
==== Front
eBioMedicine
EBioMedicine
eBioMedicine
2352-3964
Elsevier

S2352-3964(24)00374-8
10.1016/j.ebiom.2024.105338
105338
Articles
HIV-1-DNA/RNA and immunometabolism in monocytes: contribution to the chronic immune activation and inflammation in people with HIV-1
Muñoz-Muela Esperanza a
Trujillo-Rodríguez María a
Serna-Gallego Ana a
Saborido-Alconchel Abraham a
Gasca-Capote Carmen a
Álvarez-Ríos Ana b
Ruiz-Mateos Ezequiel a
Sviridov Dmitri cd
Murphy Andrew J. c
Lee Man K.S. c
López-Cortés Luis F. lflopez@us.es
af∗
Gutiérrez-Valencia Alicia aef
a Clinical Unit of Infectious Diseases, Microbiology and Parasitology, Institute of Biomedicine of Seville/Virgen del Rocio University Hospital/CSIC/University of Seville, Spain
b Department of Clinical Biochemistry, Virgen del Rocío University Hospital, Seville, Spain
c Baker Heart and Diabetes Institute, Melbourne, Australia
d Department of Biochemistry and Molecular Biology, Monash University, Clayton, Victoria, Australia
e Primary Care Pharmacist Service, Sevilla Primary Care District, Seville, Spain
∗ Corresponding author. Instituto de Biomedicina de Sevilla/Hospital Universitario Virgen del Rocío/CSIC/Universidad de Sevilla, Campus Hospital Universitario Virgen del Rocío, Calle Antonio Maura Montaner s/n, 41013, Sevilla, Spain. lflopez@us.es
f These authors have contributed equally to the present work.

11 9 2024
10 2024
11 9 2024
108 1053388 5 2024
22 8 2024
30 8 2024
Crown Copyright © 2024 Published by Elsevier B.V.
2024

https://creativecommons.org/licenses/by-nc-nd/4.0/ This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Summary

Background

Among people living with HIV-1 (PHIV), immunological non-responders (INR) experience incomplete immune recovery despite suppressive antiretroviral treatment (ART), facing more severe non-AIDS events than immunological responders (IR) due to higher chronic immune activation and inflammation (cIA/I). We analyzed the HIV-1 reservoir and immunometabolism in monocytes as a source of cIA/I.

Methods

Cross-sectional study in which 110 participants were enrolled: 25 treatment-naïve; 35 INR; 40 IR; and 10 healthy controls. Cell-associated HIV-1-DNA (HIV-DNA) and -RNA (HIV-RNA) were measured in FACS-isolated monocytes using digital droplet PCR. Intact, 5′ deleted, and 3′ deleted proviruses were quantified by the intact proviral DNA assay. Systemic inflammation, monocyte immunophenotype, and immunometabolism were characterized by immunoassays, flow cytometry, and real-time cellular bioenergetics measurements, respectively.

Findings

Monocytes from INR harbor higher HIV-RNA and HIV-DNA levels than IR. HIV-RNA was found in 14/21 treatment-naïve [2512 copies/106 TBP (331–4666)], 17/33 INR [240 (148–589)], and 15/28 IR [144 (15–309)], correlating directly with sCD163, IP-10, GLUT1high cells and glucose uptake, and inversely with the CD4+/CD8+ ratio. HIV-DNA was identified in all participants with detectable HIV-RNA, with intact provirus in 9/12 treatment-naïve [13 copies/106 monocytes (7–44)], 8/14 INR [46 (18–67)], and 9/13 IR [9 (7–24)]. INR presented glucose metabolism alterations and mitochondrial impairment; decreased coupling efficiency and BHI, and increased mitochondrial dysfunction inversely correlating with the CD4+/CD8+ ratio.

Interpretation

HIV-RNA, more than HIV-DNA, in monocytes and their altered metabolism are factors associated with the higher cIA/I that characterize INR.

Funding

This work was supported by the 10.13039/501100008530 European Regional Development Fund , 10.13039/501100004587 ISCIII , grant PI20/01646 . Other funding sources: 10.13039/501100004587 Instituto de Salud Carlos III through the Subprogram Miguel Servet (CP19/00159 ) to AGV, PFIS contracts (FI19/00304 ) to EMM, (FI21/00165 ) to ASA, and (FI19/00083 ) to CGC, and a mobility grant (MV21/00103 ) to EMM, from the 10.13039/501100004837 Ministerio de Ciencia e Innovación , Spain. AJM was granted by a CSL Centenary Award.

Keywords

Immunological non-responders
Monocytes
HIV-1 reservoir
Immunometabolism
Immune activation and inflammation
==== Body
pmc Research in context

Evidence before this study

A significant proportion of people living with HIV- 1 experience an incomplete immune recovery, despite suppressive antiretroviral treatment, known as immunological non-responders (INR). There is clear evidence of an increased chronic immune activation and inflammation (cIA/I) in these subjects, however, the precise underlying mechanisms remain incompletely understood. Persistence of the HIV-1 reservoir and immunometabolic abnormalities have been proposed as potential factors triggering cIA/I. Nevertheless, despite monocytes are essential mediators of inflammation and may be critical drivers of cIA/I, there is ongoing debate about monocytes as an HIV-1 reservoir, and little is known about their immunometabolism and their role in contributing to cIA/I, particularly in INR. Given that cIA/I are pivotal in the pathogenesis of serious non-AIDS events, leading to increased prevalence of morbility and mortality among INR, it is crucial to comprehend the factors behind cIA/I.

Added value of this study

In this study, we explore the role of monocytes as an HIV-1 reservoir, their immunometabolic features, and their implication in cIA/I, focusing on INR. Despite their short half-life, we found HIV-DNA and -RNA in circulating monocytes from more than half of the participants, and intact proviruses in 67% of them, with higher concentrations in INR compared to IR. By uncovering associations between monocyte HIV-1 reservoir and immunometabolism with monocyte activation, systemic inflammation, and immune recovery, this research provides an insight into the pathogenesis of HIV-1 and a step forward in the understanding of the INR condition.

Implications of all the available evidence

Our study agreed with the available evidence suggesting that circulating monocytes constitute an HIV-1 reservoir with ongoing transcriptional activity despite suppressive antiretroviral treatment, revealing an increase in INR compared to IR and its associations with systemic inflammation, metabolic abnormalities, and immune recovery. Further investigation aiming at developing new therapeutic strategies is required for normalizing cIA/I and decreasing the higher incidence on non-AIDS events in INR.

Introduction

Antiretroviral treatment (ART) reduces chronic immune activation and inflammation (cIA/I) inherent to HIV-1 infection; however, ART fails to completely normalize cIA/I despite persistent suppression of viremia in many people living with HIV-1 (PHIV). cIA/I constitute a critical factor in the pathogenesis of serious non-AIDS events (SNAEs) in PHIV. SNAES are more prevalent in virologically suppressed PHIV with poor immunologic restoration (15–30%), known as immunological non-responders (INR).1 SNAEs are associated with a multifactorial cIA/I, whose precise underlying mechanisms remain incompletely understood.2,3

Monocytes are essential mediators of inflammation and may be critical drivers of cIA/I and co-morbidities.4 This fact could be mediated by HIV-1 infection of monocytes themselves. Frequency of detection of cell-associated HIV-1-DNA (HIV-DNA) and replication-competent virus in circulating monocytes after stimulation have been variable.5, 6, 7, 8, 9, 10 Therefore, the role of circulating monocytes as reservoirs of HIV-1 remains controversial, particularly in PHIV on suppressive ART.

HIV-1 infection and replication in CD4+ T cells depend on increased glucose uptake through glucose transporter 1 (GLUT1) and increased glucose metabolism through aerobic glycolysis to lactate. Moreover, HIV-1 itself contributes to mitochondrial dysfunction, leading to cellular exhaustion, senescence, and apoptosis, even in PHIV on suppressive ART.11,12 Persistence of HIV-1 reservoir and immunometabolic abnormalities have been proposed as potential factors triggering cIA/I, described mainly in CD4+ T cells.13 This study aimed to assess the lesser known role of monocytes as a relevant HIV-1 reservoir, their immunometabolic features, and their contribution to cIA/I in PHIV, specifically in INR.

Methods

Study design and participants

This cross-sectional study was carried out at Virgen del Rocio University Hospital in Seville, Spain, from. The recruitment period was from March 2021 to March 2022. We included treatment-naïve (naïve) and treatment-experienced PHIV with undetectable viremia. The latter started ART with a CD4+ T-cell count <250 cells/μl and maintained an undetectable viremia in more than 95% of the determinations (punctual blips were admitted). For the purpose of this study, participants were classified as IR if, after three years of follow-up, they experienced an increase of CD4+ T-cell count >350 cells/μl. By contrast, participants were classified as INR if the CD4+ T-cell count increase was <200 cells/μl. A healthy control (HC) group was also enrolled. Exclusion criteria included opportunistic infections in the previous 36 months, active coinfections with hepatitis B or C virus, liver cirrhosis Child-Pugh class A or higher, malignancies, or drugs that might alter CD4+ T-cell count.

The primary endpoint was to evaluate differences in the HIV-1 monocyte reservoir between IR and INR, assessed by the presence of HIV-DNA and cell-associated HIV-1-RNA (HIV-RNA). Secondary endpoints were to characterize monocyte immunometabolism and the relationships with the higher cIA/I characteristics of INR. The sample size was estimated based on a study by us performed in peripheral blood mononuclear cells in which we found a 15% difference in HIV-DNA between PHIV with good and poor immune recovery. Thus, to achieve 80% power to detect differences in the contrast of the null hypothesis H0: μ1 = μ2 by means of a two-tailed Student t-test for two independent samples, considering that the significance level is 5%, it was necessary to include 33 subjects per group. As subjects were included and sampled on the same day, there was no dropouts. A control group of 25 treatment-naïve PHIV was also recruited.

Laboratory measurements

Sample collection

Peripheral blood mononuclear cells (PBMC) were isolated using BD Vacutainer® CPT™ tubes after centrifugation at 3000 rpm for 20 min. PBMC were stored with an equal volume of the ice-cold cryopreservation solution (RPMI and dimethyl sulfoxide 4:1) and autologous plasma as previously described.14 Plasma was obtained from blood samples collected into EDTA tubes after centrifugation at 3000 rpm for 20 min. Plasma was aliquoted into cryotubes and stored at −80 °C until subsequent assays to evaluate soluble markers of monocyte activation and plasma inflammation as described in the Supplementary material.

Cell isolation, immunophenotyping, and metabolic characterization

Monocyte and CD4+ T-cell isolation was carried out by fluorescence-activated cell sorting (FACS) using a BD FACS Aria Fusion (BD Biosciencies), described in Supplementary material. The purity of isolated populations were >99%, evaluated by flow cytometry analysis of the sorted cells (Supplementary Figure S1). Sorted monocytes were used for performing real-time cellular bioenergetics measurements in a Seahorse XF96 Extracellular Flux Analyzer (Agilent Technologies, Santa Clara, California, USA) as explained in the Supplementary material.

Cell samples were further labelled with flow cytometry antibodies. Part of the labelled PBMC were used for assessing monocyte immunophenotype in a Cytek™ Aurora spectral analyzer, and the remaining PBMC were washed for the subsequent staining with metabolic dyes, including l-lactate, 2-(N-(7-nitrobenz-2-oxa-1, 3-diazol-4-yl) amino)-2 deoxyglucose (2NBDG), GLUT1, MitoTracker Green (MTG) and Tetramethylrhodamine ethyl ester (TMRE) (Supplementary material). The gating strategy is shown in the Supplementary Figure S2.

Quantitation of HIV-DNA and HIV-RNA

Total HIV-DNA and -RNA were quantified by droplet digital PCR (ddPCR) using the BIO-RAD QX200 Droplet Reader (Hercules, California, USA). DNA and RNA were extracted from sorted monocytes using the E.Z.N.A.® DNA/RNA Isolation Kit (Catalog No.: R6731-01, Omega Bio-tek, Norcross, Georgia, USA). Total RNA was treated with DNAse I (Catalog No.: 2270A, Takara Bio Inc., Shiga, Japan) to eliminate DNA contamination. DNA and RNA concentrations were measured using the Qubit Assay kit (ThermoFisher Scientific) and brought to a 30 ng/μl concentration. The One-Step RT-ddPCR kit (Catalog No.: 1864021, BIORAD) was used for RNA reverse transcription and running the ddPCR. HIV-DNA and HIV-RNA were assessed as previously reported.15 The primers and probes for the viral 5′ LTR16 and gag17 regions are described in Supplementary material. The RPP30 and TBP housekeeping genes were quantified to normalize the input DNA and RNA samples, respectively. Copy numbers were calculated using Bio-Rad QuantaSoft software v.1.7.4. The lowest copy number detected was 52 for HIV-RNA and 87 for HIV-DNA.

Intact proviral DNA assay (IPDA)

Genomic DNA was extracted from FACS-isolated monocytes and CD4+ T cells using the E.Z.N.A.® DNA/RNA Isolation Kit (Catalog No.: R6731-01, Omega Bio-tek, Norcross, Georgia, USA). Intact, 5′deleted and 3′deleted HIV-1 proviruses were quantified by ddPCR using the intact proviral DNA assay (IPDA), as previously described.18 RPP30 was used as a housekeeping gene to normalize HIV-DNA copies. Intact HIV-1 copies were corrected for DNA shearing index based on RPP30 results. Data were analyzed using Bio-Rad QuantaSoft software version 1.7.4. The lowest copy number detected for intact, 5′ deleted, and 3′ deleted HIV-DNA was three copies.

Ethics statement

The study was designed and conducted according to the principles of the Declaration of Helsinki. It was approved by the Ethics Committee for Clinical Research of the Virgen Macarena and Virgen del Rocio University Hospitals, code 0170-N-21. All participants provided signed informed consent.

Statistical analysis

Results were expressed as median and interquartile ranges (IQR) for continuous variables and numbers and percentages for categorical variables. For comparisons between groups, the chi-square test and the Fisher exact test were used for binary variables, and the tests of Mann–Whitney U and Kruskal–Wallis with the Bonferroni-Dunn's correction for multiple comparisons were used to compare quantitative variables among groups. The Spearman rank correlation coefficients (ρ) were used to assess correlations between variables. To explore the relationship between statistically significant plasma inflamatory markers and time with undetectable viral load, we carried out a multiple linear regression adjusted for possible confounders, including age, sex, highest viral load before starting ART, basal or current CD4+ T-cell count, months with undetectable viral load, and number of blips in the last five years. The IBM software (SPSS v.26.0, Chicago, USA) was used for statistical analyses; p values < 0.05 were considered significant. Graphs were generated with GraphPad Prism Software, v. 9.0.0, and R Statistical (Foundation for Statistical Computing, Vienna, Austria).

Role of funders

This work was supported by the European Regional Development Fund, ISCIII, grant PI20/01646. Other funding sources were; Instituto de Salud Carlos III through the Subprogram Miguel Servet (CP19/00159) to AGV, PFIS contracts (FI19/00304) to EMM (FI21/00165) to ASA, and (FI19/00083) to CGC, and a mobility grant (MV21/00103) to EMM, from the Ministerio de Ciencia e Innovación, Spain. AJM was funded by a CSL Centenary Award. Funders had no role in study design, data collection, data analysis, interpretation, or writing of report.

Results

Study participants

One hundred PHIV and ten HC were enrolled in the study. Among PHIV, 25 were treatment-naïve and 75 ART-experienced subjects [35 INR and 40 IR]. The characteristics of the study participants are detailed in Table 1. As expected, there were differences between INR and IR in the CD4+ T-cell count (228 vs. 660 cells/μl, p < 0.001) and CD4+/CD8+ ratio (0.4 vs. 1.0; p < 0.001). Months on ART were similar in INR and IR; however, IR had undetectable viral load for a longer time.Table 1 Characteristics of the study participants.

	HC (n = 10)	IR (n = 40)	INR (n = 35)	Naïve (n = 25)	a p	b p	c p	
Male sex	5 (50)	34 (85)	32 (91.4)	21 (84)	0.010	0.732	0.489	
Age, years	41 (29–45)	48 (42–53)	54 (49–60)	38 (28–48)	<0.001	<0.001	0.010	
Weight, kg	70.0 (53.5–94.9)	76.4 (68.1–94.1)	76.0 (63.6–84.2)	68.0 (62.0–71.6)	0.094	0.030	0.381	
Nadir CD4+ T count, cells/μl	na	65 (14–163)	40 (16–64)	283 (211–565)	–	<0.001	0.146	
BMI, kg/m2	24.9 (20.2–28.7)	25.1 (22.8–30.9)	25.6 (23.3–28.3)	22.4 (20.7–23.8)	<0.001	0.005	0.345	
Zenit HIV-RNA, log10 copies/ml	na	5.4 (5.0–5.7)	5.4 (5.0–6.0)	5.3 (4.4–6.2)	–	0.900	0.753	
HIV-RNA 50–400 copies/ml, number in the last 5 years	na	0 (0–1)	0 (0–1)	na	–	0.101	0.809	
Subtype of HIV infection	na				–	0.840	0.699	
 B		37 (92.5)	31 (88.6)	23 (92.0)				
 07-BC		3 (7.5)	4 (11.4)	2 (8.0)				
Previous CDC stage C	na	13 (32.5)	18 (51.4)	0 (0)	–	<0.001	0.107	
HIV risk factor	na				–	0.013	0.276	
 MSM		19 (47.5)	12 (34.3)	18 (72)				
 Heterosexual		12 (30)	8 (22.9)	5 (20)				
 IVDU		7 (17.5)	13 (37.1)	0 (0)				
 Other		2 (5)	2 (5.7)	2 (8)				
Current ART regimen	na			na	–	–	0.052	
 InSTI-based		21 (53.8)	26 (81.3)					
 NNRTI-based		6 (15.4)	2 (6.3)					
 PI-based		12 (30.8)	4 (12.5)					
CD4+ T count, cells/μl	910 (551–1037)	660 (593–968)	228 (140–296)	283 (211–565)	<0.001	<0.001	<0.001	
CD4+ T percentage	43.8 (38.5–50.6)	37.0 (30.1–41.7)	18.1 (13.5–22.3)	19.8 (15.2–28.4)	<0.001	<0.001	<0.001	
CD8+ T count, cells/μl	501 (267–576)	769 (564–1037)	605 (390–801)	929 (562–1789)	0.001	0.006	0.015	
CD4+/CD8+ ratio	1.7 (1.1–2.1)	1.0 (0.8–1.2)	0.4 (0.3–0.4)	0.4 (0.2–0.6)	<0.001	<0.001	<0.001	
Months from diagnosis	na	174 (123–253)	223 (126–286)	na	–	–	0.303	
Months on treatment	na	160 (100–201)	162 (80–213)	na	–	–	0.521	
HIV-RNA <50 copies/ml, months	na	126 (83–189)	74 (52–119)	na	–	–	0.001	
Data are expressed as median (interquartile range) or n (%). HC, healthy controls; IR, immunological responders; INR, immunological non-responders; Naïve, treatment-naïve PHIV; MSM, men who have sex with men; IVDU, previous intravenous drug use; NNRTI, non-nucleoside reverse transcriptase inhibitors; InSTI, Integrase strand transfer inhibitor; PI, protease inhibitor; na, not applicable. p values stand for comparisons performed by Kruskal Wallis test with the Bonferroni-Dunn correction for multiple comparisons between all groups (a p) and between PHIV groups (IR, INR, treatment-naïve) (b p) or comparisons performed using Mann–Whitney U test between IR vs INR (c p).

Characterization of monocytes in PHIV

In PHIV, we observed an expansion of non-classical monocytes and an increase in the percentage of CCR5 expression in all monocyte subsets compared to HC. Moreover, monocytes from naïve expressed a higher CD38 geometric mean fluorescence intensity than those from ART-experienced participants in the three monocyte subsets (all p ≤ 0.001) (Supplementary Table S1); however, there were no significant differences in these parameters between IR and INR.

The plasma concentration of sCD14 was higher in naïve [3225.0 ng/ml (2342.0–4090.0)] than in the pooled HC, IR, and INR groups [2196.0 (1648.0–2759.0), p < 0.001], among which there were no significant differences. Furthermore, the concentrations of sCD163 were the highest in naïve [172.7 ng/ml (117.6–293.0)] followed by INR [111.4 (95.2–136.1), p = 0.005], IR [85.0 (57.2–131.4), p = 0.010 for INR vs. IR] and HC [75.6 (59.4–109.3), p = 0.662 for IR vs. HC] (Table 2).Table 2 Plasma biomarkers of monocyte activation and systemic inflammation.

	HC (n = 10)	IR (n = 40)	INR (n = 35)	Naïve (n = 25)	a p	b p	
sCD14, ng/ml	2228 (1674–2630)	2194 (1662–2935)	2185 (1606–2778)	3225 (2342–4090)	0.004	0.907	
sCD163, ng/ml	75.6 (59.4–109.3)	84.7 (57.2–131.4)	111.4 (95.2–136.1)	172.7 (117.6–293.0)	<0.001	0.010	
hs-CRP, mg/l	1.1 (0.3–1.4)	1.3 (0.7–1.9)	1.4 (0.74.4)	0.7 (0.4–2.0)	0.065	0.248	
D-dimers, μg/l	175.0 (145.0–222.5)	200.0 (100.0–270.0)	190.0 (150.0–395.0)	385.0 (210.0–657.5)	0.003	0.280	
β2-m, mg/l	1.3 (1.1–1.5)	1.8 (1.5–2.2)	1.9 (1.4–2.6)	3.1 (2.2–4.1)	<0.001	0.293	
IL-6, pg/ml	0.5 (0.2–4.5)	1.9 0.9–4.0)	3.9 (1.1–5.9)	2.1 (0.9–4.5)	0.129	0.127	
IL-8, pg/ml	2.5 (2.0–3.7)	2.8 (1.7–4.9)	2.7 (0.9–4.9)	2.0 (1.0–3.0)	0.228	0.496	
TNF-α, pg/ml	16.1 (10.5–36.2)	29.6 (17.8–50.4)	25.9 (18.6–46.0)	34.4 (25.5–61.8)	0.054	0.764	
IFN-γ, pg/ml	2.6 (0.5–7.8)	3.7 (0.9–7.5)	1.5 (0.1–4.9)	3.7 (0.7–6.6)	0.196	0.069	
IP-10, pg/ml	163.6 (143.2–216.6)	240.2 (141.1–354.1)	306.6 (191.5–539.9)	796.0 (504.8–1175.3)	<0.001	0.065	
MIP-1α, pg/ml	37.9 (30.6–65.0)	33.4 (20.0–50.4)	34.8 (23.7–51.3)	39.9 (31.8–61.0)	0.103	0.308	
MIP-1β, pg/ml	29.2 (20.3–40.5)	29.4 (24.6–42.0)	29.9 (18.3–37.7)	22.1 (18.3–26.5)	0.019	0.453	
Data are expressed as median (interquartile range). HC, healthy controls; IR, immunological responders; INR, immunological non-responders; Naïve, treatment-naïve PHIV. p values stand for comparisons performed by Kruskal Wallis test with the Bonferroni-Dunn correction for multiple comparisons between all groups (a p) or comparisons performed using Mann–Whitney U test for IR vs INR (b p).

Plasma inflammatory markers

Systemic inflammation was evaluated by measuring the plasma concentrations of hs-CRP, D-dimers, β2-m, IL-6, IL-8, TNF-α, IFN-γ, IP-10, MIP-1α and MIP-1β whose results are shown in Table 2. All participants showed similar concentrations of hs-CRP, IL-8, IFN-γ, MIP-1α, and MIP-1β. However, higher levels of D-dimers (p = 0.001), β2-m (p < 0.001), IP-10 (p < 0.001), and MIP-1α (p = 0.034) was found in plasma from naïve compared to treatment-experienced participants. INR had higher plasma concentrations of β2-m (p < 0.001) and IP-10 (p = 0.014) than HC and higher plasma levels of IP-10 compared to IR (p = 0.065). IL-6 levels in INR were twice as high compared to naïve and IR, although statistical significance was reached only when compared to HC (p = 0.048). IL-6 concentrations were weakly correlated with the percentage of non-classical monocytes (ρ = 0.230, p = 0.033) (Supplementary Figure S3a). Furthermore, several inflammatory markers correlated with each other, as shown in Supplementary Figure S3b. A multiple linear regression analysis adjusted for possible confounders, including age, sex, highest viral load before starting ART, basal or current CD4+ T-cell count, months with undetectable viral load, and number of blips in the last five years, showed that none of these parameters were associated with IP-10 or sCD163 (data not shown).

Cellular-associated HIV-1-RNA and -DNA in monocytes

HIV-RNA was quantified in monocytes and detected in a similar frequency of participants in every PHIV group: naïve (14/21; 66.7%), INR (17/33; 51.5%), and IR (15/28; 53.6%), p = 0.520. There were no differences in the clinical characteristics between monocytes with and without detectable HIV-RNA. However, the HIV-RNA concentrations were quite different between groups with the highest values in naïve [2545 (351–4482) copies/106 TBP] followed by INR [240 (148–610), p = 0.015 for naïve vs. INR], and IR [145 (79–309)], p = 0.086 for INR vs. IR (Fig. 1a). The levels of HIV-RNA correlated weakly with the IP-10 levels (ρ = 0.305, p = 0.044) and monocyte activation (sCD163, ρ = 0.280, p = 0.059), and inversely with the CD4+/CD8+ ratio (ρ = −0.289, p = 0.054) and months of treatment (ρ = −0.298, p = 0.062) (Supplementary Figure S4a).Fig. 1 HIV-1 reservoir. (a) Monocyte HIV-RNA in IR (n = 28), INR (n = 33) and N (n = 21). (b) Total HIV-DNA in CD4+ T cells (empty shapes) and monocytes (full shapes), in IR (n = 13), INR (n = 14) and N (n = 12). (c) Genome integrity of individual proviruses from monocytes in IR (n = 13), INR (n = 14) and N (n = 12). Data are expressed as median and interquartile range. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001. Comparisons among groups (two by two) were performed using Mann–Whitney U test. IR, immunological responders; INR, immunological non-responders; N, treatment-naïve PHIV.

In contrast, ddPCR targeting the 5′ LTR and gag regions only detected HIV-DNA in monocytes from 4 out of 36 IR, 1 out of 35 INR, and none of naïve. Searching for an explanation for the discordance between HIV-RNA and -DNA, we performed a deeper study by a more sensitive technique, IPDA.

Intact proviral DNA assay in monocytes

IPDA was conducted in CD4+ T cells and monocytes from 39 participants with available samples (naïve, 12; IR, 13; and INR, 14) in whom HIV-RNA was present; HIV-DNA was detected in all these samples. The levels of total HIV-DNA in CD4+ T cells were 2418 copies/106 cells (961–5011) in naïve, 2653 (773–3821) in INR (p = 0.637), and 996 (637–2152) in IR (p = 0.042 for INR vs. IR). In monocytes, the total HIV-DNA concentrations were much lower than in CD4+ T cells, with no significant differences between naïve [99 copies/106 cells (66–123)] and INR [87 (54–136)], p = 0.797. However, IR showed lower levels [40 (20–117), p = 0.019] compared to naïve and INR (Fig. 1b). HIV-DNA levels correlated with the months on ART (ρ = −0.380, p = 0.029) (Supplementary Figure S4a).

After analyzing the genome integrity of HIV-DNA in monocytes, intact provirus was detected in similar proportions in naïve [9/12 (75%)], INR [8/14 (57.1%)], and IR [9/13 (69.2%)], p = 0.611. However, INR had higher concentrations [46 copies/106 monocytes (18–67)] than naïve [13 (7–44)] and IR [10 (7–24)], p = 0.037. There were no significant differences between the last two groups (p = 0.368). Regarding defective provirus, all participants had 5′ deleted HIV-DNA, with naïve and INR showing similar levels [55 copies/106 monocytes (37–105) vs. 56 (16–92), respectively, p = 0.328] while IR exhibited the lowest values [14 (6–43), p = 0.006]. Conversely, 3′ deleted HIV-DNA were present in 7/12 (58.3%) naïve, 10/14 (71.4%) INR, and 7/13 (53.8%) IR, p = 0.620. Their levels were similar between naïve [20 copies/106 monocytes (6–31)] and INR [21 (7–30)], p = 0.845, and, once again, tend to be higher than in IR [11 (7–23)] although the difference did not reach statistical significance, p = 0.120 (Fig. 1c). The ratio between intact and total provirus was similar in naïve [15% (1–46)], INR [13% (0–49%)], and IR [32% (0–56)], p = 0.719. However, the estimated HIV-1 transcriptional activity measured as the HIV-RNA/DNA ratio was higher in naïve [9.94 (4.4–45.2)] than in INR [3.3 (1.9–8.5), p = 0.043], and IR [2.2 (0.9–7.8), p = 0.017], but similar between INR and IR, p = 0.380 (Supplementary Figure S4b).

The levels of intact provirus in monocytes were positively associated with the frequency of inflammatory monocytes (CD16+) (ρ = 0.414, p = 0.017) and total HIV-DNA in CD4+ T cells (ρ = 0.464, p = 0.022), and negatively with the percentage of the classical subset (ρ = −0.386, p = 0.027). Similarly, total and defective provirus were negatively associated with the CD4+/CD8+ ratio [(ρ = −0.414, p = 0.012) and (ρ = −0.340, p = 0.040)] and months on ART [(ρ = −0.380, p = 0.029) and (ρ = −0.399, p = 0.021)]. However, there were no correlations between the concentrations of intact or defective provirus with the different plasma inflammatory markers or sCD163 (Supplementary Figure S4a).

Monocyte immunometabolism

The immunometabolism of monocytes was evaluated in a subgroup of participants, including 10 naïve, 8 INR, 7 IR, and 9 HC.

Glucose metabolism

The main parameters of the monocyte glucose metabolism are shown in Fig. 2. Naïve expressed more frequency of GLUT1high in all monocyte subsets compared to HC, IR, and INR (all p < 0.022). In classical and intermediate monocytes there were no differences between INR, IR, and HC; however, in the non-classical subset, INR expressed higher frequency of GLUT1high than HC (p = 0.043) (Fig. 2a).Fig. 2 Monocyte glucose metabolism. Metabolic determinations in classical, intermediate and non-classical monocyte subsets of (a) GLUT1high expression and (b)l-lactate/2NBDG ratio, in HC (n = 9), IR (n = 7), INR (n = 8) and N (n = 10). Data are expressed as median and interquartile range. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001. Comparisons among groups (two by two) were performed using Mann–Whitney U test. HC, healthy controls; IR, immunological responders; INR, immunological non-responders; N, treatment-naïve PHIV.

In total monocytes and in the intermediate subset, GLUT1high expression was correlated with HIV-RNA (ρ = 0.497, p = 0.070 and ρ = 0.626, p = 0.022, respectively) (Supplementary Figure S5a). Other associations were found between GLUT1high expression in total monocytes and the plasma levels of IP-10 (ρ = 0.363, p = 0.041), β2-m (ρ = 0.387, p = 0.024), and CD38 expression (ρ = 0.353, p = 0.051), and inverse with the CD4+/CD8+ ratio (ρ = −0.478, p = 0.005). GLUT1high expression was also associated with glucose uptake (ρ = 0.475, p = 0.005), which tended to be higher in naïve and INR [63.2% (80.3–46.4) and 64.5 (44.6–88.0), respectively] than IR [51.0 (31.7–73.7)], although without reaching statistical significance (p = 0.250) probably due to the low number of IR analyzed. However, naïve and INR displayed a lower l-lactate/2NBDG ratio than in IR and HC (Fig. 2b), which was inversely correlated with the expression of GLUT1high (ρ = −0.488, p = 0.003) (Supplementary Figure S5c), total (ρ = −0.516, p = 0.071) and defective (ρ = −0.516, p = 0.071) provirus (Supplementary Figure S5d), and positively with CD4+ T-cell count (ρ = 0.476, p = 0.005) (Supplementary Figure S5b).

Mitochondrial function

Mitochondrial metabolism was assessed by i) the percentage of dysfunctional mitochondria (MTG+TMRE–) in monocyte subsets and ii) the real-time cellular bioenergetics measured by the coupling efficiency percentage (proportion of the mitochondria oxygen consumed for synthesizing ATP), the bioenergetic health index (BHI: a marker of mitochondrial function), basal oxygen consumption rate (OCR), and basal extracellular acidification rate (ECAR), in purified total monocytes.

Total monocytes from INR [18.7% (3.0–46.3)] and naïve [14.6 (3.5–27.8)] exhibited a higher percentage of dysfunctional mitochondria than those from IR [7.6 (4.1–14.2)] and HC [3.0 (1.5–9.8), p = 0.040] (Fig. 3a). A negative correlation was observed between mitochondrial dysfunction and the l-lactate/2NBDG ratio in total monocytes (ρ = −0.639, p < 0.001), and in every subset [classical (ρ = −0.637, p < 0.001), intermediate (ρ = −0.702, p < 0.001) and non-classical monocytes (ρ = −0.464, p = 0.006)] (Supplementary Figure S5e).Fig. 3 Monocyte mitochondrial function. (a) Metabolic determinations in classical, intermediate and non-classical monocyte subsets of dysfunctional mitochondria (MTG+ TMRE−) in HC (n = 9), IR (n = 7), INR (n = 8) and N (n = 10). Real-time quantification in FACS-isolated monocytes of the cellular bioenergetics parameters (b) coupling efficiency and (c) BHI, in HC (n = 8), IR (n = 6), INR (n = 3) and N (n = 2). Data are expressed as median and interquartile range. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001. Comparisons among groups (two by two) were performed using Mann–Whitney U test. HC, healthy controls; IR, immunological responders; INR, immunological non-responders; N, treatment-naïve PHIV; MTG, MitoTracker Green; TMRE, tetramethylrhodamine ethyl ester; BHI, bioenergetic health index.

Regarding cell bioenergetics, monocytes from naïve showed the worst coupling efficiency and BHI [29% (range, 23–35) and 1.1 (0.5–1.63)], followed by INR [45% (33–55), and 3.5 (0.4–5.2), p = 0.249 and p = 0.564 for naïve vs. INR, respectively] compared to IR [60% (45–62) and 6.2 (1.8–8.0), p = 0.039 and p = 0.071 for INR vs. IR, respectively], and HC [55% (range, 40–70) and 6.1 (1.4–15.3), p = 0.699 and 0.897, respectively, for IR vs. HC] (Fig. 3b and c). No differences were found in basal OXPHOS (OCR) or glycolysis (ECAR) between groups (Supplementary Figure S6).

There were inverse associations between mitochondrial dysfunction and coupling efficiency (ρ = −0.442, p = 0.051), and BHI (ρ = −0.513, p = 0.021) (Supplementary Figure S5e), reinforcing the hypothesis of an impaired mitochondrial function in INR. Furthermore, coupling efficiency was inversely correlated with the monocyte activation marker sCD163 (ρ = −0.463, p = 0.040) (Supplementary Figure S5b).

Discussion

This study supports the hypothesis of an incomplete suppression of viral replication by ART and that monocyte may act, at least, as a transient HIV-1 reservoir, given the short lifespan of these cells, with higher HIV-DNA and -RNA levels in INR compared to IR. Monocyte HIV-RNA, rather than HIV-DNA, was associated with monocyte activation and inflammation and, inversely, with the CD4+/CD8+ ratio. Furthermore, monocytes of INR exhibit glucose metabolism alterations and an increased mitochondrial dysfunction compared to IR.

Our results are in agreement with previous studies showing an expansion of non-classical monocytes in PHIV compared to HC,19 and an increased percentage of CCR5 expression in all monocyte subsets. Although there was no difference between INR and IR in the surface markers analyzed, monocytes from INR showed more activation evidenced by higher sCD163 and IP-10 plasma concentrations than IR, which in turn inversely correlated with the CD4+/CD8+ ratio and months of treatment.

While there is a consensus that monocyte-derived macrophages constitute a true HIV-1 reservoir, the suggestion of monocytes as a reservoir for HIV-1 has been controversial, given their short lifespan.5 Moreover, while several studies evaluated the presence of HIV-1 in monocytes,20 many of them were unable to ensure the absence of CD4+ T-cell contamination, and just a few demonstrated integrated DNA and production of replication-competent HIV-1 after stimulation, even after years of undetectable viremia on ART.7,8,21, 22, 23, 24 Recently, Massanella et al.6 found HIV-1 DNA by nested real-time PCR in 27% and 33% monocytes from 29 subjects before and after one year of ART with median levels of integrated HIV-1 DNA of 35 copies/106 cells (4–84) and 18 copies/106 cells (8–174) after correcting for CD4+ T-cell contamination. Using ddPCR targeting the 5′ LTR and gag regions, we found HIV-RNA in monocytes from more than half of PHIV, both naïve and with undetectable viremia, with higher levels in INR vs. IR. However, this assay only allowed HIV-DNA detection in 5 out of 85 subjects. At the same time, by using IPDA as a more sensitive approach, we found HIV-DNA in monocytes from all subjects positive for HIV-RNA. Intact proviruses were present in 75% of naïve and around 60% of IR and INR, while all subjects contained defective proviruses. Our results are in line with those of Veenhuis et al., who have been the only ones using IPDA to analyze HIV-DNA in monocytes from 30 virally suppressed PHIV.5 They found intact provirus in 40% and defective provirus in 90% of them. Likewise, they observed that intact proviruses were replication-competent after monocyte differentiation in a quantitative viral outgrowth assay. Only 5% of proviruses within CD4+ T cells are believed to be intact and, therefore, replication-competent.25 Our data reveal that this ratio may be significantly higher in circulating monocytes, roughly 18%. Nevertheless, HIV-DNA levels were lower in monocytes compared to CD4+ T cells, which could be partly due to reduced expression of CD4 and CCR5 and a higher effect of host restriction mechanisms limiting HIV-1 infection, such as SAMHD1 and APOBEC3.20,26 HIV-RNA, intact, and defective provirus levels were higher in INR than IR and negatively associated with the CD4+/CD8+ ratio and months on ART. However, it is worth noting that only HIV-RNA was associated with sCD163 and IP-10 levels. Remarkably, intact provirus was related to the frequency of inflammatory monocytes (CD16+), consistent with a higher susceptibility of CD16+ monocytes to HIV-1 infection.27 In addition, as monocytes could host HIV-1 variants that differ genetically from those found in CD4+ T cells24 and traffic to tissues, they may play a crucial role in increasing reservoir levels and promoting genetic diversity within viral sanctuaries, significantly impacting in the HIV-1 pathogenesis.

Similar to previous findings,28 we have found the maximum frequency of total monocytes expressing GLUT1high in naïve. At the same time, INR exhibited higher frequency of GLUT1high than IR and HC in the non-classical subset, without differences between the last two groups. GLUT1high expression was correlated with glucose uptake, plasma levels of IP-10, β2-m, CD38 expression, and HIV-RNA levels, and inversely with the CD4+/CD8+ ratio. These findings highlight the GLUT1-mediated metabolic pathway in monocytes as a critical mediator of inflammation, which may have a role in HIV-1 infectivity and replication.29 The higher glucose uptake with a less l-lactate production observed in INR compared to IR could be a compensatory response, potentially triggered by the increased prevalence of dysfunctional mitochondria and a notable decline in the coupling efficiency and BHI. Noteworthy, coupling efficiency correlated with sCD163 and mitochondrial dysfunction; the latter was associated with CD4+/CD8+ ratio, relating mitochondrial damage to monocyte activation and immune recovery. Mitochondrial impairment in CD4+ T cells from ART-experienced PHIV30 has been described as a result of an excessive glucose dependency known as metabolic reprogramming.31 However, our data does not fully support this idea in monocytes as we observed no disparity in basal OXPHOS nor glycolysis in PHIV. This discrepancy is likely due to the lack of cell activation in our assays, which were focused on characterizing steady state features. Indeed, Zhao et al.30 also reported no differences in OCR or ECAR in unstimulated CD4+ T cells.

Our study has limitations, such as not analyzing whether HIV-DNA is integrated and replicative. Another limitation is the small number of people analyzed for cell bioenergetics due to the technical complexity of these assays. On the other hand, GLUT1 expression was assessed on the cell surface by flow cytometry, and additional quantification of intracellular GLUT1 could have strengthened the differences among the groups.

In conclusion, these data support the hypothesis that monocytes harbor HIV-DNA and -RNA constituting at least a transient HIV-1 reservoir, higher in virologically suppressed INR than IR, associated with monocyte activation, inflammation, GLUT1high expression, glucose uptake, and inversely with the CD4+/CD8+ ratio. The HIV-1 reservoir found in monocytes from INR and their altered metabolism may be considered factors associated with the higher cIA/I that characterize these subjects.

Contributors

AGV contributed to funding acquisition; EMM, AGV and LFLC conceptualized and designed the study; LFLC contributed to the recruitment and management of participants; ASG collected participants samples; EMM, MTR, CGC and MKSL contributed to the optimization of the experiments; EMM, MTR, ASG, ASA and AIAR performed the experiments; EMM contributed to data analysis and interpretation; EMM and MTR drafted the original and final manuscript; LFLC, AGV, ERM, DS, and AJM contributed to the review and editing the manuscript; EMM and MTR verified the underlying data. All authors reviewed and approved the final version of this manuscript.

Data sharing statement

Data are available from the corresponding authors on reasonable request.

Declaration of interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A Supplementary data

Supplementary file

Acknowledgements

This study would not have been possible without the collaboration of participants and nursing staff.

Appendix A Supplementary data related to this article can be found at https://doi.org/10.1016/j.ebiom.2024.105338.
==== Refs
References

1 Yan L. Xu K. Xiao Q. Cellular and molecular insights into incomplete immune recovery in HIV/AIDS patients Front Immunol 14 2023 1 15
2 Lv T. Cao W. Li T. HIV-related immune activation and inflammation: current understanding and strategies J Immunol Res 2021 2021 1 13
3 Pacheco Y.M. Jarrin I. Rosado I. Increased risk of non-AIDS-related events in HIV subjects with persistent low CD4 counts despite cART in the CoRIS cohort Antiviral Res 117 2015 69 74 25766861
4 Bai R. Li Z. Lv S. Persistent inflammation and non-AIDS comorbidities during ART: coming of the age of monocytes Front Immunol 13 2022
5 Veenhuis R.T. Abreu C.M. Costa P.A.G. Monocyte-derived macrophages contain persistent latent HIV reservoirs Nat Microbiol 8 5 2023 833 844 36973419
6 Massanella M. Bakeman W. Sithinamsuwan P. Infrequent HIV infection of circulating monocytes during antiretroviral therapy J Virol 94 1 2019 1 14
7 Zhu T. Muthui D. Holte S. Evidence for human immunodeficiency virus type 1 replication in vivo in CD14 + monocytes and its potential role as a source of virus in patients on highly active antiretroviral therapy J Virol 76 2 2002 707 716 11752161
8 Calcaterra S. Cappiello G. Di Caro A. Garbuglia A.R. Benedetto A. Comparative analysis of total and integrated HIV-1 DNA in peripheral CD4 lymphocytes and monocytes after long treatment with HAART J Infect 43 4 2001 239 245 11869061
9 Crowe S. Zhu T. Muller W.A. The contribution of monocyte infection and trafficking to viral persistence, and maintenance of the viral reservoir in HIV infection J Leukoc Biol 74 5 2003 635 641 12960232
10 Cattin A. Wiche Salinas T.R. Gosselin A. HIV-1 is rarely detected in blood and colon myeloid cells during viral-suppressive antiretroviral therapy AIDS 33 8 2019 1293 1306 30870200
11 Blagov A.V. Sukhorukov V.N. Guo S. Zhang D. Popov M.A. Orekhov A.N. Impaired mitochondrial function in T-lymphocytes as a result of exposure to HIV and ART Cells 12 7 2023 1072 37048145
12 Loisel-Meyer S. Swainson L. Craveiro M. Glut1-mediated glucose transport regulates HIV infection Proc Natl Acad Sci U S A 109 7 2012 2549 2554 22308487
13 Teer E. Mukonowenzou N.C. Essop M.F. The role of immunometabolism in HIV-1 pathogenicity: links to immune cell responses Viruses 14 8 2022 1 8
14 Masson J.J.R. Ostrowski M. Duette G. The multiparametric analysis of mitochondrial dynamics in T cells from cryopreserved peripheral blood mononuclear cells (PBMCs) Methods mol biol 2020 215 224
15 Clarridge K.E. Blazkova J. Einkauf K. Effect of analytical treatment interruption and reinitiation of antiretroviral therapy on HIV reservoirs and immunologic parameters in infected individuals PLoS Pathog 14 1 2018 e1006792
16 Ishizaka A. Sato H. Nakamura H. Short intracellular HIV-1 transcripts as biomarkers of residual immune activation in patients on antiretroviral therapy J Virol 90 12 2016 5665 5676 27030274
17 Palmer S. Wiegand A.P. Maldarelli F. New real-time reverse transcriptase-initiated PCR assay with single-copy sensitivity for human immunodeficiency virus type 1 RNA in plasma J Clin Microbiol 41 10 2003 4531 4536 14532178
18 Bruner K.M. Wang Z. Simonetti F.R. A quantitative approach for measuring the reservoir of latent HIV-1 proviruses Nature 566 7742 2019 120 125 30700913
19 Campbell J.H. Hearps A.C. Martin G.E. Williams K.C. Crowe S.M. The importance of monocytes and macrophages in HIV pathogenesis, treatment, and cure AIDS 28 15 2014 2175 2187 25144219
20 Wong M.E. Jaworowski A. Hearps A.C. The HIV reservoir in monocytes and macrophages Front Immunol 10 2019 1435 31297114
21 Lambotte O. Taoufik Y. de Goër M.G. Wallon C. Goujard C. Delfraissy J.F. Detection of infectious HIV in circulating monocytes from patients on prolonged highly active antiretroviral therapy J Acquir Immune Defic Syndr 23 2 2000 114 119 10737425
22 Sonza S. Mutimer H.P. Oelrichs R. Monocytes harbour replication-competent, non-latent HIV-1 in patients on highly active antiretroviral therapy AIDS 15 1 2001 17 22 11192864
23 Cui J. Meshesha M. Churgulia N. Replication-competent HIV-1 in human alveolar macrophages and monocytes despite nucleotide pools with elevated dUTP Retrovirology 19 1 2022 21 36114511
24 Xu Y. Zhu H. Wilcox C.K. Blood monocytes harbor HIV type 1 strains with diversified phenotypes including macrophage-specific CCR5 virus J Infect Dis 197 2 2008 309 318 18173363
25 Bruner K.M. Murray A.J. Pollack R.A. Defective proviruses rapidly accumulate during acute HIV-1 infection Nat Med 22 9 2016 1043 1049 27500724
26 Chitrakar A. Sanz M. Maggirwar S.B. Soriano-Sarabia N. HIV latency in myeloid cells: challenges for a cure Pathogens 11 6 2022 611 35745465
27 Ellery P.J. Tippett E. Chiu Y.L. The CD16+ monocyte subset is more permissive to infection and preferentially harbors HIV-1 in vivo J Immunol 178 10 2007 6581 6589 17475889
28 Palmer C.S. Anzinger J.J. Zhou J. Glucose transporter 1–expressing proinflammatory monocytes are elevated in combination antiretroviral therapy–treated and untreated HIV+ subjects J Immunol 193 11 2014 5595 5603 25367121
29 Sáez-Cirión A. Sereti I. Immunometabolism and HIV-1 pathogenesis: food for thought Nat Rev Immunol 21 1 2021 5 19 32764670
30 Zhao J. Schank M. Wang L. Mitochondrial functions are compromised in CD4 T cells from ART-controlled PLHIV Front Immunol 12 2021 1 15
31 Loucif H. Dagenais-Lussier X. Beji C. Telittchenko R. Routy J.P. van Grevenynghe J. Plasticity in T-cell mitochondrial metabolism: a necessary peacekeeper during the troubled times of persistent HIV-1 infection Cytokine Growth Factor Rev 55 2020 26 36 32151523
