Breeding pairs of two lupus-prone mouse lines, lupus MRL/lpr (#000485) and control MRL/MpJ (#000486), and lupus NZBWF1/J (#100008) and control NZW/LacJ (#001058), were purchased from The Jackson Laboratory. Mice were bred and maintained in specific pathogen-free conditions in the animal facilities. They were housed in a controlled environment at 20-22^oC with a 12 hr light/12 hr dark cycle. Food and water were provided ad libitum to all animals. Only female mice were used in the experiments described here. MRL and MRL/lpr mice were harvested at age of 7 weeks (baseline, before production of autoantibodies and proteinuria) and 14 weeks (active disease with proteinuria). NZW/LacJ and NZBWF1/J mice were harvested between 36 to 40 weeks of age (active disease with proteinuria). All animal care and experimental protocols complied with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and relevant ethical regulations, and were approved by the institutional Animal Care & Use Committee of Augusta University (AU).

Body composition of fat and lean mass was measured by nuclear magnetic resonance (NMR) spectroscopy (Bruker Minispec LF90II, Bruker, Billerica, USA) as previously reported and normalized to total body weight (28). The volume of oxygen consumption (VO₂), carbon dioxide production (VCO₂), spontaneous motor activity and food intake were measured using the Comprehensive Laboratory Monitoring System (CLAMS) (Columbus Instruments, Columbus, USA) (28). The respiratory exchange ratio (RER) was calculated from the ratio of VCO₂ to VO₂. Mice were individually placed into the sealed chambers with free access to food and water. The study was carried out in a room set at 22^oC with 12-12 hr (6:00 am ~ 6:00 pm) light-dark cycles, and the measurements were carried out continuously for 72 hours after a 24 h acclimatization period. The data were averaged over 72 hr. Rectal body temperatures were measured by a BAT-12 thermometer (Physitemp, USA). Blood pressure was recorded in conscious, restrained mice at age of 7 weeks and 14 weeks using the CODA tail cuff blood pressure monitoring system (Kent Scientific, USA).

At age of 14 weeks, the mice were fasted for 5 hours and blood collected from the inferior vena cava. The fasting whole blood glucose was measured using an Alpha TRAK glucometer. The fasting plasma insulin, leptin and resistin levels were assessed by Luminex Assay using MILIIPLEX Metabolic Hormones Expanded Panel (Cat. # MMHE-44K-07, Millipore, Burlington, USA). Plasma adiponectin was assessed by Luminex Bio-Plex Pro Mouse Diabetes Adiponectin Assay (Cat. # 171F7002M, Biorad, Hercules, USA), according to manufacturer’s instructions. For detection of adiponectin, plasma samples were diluted 6400 times. The chemocytokine levels were measured by Flow Based Bead Array using mouse inflammation panel (Cat. # 740446, BioLegend, USA). Levels of the mast cell-specific protease MCPT-6 in the plasma were measured using the Mouse Mast Cell Protease-6/Mcpt6 ELISA Kit according to the manufacturer’s instructions (Thermofisher, USA). The plasma total cholesterol and triglyceride levels were measured with LabAssay^TM Cholesterol and LabAssay^TM Triglyceride Kits (FujiFilm Healthcare, Lexington, MA, USA) as previously described (29). Plasma lipoprotein profiles were determined as described previously (29). Briefly, pooled plasma from 4 MRL/lpr mice and 5 MRL control mice were subjected to FPLC gel filtration on two Superose 6 columns connected in series. For assessment of the production of autoantibodies in lupus mice, serum anti-double stranded DNA (anti-dsDNA) antibodies were quantified by ELISA according to the manufacturer’s instruction (Alpha Diagnostics). Mice were screened for proteinuria weekly with Uristix-4 (Siemens), and elevations confirmed by measuring urinary albumin and creatinine concentrations using a mouse albumin enzyme-linked immunosorbent assay (ELISA; Bethyl Laboratories) and a creatinine Assay Kit (BioAssay Systems) following the manufacturer’s protocols. Urinary albumin-to-creatinine ratios (UACR) were then calculated.

Mice were fasted for 5 hr followed by intraperitoneal injection of glucose at 1g/kg body weight. Glucose levels were measured via tail vein by an Alpha TRAK glucometer at baseline and every 15 min up to 2 hr following glucose injection as previously described (28).

Thoracic aortas were isolated and dissected into rings of 2-3 mm in length, with PVAT either removed or left intact. Isometric tension was recorded using a wire myograph system 620 (Danish Myo Technology A.S, Denmark). The rings with or without PVAT were equilibrated for 60 min and contracted two times with 120 mM KCl. Concentration-response curves were obtained after contracting vessels with phenylephrine (PE; 10^-6 M), followed by acetylcholine (Ach; 10^-7 M) to test endothelial integrity. The rings with or without PVAT were submaximally pre-contracted with PE at EC75 concentration and allowed to reach a stable tension. To examine endothelium-dependent relaxation, Ach (10⁻⁹ M to 10⁻⁴ M) was added cumulatively to the bath and a curve was generated. Finally, endothelium-independent relaxation (vascular smooth muscle response) was assessed by washing out PE and Ach and then repeating the experiment with PE contraction and cumulative addition of sodium nitroprusside (SNP) (10⁻¹⁰ M to 10⁻⁵ M) to the bath. Ach and SNP relaxation were expressed as percentage of PE contraction.

At age of 14 weeks, retroperitoneal (posterior to the kidneys), subcutaneous, gonadal, retroperitoneal, and brown adipose tissues, as well as mesenteric and thoracic perivascular adipose tissues from MRL and MRL/lpr mice, were carefully harvested after perfusion with phosphate buffered saline. The adipose tissues were gently blotted and individually weighed.

Thoracic aortas (with PVAT) and kidneys were fixed in 10% neutral formalin prior to being embedded in paraffin and sectioned into 5 μm thick serial cross sections by Augusta University Histopathology Laboratory. Tissues were stained with hematoxylin and eosin (H&E) and images captured using DP74 microscope camera (Olympus). To quantify adipocyte size and size distribution, uncompressed tif files were analyzed for adipocyte area by the Adiposoft plugin (v.1.16) from National Institutes of Health (NIH) ImageJ software, with each cell being individually identified by lipid inclusion (empty fields) in the tissue. Adipocytes ranging from 25-5000 μm² in area were included. Three slides were prepared from each mouse and ~ 300-450 adipocytes were analyzed. Relative frequency of adipocyte area was calculated. To evaluate vascular remodeling, aortic sections were stained using Masson’s Trichrome (MT) and Verhoeff-Van Gieson (VVG). Three sections from each mouse at 300-μm intervals were analyzed. The lumen, internal elastic lamina (IEL), external elastic lamina (EEL) and adventitia were defined, and the medial area was calculated (tissue between IEL and EEL) using NIH ImageJ. Immunohistochemical (IHC) staining was performed on paraffin-embedded sections of thoracic aortas (with PVAT) following standard procedures by incubating the sections with a primary antibody against uncoupling protein 1 (UCP1, 1:800; Abcam, ab234430, MA) or CD45 (1:1000; Abcam, ab281586, MA) at 4^oC overnight. After washing, the sections were incubated with ImmPRESS-AP Horse Anti-Rabbit IgG Polymer kit (Vector Laboratories, Burlingame, CA) and developed. To examine mast cell distribution in the PVAT, the slides were stained with 0.1% Toluidine Blue solution (PH 2.3) (Sigma; 89640-5G) for 2 minutes, followed by rinsing in distilled water for 3 exchanges and then dehydrated rapidly by applying 95% and 100% alcohol. Images were captured using DP74 (Olympus). Areas of positive staining were quantified using NIH ImageJ.

Total RNA was isolated from PVAT using the RNeasy Lipid Tissue Mini Kit (QIAGEN) according to the manufacturer’s instruction. Purity of total RNA was determined as 260/280 nm absorbance ratio with expected values between 1.8 to 2.2 using Biodrop Duo (Biochrom, Holliston, MA). cDNA was synthesized from 500 ng total RNA using the OneScript® Hot cDNA Synthesis Kit (abm G594, Richmond, Canada). Quantitative reverse transcription PCR (RT-PCR) was performed with BlasTaq™ 2x qPCR Master Mix (abm G892, Richmond, Canada) and the StepOne Plus™ Real-Time PCR System (Applied Biosystems, Waltham, MA) following the manufacturer’s instructions. Relative mRNA levels were determined using acidic ribosomal phosphoprotein P0 (Arbp) as an endogenous control gene for adipogenic/thermogenic and inflammatory markers, and Gapdh an endogenous control gene for immune cell markers. Sequences of the primers used for PCR amplification are listed in Supplementary Material Table S1 .

All data were presented as mean ± standard error of the mean (SEM). To calculate statistical significance, a nonparametric Mann-Whitney test was used after determining the distribution and variance of the data. One-way analysis of variance (ANOVA) followed by Tukey’s multiple comparisons test was used when more than two independent groups were compared. All t-tests were two-tailed, and a value of P<0.05 was considered statistically significant. All statistical analyses were conducted with Prism 9 software (Graphpad, La Jolla, CA).

MRL/lpr mice recapitulate clinical manifestations and immune dysregulation observed in human lupus, including skin lesions, lymphadenopathy, splenomegaly, elevated autoantibodies such as anti-dsDNA antibodies, and renal disease at age of 14 to 20 weeks ( Supplementary Material Figure S1 ). Body weight was measured weekly throughout the experiment, and body fat composition was determined using NMR spectroscopy. While both MRL/lpr and control MRL mice had similar body weights throughout the study ( Figures 1A, B ), MRL/lpr mice with active lupus displayed reduced fat mass ( Figure 1B ), characterized by decreased weights of subcutaneous, gonadal, and retroperitoneal (back of kidney) adipose tissues, as well as mesenteric PVAT ( Figures 1C–E, G, H ). There were no differences, however, in weights of brown adipose tissue or thoracic aortic PVAT between lupus and control mice ( Figures 1F, G ). Interestingly, in contrast to control mice, fat mass and percentage did not change over time in lupus mice ( Figure 1B ).

In order to investigate metabolic phenotype in our lupus model, we performed indirect calorimetry test using comprehensive laboratory animal monitoring system (CLAMS). Oxygen consumption was increased in MRL/lpr mice ( Figure 2A ), while CO₂ production ( Figure 2B ) and RER (calculated as the ratio of CO₂ production to O₂ consumption) ( Figure 2C ) were similar to that observed in control mice. The increased oxygen consumption in MRL/lpr mice was accompanied by increases in heat production/energy expenditure ( Figure 2D ) and locomotor activity ( Figure 2E ), while the rate of food intake ( Figure 2F ) was not significantly different between lupus and control mice. These results suggest that lupus mice are hypermetabolic, without alterations in energy substrate utilization. The increased heat generation occurring throughout the measurement period, even during times of relative inactivity ( Figure 2D ), implies that the increased energy expenditure in MRL/lpr mice is in part metabolically-driven, and not solely due to increased locomotor activity. Interestingly, despite the hypermetabolic state, MRL/lpr mice had lower core temperature compared to control mice in ambient environment (20-22°C) ( Supplementary Material Figure S2 ), which may reflect insufficient insulating capacity due to loss of subcutaneous fat.

Although systolic blood pressure, measured using tail cuff, was increased over time (up to 14 weeks), there was no significant difference in blood pressure between lupus and control mice ( Figure 3A ). Next, we measured fasting glucose and insulin levels and estimated insulin sensitivity using the insulin resistance index (HOMA-IR), calculated using the University of Oxford HOMA calculator software as: [fasting glucose (mmol/L) × fasting insulin (µIU/mL)] ÷ 22.5. No significant differences were detected between lupus and control mice in fasting glucose ( Figure 3B ), fasting insulin ( Figure 3C ), or HOMA-IR ( Figure 3D ). Moreover, we performed glucose tolerance tests (GTTs), which showed similar results in lupus and control mice ( FigureS 3E, F ). In addition, there were no significant differences in levels of plasma total triglycerides ( Figure 3I ) or cholesterol ( Figure 3J ) between lupus and control mice. However, MRL/lpr mice displayed higher levels of very-low-density lipoprotein (VLDL)-triglyceride [MRL/lpr (21.9 mg/dL) vs MRL (4.6 mg/dL)], but not VLDL-cholesterol, compared with control mice ( Figures 3G, H ). Moreover, MRL/lpr mice showed increased levels of intermediate-density lipoprotein (IDL)/low-density lipoprotein (LDL)-cholesterol [(MRL/lpr (5.0 mg/dL) vs MRL (0.9 mg/dL)] and reduced high-density lipoprotein (HDL)-cholesterol [(MRL/lpr (20.6 mg/dL) vs MRL (28.3 mg/dL)] ( Figure 3H ). No significant differences in lipid contents of other plasma lipoprotein species, including IDL/LDL-triglyceride, HDL-triglyceride and VLDL-cholesterol, were noted in MRL/lpr mice ( Figures 3G, H ). These findings suggest that MRL/lpr mice exhibit selective abnormalities in lipoprotein profile while maintaining normal baseline parameters of other systemic cardiovascular and metabolic health, such as blood pressure, glucose and insulin tolerance.

To evaluate vascular function, thoracic aortas were isolated with or without PVAT from MRL/lpr and control mice at age of 14 weeks, and isometric tension was recorded using a wire myograph system. Consistent with previous reports, MRL/lpr mice exhibited impaired endothelium-dependent vasorelaxation to acetylcholine (Ach) in thoracic aortas ( Figures 4B, D-F ). To investigate the role of PVAT in regulating endothelial function, we tested aortas with or without associated PVAT. Interestingly, the endothelial dysfunction in MRL/lpr mice was further augmented by the presence of PVAT ( Figure 4 ) as evidenced by decreased Emax (D), pEC50 (E), and AUC (F), while the presence of PVAT tended to enhance endothelium-dependent relaxation in control mice ( Figures 4A-F ). In contrast, endothelium-independent vasorelaxation to sodium nitroprusside (SNP) was unaffected by presence of PVAT in MRL/lpr or control mice ( Supplementary Material Figure S3 ). We further evaluated endothelial function in a different lupus-prone mouse model (NZBWF1/J mice). As was observed in MRL/lpr mice, endothelial dysfunction was exacerbated by the presence of PVAT in lupus NZB/W mice. Note that the NZB/W mouse study was conducted at 40 weeks of age, at which time PVAT appeared to lose its anti-contractile function in the control mice ( Supplementary Material Figure S4 ). These findings suggest that the detrimental effects of PVAT on endothelial function in MRL/lpr mice were not related to Fas-Fasl signaling associated with the MRL/lpr background. Collectively, these findings suggest that PVAT exacerbates endothelial dysfunction in the setting of active lupus.

PVAT is brown-like in healthy mice, and whitening of PVAT is associated with abnormal vascular homeostasis. To characterize the morphological features of PVAT from lupus mice, histology and immunohistochemical analysis was performed. We analyzed adipocyte size distribution in PVAT. Adipocytes were significantly hypertrophic in the PVAT of MRL/lpr mice compared to control mice (1601 ± 91.03 μm² vs 396.1 ± 27.14 um², respectively, Figures 5A, B ). In addition, while control mice exhibited a beige adipocyte dominant pattern, characterized by the majority of adipocytes being < 250 μm², lupus mice exhibited a flattening and a shift to the right of the adipocyte size distribution curves, indicating an increased proportion of large adipocytes ( Figure 5C ). Furthermore, collagen deposition ( Figures 5D, H , Masson Trichrome staining – blue color) was increased in PVAT from lupus mice, which was most evident at the adventitial border zone. Additionally, adventitial hyperplasia ( Figures 5E, I ) was also observed, whereas no difference was seen in elastin staining or media-lumen ratio ( Figures 5E, J ). Notably, as compared to control mice, expression of UCP1, a beige adipocyte marker, ( Figures 5F, K ), was reduced, while leukocyte recruitment as examined by CD45 immunostaining ( Figures 5G, L ) was increased, in PVAT from lupus mice. While mast cells play an important role in the pathogenesis of SLE, there was no difference in mast cell recruitment to PVAT of lupus mice compared with control mice, as determined by toluidine blue staining ( Supplementary material Figure S5A ) or mast cell-specific gene expression (CD117, Supplementary material Figure S5B ). Furthermore, plasma levels of mast cell specific protease-6 (Mcpt6) were similar in lupus mice and control mice ( Supplementary Material Figure S5D ), further suggesting that mast cells are most likely not responsible for the phenotypic changes in PVAT observed in lupus mice.

Next, we quantified expression of inflammatory, adipogenic, and metabolic genes in PVAT from lupus mice. Interestingly, adiponectin expression ( Figure 6A ) was significantly reduced in PVAT from lupus mice, concordant with decreased plasma levels of adiponectin ( Supplementary Material Figure S4A ). While mRNA expression of pro-inflammatory leptin in PVAT was similar in lupus mice versus control mice ( Figure 6A ), plasma leptin levels were decreased in lupus mice ( Supplementary Material Figure S4A ). Adipogenic (PPARγ) and thermogenic (UCP1) marker gene expression was significantly decreased in PVAT from lupus mice ( Figure 6A ). Importantly, expression of pro-inflammatory cytokines (IL-1β, IL-6, TNFα and IFNγ) ( Figure 6B ) was increased in PVAT from lupus mice, concomitant with elevated plasma levels of TNFα, IFNγ and IL-6 ( Supplementary Material Figure S3B ). In addition, multiple chemokines, including CCL2, CCL5, and CXCL10, were expressed at higher levels in PVAT and plasma from lupus mice ( Figure 6C , Supplementary Figure S3C ). The expression of markers of T cells (CD4, CD8α), B cells (CD19) and macrophages (CD68) was elevated in PVAT from lupus mice ( Figure 6D ). These findings suggest that inflammation originating in PVAT could induce whitening, impair thermogenic and metabolic activities, and promote the local release of factors that perturb endothelial function in lupus mice.