Beiging of perivascular adipose tissue regulates its inflammation and vascular remodeling

The study investigates how perivascular adipose tissue (PVAT) regulates vascular inflammation and remodeling after vascular injury, a key process in atherosclerosis. While endothelial dysfunction is known to initiate inflammatory cell recruitment and pathological remodeling, emerging evidence suggests PVAT can influence vascular pathology via outside-in signaling. Adipose tissues can undergo a white-to-brown/beige phenotypic transition (“beiging”) regulated by factors such as PRDM16, NFIA, and EBF2. The authors hypothesize that vascular injury triggers PVAT beiging, which modulates macrophage-mediated inflammation to promote resolution and protect against pathological remodeling. They aim to define the causality and mechanisms, including identifying beige adipocyte-derived factors that regulate macrophage phenotypes and testing translational relevance in human acute aortic dissection (AAD).

Prior work establishes that inflammation drives atherosclerosis and pathological remodeling, with endothelial dysfunction central to leukocyte recruitment. PVAT, the adipose tissue surrounding vessels, has been implicated in vascular disease via paracrine outside-in effects. Beiging of adipose tissue occurs with cold or β3-adrenergic receptor (β3AR) stimulation and is regulated by transcriptional factors including PRDM16, NFIA, and EBF2, with evidence in rodents and humans. Clinical imaging studies indicate PVAT phenotype changes correlate with coronary inflammation. However, whether and how PVAT phenotypic changes causally modulate vascular inflammation and remodeling in injury settings remained unclear. Conflicting reports on systemic beiging (e.g., cold, β3AR agonists) and atherosclerosis likely reflect systemic metabolic effects versus local vascular actions. This study addresses these gaps by focusing on local PVAT beiging in response to injury and identifying its mediators.

• Animal models: Male or ovariectomized female C57BL/6J mice were used. Endovascular injury of femoral arteries (FAs) was induced by wire insertion. Macrophage depletion was performed using intraperitoneal clodronate liposomes.
• PVAT beiging manipulation: Genetic inhibition via adipocyte-specific Prdm16 deletion (Adipoq-Cre; Prdm16flox/flox). Local pharmacologic modulation was achieved by applying pluronic F-127 gel to FA PVAT containing β3AR agonist (CL316243) or antagonist (SR59230A), or siRNAs targeting Prdm16 or Nrg4.
• Gene expression and histology: Bulk RNA-seq of outer tissues (PVAT-rich) post-injury; qRT-PCR for immune and adipocyte markers; in situ hybridization (RNAscope) for Ucp1 and Nrg4; immunohistochemistry for F4/80, UCP1, iNOS, CD206, CIDEA, NRG4; western blot for UCP1. Morphometry measured intima-media areas and adipocyte size.
• In vitro assays: PVAT stromal vascular fraction preadipocytes isolated and differentiated to beige adipocytes with standard cocktail (insulin, IBMX, indomethacin, dexamethasone, T3, rosiglitazone). Conditioned media were applied to RAW 264.7 macrophages or bone marrow-derived macrophages (BMDMs) pre-activated to M1 (IFNγ + LPS) or M2 (IL-4). Effects on macrophage gene expression (Cd86, Mrc1, Tnf, Serpine1, Ccl2, Il1b) and cell growth (CellTiter-Glo) were assessed. siRNA knockdown in adipocytes targeted Prdm16, Nrg4, Nfia; ErbB4 knockdown in macrophages tested receptor dependence. Recombinant murine NRG4 was applied to classically activated macrophages.
• scRNA-seq re-analysis: Public datasets (GSE133486) of inguinal WAT with CL316243 or cold exposure were integrated (Seurat; Harmony) to identify beige adipocyte clusters and enriched genes. Gene ontology analysis (DAVID) was used. Bulk deconvolution (GSE129083) with MuSiC validated beige cell emergence.
• Ucp1 knockout: Vascular remodeling after injury compared in Ucp1−/− vs wild-type mice.
• AAD models and human studies: Murine AAD induced by AngII, BAPN, and L-NAME; CL316243 administered via osmotic minipumps to test survival and PVAT macrophage polarization (flow cytometry). Human descending aorta PVAT from AAD autopsies versus controls underwent H&E and immunostaining for UCP1, CIDEA, NRG4.
• Statistics: Two-tailed t-tests or Mann–Whitney U-tests for two-group comparisons; one-way ANOVA with Tukey–Kramer post-hoc for multi-group; Kaplan–Meier with log-rank for survival; significance at P < 0.05.

• Injury triggers PVAT inflammation and macrophage accumulation: Immune cell genes (Cd11b, Cd11c, Mrc1, Adgre1) were upregulated in PVAT at 24 h post-injury; F4/80+ macrophages accumulated in PVAT early (day 3) and later infiltrated vessel wall (day 14). Inflammatory cytokines (Tnf, Serpine1, Il1a, Il1b) increased in PVAT.
• PVAT beiging after injury: BAT/beige markers (Ucp1, Cidea, Cox8b, Ppargc1a, Elovl3, Dio2) increased in PVAT by RNA-seq and qRT-PCR; UCP1 upregulation confirmed by in situ hybridization, IHC, and western blot localized to outer PVAT, with adipocytes becoming smaller and multilocular. Macrophage depletion with clodronate reduced UCP1 induction, implicating macrophages in beiging. β3AR signaling mediated beiging, as local antagonist (SR59230A) attenuated injury-induced PVAT beiging.
• Beiging modulates vascular remodeling and macrophage polarization: Adipocyte-specific Prdm16 deletion or local Prdm16 siRNA suppressed PVAT beiging and exacerbated neointimal thickening at day 14. Conversely, local β3AR agonist (CL316243) induced PVAT beiging and reduced intimal thickening. Beiging suppression increased iNOS/CD206 ratio in PVAT macrophages (prolonged pro-inflammatory state), while beiging activation shifted toward anti-inflammatory CD206+ macrophages; early-phase differences were not observed (day 3).
• Beige adipocytes reprogram macrophages via secreted factors: Conditioned media from PVAT-derived beige adipocytes decreased Cd86/Mrc1 ratio and reduced inflammatory cytokines (Tnf, Serpine1, Ccl2, Il1b) in RAW 264.7 cells and BMDMs; it selectively reduced growth of IFNγ+LPS-activated (M1) macrophages but not IL-4 (M2) macrophages. These effects were lost when adipocyte Prdm16 was knocked down.
• UCP1 is dispensable for remodeling protection: Ucp1−/− mice showed vascular wall thickening comparable to wild-type after injury, indicating UCP1 itself is not required for the protective effect of PVAT beiging.
• NRG4 identified as key beige adipocyte factor: scRNA-seq integration identified a beige adipocyte cluster enriched for Ucp1, Cidea, Ppargc1a, and secreted factor Nrg4 (receptor-binding GO term). Nrg4 was upregulated in PVAT after injury and during beige differentiation in vitro; co-localized with Ucp1+ beige adipocytes in situ; its induction depended on Prdm16 and Nfia but not UCP1. Beige adipocyte conditioned media failed to induce macrophage alternative activation or suppress inflammatory genes when Nrg4 was knocked down; recombinant NRG4 reduced Cd86/Mrc1 ratio and cytokines in classically activated macrophages; ErbB4 knockdown in macrophages abrogated these effects. Beige adipocyte Nrg4 knockdown also abolished suppression of M1 macrophage growth. In vivo, local Nrg4 siRNA increased intimal thickening post-injury.
• Human and murine AAD relevance: Human AAD PVAT showed increased macrophage accumulation and elevated UCP1, CIDEA, and NRG4 compared to control PVAT (e.g., UCP1-positive area higher, P=0.0070). In murine AAD, CL316243 improved survival versus vehicle (log-rank P=0.0434) and shifted PVAT macrophage polarization toward M2, reducing M1/M2 ratio (P≈0.0383).

The findings demonstrate that vascular injury induces local PVAT beiging, which shapes the inflammatory milieu to favor resolution and limits pathological vascular remodeling through an outside-in paracrine mechanism. Beiging requires macrophage infiltration and β3AR signaling. Functionally, beige PVAT suppresses prolonged M1 macrophage activity and promotes M2 polarization, reducing neointimal hyperplasia. Mechanistically, NRG4, highly expressed by beige adipocytes, signals via macrophage ErbB4 to repress inflammatory gene expression and selectively reduce growth/survival of classically activated macrophages, while sparing alternatively activated macrophages. UCP1 is not necessary for these anti-remodeling effects, indicating non-thermogenic beige adipocyte signaling is central. The translational relevance is supported by increased beiging markers and NRG4 in human AAD PVAT and by improved survival and macrophage reprogramming with β3AR-driven beiging in an AAD murine model. Collectively, PVAT beiging acts as a protective switch that transitions injury-induced inflammation toward resolution, thereby mitigating vascular remodeling.

This study establishes that vascular injury triggers beiging of adjacent PVAT, which in turn limits inflammation and pathological vascular remodeling via NRG4–ErbB4-mediated reprogramming of macrophages toward an alternatively activated state. Genetic or local pharmacologic suppression of beiging exacerbates remodeling, whereas local β3AR activation enhances beiging, dampens inflammation, and improves outcomes, including survival in an AAD model. UCP1 is dispensable, highlighting paracrine signaling as the key protective mechanism. PVAT beiging and NRG4 represent promising therapeutic targets for vascular injury, atherosclerosis, and acute aortic syndromes. Future work should define the lineage and origin of PVAT beige cells, dissect additional upstream beiging pathways (e.g., succinate metabolism, BMP4), and evaluate targeted strategies to modulate PVAT beiging and NRG4 signaling in clinical settings.

• Lineage tracing to define the origin of PVAT beige adipocytes was not performed; smooth muscle–derived adipocytes are a possibility but unconfirmed.
• While β3AR signaling and macrophages were implicated, other beiging pathways (e.g., succinate, BMP4) were not explored.
• Studies predominantly used male or ovariectomized female mice; potential sex hormone influences in intact females were not fully assessed.
• Local delivery approaches (pluronic gel) minimize systemic effects but may not fully reflect clinical feasibility; off-target effects cannot be completely excluded.
• Human data derive from a small retrospective autopsy cohort; causality and temporal dynamics in humans remain to be established.
• UCP1 independence was shown for remodeling, but other beige effector pathways beyond NRG4 were not systematically evaluated.