Chronic binge alcohol dysregulates omental adipose tissue extracellular matrix in simian immunodeficiency virus infected macaques

Abstract

Background:

Increased survival, prolonged antiretroviral treatment (ART), and lifestyle choices, including alcohol misuse, increase the risk for comorbid conditions, including cardiometabolic comorbidities among people with HIV (PWH). Published studies indicate that dysregulated adipose tissue phenotype, particularly of the visceral adipose depot, contributes to metabolic dysregulation. Using a non-human primate model of Simian Immunodeficiency Virus (SIV) infection, we previously demonstrated that chronic binge alcohol (CBA) administration to ART-treated rhesus macaques decreases whole-body glucose-insulin dynamics, increases omental adipose tissue (OmAT) collagen content, decreases OmAT adipocyte size and alters pancreatic endocrine function. The objective of the current study was to delineate the depot-specific effects of CBA on visceral (VAT) and subcutaneous adipose tissue (SAT) extracellular matrix (ECM) phenotype, the potential mechanisms involved in AT ECM remodeling, and the implications of increased tissue stiffness on AT metabolic alterations in female SIV-infected macaques.

Methods:

Omental and subcutaneous adipose samples were obtained from female SIV-infected, ART-treated macaques that received intragastric administration of CBA (12-15 g/kg/wk, CBA/SIV) or water (VEH/SIV) for 14.5 months.

Results:

CBA preferentially altered ECM phenotype in OmAT, a VAT depot. The CBA-associated changes included increased ECM accumulation, increased collagen I to III ratio, a pro-fibrotic milieu, and decreased matrix metalloproteinase 13 activity. These changes were associated with smaller adipocyte size, decreased triglyceride content, decreased gene expression of perilipins and a potential dysregulation of peroxisome proliferator-activated receptor gamma signaling.

Conclusions:

Collectively, these findings suggest that CBA-mediated ECM remodeling “traps” adipocytes within a stiff environment that we propose disrupts adipocyte metabolic programming and may increase the risk for metabolic comorbidities.

Graphical Abstract

Introduction

Highly effective antiretroviral therapy (ART) has significantly increased the life expectancy of people with human immunodeficiency virus (HIV) infection (PWH) (Marcus et al., 2020, Adam Trickey and Lei Zhang and Caroline, 2022). Increased survival of PWH on ART is compounded by metabolic comorbidities associated with a higher risk for insulin resistance (IR), glucose intolerance, and type 2 diabetes mellitus (T2DM) (Hernandez-Romieu et al., 2017, Kyaw et al., 2019, Mondy et al., 2007). Though the mechanisms remain poorly understood, metabolic dysregulation in HIV can be exacerbated by lifestyle factors including chronic alcohol misuse. Estimates indicate that PWH in the United States engage in chronic alcohol misuse at higher rates than the uninfected population (Bohm et al., 2021, Duko et al., 2019, Marshall et al., 2017). Chronic alcohol misuse is associated with accelerated disease progression, disease severity, and increased development of metabolic comorbidities in PWH (Baum et al., 2010, Marshall et al., 2017, Molina et al., 2018).

HIV, ART, and alcohol misuse have all been shown to negatively impact adipose tissue (AT) phenotype. Chronic alcohol feeding in rodents decreases overall AT mass and adipocyte size (Sun et al., 2012). In humans, chronic alcohol misuse is associated with decreased total adiposity and increased visceral adipose tissue (VAT) mass (Addolorato et al., 2000). Additionally, in men who engage in alcohol misuse, an elevated waist-to-hip ratio and increased VAT were associated with worse metabolic clinical outcomes (Kim et al., 2012). Abdominal subcutaneous adipose tissue from PWH on ART who developed lipodystrophy had a higher proportion of smaller adipocytes and lower expression of adipogenic/lipogenic mediators such as sterol regulatory element binding protein-1c (SREBP1c) and peroxisome proliferator-activated receptor gamma (PPARG) (Bastard et al., 2002). These alterations in AT mass, distribution, and morphology seen with alcohol misuse and in PWH could be due to impaired AT extracellular matrix (ECM) phenotype and metabolic homeostasis (i.e., glucose and lipid metabolism).

Clinical studies show increased fibrosis in omental adipose tissue (OmAT), a VAT depot, of obese subjects compared to other adipose depots and compared to lean subjects (Divoux et al., 2010, Michaud et al., 2016). In humans and cynomolgus macaques, HIV and simian immunodeficiency virus (SIV) increase AT collagen deposition and impair adipose-derived stem cell (ADSC) differentiation (Gorwood et al., 2019). Despite evidence suggesting that AT ECM is a functionally relevant constituent of AT physiology, it has received limited attention. Published data from the Louisiana State University Health Sciences Center (LSUHSC) Comprehensive HIV-AIDS/Alcohol Research Center (CARC) demonstrated that chronic binge alcohol (CBA) increased OmAT collagen content in male SIV-infected, ART-treated rhesus macaques (Ford et al., 2018).

Alcohol misuse produces marked alterations in adipose lipid metabolism. Chronic alcohol increases lipolysis in white adipose tissue and primary adipocytes in rodents (Crowell et al., 2016, Kang et al., 2007, Zhang et al., 2015, Zhong et al., 2012). Alcohol-induced WAT lipolysis contributes to alcohol-related hepatic steatosis via reverse triglyceride transport, particularly from VAT, which drains directly to the liver via the portal circulation (Wei et al., 2013, Zhong et al., 2012). Preclinical studies have demonstrated that alcohol-related hepatic steatosis is associated with a reduction in AT mass (Kang et al., 2007). Several published studies suggest that lipid metabolism in VAT is more susceptible to alcohol toxicity compared to subcutaneous adipose tissue (SAT)(Zhang et al., 2015, Zhong et al., 2012). While chronic alcohol effects on AT metabolism have been extensively investigated, the combined effects of HIV, ART, and chronic alcohol on AT ECM remodeling and metabolic changes have yet to be elucidated. Using a relevant preclinical model of HIV, the goal of this study was to characterize the effects of CBA on adipose depot-specific (VAT and SAT) ECM phenotype, identify potential mechanisms responsible for AT ECM dysregulation, and the implications of increased tissue stiffness on AT metabolic phenotype. We hypothesized that chronic alcohol dysregulates adipose tissue ECM phenotype, promoting a pro-fibrotic milieu, and a fibrotic ECM phenotype impairs adipose metabolic phenotype in SIV-infected, ART-treated female macaques.

Methods

Animal Experiments

All experiments described in this study were approved by the Institutional Animal Care and Use Committee at Louisiana State University Health Sciences Center (LSUHSC, New Orleans, LA) and adhered to the National Institutes of Health guidelines for the care and use of experimental animals. This cohort of female macaques was also used to study whole body insulin responsiveness, pancreatic endocrine function, OmAT proteome composition, liver pathophysiology, and immune cell mitochondrial homeostasis (Gallegos et al., 2024, McTernan et al., 2022, Poret et al., 2021, Simon et al., 2021).

Adult (6–9 yrs. old) female Indian rhesus macaques (Macaca mulatta) were randomized to either (1) isovolumetric water-administered (VEH) or (2) chronic binge alcohol-administered (CBA) groups. All macaques were surgically implanted with gastric catheters and administered alcohol at a concentration of 30% (w/v) in water (30 min infusion; 13-14 g/kg/week of alcohol; 5d/week). Macaques achieved a peak blood alcohol level of 50-60 mM (~230 mg%), 2 h after alcohol initiation. After 3 months of CBA or VEH administration, all macaques were infected intravaginally with SIV_mac251. At 2.5 months post-SIV infection, coinciding with viral set-point, all macaques were initiated on daily subcutaneous injections of 20 mg/kg of Tenofovir (TFV, 9-[^®-2-(phosphonomethoxy) propyl] adenine, PMPA) and 30 mg/kg of Emtricitabine (FTC), provided by Gilead Sciences Inc. (Foster City, CA). This dose and drug combination effectively suppresses viral load and results in minimal toxicity in normal healthy macaques from infancy to adulthood and does not result in liver or renal toxicity in SIV-infected macaques (Molina et al., 2014). Nine months post ART-initiation, 24 h after the last alcohol administration, and after an overnight fast, all macaques were euthanized according to the American Veterinary Medical Association’s guidelines. Each macaque was administered ketamine xylazine (10 mg/kg) followed by subcutaneous administration of buprenex (0.2-0.4 mg/kg). Propofol (1-2mg/kg) was administered intravenously at an infusion rate of 24-37 mg/kg/h throughout the perfusion procedure. Perfusion was achieved for 2-5 min using ice cold Ringer's solution administered using a perfusion/embalming machine. After the perfusion, death was confirmed with intravenous administration of pentobarbital (1.8-10 mg/kg). OmAT and SAT samples were excised at necropsy and immediately flash frozen or formalin-fixed for further analysis. The total time for CBA or VEH administration was 14.5 months, SIV infection for 11.5 months and 9 months on ART. Euthanasia prior to study end point was performed based on any one of the following criteria: Loss of 25% of body weight, complete anorexia for 4 days, major organ failure or medical conditions unresponsive to treatment and surgical complications unresponsive to immediate intervention. Two macaques were euthanized before study endpoint and excluded from histological and molecular analysis.

Adipose tissue histological analysis

Formalin-fixed OmAT and SAT samples were paraffin embedding and sectioned at a thickness of 5 μM using a microtome. OmAT and SAT sections were stained for collagen expression, adipocyte diameter, and adipocyte number using picrosirius red (PSR) staining. Slides were fixed in ice cold 100% acetone for 5 min and allowed to air dry. Slides were washed in 1X phosphate-buffered saline (PBS, pH 7.4) three times for 10 min each and stained in Bouin’s solution for 30 min (Sigma-Aldrich). Slides were then rinsed in water and stained with PSR (Sigma Direct Red 80, Sigma-Aldrich) for 1 hr. Slides were rinsed twice in acidic water (5 ml glacial acetic acid in 1 liter of water), once in picric alcohol (10% picric acid, 20% ethanol), once in water, and twice in 95% ethanol for 1 min each. Slides were cleared in xylene and mounted with Permount medium (Sigma-Aldrich). OmAT and SAT sections were imaged in a blinded manner using an Olympus BX51 TRF Microscope. Ten representative images were taken for each sample at 20X magnification. PSR expression, adipocyte diameter and number were quantified using Image J. Percent PSR from 10 representative images was averaged. Adipocyte diameter was determined by averaging measurements taken from 100 adipocytes across 10 different PSR images per macaque (10 adipocytes per image). Adipocyte number was determined by averaging the number of cells per area (180349.131 um) for 10 different images per macaque.

Quantitative PCR for ECM and metabolic gene expression

A Custom RT² Profiler PCR Array (Qiagen) was used to measure mRNA expression of 41 ECM-related genes. RNA was extracted from OmAT and SAT using RNeasy Mini Kits (Qiagen). cDNA was synthesized using RT² First Strand Kit (Qiagen). RT² SYBR Green qPCR Master Mix was used for the reaction following manufacturer’s instructions. Amplification was performed with Bio-Rad CFX96 (Bio-Rad). Using Qiagen’s Gene Globe RT² Profiler PCR Analyzer, EMILIN1 and MMP11 were selected as the housekeeping genes based on consistent expression across experimental conditions. Values are expressed as fold-change relative to VEH according to the ΔΔCt method. All ECM genes are listed in Table 1.

Table 1:

RT2 Profiler PCR extracellular matrix (ECM) Array gene list and primer ID.

Gene Symbol	Qiagen Primer ID
ECM Fibers
COL1A1	PPQ00101A
COL1A2	PPQ05342A
COL3A1	PPQ00035A
COL4A1	PPQ05288A
COL6A3	PPQ01467A
COL12A1	PPQ16764A
FBLN1	PPQ14030A
DCN	PPQ11670A
FN1	PPQ00312A
HSPG2	PPQ09888A
SDC1	PPQ06152A
SDC4	PPQ12295A
LUM	PPQ11600A
THBS1	PPQ08721A
LAMB2	PPQ10343A
OPN	PPQ08393A
TNC	PPQ06922A
POSTN	PPQ02047A
SERPINE1	PPQ14218B
SPARC	PPQ13237A
FGB	PPQ04107A
LGALS3	PPQ03296A
ECM Enzymes
SERPINA12	PPQ06267A
SERPINA13	PPQ00011A
LOX	PPQ05109A
ADAMTS1	PPQ10267A
ADAMTSL4	PPQ15456A
Integrins
ITGAV	PPQ10251A
ITGA5	PPQ00211A
ITGB1	PPQ69220A
ITGB2	PPQ68669B
MMP/TIMP
MMP 13	PPQ07771A
MMP 14	PPQ13263A
TIMP1	PPQ00244A
TIMP2	PPQ00253A
TIMP3	PPQ00276A
Pro-fibrotic Mediators
PDGFRA	PPQ69217A
TGFB	PPQ00038C
FGF2	PPQ00039A
SAA1	PPQ01543A
CTGF	PPQ14642A

Relative gene expression of metabolic genes was determined using RT2 SYBR^® Green qPCR Master mix and specific primers. These include enzymes involved in lipogenesis, lipoprotein lipase (LPL), acetyl-CoA carboxylase β (ACCB), fatty acid synthase (FASN), and diacylglycerol acyltransferase (DGAT); regulators of lipogenesis, peroxisome proliferator-activated receptor γ (PPARG), carbohydrate response element binding protein (ChREBP) and sterol regulatory-element binding protein-1c (SREBP1c); enzymes involved in lipolysis, adipose triglyceride lipase (ATGL), hormone sensitive lipase (HSL), and monoacylglycerol lipase (MAGL); and perilipins 1 – 4 (PLIN 1 – 4). Primer sequences are included in Table 2.

Table 2:

PCR primer sequences for genes involved in metabolism.

Gene	Forward Primer	Reverse Primer
RPS13	GACGTGAAGGAGCAGATTT	ACGTACTTGTGCAACACC
LPL	GCAGGAAGTCTGACCAATAAG	CCTCTCGTGAATGTGTGTAAG
ACCB	CACGGATATCATCGGGAAAG	GTGACGATCTCTTCGTAAGC
FASN	CAACCTGATAGTGAGTGGGAAG	CAGACGCAGCTCCTTGTAAA
DGAT	GACCTACCGCGATCTCTATTA	GGTGAAGAACAGCATCTCAA
PPARG	TGAAGGATGCAAGGGTTTCTTCCG	CGCCCAAACCTGATGGCATTATGA
ChREBP	CAGCCGAGTACATCCTTATG	CTGCTGGCACAGGTTAAT
SREBP1	CAAGGCCATCGACTACATTC	GACACCAGATCCTTCAGAGA
ATGL	CTTCAACTCCAAGGACGAG	AATGCCACCATCCACATAG
HSL	GGCACAGATTCCCTCAAGAA	TGGAGGAGCTAAGTGTCTCA
MAGL	CACTTGGTCCCAGCAATAA	GGGTCACGTTTGCACTAA
PLIN 2	CAGCTCCATTCTACTGTTCAC	CCCATGAGAGGTAGAGCTTAT
PLIN 3	GTCTGCTTGGGAAGCTAAG	CTCTACTGACTGATAGGGAGAG
PLIN 4	GGAGCTACTTCGTTCGTTTAG	GTATCTGCTTGGAACTG

Matrix metalloprotease (MMP) Activity Assay

MMP activity was determined using a fluorescence resonance energy transfer (FRET)-based MMP activity assay kit according to the manufacturer's instructions (ab112147; Abcam). OmAT and SAT (2.5 mg) were used to determine total MMP 1 and MMP 13 activity. In brief, 25 μL of OmAT or SAT tissue lysates were added to 25 μL of 2 mM p-aminophenylmercuric acetate solution and incubated at 37°C for 3 hours for MMP 1 activity and 40 min for MMP 13 activity. MMP red substrate (50 μL) was added and incubated at room temperature for 1 hour. Fluorescence was measured at Ex/Em = 540/590 nM (Tecan infinite 200). Results for MMP Activity are expressed as relative fluorescence units (RFU) per mg of tissue.

Western blot analysis for collagen and insulin signaling protein expression.

OmAT samples were homogenized in Tissue Protein Extraction Reagent (Thermo Fisher Scientific) buffer with protease inhibitor cocktail and phosphatase inhibitor cocktails II and III (Sigma-Aldrich). Protein concentration was determined by a bicinchoninic acid assay (Pierce BCA Assay, Thermo Fisher Scientific), according to manufacturer’s instructions. Protein (20 μg) was separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (Bio-Rad Mini-PROTEAN TGX Stain-Free Precast Gels and Criterion TGX Precast Gels, 4-15%) and transferred to polyvinylidene fluoride membranes (EMD Millipore, Billerica, MA) for analysis. Antibodies for Collagen I (600-401-103, Rockland Immunochemicals, 1:1000) and collagen III (600-401-105, Rockland Immunochemicals, 1:1000) were used to determine protein expression. PVDF membranes were incubated in chemiluminescent substrate (Luminata Crescendo Western HRP substrate, EMD Millipore; Billerica, MA) and imaged using Amersham Imager 680. Image J (National Institutes of Health, Bethesda, MD) was used to quantify protein band density. Collagen protein expression was normalized to the loading control, beta-tubulin (2148S, Cell Signaling Technology, 1:1000).

Triglyceride and Free Fatty Acid Analysis

OmAT triglyceride concentration was determined using a quantitative colorimetric assay kit according to the manufacturer’s instructions (ab65336; Abcam). Briefly, 100 mg of OmAT was used for triglyceride extraction. Triglyceride extract was diluted 10-fold, and 50 μL of diluted triglyceride extract was added to 2 μL of lipase and 50 μL of triglyceride reaction mix. Reaction was incubated at room temperature for 60 min protected from light. Triglyceride concentration was determined using a microplate reader (Bio-Rad, Benchmark Plus) with triglyceride concentration being proportional to the absorbance values at 570 nm. Data are expressed as μg of triglyceride per mg of tissue.

OmAT free fatty acid concentration was determined using a quantitative colorimetric assay kit according to the manufacturer’s instructions (ab65341; Abcam). Briefly, 10 mg of OmAT was used for free fatty acid extraction. Free fatty acid extract (25 μL) was added to 2 μL of Acyl-CoA Synthetase (ACS) Reagent and incubated at 37°C for 30 min. Fatty acid reaction mix (50 μL) was added and incubated at 37°C for 30 min protected from light. Free fatty acid concentration was determined using a microplate reader with fatty acid concentration being proportional to the absorbance values at 570 nm. Data are expressed as μg of free fatty acid per mg of tissue.

Statistical Analyses

All data are presented as means ± SEM, where at study endpoint n = 6–8 for each treatment group [VEH/SIV (n = 6-8) and CBA/SIV (n=6)]. Normality tests were used to assess Gaussian distribution of data. All normally distributed data were analyzed by unpaired samples t test between VEH and CBA groups. Data that were not normally distributed were analyzed using Mann-Whitney statistical test. Spearman correlation was used for PSR vs adipocyte diameter relationship and ECM correlation matrix. All analyses were conducted using Prism Graph Pad Version 9 (La Jolla, CA). Statistical significance was established at P < 0.05.

Results

Collagen content, adipocyte diameter and number in OmAT

Collagen content, adipocyte diameter, and adipocyte cell count were determined using PSR staining. CBA significantly increased PSR staining density in OmAT as shown in representative images and the graphs ([p=0.0129, df (5,5)]; Figure 1A-B). CBA significantly decreased omental adipocyte diameter ([p=0.0484, df (5,5)]; Figure 1A and 1D). CBA did not alter omental adipocyte number ([p=0.4848, Mann-Whitney U=13]; Figure 1F). Spearman correlation revealed a significant negative association between percent PSR and adipocyte diameter ([p<0.0001, spearman r = −0.9301]; Figure 1H).

Figure 1.

Omental adipose tissue (OmAT) and Subcutaneous adipose tissue (SAT) histology. OmAT and SAT isolated from vehicle (VEH)- and chronic binge alcohol (CBA)-administered macaques were stained with picrosirius red (PSR). A. Representative brightfield OmAT PSR-stained images for VEH/SIV (top) and CBA/SIV (bottom) for OmAT and SAT, respectively. B. CBA administration significantly increased percent PSR collagen staining in OmAT of CBA/SIV compared to VEH/SIV macaques. C. CBA administration significantly increased percent PSR collagen staining in SAT of CBA/SIV compared to VEH/SIV macaques. Each data point represents the average percent PSR measurement from 10 different images per macaque. D. OmAT adipocyte cell size was decreased in CBA/SIV compared to VEH/SIV macaques. E. No significant difference was detected in SAT adipocyte cell size between CBA/SIV and VEH/SIV macaques. Each data point represents the average adipocyte diameter of 100 adipocytes measured from 10 different PSR images per macaque, 10 adipocytes per image. F. No significant difference in OmAT adipocyte number between CBA/SIV and VEH/SIV macaques. G. No significant difference in SAT adipocyte number between CBA/SIV and VEH/SIV macaques. Each data point represents average number of cells per area for 10 different images per macaque. H. Significant negative correlation of OmAT percent PSR and adipocyte diameter. I. Subcutaneous percent PSR staining does not correlate with adipocyte diameter. *p<0.05, unpaired t-test. Data are expressed as mean ± SEM. N=6/group.

Collagen content, adipocyte diameter and number in SAT

CBA significantly increased PSR expression in SAT, as shown in representative images and the graphs ([p=0.0411, Mann-Whitney U=5]; Figure 1A-C). CBA did not alter adipocyte diameter ([p=0.9372, Mann-Whitney U=17]; Figure 1A and 1E) and adipocyte number ([p=0.2813, df (5,5)]; Figure 1G). There were no significant associations between subcutaneous percent PSR and adipocyte diameter ([p=0.8004, spearman r = −0.0839]; Figure 1I).

ECM-related gene expression in OmAT

Expression of 41 ECM-related genes were measured using a PCR array. These genes include various ECM fibers, ECM enzymes, integrins, MMPs and TIMPs, and pro-fibrotic mediators. CBA significantly increased OmAT mRNA expression of 3 of the 21 measured ECM fibers relative to VEH/SIV macaques (Figure 2A). Thrombospondin 1 (THBS1) [p=0.0427, df (7,5)], laminin β2 (LAMB2) [p=0.0473, Mann-Whitney U=8.5], and secreted protein acidic and rich in cysteine (SPARC) ([p=0.0143, df (7,5)] were increased in OmAT of CBA/SIV compared to VEH/SIV macaques (Figure 2A). Fibrinogen β chain (FGB) was decreased in OmAT of CBA/SIV macaques compared to VEH/SIV macaques ([p=0.0282, df (7,5)]; Figure 2A).

Figure 2.

Omental adipose tissue (OmAT) extracellular matrix (ECM) gene expression. A: Chronic binge alcohol (CBA) increased thrombospondin 1 (THBS1), laminin beta 2 (LAMB2) and secreted protein acidic and rich in cysteine (SPARC) mRNA expression in OmAT of CBA/SIV compared to VEH/SIV macaques. CBA decreased fibrinogen beta chain (FGB) mRNA expression in OmAT. B: CBA increased a disintegrin and metalloproteinase with thrombospondin motifs like 4 (ADAMTSL4) mRNA expression in OmAT of CBA/SIV compared to VEH/SIV macaques. Serine protease inhibitor A3 (SERPINA3) mRNA expression in OmAT of CBA/SIV macaques was decreased relative to VEH/SIV macaques. C: No significant difference was detected in matrix metalloproteinases (MMP) and tissue inhibitor of metalloproteinases (TIMP) gene expression between CBA/SIV and VEH/SIV macaques, D: Integrin beta 1 (ITGB1) mRNA expression was increased in CBA/SIV compared to VEH/SIV macaques. E: Platelet derived growth factor receptor alpha (PDGFRA), transforming growth factor beta 1 (TGFB1), and fibroblast growth factor 2 (FGF2) mRNA expression were increased in CBA/SIV compared to VEH/SIV macaques. *, p<0.05, unpaired t-test; **, p<0.01, unpaired t-test; #, p<0.05, Mann-Whitney. Data are expressed as mean ± SEM. N=6, CBA/SIV. Solid black line represents average mRNA expression for VEH/SIV (n=8) group.

CBA significantly affected OmAT mRNA expression of 2 of the 6 measured ECM enzymes and inhibitors relative to VEH/SIV macaques (Figure 2B). CBA increased mRNA expression of a disintegrin and metalloproteinase with thrombospondin motifs like 4 (ADAMTSL4) in the OmAT compared to VEH/SIV macaques ([p=0.0148, df (7,5)]; Figure 2B). Serine protease inhibitor A3 (SERPINA3) was decreased in OmAT of CBA/SIV macaques relative to VEH/SIV macaques ([p=0.0543, Mann-Whitney U=9]; Figure 2B).

CBA did not significantly alter expression of MMP13, MMP14, TIMP1, TIMP2 or TIMP3 in OmAT (Figure 2C). Integrin β1 (ITGB1) expression was increased in OmAT of CBA/SIV macaques compared to VEH/SIV macaques ([p=0.0458, df (7,5)]; Figure 2D).

CBA significantly increased OmAT expression of 3 of the 5 measured pro-fibrotic mediators relative to VEH treated macaques (Figure 2E). Platelet-derived growth factor receptor α (PDGFRA) [p=0.0084, df (7,5)], transforming growth factor β1 (TGFB1) [p=0.0450, df (7,5)], and fibroblast growth factor 2 (FGF2) [p=0.0182, df (7,5)] were increased in OmAT of CBA/SIV macaques compared to VEH/SIV macaques (Figure 2E).

ECM-related gene expression in SAT

There were no significant differences in mRNA expression of ECM fibers (Figure S1A), ECM enzymes and inhibitors (Figure S1B), MMPs and TIMPs (Figure S1C), Integrins (Figure S1D), and pro-fibrotic mediators (Figure S1E) in SAT of CBA/SIV macaques compared to VEH/SIV macaques.

MMP13 activity in OmAT and SAT

CBA significantly decreased MMP13 activity in OmAT of CBA/SIV macaques compared to VEH/SIV macaques ([p=0.0488, df (5,5)]; Figure 3A). There was no significant difference in OmAT MMP1 activity between CBA/SIV and VEH/SIV macaques (data not shown). There was no significant difference in SAT MMP13 activity between CBA/SIV and VEH/SIV macaques (Figure 3B). There were no differences in SAT ECM gene expression or MMP activity, therefore, additional analysis was limited to OmAT.

Figure 3.

Enzymatic activity of matrix metalloproteinase 13 (MMP 13) in omental (OmAT) and subcutaneous (SAT) adipose tissue. A: MMP13 activity was significantly decreased in OmAT of chronic binge alcohol (CBA)/SIV compared to VEH/SIV macaques. B: No significant difference was detected in SAT MMP13 activity between CBA/SIV and VEH/SIV. *, p<0.05, unpaired t-test. Data are expressed as mean ± SEM. N=6/group.

Collagen I and collagen III protein expression in OmAT

There was no significant difference in OmAT collagen I protein expression between CBA/SIV and VEH/SIV macaques (Figure 4A-B). In contrast, collagen III protein expression was significantly increased in OmAT of CBA/SIV compared to VEH/SIV macaques ([p=0.0411, Mann-Whitney U=5]; Figure 4A and 4C). As a result, the ratio between collagen I and collagen III was significantly higher in OmAT of CBA/SIV compared to VEH/SIV macaques ([p=0.0315, df (5,5)]; Figure 4D).

Figure 4.

Collagen I and III protein expression in OmAT. A: Representative Western blot images for collagen I, collagen III, and beta-tubulin. B No significant difference was detected in collagen I protein expression between CBA/SIV and VEH/SIV. C: Collagen III protein expression was increased in CBA/SIV compared to VEH/SIV macaques. D: Collagen I: Collagen III protein expression ratio was significantly increased in CBA/SIV compared to VEH/SIV macaques. *, p<0.05, unpaired t-test; #, p<0.05, Mann-Whitney. Data are expressed as mean ± SEM. N=6/group.

OmAT ECM correlation matrix

Correlations among the 41 ECM-related genes measured with PSR, adipocyte cell size, and MMP13 activity in OmAT were tested. Spearman correlations revealed significant positive associations between percent PSR and OmAT mRNA expression of LAMB2 [p=0.050, spearman r = 0.644], SPARC [p=0.023, r = 0.721], PDGFRA [p=0.020, r = 0.733], and serum amyloid amylase 1 (SAA1) [p=0.020, r = 0.733] (Figure 5). Percent PSR was negatively correlated with SERPINA3 mRNA expression [p=0.029, r = −0.699] in OmAT (Figure 5). OmAT cell size was positively correlated with SERPINA3 mRNA expression [p=0.047, r = 0.650] (Figure 5). OmAT cell size was negatively correlated with OmAT mRNA expression of SPARC [p=0.017, r = −0.745], PDGFRA [p=0.044, r = −0.661], and SAA1 [p=0.027, r = −0.709] (Figure 5). OmAT MMP 13 activity was negatively correlated with mRNA expression of collagen 1 α1 (COL1A1) [p=0.021, r = −0.729], collagen 1 α2 (COL1A2) [p=0.012, Spearman r = −0.772], collagen 3 α1 (COL3A1) [p=0.014, r = −0.760], fibronectin (FN1) [p=0.015, r = −0.754], ITGAV [p=0.018, Spearman r = −0.742], tissue inhibitor of MMP (TIMP2) [p=0.046, r = −0.652], and PDGFRA [p=0.033, r = −0.687] (Figure 5).

Figure 5.

Omental adipose tissue (OmAT) ECM correlation matrix. Correlations with a significant p-value (<0.05) are marked with an asterisk (*) and their correlation coefficient is listed. Blue indicates a positive correlation, while red indicates a negative correlation.

Triglyceride and free fatty acid content in OmAT

Free fatty acid and triglyceride content in OmAT were measured using commercial ELISA kits. OmAT triglyceride content was decreased in CBA/SIV macaques compared to VEH/SIV macaques ([p=0.0277, df (7,5)]; Figure 6A). CBA/SIV did not affect OmAT free fatty acid content ([p=0.4136, Mann-Whitney U=17]; Figure 6B).

Figure 6.

Omental adipose tissue (OmAT) triglyceride and free fatty acid content. A: Triglyceride concentration was significantly decreased in chronic binge alcohol (CBA)/SIV compared to VEH/SIV macaques. B: No significant difference was detected in OmAT free fatty acid concentration between CBA/SIV and VEH/SIV macaques. *, p<0.05, unpaired t-test. Data are expressed as mean ± SEM. N=6/group.

Gene expression of lipogenic mediators in OmAT

Gene expression of key lipogenic mediators was measured using RT-qPCR. PPARG expression was increased in OmAT of CBA/SIV compared to VEH/SIV macaques ([p<0.001, df (7,5)] Figure 7A). CBA decreased SREBP1c in OmAT of CBA/SIV macaques compared to VEH/SIV macaques ([p=0.0148, df (7,5)]; Figure 7B). No significant difference was detected in ChREBP mRNA expression in OmAT from CBA/SIV macaques and VEH/SIV macaques (Figure 7C).

Figure 7.

Omental adipose tissue (OmAT) gene expression of lipogenic mediators and perilipins. A: Peroxisome proliferator-activated receptor gamma (PPARG) gene expression was significantly increased in chronic binge alcohol (CBA)/SIV compared to VEH/SIV macaques. B: Sterol regulatory element binding protein 1c (SREBP1c) gene expression was significantly decreased in CBA/SIV compared to VEH/SIV macaques. C: No significant difference was detected in carbohydrate response element binding protein (ChREBP) gene expression between CBA/SIV and VEH/SIV macaques D-F: Perilipin 2 – 4 gene expression was increased in CBA/SIV compared to VEH/SIV macaques. *, p<0.05, unpaired t-test; ****, p<0.0001, unpaired t-test. #, p<0.05, Mann-Whitney; ##, p<0.01, Mann-Whitney. Data are expressed as mean ± SEM. N=6-8/group.

Gene expression of perilipins in OmAT

CBA decreased gene expression of PLIN2 [p=0.0180, Mann-Whitney U=6], PLIN3 [p=0.003, Mann-Whitney U=2.5], and PLIN4 [p=0.0284, df (7,5)] in OmAT of CBA/SIV compared to VEH/SIV macaques (Figure 7D-F). PLIN1 expression was not detected in the samples and was excluded from analysis.

Gene expression of lipogenic or lipolytic enzymes in OmAT

CBA/SIV did not alter gene expression of lipogenic or lipolytic enzymes in OmAT (Figures S2A-D and S3A-C).

Discussion

We investigated the impact of chronic binge alcohol (CBA) on OmAT and SAT ECM phenotype and the consequences of alcohol-mediated changes in ECM on adipose tissue metabolic capacity in SIV-infected, ART-treated female macaques. Our results indicate that CBA/SIV increased collagen content in both the omental and subcutaneous adipose but produced some site-specific alterations. In OmAT, CBA administration reduced adipocyte size, increased collagen expression, pro-fibrotic gene expression, collagen I to collagen III ratio, and decreased MMP 13 degradative activity, suggesting a more profound impact of CBA on OmAT ECM-phenotype. Our results also demonstrate that CBA decreased OmAT triglyceride content, decreased PLIN gene expression, and disrupted gene expression of lipogenic regulators without altering lipogenic or lipolytic enzyme gene expression in SIV infection. Taken together, these findings provide strong support for a CBA-mediated increase in OmAT stiffness and dysregulation of markers of OmAT lipid homeostasis and lipid storage in SIV-infection. Based on these data, we speculate that a reduction in OmAT MMP 13 activity may be a plausible mechanism contributing to CBA-mediated ECM accumulation in SIV-infection.

The current findings reflecting collagen accumulation and reduced adipocyte diameter in the OmAT of female SIV-infected macaques are in accordance with previously published data from male SIV-infected rhesus macaques (Ford et al., 2018). We provide evidence for increased adipose tissue stiffness as reflected by decreased collagen III protein expression and increased collagen I to collagen III ratio. Collagen I produces physically thicker fibers and confers more tensile strength and tissue rigidity (Shoulders and Raines, 2009). Collagen III is thinner and contributes to tissue elasticity or compliance, which is the reciprocal of stiffness. An increase in the ratio of collagen I to collagen III is used as a marker of tissue rigidity, as has been shown in cardiac and muscle fibrosis (Collier et al., 2012, Hindle et al., 2009), and in chronic alcohol-mediated increase in liver stiffness (Mueller et al., 2010).

Our results provide the first insight into chronic alcohol effects on ECM composition in SIV-infection, and the results support our histological findings in the OmAT and SAT. Specifically, the ECM array indicates a CBA-mediated increase in mRNA expression of THBS1, LAMB2, and SPARC. THBS1, a major regulator of fibrotic signaling, is a large adhesive glycoprotein expressed mainly in VAT and is associated with obesity, adipose inflammation, and insulin resistance (Varma et al., 2008). Human hepatic stellate cells (HSC) isolated from people with alcohol-related liver disease, and peripheral blood mononuclear cells (PBMCs) isolated from rhesus macaques with a heavy drinking phenotype have increased THBS1 expression (Liu et al., 2020). (Sureshchandra et al., 2019). SPARC, also known as Osteonectin, is a collagen-binding matricellular protein (Jundt et al., 1987) that is secreted by adipocytes into the ECM and is involved in adipogenesis (Chavey et al., 2006, Nie and Sage, 2009). SPARC contributes to the pathogenesis of fibrosis in the kidney and the liver (Camino et al., 2008) and HSCs expressing SPARC are identified as myofibroblasts (Mazzolini et al., 2018). HSCs from people with alcohol-related liver disease and mouse HSCs from a model of alcohol-related hepatitis and fibrosis express increased SPARC (Liu et al., 2020). Collectively, these data suggest SPARC is a conserved pathway of alcohol-mediated fibrosis across species and may potentially be a marker of a CBA-mediated OmAT adipocyte phenotypic shift to an osteogenic or myofibroblast-like profile in SIV/HIV infection.

The ECM enzyme ADAMTSL4 was significantly increased and SERPINA3 was decreased by CBA in OmAT. There is limited literature on SERPINA3 in the context of alcohol and in adipose tissue and metabolic disease. SERPINA3 gene expression was increased in post-mortem hippocampal tissue of people with alcohol use (McClintick et al., 2013) and in people with alcohol dependence (Zhang et al., 2021a). Recent evidence in the context of high-fat diet-induced obesity in mice demonstrated that SERPINA3 deficiency led to more severe obesity with glucose intolerance and insulin resistance (Li et al., 2022). Thus, it appears that SERPINA3 may be modulating divergent pathways with alcohol and obesity.

ITGB1 expression was significantly increased in OmAT of CBA/SIV female macaques. Evidence from adipose-specific ITGB1 knockout mice suggests that ITGB1 interacts with the insulin receptor to regulate insulin action and systemic metabolism (Ruiz-Ojeda et al., 2021). In the context of alcohol, ITGB1 expression is significantly increased in circulating T-lymphocytes from people who chronically consume alcohol (Sacanella et al., 1999), and increased hepatocyte ITGB1 expression has been reported in people with alcohol-related liver disease (Chedid et al., 1993). Published studies suggest that under normal, physiological conditions ITGB1 may mediate positive effects on AT and whole-body metabolic health (Ruiz-Ojeda et al., 2021, Zong et al., 2009). However, our current results and the literature on ITGB1 in hepatic fibrosis and alcohol-related liver disease suggest that pro-fibrotic effects may prevail in conditions of increased ITGB1 expression (Nejjari et al., 2001, Patsenker and Stickel, 2011, Chedid et al., 1993).

In addition to increased ECM fiber expression, our results indicate a CBA-mediated increase in OmAT profibrotic mediators, specifically PDGFRA, TGFB, and FGF2 in SIV-infection. While TGFB is a known mediator for tissue pro-fibrotic milieu, the effect on adipocytes and their progenitor cells remains controversial. As previously mentioned, THBS1, which was upregulated with CBA/SIV, is a potent activator of TGFB. Inhibition of PDGFR tyrosine kinase activity reduces fibrosis in animal models of skin, kidney and atrial lung fibrosis (Abdollahi et al., 2005, Akhmetshina et al., 2009, Chen et al., 2011), and PDGFRA heterozygous mice are resistant to hepatic fibrosis (Hayes et al., 2014). In the context of these published findings, our results are consistent with the hypothesis that tissue fibrosis requires upregulation and activation of PDGFRA. Moreover, WAT PDGFRA signaling opposes adipogenesis and promotes the transition of adipose progenitor cells to ECM-synthesizing pro-fibrotic cells (Iwayama et al., 2015). Collectively, the results of our ECM array indicate that CBA promotes a profibrotic milieu, specifically in OmAT, of female SIV-infected, ART-treated rhesus macaques.

Though mRNA expression of MMPs and TIMPs was not different between CBA/SIV and VEH/SIV macaques, MMPs are known to undergo post-translational processing to the active enzyme, and MMP 1 and MMP 13 activity were measured in OmAT. CBA significantly decreased MMP 13 activity. MMP 13 has a broad substrate specificity and can degrade ECM fibers including collagens I, II and III, and fibronectin, and ECM components, such as SPARC and LAMB (Hsu and Lai, 2007, Kehlet et al., 2018, Knäuper et al., 1996, Yamamoto et al., 2016). In this study, MMP 13 activity was negatively correlated with several ECM-related genes, most notably, collagens I and III, fibronectin, TIMP 2, and PDGFRA. We believe these findings indicate that the OmAT pro-fibrotic phenotype in CBA/SIV macaques is largely associated with, and potentially mediated through, decreased MMP 13 activity.

We examined the changes in markers of lipid homeostasis of OmAT isolated from CBA/SIV/ART and VEH/SIV/ART macaques. CBA decreased OmAT triglyceride content, decreased PLIN gene expression, and disrupted gene expression of lipogenic regulators without altering lipogenic or lipolytic enzyme gene expression. Together, these results suggest that a consequence of ECM dysregulation is adipose tissue lipid dyshomeostasis and reduced lipid storage capacity. Perilipins surround lipid droplets as a barrier to lipolytic degradation and are an important target of PPARG in adipocytes (Arimura et al., 2004). In combination with our results showing decreased OmAT adipocyte size and triglyceride content, decreased perilipin expression may be indicative of decreased CBA/SIV-mediated OmAT lipid storage capacity.

Published studies show that chronic alcohol perturbs WAT metabolic homeostasis. PPARG and SREBP1c exhibit reciprocal communication with the ability to activate each other and synergistically increase lipogenesis in WAT (Festuccia et al., 2009). Contrary to most reports, our results indicate a robust increase in PPARG gene expression in OmAT of CBA/SIV macaques. Furthermore, in contrast to the stimulatory effect of PPARG on SREBP1c, CBA/SIV decreased OmAT SREBP1c gene expression. Evidence from rodent models indicates that chronic alcohol consumption promotes hepatic lipid accumulation and steatosis via mechanisms involving increased PPARG, SREBP, and ChREBP (Marmier et al., 2015, You and Crabb, 2004, Zhang et al., 2021b). Evidence from studies in obesity and type 2 diabetes showing decreased SREBP1c and lipogenesis in AT and increased SREBP1c expression and lipogenesis in the liver (Wondmkun, 2020) which may be representative of the metabolic phenotype seen in CBA/SIV macaques.

Our study is not without limitations. We have used cross sectional analysis of adipose tissue ECM and metabolic changes (at study endpoint) which may not fully reflect the entire spectrum of alterations in adipose tissue phenotype or the contributing mechanisms. Ongoing studies will examine AT changes longitudinally to identify mechanisms that lead to ECM dysregulation. Additional limitations include lack of OmAT samples obtained during the fed state, which could have potentially identified additional metabolic alterations, such as those related to insulin-dependent glucose homeostasis. Our published data from this cohort of animals do not show a CBA/SIV-mediated increase in liver triglyceride or NEFA levels, nor an increase in plasma triglyceride levels (Gallegos et al., 2024). There were also no significant correlations between adipose triglyceride levels and systemic or liver triglyceride levels, and whole-body glucose insulin dynamics (Simon et al., 2021). It appears that the observed changes in adipose tissue observed here have not produced overt effects in systemic circulation or in the liver. Several factors can be invoked as relevant. All macaques are relatively young and maintained on a nutritionally balanced diet, supplemented with fruits and vitamins. The study also does not determine the effects of SIV infection, and it is possible that the alcohol-mediated changes are confounded by SIV-infection. Despite these limitations, our results provide the framework for future studies to directly investigate the relationship between CBA/SIV-mediated changes in markers of ECM remodeling and lipid homeostasis Specifically, whether alcohol-induced changes in ECM promote changes in metabolic phenotype or vice versa, and how diet and SIV/HIV may impact adipose ECM and metabolic phenotype are being investigated in ongoing studies.

Overall, our results demonstrate that CBA preferentially alters OmAT ECM phenotype in SIV infection as summarized in Figure 8. Specifically, CBA administration to SIV-infected, ART-treated macaques results in alterations in OmAT ECM phenotype marked by increased ECM accumulation, a pro-fibrotic milieu, decreased MMP 13 activity and potentially AT stiffness. Furthermore, CBA significantly decreased triglyceride content, gene expression of perilipins, and dysregulated PPARG signaling. Collectively, these results suggest that CBA-mediated ECM accumulation traps adipocytes within a stiff environment potentially leading to disruption of adipocyte metabolic programming in HIV/SIV.

Figure 8.

Summary of chronic binge alcohol- (CBA) mediated effects on omental adipose tissue (OmAT) extracellular matrix (ECM) phenotype and markers of lipid homeostasis in simian immunodeficiency virus infection. Created with BioRender.com.