Skeletal muscle bioenergetic health and function in people living with HIV: association with glucose tolerance and alcohol use

Abstract

At-risk alcohol use is prevalent and increases dysglycemia among people living with human immunodeficiency virus (PLWH). Skeletal muscle (SKM) bioenergetic dysregulation is implicated in dysglycemia and type 2 diabetes. The objective of this study was to determine the relationship between at-risk alcohol, glucose tolerance, and SKM bioenergetic function in PLWH. Thirty-five PLWH (11 females, 24 males, age: 53 ± 9 yr, body mass index: 29.0 ± 6.6 kg/m²) with elevated fasting glucose enrolled in the ALIVE-Ex study provided medical history and alcohol use information [Alcohol Use Disorders Identification Test (AUDIT)], then underwent an oral glucose tolerance test (OGTT) and SKM biopsy. Bioenergetic health and function and mitochondrial volume were measured in isolated myoblasts. Mitochondrial gene expression was measured in SKM. Linear regression adjusting for age, sex, and smoking was performed to examine the relationship between glucose tolerance (2-h glucose post-OGTT), AUDIT, and their interaction with each outcome measure. Negative indicators of bioenergetic health were significantly (P < 0.05) greater with higher 2-h glucose (proton leak) and AUDIT (proton leak, nonmitochondrial oxygen consumption, and bioenergetic health index). Mitochondrial volume was increased with the interaction of higher 2-h glucose and AUDIT. Mitochondrial gene expression decreased with higher 2-h glucose (TFAM, PGC1B, PPARG, MFN1), AUDIT (MFN1, DRP1, MFF), and their interaction (PPARG, PPARD, MFF). Decreased expression of mitochondrial genes were coupled with increased mitochondrial volume and decreased bioenergetic health in SKM of PLWH with higher AUDIT and 2-h glucose. We hypothesize these mechanisms reflect poorer mitochondrial health and may precede overt SKM bioenergetic dysregulation observed in type 2 diabetes.

INTRODUCTION

About 1.2 million people are currently living with human immunodeficiency virus (HIV) in the United States (1), and antiretroviral therapy (ART) has increased life expectancy among people living with HIV (PLWH) to nearly that of seronegative individuals (2). Accordingly, aging PLWH face multimorbidity including increased risk for insulin resistance (3) that may be complicated by lifestyle factors such as at-risk alcohol use. It is estimated that ∼30% of PLWH have concurrent at-risk alcohol use (4, 5).

At-risk alcohol use has multisystemic adverse consequences (6) and increases the risk for comorbidities among PLWH (7). Using a chronic binge alcohol (CBA) administration paradigm to model at-risk alcohol use and simian immunodeficiency virus (SIV) infection to model HIV disease, our laboratory has previously observed that at-risk alcohol decreases the acute insulin response to glucose and disposition index (8). Among PLWH, alcohol use increased risk for dysglycemia (9), was directly related to 2-h glucose levels after an oral glucose tolerance test (OGTT) (9), and decreased homeostatic model assessment (HOMA) β cell function (10). Together, these findings suggest that at-risk alcohol use dysregulates glucose homeostasis in the context of SIV/HIV infection.

Functional skeletal muscle (SKM) mass is necessary for metabolic health (11, 12) and relies on mitochondrial (13, 14) and glycolytic (15, 16) function for its maintenance. Dysfunction of bioenergetic pathways may mechanistically contribute to the accentuated muscle wasting we have previously observed with CBA in male rhesus macaques with end-stage simian acquired immune deficiency syndrome (SAIDS) (17) and to impaired myoblast regeneration capacity (18). Work from our laboratory showed that CBA administration, SIV infection, and ART treatment decrease myoblast maximal mitochondrial respiration and SKM succinate dehydrogenase activity in male rhesus macaques (19). Furthermore, microarray analyses revealed that CBA decreased expression of glycolytic genes in SKM (20). This preclinical evidence demonstrates impaired SKM bioenergetic health, but this has not yet been translated to PLWH.

Insulin resistance in type 2 diabetes mellitus (T2D) is characterized by bioenergetic dysfunction (21, 22), but the relationship between SKM bioenergetic function and glucose tolerance in people without overt T2D is not well established, especially among PLWH. Moreover, alcohol dysregulates mitochondrial and glycolytic metabolism in SKM and myoblasts (muscle precursor cells) (18, 23). However, dose or concentration, duration, and other factors may affect the magnitude and direction of these changes. Because PLWH have greater rates of cardiometabolic comorbidities (24, 25) and prevalence of at-risk alcohol use (4, 5) than the general population, determining the relationship between SKM mitochondrial and glycolytic metabolism with at-risk alcohol and glucose tolerance may be particularly meaningful in this population. Therefore, the present study examined the relationship between at-risk alcohol, glucose tolerance, and myoblast/SKM bioenergetic function in PLWH.

METHODS

Participants

The parent study, Alcohol & Metabolic Comorbidities in PLWH: Evidence Driven Interventions (ALIVE-Ex; NCT03299205, registered October 3, 2017), is divided into two phases; the first phase was carried out to assess the relationship between alcohol use, glucose tolerance, and SKM dysfunction, and the second phase includes an exercise intervention to improve glycemic control and SKM metabolic function in PLWH. The present study includes the 35 in-care PLWH from the Greater New Orleans, LA metropolitan area (24 males, 11 females; age: 53 ± 9 yr; BMI: 29.0 ± 6.6 kg/m²; means ± SD) participating in the first phase of the ALIVE-Ex study who were medically cleared and underwent the muscle biopsy procedure. Participant characteristics and distribution based on Alcohol Use Disorders Identification Test (AUDIT) score are summarized in Table 1. Recruitment details, demographics of the full cohort in the first phase of the parent study, medical history, alcohol assessment, and oral glucose tolerance test procedure have been previously described (9). Inclusion criteria for the ALIVE-Ex study included fasting plasma glucose 95–124 mg/dL and diagnosed HIV infection. Exclusion criteria included T2D diagnosis, metformin or insulin use, current pregnancy, allergy to lidocaine, any bleeding disorder, or use of blood thinners that would contraindicate muscle biopsy. The study was approved by the Louisiana State University Health Sciences Center Institutional Review Board (IRB #736) and complied with the US Federal Policy for the Protection of Human Subjects (45 CFR, Part 46) and the ethical principles of the Declaration of Helsinki. Each individual was given an explanation of the study goals, procedures, risks, and benefits and verbally demonstrated understanding before providing written informed consent to participate. The first subject was enrolled on November 20, 2017.

Table 1.

Participant characteristics by Alcohol Use Disorders Identification Test (AUDIT) score

		AUDIT
	All (n = 35)	<5 (n = 21)	≥5 (n = 14)	P Value
Sex, % (n)				0.298
Female	31.4 (11)	38.1 (8)	21.4 (3)
Male	68.6 (24)	61.9 (13)	78.6 (11)
Race, % (n)				0.143
Black	79.4 (27)	74.1 (15)	92.3 (12)
White	20.6 (7)	28.6 (6)	7.7 (1)
Age, yr (range: 28–69 yr), % (n)				0.821
20 to <30 yr	2.9 (1)	4.8 (1)	0.0 (0)
30 to <40 yr	8.6 (3)	9.5 (2)	7.1 (1)
40 to <50 yr	22.9 (8)	19.1 (4)	28.6 (4)
50 to <60 yr	45.7 (16)	42.9 (9)	50.0 (7)
60+ yr	20.0 (7)	23.8 (5)	14.3 (2)
Income, % (n)				0.139
<$20,000	91.4 (32)	85.7(18)	100.0 (14)
$20,000–$39,999	8.6 (3)	14.3 (3)	0.0 (0)
Education, % (n)				0.414
<High school	37.1 (13)	42.9 (9)	28.6 (4)
High school graduate	37.1 (13)	33.3 (7)	42.9 (6)
Some college or trade school	20.0 (7)	14.3 (3)	28.6 (4)
4-yr College/graduate school	5.7 (2)	9.5 (2)	0 (0)
Body mass index (BMI), % (n)				0.120
<18.5	2.9 (1)	0.0 (0)	7.1 (1)
18.5–24.9	22.9 (8)	33.3 (7)	7.1 (1)
25–29.9	42.9 (15)	28.6 (6)	64.3 (9)
30–34.9	8.6 (3)	9.5 (2)	7.14 (1)
35+	22.9 (8)	28.6 (6)	14.3 (2)
Viral load, % (n)				0.288
<200 Copies/mL	96.0 (24)	91.7 (11)	100.0 (13)
200+ Copies/mL	4.0 (1)	8.3 (1)	0.0 (0)
Smoking status, % (n)				0.053
Nonsmoker	48.6 (17)	61.9 (13)	28.6 (4)
Current smoker	51.4 (18)	38.1 (8)	71.4 (10)
Phosphatidylethanol (PEth), % (n)				0.011
Positive (≥8 ng/mL)	60.0 (21)	43.9 (9)	85.7 (12)
Negative (<8 ng/mL)	40.0 (14)	57.1 (12)	14.3 (2)
Glucose tolerance status, % (n)				0.037
Normal	57.1 (20)	71.4 (15)	35.7 (5)
Impaired	42.9 (15)	28.6 (6)	64.3 (9)

ALIVE-Ex study (2018–2020). Bold values indicate P < 0.05.

Oral Glucose Tolerance Test and Blood Collection

After an overnight fast, glucose was measured and verified by capillary blood using a glucometer (True Metrix Pro Blood Glucose Monitoring System, McKesson) during participant screening as previously described (9). Participants with a fasting glucose between 95 and 124 mg/dL returned to the laboratory and underwent an oral glucose tolerance test (OGTT). Participants consumed a 75-g bolus of glucose (Trutol, Thermo Fisher Scientific) and rested comfortably in a seated position for 2 h. Blood samples were collected from an antecubital vein using standard venipuncture technique into EDTA-coated tubes before and 2 h after glucose ingestion. Plasma was separated (750 g, 15 min) and stored in several aliquots at −80°C until analysis for glucose in duplicate using an Analox (GM7 Micro-Stat Analyzer, Analox Instruments Ltd., UK). For calculation of gene expression, participants were categorized as having normal (<140 mg/dL) or impaired (140 to <200 mg/dL) glucose tolerance based on their plasma glucose measured 2 h after glucose ingestion. For regression analyses, 2-h glucose levels were considered a continuous measure of glucose tolerance, with a lower value indicating greater tolerance.

Demographics and Alcohol Use Assessment

At the oral glucose tolerance test (OGTT) visit, participants provided their demographic information. The AUDIT (26) was used to assess alcohol use. AUDIT < 5 was considered low-risk alcohol use and AUDIT ≥ 5 was considered at-risk alcohol use for gene expression calculations. Continuous AUDIT scores were used for regression analyses. This self-report alcohol use assessment was complemented by phosphatidylethanol (PEth) analysis as a biomarker of recent alcohol use (27). A dried blood spot was prepared from the pre-OGTT blood sample for PEth measurement (United States Drug Testing Laboratories, Inc., Des Plaines, IL). PEth values were used as a continuous variable.

Muscle Biopsy

For the muscle biopsy, participants returned to the laboratory after an overnight fast and were administered a standardized meal of Carnation Breakfast Essential Nutritional Drink (chocolate or vanilla, 240 kcals: 17% from protein, 15% from fat, 68% from carbohydrates; Nestlé, Vevey, Switzerland) 90 min before the muscle biopsy procedure. Biopsy supplies (Temno Biopsy Needle, 14 G, CIA7003702; Basic Biopsy Tray, CIA7005455) were purchased from Central Infusion Alliance (Skokie, IL). A small area of skin on the thigh over the vastus lateralis approximately halfway between the anterior superior iliac spine and the superior lateral patella was cleaned using povidone-iodine swab sticks and local anesthesia was applied (2% lidocaine without epinephrine). Then, a small incision was made through the skin and fascia using a sterile scalpel and a sterile, single-use microbiopsy needle was introduced into the muscle under ultrasound guidance. Six samples (∼25 mg each) were obtained through the same incision site. After completion of the biopsy, pressure was applied to the incision site for 5 min to prevent bleeding, and the site was closed using a sterile bandage. Participants were provided with written instructions, antibiotics, and gauze for care of the biopsy site. A staff nurse followed up with each participant 24–36 h following the biopsy procedure to inquire about the incision site. A portion of the muscle sample (∼75 mg) was quickly cleaned of residual blood, fat, and connective tissue and subsequently flash frozen; a portion (∼50 mg) was retained in sterile culture media for myoblast isolation (described in Myoblast Isolation and Expansion) and a portion (∼25 mg) was immediately fixed in zinc-buffered formalin.

Myoblast Isolation and Expansion

Primary myoblasts were isolated from vastus lateralis muscle samples (∼50 mg) as previously described (18, 28–31). Samples were washed three times in Ham’s F-12 nutrient mixture (GE Healthcare Life Sciences, Marlborough, MA), minced, and trypsin digested (0.25% trypsin EDTA diluted 1:4 in Ham’s F-12). Digested muscle tissue was plated for 5 h in growth media [Ham’s F-12 nutrient mixture with 10% fetal bovine serum, 2% l-glutamine, and 2.5 ng/mL recombinant human fibroblast growth factor (R&D systems, Minneapolis, MN)] to allow fibroblasts to adhere to the plate. Supernatant containing muscle tissue and myoblasts was transferred to a fresh plate with growth media, and myoblast colonies were allowed to grow for 1 wk. Thereafter, media was changed every other day. Myoblasts were passaged at 80%–90% confluence and cryopreserved after each passage. Cells were taken to passage 3 (P3) to P4 for use in experiments.

Mitochondrial Function

Mitochondrial oxygen consumption rate (OCR) was measured using a Mito Stress Test (18) and Seahorse XFe96 technology (Agilent Technologies, Santa Clara, CA). Myoblasts at P3 to P4 were seeded on a collagen-coated 96-well Seahorse plate in triplicate at a density of 35,000 cells/well and maintained under standard cell culture conditions (5% CO₂, 37°C). After 24 h, growth media was replaced with XF Assay Medium (pH 7.4) with sodium pyruvate (1 mM), L-glutamine (2 mM), and glucose (10 mM) and incubated at 37°C without CO₂ for 1 h before measuring myoblast oxygen consumption rate (OCR) with an XFe96 Extracellular Flux Analyzer (Agilent) according to manufacturer’s instructions. Respiratory parameters were assessed using the Mito Stress Test Kit (Agilent) by the sequential addition of oligomycin (1.5 µM), carbonyl cyanide‐p‐trifluoromethoxyphenylhydrazone (FCCP; 2 μM), and rotenone/antimycin A (0.5 μM). Resulting OCR measurements were normalized to cell count obtained by staining nuclei with Hoescht dye (2 µM; Thermo Fisher Scientific, Waltham, MA) and visualizing on a BioTek Cytation 1 cell imaging multimode reader (BioTek, Winooski, VT).

Glycolytic Function

To quantify glycolytic function, capacity, and reserve, a Glycolysis Stress Test was performed using Seahorse XFe96 technology (Agilent; 32). Myoblasts were seeded as described in Mitochondrial Function. After 24 h, growth media was replaced with glucose- and pyruvate-free XF Assay Medium (pH 7.4) with L-glutamine (2 mM). Then, myoblasts were incubated at 37°C without CO₂ for 1 h before measuring myoblast extracellular acidification rate (ECAR) at baseline and after the sequential addition of glucose (10 mM), oligomycin (1.5 μM), and 2-deoxyglucose (2-DG; 50 mM). Nuclei were stained using Hoescht dye (2 µM; Thermo Fisher Scientific) and visualized on a Cytation 1 cell imaging multimode reader (BioTek) for normalization to cell count.

Mitochondrial Volume Assessment

To quantify functional myoblast mitochondrial volume, MitoTracker Deep Red FM probes (Invitrogen, Carlsbad, CA) were used (33, 34). Myoblasts were incubated with growth media containing MitoTracker Deep Red FM (100 nM) under standard cell culture conditions for 30 min. An unstained control was simultaneously incubated in growth media without MitoTracker. Cells were pelleted, washed with phosphate-buffered saline (PBS), and prepared for flow cytometry using the PerFix-nc Kit (Beckman Coulter, Brea, CA) according to manufacturer instructions. Briefly, ∼500,000 cells in 100-µL PBS were incubated at room temperature for 15 min with 5 µL of fixative. Then, cells were washed and resuspended in 300 µL final reagent. Ten-thousand events per sample were acquired and excited by a 633-nm wavelength laser line using flow cytometry (BD FACSCanto II, Becton, Dickinson and Company, Franklin Lakes, NJ). Fluorescence was analyzed using FlowJo software (v. 10.7.1, Becton, Dickinson and Company). Autofluorescence of the unstained control sample was subtracted from the mean fluorescence intensity (MFI) of each experimental sample for analysis.

RNA Isolation and Quantitative Real-Time Polymerase Chain Reaction

To assess expression of mitochondrial genes, total RNA was extracted from flash frozen SKM samples (∼10 mg) using the miRNeasy Mini Kit (Qiagen, Valencia, CA) according to manufacturer’s instructions. cDNA was synthesized from 0.15 µg of RNA using the Quantitect reverse transcription kit (Qiagen) according to manufacturer’s instructions and diluted 1:4 in nuclease-free water. Custom primers designed to span exon-exon junctions were purchased from Integrated DNA Technologies (Coralville, IA; Table 2). Final reactions contained cDNA (7.5 ng), primers (500 nM), SyBr green (Quantitect SyBr Green PCR kit, Qiagen), and nuclease-free water to 20 µL. qPCR reactions were carried out in duplicate using a CFX96 thermal cycler (Bio-Rad, Hercules, CA) with ribosomal protein S13 (RPS13) as the endogenous control (18, 31, 35, 36). Data were then analyzed using the 2^−ΔΔCt method. Results for target genes are calculated as fold change versus the normal glucose tolerance, low-AUDIT group.

Table 2.

List of primers for qPCR analysis (primers from IDT Technologies)

Gene	Forward Primer	Reverse Primer
TFAM	CTCCTGAGTAGCTGGGATTA	TCAGCTCCTACAGCAACTA
PGC1A	CCACCACTCCTCCTCATAA	ACTGTACCTGGGCTTCTT
PGC1B	GGCAACTCCAAGGTCTATTC	ACGAGATGGTTGCCAATG
PPARA	GATTGATAAGGGTCCCATGAC	GGATGTCCTAACCCTAACTTTC
PPARG	GACTTCTCCAGCATTTCTACTC	GGTACTCTTGAAGTTTCAGGTC
PPARD	TGATGCCCAGTACCTCTT	TCTCGGTTTCGGTCTTCT
MFN1	CGATGCACCGATGAAGTAA	GCAAGGGATCAGTGTATGTAG
MFN2	CATGCAGCAGGACATGATAG	TAGGTCATAGTTGAGGGAGAAG
DRP1	GTCCTCGTCCTGCTTTATTT	CATGAACCAGTTCCACACA
MFF	AAGTAGCACCAAAC	CCTGCTACAACAATCCTCTC
UCP2	CTCTGGCCTTCACAACATC	GGCAAATGGTATCCCTTCTC
UCP3	ACTATGGACGCCTACAGAA	CTCAGCACAGTTGACGATAG
RPS13	CTTCACAGATCGGTGTAATCC	ATCAGGAGCAAGTCCCTTA

DRP, dynamin-related protein; MFF, mitochondrial fission factor; MFN, mitofusin; PGC, peroxisome proliferator-activated receptor-γ (PPAR) coactivator; RPS13, ribosomal protein S13; TFAM, mitochondrial transcription factor A; UCP, uncoupling protein.

Statistical Analyses

Descriptive statistics were used to characterize the sample population overall. χ² and Wilcoxon rank-sum tests were used to compare the means of the continuous measures by AUDIT score <5 or ≥5. Spearman’s correlations were used to examine the relationship between alcohol use measures (AUDIT and PEth). As anticipated, AUDIT and PEth were strongly associated (Spearman’s ρ = 0.64, P < 0.001). Spearman’s correlations were used to determine whether a significant relationship existed between sex, age, smoking status, and body mass index (BMI) and each outcome measure of interest. Outcome measures were log transformed before analyses. Multiple linear regression was used to determine the relationships between alcohol use, glucose tolerance, and each outcome measure. Sex, age, and smoking status, but not BMI, were significantly associated with at least one outcome measure; therefore, subsequent regression models were adjusted for age, sex, and smoking status. Four separate regression models were run for each dependent variable: model 1 examined the relationship with glucose tolerance, model 2 examined the relationship with AUDIT, and model 3 examined the relationship with the interaction of glucose tolerance and AUDIT. The fourth model (results not shown) included both AUDIT and glucose tolerance without the interaction term. Analyses were completed using SPSS Statistics software (version 25, IBM, Armonk, NY).

RESULTS

Participant Characteristics

Participants were majority male (∼69%), Black (∼79%), middle aged (53 ± 9 yr), and virally suppressed (<200 copies/mL; 96%). Sex, race, age, income, education, BMI, and mean 2-h glucose after OGTT (AUDIT < 5: 121.8 ± 32.9 mg/dL, AUDIT ≥ 5: 140.1 ± 35.5 mg/dL) did not differ between participants with lower (<5) or higher (≥5) AUDIT scores. Although smoking was not different between groups, there was a strong trend for current smoking to be more prevalent among participants with AUDIT ≥ 5 (P = 0.053). A positive PEth test, a measure of recent alcohol use, and 2-h glucose in the “impaired” category were significantly (P < 0.05) more prevalent among participants with AUDIT scores ≥ 5 (Table 1).

Mitochondrial Function

Proton leak, measured as the difference between OCR after oligomycin treatment and rotenone/antimycin A treatment, was significantly associated (P < 0.05) with higher 2-h glucose levels and higher AUDIT score (Table 3). Increasing nonmitochondrial oxygen consumption, measured as OCR after rotenone/antimycin A treatment, was associated with increasing AUDIT scores, and bioenergetic health index (37) [BHI; (ATP-linked OCR × spare capacity)/(proton leak × nonmitochondrial OCR)] was lower with higher AUDIT scores. No significant associations were observed between 2-h glucose, AUDIT, or their interaction and measures of mitochondrial respiration (basal OCR, spare capacity, ATP-linked OCR, or coupling efficiency). For visualization of OCR throughout the assay, OCR traces during the Mitochondrial Stress Test stratified by glucose tolerance and alcohol use categories are shown in Fig. 1.

Table 3.

Mitochondrial Stress Test parameters

	Basal OCR		Spare Capacity		ATP-Linked OCR		Proton Leak		Nonmitochondrial OCR		Coupling Efficiency		BHI
	β	P	β	P	β	P	β	P	β	P	β	P	β	P
Model 1
GLUC	0.18	0.88	0.15	0.44	0.09	0.67	0.38	0.05	0.35	0.06	−0.27	0.17	−0.22	0.24
Model 2
AUDIT	0.09	0.64	0.02	0.92	−0.03	0.89	0.41	0.02	0.35	0.04	0.07	0.70	−0.38	0.03
Model 3
GLUC	0.02	0.95	0.08	0.78	0.04	0.9	−0.06	0.81	−0.02	0.92	−0.42	0.15	0.11	0.68
AUDIT	−2.46	0.49	−1.43	0.66	−1.32	0.71	−5.11	0.09	−4.37	0.15	−1.08	0.74	2.89	0.36
GLUC × AUDIT	2.54	0.48	1.42	0.67	1.28	0.73	5.54	0.07	4.72	0.13	1.30	0.70	−3.31	0.30

Linear regression β coefficients of 2-h glucose levels after an oral glucose tolerance test (GLUC; model 1), alcohol use disorders information test (AUDIT) score (model 2), and their interaction (model 3) and mitochondrial stress test parameters measured in myoblasts adjusted by age, sex, and smoking status in people living with HIV (PLWH) in the ALIVE-Ex study (2018–2020). Bold text indicates P < 0.05. BHI, bioenergetic health index; OCR, oxygen consumption rate.

Figure 1.

Mitochondrial Stress Test. Note: Oxygen consumption rate (OCR) normalized to cell count at baseline (BL; measurements 1–3) and after the sequential addition of oligomycin (Oligo; measurements 4–6), carbonyl cyanide‐p‐trifluoromethoxyphenylhydrazone (FCCP; measurements 7–9), and rotenone/antimycin A (Rot + AA; measurements 10–12) in myoblasts from participants with normal or impaired glucose tolerance and low (<5) or high (≥5) Alcohol Use Disorders Identification Test (AUDIT) scores. n = 35 participants, means ± SE.

Glycolytic Function

No significant associations were observed between 2-h glucose, AUDIT, and measures of glycolysis (nonglycolytic acidification, glycolysis, glycolytic capacity, and glycolytic reserve; Table 4). There was a nonsignificant trend (P = 0.053) for a negative association between baseline glycolysis and the interaction between 2-h glucose and AUDIT. For visualization of ECAR throughout the assay, ECAR traces during the glycolytic stress test stratified by glucose tolerance and alcohol use categories are shown in Fig. 2.

Table 4.

Glycolysis Stress Test parameters

	Nonglycolytic Acidification		Glycolysis		Glycolytic Capacity		Glycolytic Reserve
	β	P	β	P	β	P	β	P
Model 1
GLUC	0.24	0.19	0.03	0.86	0.07	0.72	0.11	0.57
Model 2
AUDIT	0.28	0.10	0.12	0.50	0.15	0.41	0.15	0.42
Model 3
GLUC	−0.10	0.70	0.35	0.20	0.3	0.29	0.07	0.82
AUDIT	−3.90	0.19	6.38	0.05	5.09	0.12	0.29	0.94
GLUC × AUDIT	4.21	0.17	−6.38	0.05	−5.05	0.13	−0.16	0.97

Linear regression β coefficients of 2-h glucose levels after an oral glucose tolerance test (GLUC; model 1), alcohol use disorders information test (AUDIT) score (model 2), and their interaction (model 3) and glycolysis stress test parameters measured in myoblasts adjusted by age, sex, and smoking status in people living with HIV (PLWH) in the ALIVE-Ex study (2018–2020). Bold text indicates P < 0.05.

Figure 2.

Glycolysis Stress Test. Note: Extracellular acidification rate (ECAR) normalized to cell count at baseline (BL; measurements 1–3) and after the sequential addition of glucose (measurements 4–6), oligomycin (Oligo; measurements 7–9), and 2-deoxyglucose (2-DG; measurements 10–12) in myoblasts from participants with normal or impaired glucose tolerance and low (<5) or high (≥5) Alcohol Use Disorders Identification Test (AUDIT) scores. n = 35 participants, means ± SE.

Mitochondrial Volume

There was no significant association of mitochondrial volume, measured by MitoTracker, with 2-h glucose or AUDIT independently. However, mitochondrial volume was positively associated with the interaction between 2-h glucose and AUDIT (Table 5).

Table 5.

Mitochondrial volume

	Mitochondrial Volume
	β	P
Model 1
GLUC	0.23	0.26
Model 2
AUDIT	0.17	0.36
Model 3
GLUC	−0.25	0.37
AUDIT	−7.27	0.03
GLUC × AUDIT	7.53	0.03

Linear regression β coefficients of 2-h glucose levels after an oral glucose tolerance test (GLUC, model 1), alcohol use disorder information test (AUDIT) score (model 2), and their interaction (model 3) and mitochondrial volume (MitoTracker mean fluorescence intensity) measured in myoblasts adjusted by age, sex, and smoking status in people living with HIV (PLWH) in the ALIVE-Ex study (2018–2020). Bold text indicates P < 0.05.

SKM Mitochondrial Gene Expression

Expression of genes controlling mitochondrial biogenesis [mitochondrial transcription factor A (TFAM) and peroxisome proliferator-activated receptor-γ (PPAR) coactivator (PGC)1B] was significantly decreased with higher 2-h glucose (Table 6). Expression of genes associated with lipid handling was significantly decreased with higher 2-h glucose (PPARG) and with the interaction between 2-h glucose and AUDIT (PPARG and PPARD). Expression of genes controlling mitochondrial dynamics was significantly decreased with higher 2-h glucose [mitofusin (MFN)1] and AUDIT [MFN1, dynamin-related protein (DRP)1, and mitochondrial fission factor (MFF)]. There was a statistically significant interaction between 2-h glucose and AUDIT (MFF). No significant relationship was observed between 2-h glucose, AUDIT, and PGC1A, PPARA, MFN2, or gene expression of uncoupling proteins UCP2 and UCP3.

Table 6.

Skeletal muscle mitochondrial gene expression

	TFAM		PGC1A		PGC1B		PPARA		PPARG		PPARD
	β	P	β	P	β	P	β	P	β	P	β	P
Model 1
GLUC	−0.46	0.01	−0.27	0.15	−0.41	0.01	−0.26	0.11	−0.37	0.04	−0.18	0.37
Model 2
AUDIT	−0.14	0.46	−0.03	0.87	−0.09	0.57	−0.06	0.68	−0.30	0.08	−0.21	0.26
Model 3
GLUC	−0.52	0.07	−0.26	0.37	−0.28	0.21	−0.24	0.32	0.13	0.59	0.32	0.24
AUDIT	−0.7	0.83	0.92	0.78	2.82	0.28	0.58	0.84	7.05	0.02	7.23	0.03
GLUC × AUDIT	0.75	0.82	−0.86	0.80	−2.81	0.29	−0.56	0.85	−7.40	0.01	−7.56	0.03

	MFN1		MFN2		DRP1		MFF		UCP2		UCP3
	β	P	β	P	β	P	β	P	β	P	β	P
Model 1
GLUC	−0.46	0.02	0.1	0.64	−0.26	0.19	−0.03	0.87	−0.04	0.85	−0.24	0.20
Model 2
AUDIT	−0.43	0.02	−0.23	0.22	−0.55	<0.01	−0.47	<0.01	−0.03	0.87	0.08	0.64
Model 3
GLUC	−0.22	0.42	0.41	0.17	−0.01	0.98	0.58	0.03	0.34	0.23	0.01	>0.99
AUDIT	1.57	0.61	2.89	0.40	0.14	0.96	6.23	0.04	6.37	0.06	5.79	0.07
GLUC × AUDIT	−1.93	0.55	−3.26	0.35	−0.68	0.82	−6.91	0.02	−6.53	0.06	−5.71	0.08

Linear regression β coefficients of 2-h glucose levels after an oral glucose tolerance test (GLUC; model 1), alcohol use disorders information test (AUDIT) score (model 2), and their interaction (model 3) and expression of mitochondrial genes measured in skeletal muscle adjusted by age, sex, and smoking status in people living with HIV (PLWH) in the ALIVE-Ex study (2018–2020). Bold text indicates P < 0.05. DRP: dynamin-related protein; MFF, mitochondrial fission factor; MFN, mitofusin; PPAR, peroxisome proliferator-activated receptor; PGC, peroxisome proliferator-activated receptor-γ coactivator; TFAM, mitochondrial transcription factor A; UCP, uncoupling protein.

DISCUSSION

Skeletal muscle tissue plays a key role in whole body metabolism and glucose homeostasis; therefore, factors that influence SKM energetic health have multisystemic implications. In the present study, there was a negative interaction of 2-h glucose level and AUDIT score with SKM expression of genes controlling mitochondrial fission and lipid handling and this was associated with greater myoblast mitochondrial volume. Bioenergetic health index (BHI) accounts for positive (ATP-linked oxygen consumption and spare capacity) and negative (proton leak and nonmitochondrial oxygen consumption) mitochondrial respiratory parameters (37). In myoblasts from PLWH, BHI was lower with increasing AUDIT score. Increased 2-h glucose and AUDIT were independently related to decreased SKM expression of genes regulating mitochondrial biogenesis and dynamics. Overall, our study provides evidence that the mitochondrial profile in PLWH is dysregulated with increased 2-h glucose and at-risk alcohol use when adjusted for age, sex, and smoking status.

We previously observed decreased glycolytic function and, conversely, increased maximal mitochondrial respiration in myoblasts cultured with 50 mM ethanol, a physiologically relevant dose (18). In contrast, myoblasts isolated from SKM of chronic binge alcohol-administered, SIV-infected male rhesus macaques had decreased maximal mitochondrial respiration compared with controls (19). In the current study, proton leak-associated and nonmitochondrial oxygen consumption were higher and overall bioenergetic health lower with AUDIT in myoblasts from PLWH, with no statistically significant changes in other bioenergetics measures such as basal, maximal, or ATP-linked oxygen consumption. There may be a threshold of intensity of alcohol exposure (concentration × frequency) required to produce overt bioenergetic deficits. Furthermore, whereas samples from our previous preclinical studies were derived from animals fed high-quality, healthy diets, with normal fasting glycemia, participants in the present study had elevated fasting glucose (95–124 mg/dL) and habitual diets were not controlled. Therefore, it is possible that differences may be elucidated with the inclusion of more participants at either end of the fasting glucose tolerance spectrum, from normal fasting glucose to fasting hyperglycemia indicative of T2D.

Increased mitochondrial content is often an adaptive process (e.g., in response to exercise) that can increase aerobic energy production capacity via increased content of electron transport chain complexes on the inner mitochondrial membrane (38). This adaptive increase is driven by increased expression of master regulators of mitochondrial biogenesis (e.g., PGC1α and PGC1β) along with transcription of the mitochondrial genome by TFAM (39, 40). In the present study, greater mitochondrial volume was not driven by increased expression of regulators of mitochondrial biogenesis as gene expression of PGC1B and TFAM was decreased with lower glucose tolerance (higher 2-h glucose). Moreover, the greater mitochondrial volume did not coincide with improved mitochondrial function. There was no concomitant increase in gene expression of uncoupling proteins, changes in ATP-linked oxygen consumption, or decreased maximal oxygen consumption. Rather, it appears that the observed increase in mitochondrial content may reflect accumulation of suboptimal portions of the mitochondrial network that would typically have been eliminated through mitochondrial fission and subsequent mitophagy under normal physiological conditions. Accordingly, bioenergetic health was decreased with AUDIT, and proton leak, a negative indicator of bioenergetic health (37), was increased with lower glucose tolerance. These data suggest that the overall health of the mitochondrial network is compromised in PLWH with lower glucose tolerance and at-risk alcohol use.

Mitochondrial dynamics maintain quality and morphology of the mitochondrial network. Inhibition of either mitochondrial fusion (41) or fission (42) impairs mitochondrial function, particularly electron transport chain activity. Mitochondrial fission is essential for quality control of the mitochondrial network, limiting mitochondrial size (43), and precedes mitophagy (44), which eliminates damaged or dysregulated mitochondria. In the present study, SKM expression of DRP1 was decreased with alcohol, and gene expression of an essential DRP1 recruitment factor (45), MFF, was decreased with the interaction of increased AUDIT and 2-h glucose. It is possible that the observed increase in mitochondrial volume in the present study may have been due, in part, to decreased fission machinery. This hypothesis is consistent with previous findings that ethanol increased the formation of enlarged hepatic mitochondria with aberrant Drp1 pathway signaling in a murine model of chronic ethanol administration (46).

Separately, chronic ethanol feeding increased total SKM mitochondrial area in mice with a paradoxical decrease in fusion protein Mfn1 but not Mfn2 (47), which is consistent with gene expression results in the present study. These preclinical data support a role for altered mitochondrial dynamics in enlarged SKM mitochondria in human participants with alcohol use disorder (48). In the present study, participants ranged in severity of at-risk alcohol use and mitochondrial volume was not associated with AUDIT alone. Instead, increased mitochondrial volume was observed with the combination of higher AUDIT and lower glucose tolerance. Preclinical models have shown that MFN2 expression protects against glucose intolerance (49), whereas deficiency of fusion protein optic atrophy 1 (OPA1) protects against high-fat diet-mediated glucose intolerance (50), suggesting that mitochondrial fusion-related gene expression might be implicated in the development of T2D although the direction is less clear. Furthermore, people with overt T2D have smaller and more fragmented mitochondria than nondiabetic counterparts (51). The relationship between mitochondrial volume and the interaction between alcohol and glucose tolerance in the present study differs from that observed in T2D patients and may reflect a compensatory mechanism to maintain bioenergetic function in our cohort of PLWH. Although participants were largely virally suppressed, mitochondrial dynamics parameters may be complicated by their HIV status. Detailed analyses of mitochondrial dynamics are the subject of ongoing investigations.

Although no overt insufficiency in the capacity for mitochondrial ATP synthesis was observed in the present study, the interaction of AUDIT and lower glucose tolerance was associated with decreased PPARG and PPARD gene expression, suggesting dysregulated fuel utilization. This family of genes encodes transcription factors required for lipid handling, and in SKM, they are particularly important for fatty acid oxidation and insulin sensitivity (52–54). Previous work demonstrated that semiweekly administration of a PPARα, δ, or γ agonist protects against chronic ethanol-induced decreases in insulin-like growth factor 1 (IGF-1) receptor binding and insulin-responsive gene expression [aspartyl‐asparaginyl‐β‐hydroxylase (AAH)] in the liver, with concomitantly decreased steatosis (55). In SKM, activation of PPARδ rescued ethanol-mediated decreases in Akt phosphorylation, intramuscular triglyceride accumulation, and myotube glucose uptake (56). SKM insulin resistance is a key feature of T2D and precedes diabetes onset, and the PPAR family of transcription factors is targeted by antidiabetic medications that directly improve SKM insulin sensitivity (i.e., thiazolidinediones) (57). Therefore, it is notable that our observed decreases in expression of these genes with the interaction of AUDIT and 2-h glucose occurred in participants without overt T2D, suggesting that altered SKM lipid handling via PPAR signaling may occur before overt pathology. Our study was not powered to examine differences in specific fuel substrate oxidation, and studies that will determine the impact of alcohol use are warranted.

Perspectives and Significance

Overall, mitochondrial-related gene expression decreased with lower glucose tolerance and higher AUDIT score, and these changes were coupled with increased mitochondrial volume and decreased bioenergetic health in SKM of PLWH with elevated fasting glucose. However, the increase in mitochondrial volume does not appear to be adaptive. We hypothesize that the changes observed in genes regulating biogenesis, dynamics, and lipid handling may underlie poorer bioenergetic health and may precede overt energetic dysregulation observed with T2D (Fig. 3). These data underscore the need for effective interventions such as exercise to improve SKM and systemic metabolic health among PLWH with dysglycemia and at-risk alcohol use.

Figure 3.

Summary of findings. Note: Observed relationship between Alcohol Use Disorders Identification Test (AUDIT) score, 2-h glucose level, and outcome measures including expression of genes controlling mitochondrial biogenesis, fusion, fission, and lipid handling; mitochondrial volume; and indices of bioenergetic health, indicating a dysregulated mitochondrial profile as AUDIT score and 2-h glucose increase in PLWH. Image created with BioRender.com.

GRANTS

This work was funded by National Institutes of Health (NIH)/ National Institute on Alcohol Abuse and Alcoholism (NIAAA) Grants F32AA027982 (to D.E.L.), UH2AA026198 (to P.E.M.), and P60AA009803 (to P.E.M.).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

D.E.L., L.S., and P.E.M. conceived and designed research; D.E.L., J.A.Z., J.A., and R.H.M. performed experiments; D.E.L. and T.F.F. analyzed data; D.E.L. and L.S. interpreted results of experiments; D.E.L. prepared figures; D.E.L. drafted manuscript; D.E.L., T.F.F., S.D.P., L.S., and P.E.M. edited and revised manuscript; D.E.L., T.F.F., S.D.P., J.A.Z., J.A., R.H.M., L.S., and P.E.M. approved final version of manuscript.