such as VAMs (cluster 0) and LAMs (cluster 10), and two tissue-resident macrophage clusters not previously identified in the adipose using unbiased scRNA-seq profiling of F4/80⁺CD11b⁺ cells: cluster 5 (NAMs) and cluster 4 (AAMs).

Aging remodels resident adipose tissue macrophages

The impact of aging on different resident ATM populations and their transcriptome remains to be established. Our unbiased scRNA-seq profiling of total resident (noncirculating) F4/80⁺CD11b⁺ cells showed that aging induced perturbations in greater than half of the clusters identified (Fig. 2a,b).

The clusters likely representing precursor or proliferating populations, namely MP-1 and cATM, remained largely unaffected by age in male mice (Fig. 2a,b). The exceptions were MP-2 and MP-3, which showed a twofold reduction with age (Fig. 2a,b). ATM-1 and ATM-2 remained largely constant between conditions, suggesting that these clusters might be continually seeded from infiltrating monocytes (Fig. 2a,b). VAMs decreased threefold with age in male mice, but did not change in female mice (Fig. 2a,b). AAMs increased from 1% in young male mice to nearly 20% in aged male mice (Fig. 2a,b). NAMs did not seem to change with age, whereas LAMs and IAMs exhibited an increase with age (Fig. 2a,b).

To independently test age-related changes from our scRNA-seq data, we built a cell surface marker panel to identify and quantify VAM, ATM-2, AAM, NAM, DC-1, LAM and MP-1 clusters via flow cytometry (CD45, CD11b, CD11c, CD226, FOLR2, CD163, CD38, CD169 and CD9) (Fig. 2c,d and Extended Data Fig. 5). Our scRNA-seq data showed that cluster 5 (NAMs, CD169⁺CD11c⁻) was enriched for Siglec1 (CD169), but not Itgax (CD11c) (Fig. 2c,d and Extended Data Fig. 5). Cluster 13 (MP-1, CD169⁺CD11c⁺) also expressed Siglec1, but unlike cluster 5 expressed Itgax (CD11c) (Fig. 2c,d and Extended Data Fig. 5). Clusters 3 and 6 were both enriched for Cd226 (CD226). Cluster 6 (DC-1, CD169⁻CD226⁺CD163⁺CD11c⁺), but not cluster 3 (ATM-2, CD169⁻CD226⁺CD163⁻CD11c⁺) showed significant enrichment for Cd163 (CD163) (Fig. 2c,d and Extended Data Fig. 5). Cluster 10 (LAMs, CD169⁻CD226⁻CD9⁺CD11c⁺) lacked CD226 and CD163 expression but was enriched for CD9 (Cd9) (Fig. 2c,d and Extended Data Fig. 5). Both cluster 0 (VAMs, CD38⁺FOLR2⁺CD169⁻CD226⁻CD11c⁻) and cluster 4 (AAMs, CD38⁺FOLR2⁻CD169⁻CD226⁻CD11c⁻) expressed Cd38 (CD38) and Folr2 (folate receptor β; FOLR2) (Fig. 2c,d and Extended Data Fig. 5). Our flow cytometric analysis of VAT showed a decrease in cluster 0 (VAMs) in male but not female mice (Fig. 2e,f). Cluster 10 (LAMs) and cluster 4 (AAMs) increased with age in both sexes (Fig. 2e,f). Cluster 5 (NAMs) was more abundant in young female mice compared to male mice, and decreased with age in female but not male mice (Fig. 2e,f). In contrast, cluster 13 (MP-1) increased in female but not in male mice (Fig. 2e,f). Notably, we observed a decrease in cluster 3 (ATM-2) and cluster 6 (DC-1) in male but not female mice (Fig. 2e,f).

In summary, these data show that ATMs undergo remodeling in response to aging in a sex-dependent manner, and support the existence of a transcriptionally distinct CD38⁺ population (AAMs) in the VAT that expands significantly during aging^27,28.

Inflammaging-associated secretory phenotype of resident adipose tissue macrophages

As cellular senescence is closely linked to inflammaging, we sought to determine whether senescent cell signatures could be detected at the single-cell level resolution in tissue-resident macrophages. Notably, senescence-associated secretory phenotype (SASP) response genes such as C-C and C-X-C family of chemokines, metalloproteases, interleukin (IL)-1α, IL-1β, IL-6, IL-18, tumor necrosis factor (TNF), high mobility group box 1 (HMGB1) and growth differentiation factor 15 (GDF15) are neither specific nor exclusive to senescent cells³. In fact, compared to rare senescent cells in vivo, abundant tissue macrophages are canonical producers of most SASP factors or inflammaging mediators. To investigate senescence-like features in macrophages, we applied the prior published SenMayo gene set to calculate a GSES for each cell²⁹. Our analysis revealed that NAMs (cluster 5) and to a lesser extent AAMs (cluster 4) and MP-1 (cluster 13) were enriched for senescence-associated genes (Fig. 3a). Additionally, Gene Ontology (GO) analysis of all clusters suggested NAMs were enriched for genes linked to dysfunctional, aged, ‘senescent-like’ cells and genes involved in the cellular response to unfolded proteins, phagolysosome assembly, and positive regulation of gliogenesis (Fig. 3b); however, this approach does not capture definitive markers of true senescence, such as those associated with permanent cell cycle arrest and resistance to apoptosis.

**Fig. 3: Adipose NAMs are characterized by CD169-expression and display distinct morphology and localization.**

Taken together, our data support the hypothesis that the SASP-like features observed in tissue-resident macrophages likely represent an inflammaging-associated secretory phenotype (IASP). This macrophage-specific IASP reflects a chronic inflammatory and tissue remodeling response driven by aged cells accumulating damage-associated molecular patterns.

Adipose NAMs are enriched for CD169 and display unique morphology and localization

As our findings suggested that cluster 5 was enriched for genes linked to senescent cells, and CD169⁺CD11c⁻ NAMs declined with age in female mice, we sought to further characterize these cells across lifespan. Analysis of DEGs for cluster 5 showed that NAMs were enriched for genes involved in β-adrenergic signaling (Adrb2, Arrb2, Maoa and Aldh2), IL-10 signaling (Il10rb and Il10), sialic-acid modifying enzymes and binding proteins (St3gal6, Neu1, Siglece, Siglec1 and Siglech), chemokines/chemoattractants (Cxcl2, Ccl12, Cxcl12, Ccl2, Ccl4 and Ccl3) and other genes enriched in NAMs from the lung and sciatic nerve (Csf1r, Ms4a7, Mgl2, Sdc4, Aoah, Xist, St3gal6, Tanc2 and Cdr2) (Fig. 3c and Supplementary Table 2)^{3,12,13,20,21}. NAMs were also enriched for several genes involved in stress response (Hspa1a, Hspa1b, Hspa8, Dnaja1, Dnajb1, Hsph1, Ier3, Ier5 and Atf3), complement signaling (C4b, C5ar1, C5ar1 and C5ar2), prostaglandin metabolism (Hpgd and Hpgds), and binding/transport genes (Trf, Cp, Slc40a1 and Slco2b1), suggesting a potential role for NAMs in the regulation of cellular stress, regulation of inflammation, and iron (Fig. 3c and Supplementary Table 2). Notably, NAMs expressed the leukocyte immunoglobulin-like receptor genes Lilr4b and Lilra5, the latter of which has been associated with late-onset sporadic Alzheimer’s disease (Fig. 3c)³⁰.

To examine NAMs further, we performed in vivo two-photon (2P) microscopy using double-fluorescent reporter (mT/mG) mice. mT/mG is a cell membrane-targeted, two-color fluorescent Cre-reporter allele which only expresses membrane tdTomato (mT) in cells/tissues before Cre recombination. Following Cre recombination, the mT sequence is excised from Cre-expressing cells and tissues and replaced with membrane eGFP (mG). This can be visualized as change of color from red (in non-recombined cells) to green (in recombined cells)³¹. Using pan-myeloid dual reporter LysM^Cre:Rosa26^mT/mG mice (referred to here as LysM:mTmG), we could visually appreciate the diversity of myeloid cells (LysM:mGFP⁺) surrounding different structures (mTomato⁺) such as adipocytes, blood vessels and nerves that are easily distinguished by their morphology, with adipocytes being round, nerves appearing dense and curved/sinusoidal and vessels appearing hollow and straight (Supplementary Video 1). Via in vivo 2P microscopy, we could visualize dynamic LysM:mGFP⁺ cells moving rapidly through or near the vasculature, and slightly more elongated LysM:mGFP⁺ cells situated atop nerves that were largely static but surveilled nerves via their pseudopodia (Supplementary Video 1). To characterize Siglec1 or CD169⁺ NAMs more closely, we used CD169^Cre:Rosa26^mT/mG dual reporter mice (referred to here as CD169:mTmG). Whole-mount immunofluorescence (IF) imaging of VAT from adult CD169:mTmG female mice stained with the pan-neuronal marker TUBB3 showed elongated CD169⁺ cells colocalizing on the surface of TUBB3⁺ nerve bundles and blood vessels (Fig. 3d, Supplementary Video 2 and Extended Data Fig. 6a). Confocal imaging also showed elongated CD169⁺ NAMs embedded deep within TUBB3⁺ nerve bundles in VAT, but also in the subcutaneous white adipose tissue (SAT) and BAT depots (Fig. 3d, Supplementary Video 2 and Extended Data Fig. 6b). Notably, we frequently observed CD169⁺ NAMs with extended pseudopodia contacting thin and large nerve fibers and blood vessels (Fig. 3d and Extended Data Fig. 6c). Higher-resolution imaging revealed elongated CD169⁺ NAMs with pseudopodia wrapped around nerves while simultaneously contacting blood vessels (Fig. 3d and Extended Data Fig. 6c). Serial high-resolution imaging of VAT from CD169:mTmG reporter mice stained with 4,6-diamidino-2-phenylindole (DAPI) showed multiple elongated CD169⁺ cells surrounding TUBB3⁺ nerves, suggesting that NAMs may be involved in adipose nerve maintenance and/or support their survival (Fig. 4a and Supplementary Video 3). Consistent with imaging, our scRNA-seq data showed that NAMs (cluster 5) were enriched for genes required for podosome formation, cell adhesion and cell polarity/contractility (Vcam1, Cd38, Sh3pxd2b, Myo5a, Mylk and Tmod1) and the neurotropic factors (Serpinf1 and Sema6d) (Fig. 3c). To examine these interactions in more detail, we performed transmission electron microscopy (TEM) of isolated nerves from the SAT. Unlike the VAT, the large nerves from SAT can be isolated using a dissection microscope. TEM of isolated SAT nerves showed classic nerve fascicles with endoneurium and perineurium, myelinated and unmyelinated axons, Schwann cells and blood vessels (Fig. 4b). TEM of the endoneurium showed elongated macrophages with extended pseudopodia surrounded by myelinated and unmyelinated nerve axons, suggesting these cells survey the nerve microenvironment (Fig. 4c). Intriguingly, TEM of SAT nerves also showed macrophages with internalized electron dense material, likely myelin, suggesting NAMs may play a role in myelin clearance (Fig. 4d). Consistent with this hypothesis, our scRNA-seq showed NAMs were enriched for Pmp22, a major component of myelin in the peripheral nervous system (PNS) (Fig. 3c). IF imaging of SAT and VAT stained with TUBB3 and the myelin markers myelin basic protein (MBP) or CNPase confirmed the presence of CD169⁺ NAMs on myelinated nerve bundles and deeply embedded between myelinated nerve fibers in SAT (Fig. 4e). Furthermore, in the VAT we observed colocalization of CD169⁺ cells with MBP⁺ particles adjacent to myelinated nerves, supporting a role for CD169⁺ NAMs in myelin uptake, clearance or maintenance of myelinating Schwann cells in the PNS (Fig. 4f and Supplementary Videos 4 and 5).

**Fig. 4: CD169⁺ NAMs engulf myelin and can surveil the adipose nerve microenvironment.**

Next, using in vivo 2P microscopy we assessed the real-time dynamics of CD169⁺ NAMs in SAT using CD169:mTmG reporter mice. We imaged large nerve bundles, as small thin fibers that normally run along blood vessels are not strongly labeled by mTomato (Fig. 3d). Visualization of CD169⁺ NAMs in the SAT revealed motile CD169⁺ cells on blood vessels, and non-motile CD169⁺ cells on large nerve bundles (Fig. 4g and Supplementary Videos 6 and 7). Despite being fairly static, live imaging showed CD169⁺ NAMs extend their dendrites on neurons, suggesting that they are actively surveilling the nerve microenvironment (Fig. 4g and Supplementary Video 6). Moreover, CD169⁺ cells on nerves could be seen forming balloon-like protrusions on the surface of the nerve, which may represent active phagocytosis of nerve cellular components (Fig. 4g and Supplementary Video 7).

In summary, these data demonstrate that CD169⁺ NAMs have diverse morphology reflecting their dynamic functional states and can be observed as elongated cells that contact nerves and blood vessels via their extended pseudopodia. These findings support a role for NAMs in neuroimmune interactions involved in the surveillance and maintenance of the PNS.

CD169⁺ NAMs are necessary for preventing catecholamine resistance and adipose tissue dysfunction

We next sought to understand the physiological role of NAMs in adipose tissue function and inflammaging in vivo. Resident CD169⁺ macrophages have previously been shown to associate with sympathetic nerves around the lung airway¹³. In the VAT, while CD169 was not exclusively expressed in NAMs, our scRNA-seq, imaging and flow cytometry data demonstrate that NAMs are enriched in CD169 (Fig. 2c). Given NAMs decline with age, and our data showed female VAT is enriched in CD169⁺ macrophages compared to male mice, we conducted our next studies in female mice (Fig. 2f). To study NAMs we utilized (CD169-DTR mice to deplete CD169⁺ macrophages in the VAT of 3-month-old wild-type (WT) or CD169-DTR female mice treated with diphtheria toxin (DT) for 12 days (Fig. 5a)³². Depletion of NAMs in VAT was confirmed via flow cytometry on day 13 (D13) (Fig. 5b and Extended Data Fig. 7). Depletion of NAMs led to a 5% reduction in body weight by D3 that gradually increased and was equivalent to control mice after D6 (Fig. 5c). Consistent with the latter observation, VAT weight was not significantly different between groups at D13, but there was an increase in the number of total SVF cells from the VAT of CD169-DTR mice (Fig. 5d,e). Flow cytometry of SVF from VAT showed no significant changes in the proportion of B cells, T cells, neutrophils, eosinophils, and MHCII⁺CD11c⁺ cell populations; however, there was a twofold increase in the proportion of Ly6C⁺ monocytes (Fig. 5f and Extended Data Fig. 7). Unlike the VAT, in the spleen, the tissue weight and total number of splenocytes did not change significantly following depletion of CD169⁺ cells, and the increase in Ly6C⁺ monocytes was negligible (Extended Data Figs. 7 and 8a–d). Evaluation of CCR2⁺Ly6C⁺ and CCR2⁻Ly6C⁺ monocyte proportions in the bone marrow of control and CD169-DTR mice following depletion revealed that the increase in monocyte numbers in VAT and spleen specifically correlates with an increase in CCR2⁺Ly6C⁺ monocytes (but not CCR2⁻Ly6C⁺ monocytes) in peripheral tissues such as the bone marrow (Extended Data Fig. 9a,b).

Previously, we and others showed NAMs can regulate norepinephrine bioavailability and may have prolipolytic and anti-obesogenic effects^3,12. To investigate the role for CD169⁺ NAMs in lipolysis we studied two enzymes that regulate lipolysis, HSL and ATGL, in the VAT of adult 9-month-old WT and CD169-DTR female mice killed on D8 (Extended Data Fig. 9c,d). Immunoblot analysis of VAT from ad libitum-fed and 24-h fasted mice showed depletion of CD169⁺ NAMs led to a reduction in monoamine oxidase A (MAOA) protein levels (Fig. 5g and Extended Data Fig. 9e,f). Of note, depletion of CD169⁺ NAMs also resulted in a reduction of total ATGL and HSL in the fed and fasted conditions, but levels of phosphorylated HSL (p-HSL) at serines 563 (S563) and 660 (S660) were only reduced during fasting (Fig. 5g and Extended Data Fig. 9e,f). Furthermore, expression of the β3-adrenergic receptor (Adrb3), which acts upstream of ATGL and HSL during catecholamine-induced lipolysis, was reduced in CD169-DTR mice (Extended Data Fig. 9g). Together, these data suggested that a reduction in adipose NAMs, and consequentially reduced MAOA, may promote catecholamine resistance by increasing catecholamine availability. The increase in catecholamines likely activates a negative feedback mechanism to downregulate the expression of lipolytic proteins. Consistent with this, we also observed decreased expression of the master regulator of adipocyte differentiation, Pparγ, suggesting there is an overall dysregulation of lipid metabolism following NAM depletion (Extended Data Fig. 9g). Last, we observed a reduction in the expression of oxidative phosphorylation proteins in the VAT, suggesting loss of NAMs may also alter mitochondrial function (Fig. 5h and Extended Data Fig. 9h). Together, our results confirm that NAMs are critical for regulating catecholamine signaling and suggest that they play an essential role in preventing catecholamine resistance and maintaining adipose tissue function.

Depletion of CD169⁺ NAMs during aging impairs adipose tissue metabolism and increases inflammation

To determine the function of CD169⁺ NAMs through lifespan, we aged CD169-DTR mice up to 24 months and evaluated the animals 24 days (D24) after NAM depletion (Fig. 6a). Consistent with previous data, 6-month-old CD169-DTR female mice lost 5% of their body weight by D3 but regained it by D12 and maintained their body weight until D24 (Fig. 6b). On the contrary, 24-month-old CD169-DTR female mice lost approximately 15% of their body weight by D24 and had smaller fat pads and more SVF cells compared to controls (Fig. 6b–e). To investigate whether the reduction in VAT weight was due to changes in lipolysis, we measured serum free fatty acid (FFA) levels on D24 and found that young 6-month-old CD169-DTR female mice had similar serum FFA levels, but 24-month-old female mice had lower FFA levels relative to WT controls (Fig. 6f). In addition, analysis of VAT from 24-month-old WT and CD169-DTR female mice showed that ablation of CD169⁺ NAMs also reduced total MAOA, ATGL and phosphorylated HSL (S563 and S660) protein levels, but not total HSL (Fig. 6g and Extended Data Fig. 10a). Moreover, GDF3 (Gdf3), which can inhibit catecholamine-induced lipolysis, was increased 20-fold in VAT from 24-month-old but not 6-month-old CD169-DTR female mice^3,33,34 (Fig. 6h). Expression of Adrb3 was lower between genotypes but did not reach statistical significance at D24 (Fig. 6h). Loss of CD169⁺ NAMs also led to reduction in the expression of Pparγ in aged VAT (Fig. 6h). Depletion of NAMs for 34 days in aged CD169-DTR mice led to comparable body weight loss and reduced expression of lipolysis-associated genes (Adrb3 and Pnpla2) (Extended Data Fig. 10b,c). Of note, the rate-limiting enzyme for fatty acid oxidation (FAO), Cpt1a, was increased in aged CD169-DTR mice, suggesting body weight loss may be driven by increased FAO (Extended Data Fig. 10d). Together, these data confirm CD169⁺ NAMs are required for regulation of the lipolytic response and raises the possibility that GDF3 is an important regulator of immunometabolic response in aging.

As chronic inflammation is a hallmark feature of the aging process and is directly linked to decline in tissue function, we assessed whether loss of CD169⁺ NAMs alters age-induced inflammation. Analysis of serum showed there was an elevation in monocyte chemoattractant protein-1 (MCP-1) and TNF levels in 24-month-old CD169-DTR mice, but no changes in systemic levels of IL-1β or IL-6 (Fig. 6i). Flow cytometry analyses revealed that unlike adult VAT, depletion of NAMs led to a decrease in B cells and CD4⁺ T cells and increase in CD8⁺ T cells in aged VAT (Fig. 6j). Moreover, NAM depletion in aged VAT was coupled with a 16-fold increase in monocytes, as well as increases in neutrophils, eosinophils, and MHCII⁺CD11c⁺ ATMs/DCs (Fig. 6j). Similar changes were observed 34 days post-NAM depletion (Extended Data Fig. 10e). Assessment of specific CD4⁺ T cell subsets showed that regulatory T (T_reg) cells, γδT cells and type 2 helper T (T_H2) cells were reduced following NAM depletion (Extended Data Fig. 10f,g). In comparison, we did not observe changes in tissue weight or the B cell and T cell subsets in the spleen, but there was a small increase in monocytes, neutrophils and eosinophils (Extended Data Fig. 10h,i). Together, these data demonstrate that a reduction in NAMs during aging alters the immune cell landscape in VAT, but not spleen, and suggest that loss of NAMs may potentiate age-induced inflammation.

Consistent with our hypothesis, elimination of CD169⁺ NAMs in aging increased expression of pro-inflammatory cytokines Il1b, Il18, Il6 and Tnf at 24 and 34 days in VAT (Fig. 6k and Extended Data Fig. 10j). Notably, Il10, which is generally considered anti-inflammatory, was elevated upon NAM depletion in aging (Fig. 6k and Extended Data Fig. 10j). Given that NLRP3 inflammasome activation induces inflammaging and impairs metabolic function in aging², we also assessed the expression of adaptor protein, ASC (Asc), and the main effector enzyme, Caspase-1 (Casp1). We found that both Asc and Casp1 were only increased in the VAT of aged CD169-DTR mice (Fig. 6k and Extended Data Fig. 10j). Together, these data show that CD169⁺ NAMs are required to maintain neuroimmune homeostasis and are necessary for restraining age-induced inflammation in aging.

Discussion

Following recent work from our laboratory and multiple groups^{3,8,9,10,11,12,13,14}, it has come to light that macrophages are integral components of postganglionic sympathetic nerve fibers. In this study, we utilized an iv-labeling approach to enrich for noncirculating F4/80⁺CD11b⁺ cells, and performed unbiased scRNA-seq to illuminate the transcriptional diversity of ATMs in the VAT through lifespan. We validated the transcriptional signatures associated with two ATM subclusters: VAMs and LAMs. Notably, our study defined transcriptional signatures of two macrophage subtypes: AAMs and NAMs. We identified that the NAM subcluster in VAT is enriched for Siglec1 (CD169), show CD169⁺ NAMs decrease with age in VAT and provide data suggesting that they may regulate myelin uptake and processing. Finally, using CD169-DTR mice we demonstrate that CD169⁺ NAMs are functionally important for regulating lipolysis, restraining age-induced inflammation and preventing adipose tissue dysfunction at homeostasis and during aging.

Our study supports a model whereby reduction in CD169⁺ NAMs may drive or partially explain the reduction in lipolysis and increase in visceral adiposity observed during aging. This hypothesis is supported by our data showing that CD169⁺ NAMs decrease with age in the VAT and that a reduction in NAMs in CD169-DTR female mice blunts lipolysis protein expression. However, it remains unclear whether a similar mechanism (a reduction in NAMs) might also play a role in diet-induced obesity. One study demonstrated that ablation of CD169⁺ macrophages using CD169-DTR mice fed a chow or high-fat diet (HFD) resulted in adipocyte hypertrophy in VAT, a result that would be consistent with our current hypothesis³⁵. A separate study that did not use CD169 showed that F4/80⁺CD45⁺ NAMs accumulate on sympathetic nerves from the SAT of HFD-fed mice, but it is unknown if similar changes also occur in the VAT¹². The latter study also noted an accumulation of CD11c⁺ macrophages on sympathetic adipose nerves from obese mice, suggesting the increase is due to infiltrating CD169⁻CD11c⁺ cells expressing F4/80 on their surface. Thus, it remains unclear what role NAMs play under distinct energy states (lean and obese) or in distinct fat depots (VAT and SAT).

Here we show that CD169⁺ macrophage subsets preferentially associate with nerves in the VAT, consistent with a study showing CD169 (Siglec1 or Sn) can be used to identify NAMs in the lung¹³. We also provide evidence that CD169⁺ NAMs can simultaneously contact nerves and blood vessels suggesting that they participate in the maintenance of both structures. In line with this, our scRNA-seq showed that NAMs are enriched for Vcam1, which mediates the adhesion of cells to the vascular endothelium; however, the role of CD169⁺ cells in the PNS remains less well defined. Our IF imaging and transcriptomic analysis revealed CD169⁺ NAMs engulf myelin and that NAMs are enriched for gliogenesis-related genes, supporting a role for NAMs in myelin uptake, processing or glial cell generation. Of note, two studies using genetic models of peripheral and central nervous system demyelination have shown that CD169-deficient mice exhibited reduced demyelination and infiltration of CD8⁺ T cells and macrophages^36,37. Hence, studies are needed to delineate whether CD169⁺ NAMs have defined roles in Schwann cell generation, maintenance and/or nerve demyelination. As CD169-DTR mice deplete CD169⁺ cells globally, developing conditional knockout models targeting NAMs specifically will be essential to define their role in the PNS.

Notably, our study identified a population of macrophages in the VAT, which we termed AAMs, that was virtually absent in the young condition. Using flow cytometry, we confirmed that AAMs increase with age in the VAT. We found that AAMs were enriched for expression of genes encoding several proteins referred to as ‘SASP’, but which include canonical inflammation-associated effectors (for example, C3, CD55, CCL8 and MMP9) that are not necessarily derived from senescent cells, suggesting that macrophage subsets may contribute to the production of these inflammaging-associated proteins in adipose tissue. Our data suggest that accumulation of AAMs may promote age-related dysfunction, but this requires future investigation with targeted approaches. In support of this hypothesis, we showed that AAMs can be defined by expression of the cell surface marker CD38 (a marker previously linked to a subset of macrophages that reduces NAD⁺ levels and drives functional decline in aging^27,28). Furthermore, AAMs were also enriched for Cxcl13 expression, which is involved in the formation of FALCs, which accumulate in the adipose tissue during aging³⁸.

In summary, our results transcriptionally identify CD169⁺ macrophages as an ATM subpopulation in the VAT and NAMs, and identify a role for these ATMs in catecholamine-induced lipolysis and age-induced inflammation. Furthermore, we define an ATM subpopulation, AAMs, which may play a role in driving age-related dysfunction.

Methods

Mice

The 2- and 22-month-old C57BL/6J WT male and female mice were obtained from the National Institute on Aging (NIA) Aged Rodent Colony and housed in facilities at Yale University. CD169^Cre:Rosa26^mTmG (CD169:mTmG) and LysM^Cre:Rosa26^mTmG (LysM:mTmG) reporter mice were generated in our facility by crossing CD169^Cre mice (kindly provided by K. M. Khanna, New York University) or LysM^Cre mice to Rosa26^mTmG mice (The Jackson Laboratory). Siglec1^DTR/+ (CD169-DTR) mice were generated by K. M. Khanna in NYU Langone Health. Experiments using CD169-DTR mice were conducted in adult female (3, 6 and 9 months old) or aged female (24 months old) mice. All mice used in this study were housed in specific-pathogen-free facilities in ventilated cage racks that deliver HEPA-filtered air to each cage with free access to sterile water through a Hydropac system at the Yale School of Medicine. Mice were fed ad libitum with a standard vivarium chow (Harlan 2018s) and housed under a 12-h light–dark cycle with controlled temperature and humidity conditions (approximately 22 °C and 60% humidity). Animal experiments and animal use were conducted in compliance with the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals and was approved by the Institutional Care and Use Committee at Yale University.

Diphtheria toxin administration

DT (Sigma Aldrich) was prepared in endotoxin-free PBS and administered i.p. on day 1 (40 ng g⁻¹ body weight) and then every 2–3 days (4 ng g⁻¹ body weight) until mice were killed.

Adipose tissue digestion and flow cytometry

VAT was digested in 0.8 mg ml⁻¹ collagenase II (Worthington Biochemicals), 3% BSA (Sigma), 1.2 mM calcium chloride, 1.0 mM magnesium chloride, 0.8 mM zinc chloride and 1× HBSS (Life Technologies, 14185-052) for 40–45 min in a 37 °C water bath with vigorous shaking. The SVF was filtered on a 100-μm cell strainer. Red blood cell lysis was performed using ACK lysis buffer, and cells were filtered a second time with a 70-μm cell strainer. One million SVF cells were used for flow cytometry staining. Cells were stained with live/dead Aqua viability dye (Invitrogen) or Viakrome 808 viability dye (Beckman Coulter) for 30 min, followed by Fc receptor blocking (Invitrogen) for 10 min and surface staining for 30 min with the following antibodies: CD45-PE-Cy7 (clone 30-F11, cat. no. 103114), CD11b-APC (clone M1/70 cat. no. 101212), FOLR2-PE (clone 10/FR2 cat. no. 153304), CD9-FITC (clone MZ3 cat. no. 124808), CD226-BV421 (clone TX42.1 cat. no. 133615), CD11c-BV750 (clone N418 cat. no. 117357), B220-BV711 (clone RA3-6B2 cat. no. 103255), Ly6G-FITC (clone 1A8 cat. no. 127606), Ly6C-AF700 (clone HK1.4 cat. no. 128024) (all from BioLegend); CD45.2-FITC (clone 104 cat. no. 11-0454-85), F4/80-PE-Texas Red (clone BM8 cat. no. MF48017), CD163-SB702 (clone TNKUPJ cat. no. 67-1631-82), CD169-APC-eFluor780 (clone SER-4 cat. no. 47-5755-82), CD45-APC (clone 30-F11, cat. no. 17-0451-83), CD169-PE (clone SER-4 cat. no. 12-5755-80), FOXP3-PE-Cy5.5 (clone FJK-16s cat. no. 35-5773-82), LYVE1-biotin (clone ALY7 cat. no. 13-0443-82) (all from Thermo Fisher Scientific); CD38-BUV737 (clone 90/CD38 cat. no. 741748), SiglecF-BV605 (clone E50-2440 cat. no. 740388), CD11b-BUV395 (clone M1/70 cat. no. 563553), CD3-BV480 (clone 17A2 cat. no. 565642), CD4-BUV563 (clone GK1.5 cat. no. 612923), CD8a-BUV615 (clone 53-6.7 cat. no. 613004), I-A/I-E-BUV805 (clone M5/114.15.2 cat. no. 748844), CD25-PE-CF594 (clone PC61 cat. no. 562694) (all from BD Biosciences); and F4/80-StarBrightViolet570 (clone CI:A3-1 cat. no. MCA497) (from Bio-Rad). Flow cytometry analysis was performed on a Becton Dickinson Symphony instrument, and data were analyzed using FlowJo software (TreeStar).

RNA extraction, cDNA generation and RT–qPCR

Total RNA was isolated using TRIzol followed by the RNeasy kit (QIAGEN) according to manufacturer’s instructions. Synthesis of complementary DNA was performed using the iScript cDNA Synthesis kit (Bio-Rad) according to manufacturer’s instructions. Then, 1–2 μg of cDNA was diluted to 5 ng ml⁻¹ and was amplified by specific primers in a 20-μl reaction using the Power SYBR Green PCR Master Mix (Applied Biosystems). Analysis of gene expression was carried out in a LightCycler 480 II instrument (Roche). For each gene, mRNA expression was calculated as 2^ΔCT relative to Actb and Hprt1 expression. Primer sequences used are provided in Supplementary Table 4.

Western blotting

VAT was collected and immediately snap-frozen in liquid nitrogen. VAT was pulverized using a mortar and pestle, chilled with liquid nitrogen and homogenized with 300–500 μl of RIPA buffer (Sigma) containing protease inhibitor (Sigma) and phosphatase inhibitors 2 and 3 (Sigma). Samples were left on ice for 30 min with vortexing every 5 min to disrupt membranes and then centrifuged at 13,200 rpm at 4 °C for 15 min. Supernatant was collected and protein concentration was quantified using the DC protein assay (Bio-Rad). Then, 30 μg of protein lysate containing 1× NuPAGE Sample Reducing Agent (Invitrogen) and 1× NuPAGE LDS Sample Buffer (Life Technologies) was heated at 95 °C for 10 min and then separated by SDS–PAGE using a NuPAGE 4–12% Bis-Tris Gel (Invitrogen). Proteins were transferred onto nitrocellulose membranes using a semi-dry transfer system (Bio-Rad). Membranes were then blocked in 5% BSA in 1× TBST (Bio-Rad) and incubated in primary antibodies overnight. The blots were probed with the following antibodies: phosphorylated HSL S536 (Cell Signaling; 4139S), phosphorylated HSL S660 (Cell Signaling; 45804S), total HSL (Cell Signaling; 4107S), ATGL (Cell Signaling; 2439S), MAOA (Abcam; ab126751), β-actin (Cell Signaling; 4967S), HSP-90 (Cell Signaling; 4874S) and Ndufb8, Sdhb, Uqcrc2 and Atp5a (MitoProfile Total OXPHOS, Abcam; 110413). HRP-conjugated anti-mouse (Invitrogen) or anti-rabbit (Invitrogen) secondary antibodies were used appropriately. Images were collected using a ChemiDoc MP Imaging System (Bio-Rad) and analyzed using Image Lab software (Bio-Rad).

Cytokine measurements

Blood was collected and incubated at room temperature for 30 min then centrifuged at 3,000 rpm at 4 °C for 20 min to isolate the serum. The concentrations of serum IL-1β, IL-6, GROα, TNF, MCP-1 and eotaxin were determined by six-plex magnetic bead panel (Invitrogen) following the manufacturer’s instructions with appropriate dilution. Analysis of cytokine levels was carried out in a Luminex 200 instrument (Thermo Fisher Scientific).

Free fatty acid measurements

Blood was collected and incubated at room temperature for 30 min then centrifuged at 3,000 rpm at 4 °C for 20 min to isolate the serum. Serum FFA levels were measured using the Free Fatty Acid Assay kit (Sigma MAK466) as per the manufacturer’s instructions.

Whole-mount immunofluorescence and confocal microscopy

Adipose tissues were fixed in 4% paraformaldehyde (PFA) overnight at 4 °C, washed three times with PBS for 30 min, followed by permeabilization with 5% BSA and 0.2% Triton X-100 overnight at 4 °C or permeabilization with 5% BSA and 0.5% Triton X-100 for 48 h at 4 °C. Tissues were stained with primary antibodies for 48 h at 4 °C. The following primary antibodies were used: TUBB3 (TUJ1, 1:200 dilution), CD169 (SER-4, 1:100 dilution), MBP (P82H9, 1:100 dilution) and CNP (SMI 91, 1:100 dilution). Tissues were washed and mounted on a glass slide and coverslip in mounting medium (DAKO S3023). For imaging of large nerve fibers, we isolated nerves from SAT but kept surrounding adipocytes to preserve nerve integrity and processed as described above. Samples were imaged by confocal microscopy using Leica Stellaris 5 or 8 and analyzed using ImageJ or Leica Application Suite X.

Transmission electron microscopy

Mice were perfused with 4% PFA. SAT was collected and maintained in 4% PFA. Then, 1–2-mm² pieces of tissues containing nerve fibers were dissected out and fixed in 2.5% glutaraldehyde and 2% paraformaldehyde in 0.1 M cacodylate buffer (pH 7.4) for 2 h at room temperature and overnight at 4 °C. They were then rinsed and post-fixed in 1% OsO₄ at room temperature for 1 h. After staining en bloc with 2% aqueous uranyl acetate for 30 min, the tissue was dehydrated in a graded series of ethanol to 100% propylene oxide and finally embedded in EMbed 812 resin. For TEM, sample blocks were then polymerized in 60 °C oven for 24 h. Thin sections (60 nm) were cut by a Leica ultramicrotome and post-stained with 2% uranyl acetate and lead citrate. Sample were examined with a FEI Tecnai transmission electron microscope at 80 kV of accelerating voltage, digital images were recorded with an Olympus Morada CCD camera and iTEM imaging software.

Intravital two-photon microscopy

LysM^Cre:Rosa26^mTmG (LysM:mTmG) or CD169^Cre:Rosa26^mTmG (CD169:mTmG) mice were anesthetized with ketamine/xylazine and placed in dorsal recumbency position on a heating pad. For VAT, a 1-cm abdominal incision was made to open the skin and peritoneal cavity. The VAT was gently pulled out through the incision and placed on a metal support arm. For the SAT, a 4–5-cm curved abdominal incision was made to create a lateral skin flap exposing the SAT lying on the inside of the skin. Skin and SAT were placed over a metal support arm resting next to the abdomen. A coverslip was placed over the tissue on the support arm using a micromanipulator. Sterile gauze and a warm saline infusion were utilized to ensure moistening of exposed tissue and abdominal cavity during imaging. The animals were restrained for approximately 4 h and kept under isoflurane anesthetic by a precision vaporizer. After imaging was completed, the animals were killed. Inguinal adipose tissues were visualized by 2P microscopy using LaVision Biotec TriMScope (Miltenyi Biotec) and analyzed using Imaris (Oxford Instruments) software.

Bulk RNA sequencing analysis

mRNA from VAT and BAT was purified from approximately 1 μg of total RNA with oligo-dT beads and sheared by incubation at 94 °C. RNA integrity was determined by running an Agilent Bioanalyzer gel, which measures the ratio of the ribosomal peaks. Following first-strand synthesis with random primers, second-strand synthesis was performed with dUTP for generating strand-specific sequencing libraries. The cDNA library was then end-repaired, and A-tailed, adaptors were ligated and second-strand digestion was performed by uracil-DNA-glycosylase. Indexed libraries that meet appropriate cutoffs for both are quantified by RT–qPCR using a commercially available kit (KAPA Biosystems). Insert size distribution was determined with the LabChip GX or Agilent Bioanalyzer. Samples with a yield of ≥0.5 ng μl⁻¹ were used for sequencing. Raw reads were quality-assessed with FastQC (http://www.bioinformatics.babraham.ac.uk/projects/fastqc). They were mapped to the GENCODE vM9 mouse reference genome³⁹ with STAR⁴⁰ using options –outFilterMultimapNmax 15, –outFilterMismatchNmax 6, –outSAMstrandField All, –outSAMtype BAM SortedByCoordinate and –quantMode TranscriptomeSAM. Gene expression was quantified with RSEM⁴¹. PCA was performed in R after removing a donor effect with the ComBat function from the sva R package (v.3.28.0)⁴². PCA was based on the R-log-transformed FPKM values of the top 5,000 genes with the highest average expression.

scRNA-seq analysis

The iv labeling was performed as previously described¹⁵. In brief, 2- and 22-month-old mice received 2.5 μg of murine CD45-APC antibody iv by retro-orbital injection to label circulating cells (five mice per group). Mice were killed 3 min after antibody administration, and the VAT was digested as previously described (refer to the ‘Adipose tissue digestion and flow cytometry’ section). Following adipose tissue digestion, SVF cells isolated from VAT were pooled and then stained with viability dye, CD45.2, CD45, CD3, CD19, SiglecF, F4/80 and CD11b and were sorted using a BD FACSAria II and analyzed using Chromium Next GEM Automated Single Cell 3′ cDNA kit v.3.1 (10x Genomics). Experiments on 2- and 22-month-old female mice were performed the next day using the same protocol (five mice per group). The Cell Ranger Single-Cell Software v.3.0.2 (available at 10x Genomic’s website) was used to perform sample demultiplexing and the resulting fastq files were aligned on the mm10 genome reference (10x Genomics; https://cf.10xgenomics.com/supp/cell-exp/refdata-gex-mm10-2020-A.tar.gz) with the Cell Ranger count tool. Data analysis was performed using Seurat v.4.3.0 R package, including QC, dimension reduction, clustering, cell-type identification and comparative analyses between conditions^{43,44,45,46,47}. In the step of QC, poor-quality cells with the number of expressed genes <363 (363 corresponding to the first mode in the bimodal distribution of number of expressed genes) were filtered out. We also excluded cells if their mitochondrial gene percentages were over 10%. After combining cells from all samples, we first normalized the raw count matrix using log-normalization with scale.factor 10,000 and then defined top variable genes using ‘mean.var.plot’ methods. We then applied PCA for dimensionality reduction and retained 20 leading principal components for cell clustering. The shared nearest neighbor graph was constructed using default hyperparameters, which was then used for cell clustering with the Louvain algorithm at a resolution of 0.5. Cluster-specific genes were found by Wilcoxon rank-sum test (min.pct, 0.1 and logfc.threshold, 0.1). After identifying cluster-specific genes, we annotated cell types based on canonical marker genes. Clusters 9, 12, and 14 did not express significant levels of Ptprc (CD45), Itgam (CD11b) and Adgre1 (F4/80), suggesting that these clusters were likely contaminating cells and were therefore excluded from future analyses. DEGs for conditions were identified by performing a Wilcoxon rank-sum test on each cluster for the conditions Aged and Young (min.cell, 25; min.pct, 0.01; and logfc.threshold, log₂).

Enrichment analysis

Enrichment analysis was performed using the ClusterProfiler v.4.11.0 R package on DEGs found in scRNA-seq between clusters and conditions⁴⁸. The R package msigdbr v.7.5.1 was used to retrieve terms of interest from The Molecular Signatures Database⁴⁹. Specifically, GO terms were retrieved from Mus musculus and category C5 (subcategory GO:BP, GO:CC and GO:MF) and hallmark gene sets were retrieved from M. musculus and category H. The Benjamini–Hochberg method was used for adjusting P values. The threshold cutoff for P values and q values was set at 0.05. A single-cell GSES was calculated for each cell using the ‘enrichIt’ function from the R package ‘escape’ v.1.10.0, with method set to UCell^50,51. Single-cell GSES was calculated for the following gene sets: SenMayo (SAUL_SEN_MAYO, GSEA: MM16098) and NAMs (LEVEAU_NERVE_ASSOCIATED_MACROPHAGES; Supplementary Table 3).

Trajectory analysis

The Velocyto tool was employed to generate two count matrices for each sample: one for pre-mature (unspliced) and another for mature (spliced) RNA abundances⁵². These matrices serve as the input for subsequent RNA velocity inference. Specifically, the data from four samples were combined and processed using scVelo (v.0.2.5)⁵³. Preprocessing steps include gene filtering (removing genes with a minimum shared count of less than 20), normalization, extracting top 5,000 highly variable genes and log1p transform. RNA velocities were inferred based on the steady-state model. Velocities were projected onto the UMAP cell embedding for visualization, which facilitates inference of cell state transition dynamics.

Statistics and reproducibility

The investigators were not blinded to allocation during experiments and outcome assessment. Control and experimental groups were assigned by cage using simple randomization. All experiments contained littermates and non-littermates, which were both randomly assigned to control and experimental group. Young and aged mice were obtained from the NIA Aged Rodent Colony. Statistical significance was calculated by a two-tailed Student’s t-test or analysis of variance (ANOVA) using a post hoc test for multiple comparisons; ^∗P < 0.05; ^∗∗P < 0.005; ^∗∗∗P < 0.001; ^∗∗∗∗P < 0.0001. No data were excluded from the analysis. Aged mice with extreme frailty, tumors or other age-related pathology were excluded from the study. This amounted to less than 5% of all aged mice. A confidence interval of 95% was used for all statistical tests. All data were assumed to be normally distributed, unless the s.d. was identified as significantly different between groups. Sample size for experiments was based upon previously published experiments. All statistical tests were performed using GraphPad Prism v.7 for Windows (GraphPad Software). Data are expressed as mean ± s.e.m. Biological replicates and the number of independent experiment repetition are described in the figure legends.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

All sequencing datasets are deposited in the Genome Expression Omnibus under accession number GSE294122. All other data supporting the findings of this study are available within the paper or in the Supplementary Information, or from the corresponding author upon request.

Code availability

Software and package versions used are outlined in the Methods section. The code and analysis notebooks used to generate the results presented in this study are provided in a public GitHub repository at https://github.com/egh23/ATManalysis.git. These materials include all scripts necessary to reproduce the key findings and analyses described in the paper. Access to these files ensures that readers have full access to the minimum dataset and computational methods required to interpret, verify and extend the research. No additional custom code was used beyond what is provided in the GitHub repository. The code repository has also been permanently archived on Zenodo at https://doi.org/10.5281/zenodo.15716667 (ref. ⁵⁴).

References

Franceschi, C. et al. Inflamm-aging: An evolutionary perspective on immunosenescence. Ann. NY Acad. Sci. 908, 244–254 (2000).
Article CAS PubMed Google Scholar
Youm, Y. H. et al. Canonical Nlrp3 inflammasome links systemic low-grade inflammation to functional decline in aging. Cell Metab. 18, 519–532 (2013).
Article CAS PubMed PubMed Central Google Scholar
Camell, C. D. et al. Inflammasome-driven catecholamine catabolism in macrophages blunts lipolysis during ageing. Nature 550, 119–123 (2017).
Article CAS PubMed PubMed Central Google Scholar
Goldberg, E. L. et al. IL-33 causes thermogenic failure in aging by expanding dysfunctional adipose ILC2. Cell Metab. 33, 2277–2287 e2275 (2021).
Article CAS PubMed PubMed Central Google Scholar
Lumeng, C. N. et al. Aging is associated with an increase in T cells and inflammatory macrophages in visceral adipose tissue. J. Immunol. 187, 6208–6216 (2011).
Article CAS PubMed Google Scholar
Jaitin, D. A. et al. Lipid-associated macrophages control metabolic homeostasis in a Trem2-dependent manner. Cell 178, 686–698 e614 (2019).
Article CAS PubMed PubMed Central Google Scholar
Silva, H. M. et al. Vasculature-associated fat macrophages readily adapt to inflammatory and metabolic challenges. J. Exp. Med 216, 786–806 (2019).
Article CAS PubMed PubMed Central Google Scholar
Chakarov, S. et al. Two distinct interstitial macrophage populations coexist across tissues in specific subtissular niches. Science https://doi.org/10.1126/science.aau0964 (2019).
Gabanyi, I. et al. Neuro-immune Interactions drive tissue programming in intestinal macrophages. Cell 164, 378–391 (2016).
Article CAS PubMed PubMed Central Google Scholar
Kolter, J. et al. A subset of skin macrophages contributes to the surveillance and regeneration of local nerves. Immunity 50, 1482–1497 e1487 (2019).
Article CAS PubMed Google Scholar
Muller, P. A. et al. Crosstalk between muscularis macrophages and enteric neurons regulates gastrointestinal motility. Cell 158, 1210 (2014).
Article CAS PubMed Google Scholar
Pirzgalska, R. M. et al. Sympathetic neuron-associated macrophages contribute to obesity by importing and metabolizing norepinephrine. Nat. Med. 23, 1309–1318 (2017).
Article CAS PubMed PubMed Central Google Scholar
Ural, B. B. et al. Identification of a nerve-associated, lung-resident interstitial macrophage subset with distinct localization and immunoregulatory properties. Sci. Immunol. https://doi.org/10.1126/sciimmunol.aax8756 (2020).
Medof, M. E., Walter, E. I., Rutgers, J. L., Knowles, D. M. & Nussenzweig, V. Identification of the complement decay-accelerating factor (DAF) on epithelium and glandular cells and in body fluids. J. Exp. Med. 165, 848–864 (1987).
Article CAS PubMed Google Scholar
Anderson, K. G. et al. Intravascular staining for discrimination of vascular and tissue leukocytes. Nat. Protoc. 9, 209–222 (2014).
Article CAS PubMed PubMed Central Google Scholar
Goldberg, E. L. et al. Ketogenesis activates metabolically protective γδ T cells in visceral adipose tissue. Nat. Metab. 2, 50–61 (2020).
Article CAS PubMed PubMed Central Google Scholar
Magalhaes, M. S. et al. Role of Tim4 in the regulation of ABCA1+ adipose tissue macrophages and post-prandial cholesterol levels. Nat. Commun. 12, 4434 (2021).
Article CAS PubMed PubMed Central Google Scholar
Dick, S. A. et al. Three tissue resident macrophage subsets coexist across organs with conserved origins and life cycles. Sci. Immunol. 7, eabf7777 (2022).
Article CAS PubMed Google Scholar
Zaman, R. et al. Selective loss of resident macrophage-derived insulin-like growth factor-1 abolishes adaptive cardiac growth to stress. Immunity 54, 2057–2071 e2056 (2021).
Article CAS PubMed Google Scholar
Ydens, E. et al. Profiling peripheral nerve macrophages reveals two macrophage subsets with distinct localization, transcriptome and response to injury. Nat. Neurosci. 23, 676–689 (2020).
Article CAS PubMed PubMed Central Google Scholar
Wang, P. L. et al. Peripheral nerve resident macrophages share tissue-specific programming and features of activated microglia. Nat. Commun. 11, 2552 (2020).
Article CAS PubMed PubMed Central Google Scholar
Dick, S. A. et al. Self-renewing resident cardiac macrophages limit adverse remodeling following myocardial infarction. Nat. Immunol. 20, 29–39 (2019).
Article CAS PubMed Google Scholar
Kim, K.-W. et al. MHC II+ resident peritoneal and pleural macrophages rely on IRF4 for development from circulating monocytes. J. Exp. Med. 213, 1951–1959 (2016).
Article CAS PubMed PubMed Central Google Scholar
Subramanian, A. et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl Acad. Sci. USA 102, 15545–15550 (2005).
Article CAS PubMed PubMed Central Google Scholar
Liberzon, A. et al. The Molecular Signatures Database hallmark gene set collection. Cell Syst. 1, 417–425 (2015).
Article CAS PubMed PubMed Central Google Scholar
Castanza, A. S. et al. Extending support for mouse data in the Molecular Signatures Database (MSigDB). Nat. Methods 20, 1619–1620 (2023).
Article CAS PubMed PubMed Central Google Scholar
Covarrubias, A. J. et al. Senescent cells promote tissue NAD+ decline during ageing via the activation of CD38+ macrophages. Nat. Metab. 2, 1265–1283 (2020).
Article CAS PubMed PubMed Central Google Scholar
Chini, C. C. S. et al. CD38 ecto-enzyme in immune cells is induced during aging and regulates NAD+ and NMN levels. Nat. Metab. 2, 1284–1304 (2020).
Article CAS PubMed PubMed Central Google Scholar
Saul, D. et al. A new gene set identifies senescent cells and predicts senescence-associated pathways across tissues. Nat. Commun. 13, 4827 (2022).
Article CAS PubMed PubMed Central Google Scholar
Wightman, D. P. et al. A genome-wide association study with 1,126,563 individuals identifies new risk loci for Alzheimer’s disease. Nat. Genet. 53, 1276–1282 (2021).
Article CAS PubMed PubMed Central Google Scholar
Muzumdar, M. D., Tasic, B., Miyamichi, K., Li, L. & Luo, L. A global double-fluorescent Cre reporter mouse. Genesis 45, 593–605 (2007).
Article CAS PubMed Google Scholar
Perez, O. A. et al. CD169+ macrophages orchestrate innate immune responses by regulating bacterial localization in the spleen. Sci. Immunol. 2, eaah5520 (2017).
Article PubMed PubMed Central Google Scholar
Guo, T. et al. Adipocyte ALK7 links nutrient overload to catecholamine resistance in obesity. eLife 3, e03245 (2014).
Article PubMed PubMed Central Google Scholar
Yogosawa, S., Mizutani, S., Ogawa, Y. & Izumi, T. Activin receptor-like kinase 7 suppresses lipolysis to accumulate fat in obesity through downregulation of peroxisome proliferator–activated receptor γ and C/EBPα. Diabetes 62, 115–123 (2012).
Article PubMed PubMed Central Google Scholar
Chen, Q. et al. Resident macrophages restrain pathological adipose tissue remodeling and protect vascular integrity in obese mice. EMBO Rep. 22, e52835 (2021).
Article CAS PubMed PubMed Central Google Scholar
Kobsar, I. et al. Attenuated demyelination in the absence of the macrophage-restricted adhesion molecule sialoadhesin (Siglec-1) in mice heterozygously deficient in P0. Mol. Cell. Neurosci. 31, 685–691 (2006).
Article CAS PubMed Google Scholar
Ip, C. W., Kroner, A., Crocker, P. R., Nave, K.-A. & Martini, R. Sialoadhesin deficiency ameliorates myelin degeneration and axonopathic changes in the CNS of PLP overexpressing mice. Neurobiol. Dis. 25, 105–111 (2007).
Article CAS PubMed Google Scholar
Camell, C. D. et al. Aging induces an Nlrp3 inflammasome-dependent expansion of adipose B cells that impairs metabolic homeostasis. Cell Metab. 30, 1024–1039 e1026 (2019).
Article CAS PubMed PubMed Central Google Scholar
Mudge, J. M. & Harrow, J. Creating reference gene annotation for the mouse C57BL6/J genome assembly. Mamm. Genome 26, 366–378 (2015).
Article CAS PubMed PubMed Central Google Scholar
Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013).
Article CAS PubMed Google Scholar
Li, B. & Dewey, C. N. RSEM: accurate transcript quantification from RNA-seq data with or without a reference genome. BMC Bioinform. 12, 323 (2011).
Article CAS Google Scholar
Leek, J. T., Johnson, W. E., Parker, H. S., Jaffe, A. E. & Storey, J. D. The sva package for removing batch effects and other unwanted variation in high-throughput experiments. Bioinformatics 28, 882–883 (2012).
Article CAS PubMed PubMed Central Google Scholar
Hao, Y. et al. Dictionary learning for integrative, multimodal and scalable single-cell analysis. Nat. Biotechnol. 42, 293–304 (2024).
Article CAS PubMed Google Scholar
Hao, Y. et al. Integrated analysis of multimodal single-cell data. Cell 184, 3573–3587.e3529 (2021).
Article CAS PubMed PubMed Central Google Scholar
Stuart, T. et al. Comprehensive integration of single-cell data. Cell 177, 1888–1902 e1821 (2019).
Article CAS PubMed PubMed Central Google Scholar
Butler, A., Hoffman, P., Smibert, P., Papalexi, E. & Satija, R. Integrating single-cell transcriptomic data across different conditions, technologies, and species. Nat. Biotechnol. 36, 411–420 (2018).
Article CAS PubMed PubMed Central Google Scholar
Satija, R., Farrell, J. A., Gennert, D., Schier, A. F. & Regev, A. Spatial reconstruction of single-cell gene expression data. Nat. Biotechnol. 33, 495–502 (2015).
Article CAS PubMed PubMed Central Google Scholar
Wu, T. et al. clusterProfiler 4.0: A universal enrichment tool for interpreting omics data. Innovation 2, 100141 (2021).
CAS PubMed PubMed Central Google Scholar
Dolgalev, I. MSigDB gene sets for multiple organisms in a tidy data format. R package version 7.5.1 (CRAN, 2022); https://CRAN.R-project.org/package=msigdbr
Borcherding, N. et al. Mapping the immune environment in clear cell renal carcinoma by single-cell genomics. Commun. Biol. 4, 122 (2021).
Article CAS PubMed PubMed Central Google Scholar
Andreatta, M. & Carmona, S. J. UCell: robust and scalable single-cell gene signature scoring. Comput. Struct. Biotechnol. J. 19, 3796–3798 (2021).
Article CAS PubMed PubMed Central Google Scholar
La Manno, G. et al. RNA velocity of single cells. Nature 560, 494–498 (2018).
Article PubMed PubMed Central Google Scholar
Bergen, V., Lange, M., Peidli, S., Wolf, F. A. & Theis, F. J. Generalizing RNA velocity to transient cell states through dynamical modeling. Nat. Biotechnol. 38, 1408–1414 (2020).
Article CAS PubMed Google Scholar
Gonzalez-Hurtado, E. egh23/ATManalysis: initial release of analysis code. Zenodo https://doi.org/10.5281/zenodo.15716667 (2025).

Download references

Acknowledgements

We thank X. Liu from the Yale Center for Cellular and Molecular Imaging Electron Microscopy facility for TEM and FIB-SEM imaging. We thank D. Trotta from the Yale Flow Cytometry Facility for cell sorting on BD FACSAria II. This work was supported by the American Federation for Aging Research postdoctoral fellowship (C.L.), the Aging Biology Foundation (M.N.A.) and by National Institutes of Health grants P01AG051459, AR070811 and Cure Alzheimer’s Fund (V.D.D.).

Author information

These authors contributed equally: Elsie Gonzalez-Hurtado, Claire Leveau.

Authors and Affiliations

Department of Pathology, Yale School of Medicine, New Haven, CT, USA
Elsie Gonzalez-Hurtado, Claire Leveau, Keyi Li, Manish Mishra, Rihao Qu, Taverlyn M. Shepard, Yuval Kluger & Vishwa Deep Dixit
Department of Immunobiology, Yale School of Medicine, New Haven, CT, USA
Elsie Gonzalez-Hurtado & Vishwa Deep Dixit
Department of Physiology, University of California, San Francisco, San Francisco, CA, USA
Emily L. Goldberg
Department of Comparative Medicine, Yale School of Medicine, New Haven, CT, USA
Sviatoslav Sidorov & Vishwa Deep Dixit
Department of Microbiology, New York University Grossman School of Medicine, New York, NY, USA
Payal Damani-Yokota, Stephen T. Yeung, Camille Khairallah & Kamal M. Khanna
Department of Genetics, Yale School of Medicine, New Haven, CT, USA
David Gonzalez
Department of Biochemistry, Molecular Biology and Biophysics, University of Minnesota, Minneapolis, MN, USA
Christina Camell
Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, MO, USA
Maxim N. Artyomov
Perlmutter Cancer Center, New York University Langone Health, New York, NY, USA
Kamal M. Khanna
Yale Center for Research on Aging, Yale School of Medicine, New Haven, CT, USA
Vishwa Deep Dixit

Authors

Elsie Gonzalez-Hurtado
View author publications
Search author on:PubMed Google Scholar
Claire Leveau
View author publications
Search author on:PubMed Google Scholar
Keyi Li
View author publications
Search author on:PubMed Google Scholar
Manish Mishra
View author publications
Search author on:PubMed Google Scholar
Rihao Qu
View author publications
Search author on:PubMed Google Scholar
Emily L. Goldberg
View author publications
Search author on:PubMed Google Scholar
Sviatoslav Sidorov
View author publications
Search author on:PubMed Google Scholar
Payal Damani-Yokota
View author publications
Search author on:PubMed Google Scholar
Stephen T. Yeung
View author publications
Search author on:PubMed Google Scholar
Camille Khairallah
View author publications
Search author on:PubMed Google Scholar
David Gonzalez
View author publications
Search author on:PubMed Google Scholar
Taverlyn M. Shepard
View author publications
Search author on:PubMed Google Scholar
Christina Camell
View author publications
Search author on:PubMed Google Scholar
Maxim N. Artyomov
View author publications
Search author on:PubMed Google Scholar
Yuval Kluger
View author publications
Search author on:PubMed Google Scholar
Kamal M. Khanna
View author publications
Search author on:PubMed Google Scholar
Vishwa Deep Dixit
View author publications
Search author on:PubMed Google Scholar

Contributions

E.G.-H. and C.L. performed experiments, analyzed data and wrote the paper. K.L. and R.Q. analyzed the scRNA-seq data. M.M. performed the adipose tissue digestion and flow cytometry for quantification of NAMs in young and aged male and female mice in VAT. E.L.G. performed and analyzed the bulk RNA-seq of resident versus circulating macrophages in VAT and BAT. S.S. analyzed bulk RNA-seq data of resident versus circulating macrophages in VAT and BAT. S.T.Y. and C.K. administered diphtheria toxin, monitored body weight and collected tissue samples from CD169-DTR mice. P.D.-Y. performed the flow cytometric analysis of bone marrow from WT and CD169-DTR mice. T.M.S. performed and analyzed RT–qPCR of VAT from 9-month-old CD169-DTR mice. D.G. performed the 2P in vivo imaging in CD169^Cre:Rosa^mTmG mice (CD169:mTmG). C.C. analyzed the two-photon videos in LysM^Cre:Rosa26^mTmG mice. M.N.A. provided critical advising for the bioinformatic analysis. K.M.K. provided critical reviews and supervised the depletion experiments with CD169-DTR mice. Y.K. provided critical reviews and supervised the bioinformatic analysis and interpretation of the scRNA-seq data. V.D.D. conceived the project, helped with data analysis, interpretation and preparation of the paper.

Corresponding author

Correspondence to Vishwa Deep Dixit.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Aging thanks Cecile Bénézech and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended Data Fig. 1 Resident and circulating macrophages are transcriptionally distinct.

(a) Volcano plot depicting the top differentially expressed genes (DEGs) in VAT resident macrophages (F4/80⁺CD11b⁺iv⁻) versus VAT circulating macrophages (F4/80⁺CD11b⁺iv ⁺). Statistical significance was defined at a false discovery rate (FDR) < 0.10 (10%). Genes meeting this threshold are highlighted in red. VAT: Visceral adipose tissue. (b) Volcano plot depicting the top differentially expressed genes (DEGs) in BAT resident macrophages (F4/80⁺ CD11b⁺iv⁻) versus BAT circulating macrophages (F4/80⁺CD11b⁺iv⁺). Statistical significance was defined at a false discovery rate (FDR) < 0.10 (10%). Genes meeting this threshold are highlighted in red. BAT: brown adipose tissue (c) Volcano plot depicting the top differentially expressed genes (DEGs) in VAT resident macrophages (F4/80⁺CD11b⁺iv⁻) versus BAT resident macrophages (F4/80⁺CD11b⁺iv⁻). Statistical significance was defined at a false discovery rate (FDR) < 0.10 (10%). Genes meeting this threshold are highlighted in red. (d) Pathway enrichment analysis of resident macrophages (F4/80⁺CD11b⁺iv⁻) versus circulating macrophages (F4/80⁺CD11b⁺iv⁺) in the VAT. Statistical significance was defined at a false discovery rate (FDR) < 0.10 (10%). (e) Pathway enrichment analysis of resident macrophages (F4/80⁺CD11b⁺iv⁻) versus circulating macrophages (F4/80⁺CD11b⁺iv⁺) in the BAT. Statistical significance was defined at a false discovery rate (FDR) < 0.10 (10%).

Extended Data Fig. 2 Single cell RNA-sequencing of resident F4/80⁺CD11b⁺ cells from the VAT of young and aged males and females identified 16 distinct clusters.

(a) Uniform Manifold Approximation and Projection (UMAP) representation of unsupervised clustering of 15,709 resident F4/80⁺CD11b⁺ (CD45⁺CD45.2iv⁻CD3⁻CD19⁻SiglecF⁻) cells sorted from VAT of young (2-month-old) and aged (21-month-old) male (13,134 cells) and female mice (2,575 cells) (top) and repartition of cells in each cluster as a fraction of total cells by cluster (bottom). (b) Heatmap depicting the top 25 differentially expressed genes (DEGs) per cluster. (c) Uniform Manifold Approximation and Projection (UMAP) (left) and violin plots (right) showing the expression of Ptprc (CD45), Itgam (CD11b), and Adgre1 (F4/80) genes by cluster.

Extended Data Fig. 3 Macrophage heterogeneity in adipose tissue revealed by RNA velocity and differential gene expression.

(a) Differential abundance test between male and female VAT clusters using scProportionTest. Statistical significance was defined at a false discovery rate (FDR) < 0.05 (5%) and absolute log₂(fold-change) > 0.58. Clusters meeting this threshold are highlighted in red. (b) RNA velocity projected onto the Uniform Manifold Approximation and Projection (UMAP) plot. Velocities are based on the steady-state model.(c) Volcano plot depicting the top differentially expressed genes (DEGs) in Cluster 0 versus Cluster 4. Genes with an adjusted p value > 0.05 and log₂(fold-change) > 0.5 were deemed significant.(d) Volcano plot depicting the top differentially expressed genes (DEGs) in Cluster 0 versus Cluster 5. Genes with an adjusted p value > 0.05 and log₂(fold-change) > 0.5 were deemed significant.

Extended Data Fig. 4 Cluster 5 represents bona fide nerve-associated macrophages in adipose tissue.

(a) Violin plot depicting the gene expression of Maoa, Lyve1, and Ccr2 per cluster. (b) Uniform Manifold Approximation and Projection (UMAP) depicting the gene expression of Slc6a2.(c) Uniform Manifold Approximation and Projection (UMAP) (left) and violin plot (right) showing enrichment of nerve-associated macrophage (NAM) genes and gene set enrichment score (GSES) per cell. The NAM gene set (LEVEAU_NERVE_ASSOCIAED_MACROPHAGES, Table S3) was constructed from bulk RNA sequencing and microarray datasets of NAM-enriched populations in different tissues (adipose tissue, lung, gut, skin, and peripheral nerves); a gene was considered NAM-enriched and included in the NAM gene set if it appeared in 5 or more of the published datasets assessed.(d) Uniform Manifold Approximation and Projection (UMAP) depicting the gene expression of Timd4, Lyve1, and Folr2.

Extended Data Fig. 5 Gating strategy to define resident F4/80⁺CD11b⁺ clusters in VAT using single-cell RNA-seq.

Schematic depicting the flow cytometry gating strategy used to define clusters 0, 3, 4, 5, 6, 10, and 13 identified via single-cell RNA sequencing of resident F4/80⁺CD11b⁺ (CD45⁺CD45.2iv⁻CD3⁻CD19⁻SiglecF⁻) cells from the visceral white adipose tissue (VAT). Fluorescence minus one (FMO) control(s) were included for the following cell surface markers: CD11b, SiglecF, CD169, CD11c, CD226, CD9, CD163, FOLR2, and CD38. Abbreviations for each population and defining cell surface markers are as follows: Nerve-associated-macrophages (NAM, C5): CD169⁺CD11c⁻, Myeloid Precursors-1 (MP-1, C13): CD169⁺CD11c⁺, Vessel-associated Macrophages (VAM, C0): CD38⁺FOLR2⁺CD11c⁻, Age-associated Macrophages (AAM, C4): CD38⁺FOLR2⁻CD169⁻CD11c⁻, Lipid-associated Macrophages (LAM, C10): CD9⁺CD11c⁺CD226⁻CD163⁻CD169⁻, Adipose Tissue Macrophages-1 (ATM-1, C3): CD226⁺CD11c⁺CD163⁻CD169⁻, Dendritic Cells-1 (DC-1, C6): CD163⁺CD226⁺CD11c⁺CD163⁺CD169.

Extended Data Fig. 6 Whole mount immunofluorescence confocal microscopy of CD169⁺ NAMs in VAT.

(a) Whole mount immunofluorescence confocal microscopy of visceral adipose tissue (VAT) from adult female CD169:mTmG dual reporter mice labeled with the pan-neuronal marker TUBB3 (merged image, top left). CD169 is labeled with membrane eGFP (mGFP⁺) (green, top right), non-recombined structures such as adipocytes, nerves, and blood vessels are labeled with membrane Tomato (mTomato⁺) (red, bottom left), and nerves are labeled with TUBB3 (blue, bottom right). (b) Whole mount immunofluorescence confocal microscopy of visceral (VAT, left), subcutaneous (SAT, middle), and brown adipose tissue (BAT, right) from adult female CD169:mTmG dual reporter mice labeled with the pan-neuronal marker TUBB3 (merged image, top left). CD169 is labeled with membrane eGFP (mGFP⁺) (green, top right), non-recombined structures such as adipocytes, nerves, and blood vessels are labeled with membrane Tomato (mTomato⁺) (red, bottom left), and nerves are labeled with TUBB3 (blue, bottom right).(c) Whole mount immunofluorescence confocal microscopy of visceral adipose tissue (VAT) from adult female CD169:mTmG dual reporter mice labeled with the pan-neuronal marker TUBB3 (merged image, top left). CD169 is labeled with membrane eGFP (mGFP⁺) (green, top right), non-recombined structures such as adipocytes, nerves, and blood vessels are labeled with membrane Tomato (mTomato⁺) (red, bottom left), and nerves are labeled with TUBB3 (blue, bottom right).

Extended Data Fig. 7 Gating strategy to quantify the depletion efficiency of NAMs in VAT.

Schematic depicting the flow cytometry gating strategy used to quantify the depletion efficiency of nerve-associated macrophages (NAMs) and other immune cell subsets in the visceral white adipose tissue (VAT) and spleen from wild-type (WT) and CD169-DTR mice following intraperitoneal diphtheria toxin administration in 3-, 6-, 9-, and 24-month-old female mice and at the following endpoints: day 8, day 13, day 25, and day 34.

Extended Data Fig. 8 Spleen immune cell composition remains largely unchanged following CD169⁺ cell depletion.

(a) Representative flow cytometry plots (left) and bar plot (right) quantifying nerve-associated macrophages (NAMs) from the spleen of 3-month-old WT and CD169-DTR female mice treated with DT and sacrificed on D13 (n = 6-9 per genotype). Bar plot represents the proportion of NAMs (CD169⁺CD11c⁺) as a fraction of the total live CD45⁺ subset. Data is represented as a mean +/- SEM. Statistical significance was determined via Student’s two-tailed t-test with α = 0.05 and p < 0.05. (b) Spleen tissue wet weight from 3-month-old WT and CD169-DTR female mice treated with DT and sacrificed on D13. Data is represented as a mean +/- SEM. Statistical significance was determined via Student’s two-tailed t-test with α = 0.05 and p < 0.05. (c) Splenocyte cell number from 3-month-old WT and CD169-DTR female mice treated with DT and sacrificed on D13. Data is represented as a mean +/- SEM. Statistical significance was determined via Student’s two-tailed t-test with α = 0.05 and p < 0.05. (d) Flow cytometry of the spleen from 3-month-old WT and CD169-DTR female mice treated with DT and sacrificed on D13. Bar plots represent the proportion of B cells, CD8⁺ T cells, CD4⁺ T cells, Monocytes, Neutrophils, Eosinophils, and MHCII⁺CD11c⁺ ATMs or dendritic cells (DCs) as a fraction of the total live CD45⁺ subset). Populations were defined using the cell surface markers and gating strategy outlined in Supplementary Figure 5. Data is represented as a mean +/- SEM. Statistical significance was determined via Student’s two-tailed t-test with α = 0.05 and p < 0.05.

Extended Data Fig. 9 CD169⁺ NAM depletion leads to dysregulated lipid metabolism in adipose tissue.

(a) Schematic representation of the experimental protocol used to deplete CD169-macrophages using CD169 diphtheria toxin receptor (CD169-DTR) mice. 9.5-week-old wild-type (WT) and CD169-DTR male mice (n = 3 per genotype) were injected intra-peritoneally (i.p) with 40 ng/gram diphtheria toxin (DT) on day 0 (D0) and bone marrow was collected and isolated from the femur on day 2 (D2) for analysis via flow cytometry. (b) Quantification of neutrophils, macrophages, and monocytes in WT versus CD169-DTR mice in the bone marrow (BM) cells isolated from the femur. Data is represented as a mean +/- SEM. Statistical significance was determined via non-parametric Mann-Whitney U test with p < 0.05. (c) Schematic representation of the experimental protocol used to deplete CD169-macrophages using CD169 diphtheria toxin receptor (CD169-DTR) mice. 9-month-old wild-type (WT) and CD169-DTR female mice (n = 9-11 per genotype) were injected intraperitoneal (i.p) with 40 ng/gram diphtheria toxin (DT) on day 0 (D0) followed by 4 ng/gram DT on D2 and D4. On day 7 (D7) mice either remained ad libitum-fed (n = 4–5 per genotype) or were fasted for 24-hours (n = 5-6 per genotype) and then scarified on D8. (d) Bar plot showing the proportion of CD169⁺CD11c⁻ nerve-associated macrophages (NAMs) in visceral white adipose tissue (VAT) from 9-month-old wild-type (WT) and CD169-DTR female mice treated with diphtheria toxin (DT). Mice were either fed ad libitum or fasted for 24 hours and sacrificed on day 8 (D8). NAMs are shown as a fraction of the total live CD45⁺ cell population. Data are presented as mean ± SEM. Statistical significance was assessed using a two-tailed Student’s t-test (α = 0.05); p < 0.05 was considered significant.(e) Relative protein quantification of Hormone-sensitive Lipase (HSL), HSL phosphorylated at Ser563 (p-HSL_S563), HSL phosphorylated at Ser660 (p-HSL_S660), Adipose triglyceride lipase (ATGL), and Monoamine oxidase A (MAOA) in VAT lysates. Protein was collected from 9-month-old wild-type (WT) and CD169-DTR female mice treated with diphtheria toxin (DT) and sacrificed on day 8 (D8), under ad libitum-fed conditions (n = 4 per genotype). Data are presented as mean ± SEM. Statistical significance was assessed using a two-tailed Student’s t-test (α = 0.05); p < 0.05 was considered significant.(f) Relative protein quantification of Hormone-sensitive Lipase (HSL), HSL phosphorylated at Ser563 (p-HSL_S563), HSL phosphorylated at Ser660 (p-HSL_S660), Adipose triglyceride lipase (ATGL), and Monoamine oxidase A (MAOA) in VAT lysates. Protein was collected from 9-month-old wild-type (WT) and CD169-DTR female mice treated with diphtheria toxin (DT) and sacrificed on day 8 (D8), under 24-hour fasting conditions (n = 4 per genotype). Data are presented as mean ± SEM. Statistical significance was assessed using a two-tailed Student’s t-test (α = 0.05); p < 0.05 was considered significant.(g) Relative protein quantification of oxidative phosphorylation (OXPHOS) complex subunits in VAT lysates from ad libitum-fed 9-month-old wild-type (WT) and CD169-DTR female mice treated with diphtheria toxin (DT) and sacrificed on day 8 (D8) (n = 4 per genotype). The following mitochondrial proteins were assessed: Complex I (CI): NADH:ubiquinone oxidoreductase subunit B8 (NDUFB8), Complex II (CII): succinate dehydrogenase subunit B, iron-sulfur protein (SDHB), Complex III (CIII): ubiquinol-cytochrome c reductase core protein 2 (UQCRC2), Complex IV (CIV): mitochondrially encoded cytochrome c oxidase I (MTCO1), and Complex V (CV): ATP synthase F1 subunit alpha (ATP5A). Data are presented as mean ± SEM. Statistical significance was assessed using a two-tailed Student’s t-test (α = 0.05); p < 0.05 was considered significant.(h) Relative mRNA expression of Adrb3 (top) and Pparγ (bottom) in the VAT of ad libitum-fed 9-month-old WT and CD169-DTR female mice treated with DT and sacrificed on D8 (n = 3–5 per genotype). Data is represented as a mean +/- SEM. Statistical significance was determined via Student’s two-tailed t-test with α = 0.05 and p < 0.05.

Extended Data Fig. 10 Chronic CD169⁺ NAM depletion impairs lipolysis and promotes fatty acid oxidation during aging.

(a) Relative protein quantification of Hormone-Sensitive Lipase (HSL), HSL phosphorylated at Ser563 (p-HSL_S563), HSL phosphorylated at Ser660 (p-HSL_S660), Adipose triglyceride lipase (ATGL), and Monoamine oxidase A (MAOA) in visceral white adipose tissue (VAT) lysates from ad libitum-fed 24-month-old wild-type (WT) and CD169-DTR female mice treated with diphtheria toxin (DT) and sacrificed on day 25 (D25) (n = 4 per genotype). Data are presented as mean ± SEM. Statistical significance was assessed using a two-tailed Student’s t-test (α = 0.05); p < 0.05 was considered significant. (b) Change in body weight curves for 24-month-old WT and CD169-DTR female mice (n = 5-7 per genotype) treated with DT and sacrificed on D34.(c) Relative mRNA expression of Adrb3, Pnpla2, and Hsl in the VAT of ad libitum-fed 24-month-old WT and CD169-DTR female mice treated with DT and sacrificed on D34 (n = 5-7 per genotype). Data are presented as mean ± SEM. Statistical significance was assessed using a two-tailed Student’s t-test (α = 0.05); p < 0.05 was considered significant. (d) Relative mRNA expression of Cpt1a in the VAT of ad libitum-fed 24-month-old WT and CD169-DTR female mice treated with DT and sacrificed on D34 (n = 5-7 per genotype). Data are presented as mean ± SEM. Statistical significance was assessed using a two-tailed Student’s t-test (α = 0.05); p < 0.05 was considered significant. (e) Flow cytometry data from the VAT of 24-month-old WT and CD169-DTR female mice treated with DT and sacrificed on D24. Bar plots represent the proportion of CD4⁺ regulatory T- cells (T-regs), gamma-delta T cells (gd T cells), CD4⁺ T helper 1 cells (T_H1 cells), CD4⁺ T helper 2 cells (T_H2 cells), and CD4⁺ T helper 17 cells (T_H17 cells) as a fraction of the total live CD45⁺ subset. Populations were defined using the cell surface markers and gating strategy outlined in Supplementary Figure 5. Data is represented as a mean +/- SEM. Statistical significance was assessed using a two-tailed Student’s t-test (α = 0.05); p < 0.05 was considered significant. (f) Flow cytometry data from the VAT of 24-month-old WT and CD169-DTR female mice treated with DT and sacrificed on D34. Bar plots represent the proportion of B cells, CD8⁺ T cells, CD4⁺ T cells, Monocytes, Neutrophils, Eosinophils, and MHCII⁺CD11c⁺ ATMs or dendritic cells (DCs) as a fraction of the total live CD45⁺ subset. Populations were defined using the cell surface markers and gating strategy outlined in Supplementary Figure 5. Data is represented as a mean +/- SEM. Statistical significance was assessed using a two-tailed Student’s t-test (α = 0.05); p < 0.05 was considered significant. (g) Flow cytometry data from the VAT of 24-month-old WT and CD169-DTR female mice treated with DT and sacrificed on D34. Bar plots represent the proportion of CD4⁺ regulatory T- cells (T-regs), gamma-delta T cells (gd T cells), CD4⁺ T helper 1 cells (T_H1 cells), CD4⁺ T helper 2 cells (T_H2 cells), and CD4⁺ T helper 17 cells (T_H17 cells) as a fraction of the total live CD45⁺ subset. Populations were defined using the cell markers and gating strategy outlined in Supplementary Figure 5. Data is represented as a mean +/- SEM. Statistical significance was assessed using a two-tailed Student’s t-test (α = 0.05); p < 0.05 was considered significant. (h) Tissue wet weight of the spleen from 24-month-old wild-type (WT) and CD169-DTR female mice treated with DT and sacrificed on D25 (n = 8–19 per genotype). Data is represented as a mean +/- SEM. Statistical significance was assessed using a two-tailed Student’s t-test (α = 0.05); p < 0.05 was considered significant. (i) Flow cytometry data from the spleen from 24-month-old WT and CD169-DTR female mice treated with DT and sacrificed on D25. Bar plots represent the proportion of nerve-associated macrophages (NAMs), B cells, CD8⁺ T cells, CD4⁺ T cells, Monocytes, Neutrophils, Eosinophils, and MHCII⁺CD11c⁺ ATMs or dendritic cells (DCs) as a fraction of the total live CD45⁺ subset. Populations were defined using the cell surface markers and gating strategy outlined in Supplementary Figure 5. Data is represented as a mean +/- SEM. Statistical significance was determined via Student’s two-tailed t-test with α = 0.05 and p < 0.05. (j) Relative mRNA expression of Nlrp3, Asc, Casp1, Il1b, Il18, Il10, Il6, and Tnf in the VAT from ad libitum-fed 24-month-old WT and CD169-DTR female mice treated with DT and sacrificed on D34 (n = 5-7 per genotype). Data is represented as a mean +/- SEM. Statistical significance was determined via Student’s two-tailed t-test with α = 0.05 and p < 0.05.

Supplementary information

Reporting Summary

Supplementary Tables

1–4 Table 1. QC metrics for scRNA-seq samples from resident F4/80⁺CD11b⁺ cells sorted from the visceral white adipose tissue (VAT) of 2- and 22-month-old male and female mice. Resident cells are defined as: CD45⁺CD45.2iv⁻CD3⁻CD19⁻SiglecF. Table 2. List of DEGs for each cluster. Table 3. NAM gene set list: ‘LEVEAU_NERVE_ASSOCIATED_MACROPHAGES’. Table 4. List of primer sequences used for real-time qPCR.

Supplementary Video 1

In vivo 2P imaging of VAT from a 24-month-old male LysM:mTmG dual reporter male mouse which labels LysM⁺ cells with membrane eGFP (mGFP⁺) (green) and non-LysM (non-recombined) cells/tissues such as adipocytes, nerves, and blood vessels with mTomato⁺ (red). Scale bar, 100 µm. Data are representative of n = 1 independent experiment.

Supplementary Video 2

Serial sections from whole-mount IF imaging of VAT from CD169:mTmG dual reporter female mice which labels CD169⁺ macrophages with membrane eGFP (mGFP) (green) and non-CD169 (non-recombined) cells/tissues such as adipocytes, nerves and blood vessels with membrane tdTomato (mTomato⁺) (red). VAT was stained with the pan-neuronal marker, TUBB3. Scale bar, 20 µm.

Supplementary Video 3

Serial sections from whole-mount IF imaging of VAT from CD169:mTmG dual reporter female mice which labels CD169⁺ macrophages with mGFP and non-CD169 (non-recombined) cells/tissues such as adipocytes, nerves and blood vessels with membrane tdTomato (mTomato). VAT was stained with DAPI and the pan-neuronal marker, TUBB3. Scale bar, 1 µm

Supplementary Video 4

3D surface rendering reconstruction of a CD169⁺ NAM (red) associated with an MBP⁺ myelinated nerve fiber (green) in the VAT from WT mice. Tissue was stained with antibodies against CD169 (red), MBP (green) and TUBB3 (blue). Data are representative of n = 1 independent experiment using 3-month-old WT male mice (n = 2). Scale bar, 10 µm.

Supplementary Video 5

3D surface rendering reconstruction of a CD169⁺ NAM (red) associated with an MBP⁺ myelinated nerve fiber (green) in VAT from WT mice. Tissue was stained with antibodies against CD169 (red), MBP (green) and TUBB3 (blue). Data are representative of n = 1 independent experiment using 3-month-old WT male mice (n = 2). Scale bar, 10 µm.

Supplementary Video 6

In vivo 2P imaging of inguinal white adipose tissue (SAT) from adult CD169:mTmG reporter mice which labels CD169⁺ macrophages with mGFP (green) and non-CD169 (non-recombined) cells/tissues such as adipocytes, nerves and blood vessels with membrane tdTomato (mTomato⁺) (red). Data are representative of n = 2 independent experiments using 3-month-old male mice.

Supplementary Video 7

In vivo 2P imaging of SAT from adult CD169:mTmG reporter mice which labels CD169⁺ macrophages with membrane eGFP (mGFP⁺) (green) and non-CD169 (non-recombined) cells/tissues such as adipocytes, nerves and blood vessels with membrane tdTomato (mTomato⁺) (red). Data are representative of n = 2 independent experiments using 3-month-old male mice.

Source data

Source Data Fig. 4

Single-channel images from confocal and 2P microscopy experiments from Fig. 4.

Source Data Fig. 5

Raw images from western blot in Fig. 5g,h.

Source Data Fig. 6

Raw images from western blot in Fig. 6.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Gonzalez-Hurtado, E., Leveau, C., Li, K. et al. Nerve-associated macrophages control adipose homeostasis across lifespan and restrain age-related inflammation. Nat Aging 5, 1828–1843 (2025). https://doi.org/10.1038/s43587-025-00952-9

Download citation

Received: 27 September 2024
Accepted: 30 July 2025
Published: 02 September 2025
Issue date: September 2025
DOI: https://doi.org/10.1038/s43587-025-00952-9

This article is cited by

Newfound immune cell in mice hints at why inflammation spikes with old age
- Gemma Conroy
Nature (2025)