Study participants

A retrospective and longitudinal study was carried out using peripheral blood mononuclear cells (PBMCs) and plasma samples from 46 PWH under suppressive ART and confirmed SARS-CoV-2 infection (previous SARS-CoV-2 anti-IgM/IgG+) stored in the Spanish HIV Hospital Universitario Gregorio Marañón BioBank (HIV HUGM BioBank). Samples were selected and analysed based on availability at three time-points: HIV-1/pre-SARS-CoV-2 infection time-point, called the HIV/pre-SCV2 group (median of 11.5 months prior to co-infection samples, n = 18); HIV-1/SARS-CoV-2 co-infection time-point (active or recently resolved SARS-CoV-2 infection), called the HIV/SCV2 group (co-infection time, n = 46) and HIV-1/post-SARS-CoV-2 infection time-point, called the HIV/post-SCV2 group (median of 12.5 months after co-infection samples, n = 36). Inclusion in the study required data for at least two of these time-points. Longitudinal PBMCs and plasma’ samples, as well as clinical data from PWH were provided by the HIV-HGM BioBank and their medical history from participant hospitals’ internal medicine departments, respectively. PBMCs were stored at -170º C, while plasma was stored at -20ºC, until use.

All PWH participating in the study provided their informed consent, and protocols were approved by the Ethics Committee of Hospital General Universitario Gregorio Marañón (HGUGM) (14th April, 2020. Minutes: 11/2020).

Laboratory determinations

Plasma levels of anti-S-IgG-SARS-CoV-2 and anti-N-IgG-SARS-CoV-2 were measured at the HIV/SCV2 and HIV/post-SCV2 time-points. High-sensitivity C-reactive protein (CRP), and ferritin were measured in PWH at the HIV/pre-SCV2, HIV/SCV2 and HIV/post-SCV2 time-points. All measurements were performed using Abbott anti-SARS-CoV-2 IgG assays in the Alinity-i autoanalyzer and ADVIA Chemistry XPT (Siemens Healthineers), respectively.

Cytokine and chemokine quantification

Up to 24 pro-inflammatory (IFNα2, IFNγ, IL-1β, IL-33, IL-6, IL-23, IL-8, IL-12p70, IL-18, IL-17A, TNF-α, MCP-1, Eotaxin, IP-10, IL-8, MIG, MIP-1α, MIP-1β, RANTES, TARC, ENA-78, MIP-3α, GROα and I-TAC) and anti-inflammatory (IL-10) soluble cytokines and chemokines were quantified in plasma from PWH at the HIV/pre-SCV2, HIV/SCV2 and HIV/post-SCV2 time-points. Inflammation Panel 1 and Proinflammatory Chemokine Panel 1 from LEGENDplex immunoassays and software (BioLegend) were used according to the manufacturer’s instructions. The analysis was performed on a CytoFlex S flow cytometer (Beckman Coulter).

Immunophenotyping assays

Immunophenotyping of Natural Killer (NK) cells, Dendritic cells (DCs), monocytes and T cells was performed using a previously described multiparametric flow cytometry protocol [9, 20]. Briefly, thawed and washed PBMCs were stained at the HIV/pre-SCV2, HIV/SCV2, and HIV/post-SCV2 time-points with surface markers for cell viability; lineage (CD3, CD14, CD19, CD16, CD56), maturation (CD57, TIM3), activation (CD158b), and C-type lectin-like inhibiting and activating receptors (NKG2A and NKG2C, NKG2D, respectively) for NK cells. The distribution of monocyte subpopulations was studied including markers for cell viability and lineage (CD3, CD8, HLA-DR, CD14, CD16). Similarly, PBMCs were stained with markers for cell viability, lineage (Lin2, HLA-DR, CD123, CD11c), maturation (TIM3, PDL1), and activation (CD141, CD40, CD80) for DCs. PBMCs were also stained with markers for cell viability, lineage (CD3, CD4 and/or CD8), maturation (CD45RA, CD27), activation (HLA-DR, CD137, CD69), senescence (CD57) and exhaustion (TIM3, PD1) for T cells. Isotype controls were included for TIM3, CD158b, NKG2A, NKG2C, NKG2D for NK cells; TIM3, PDL1, CD141, CD40, CD80 for DCs; and CD137, CD69, TIM3, PD1 for T cells. All details for antibody panels are summarized in Supplementary Table 1.

Viable NK cells were classified into four subsets according to the expression of CD56 and CD16 (CD56neg; CD56dim, which includes CD16high subset; and CD56high). The gating strategy for NK cells is shown in Supplementary Fig. 1A. Viable monocytes were defined in three subsets according to the expression of CD14 and CD16: classical (CD16neg CD14high), intermediate (CD16dim CD14high), and patrolling (CD16high CD14dim). A representative gating strategy for monocytes is shown in Supplementary Fig. 1B. Viable DCs were characterised into two subpopulations: myeloid dendritic cells (mDCs) defined as HLADR + CD11c + CD123- and plasmacytoid dendritic cells (pDCs) defined as HLADR + CD11c-CD123+. A schematic gating strategy for DCs can be found in Supplementary Fig. 1C. T-cell memory subsets were classified based on the expression of CD45RA and CD27 as naïve (CD45RA + CD27+), central memory (CM; CD45RA − CD27+), effector memory (EM; CD45RA − CD27−) and terminally differentiated (TemRA; CD45RA + CD27−). A representative gating strategy for T-cells can be found in Supplementary Fig. 1D.

HIV-1 and SARS-CoV-2 specific T-cell response analysis

To analyse the specific and longitudinal T cell response to HIV-1 (HIV/pre-SCV2, HIV/SCV2, and HIV/post-SCV2 time-points) and SARS-CoV-2 (HIV/SCV2 and HIV/post-SCV2 time-points), thawed PBMCs were resuspended in RPMI media supplemented with 10% heat-inactivated calf serum, 100U/ml penicillin G and 100 µl/ml streptomycin sulphate, and rested for 2 h with 10U/ml DNase I (Roche Diagnoses). In vitro stimulation for 6 h at 37 °C/5% CO2 was performed in the presence or absence of 4 µg/ml of an overlapped HIV (Gag)-specific peptide pool (PePMix HIV-1, JPT Peptide Technologies) and 1 µg/ml of SARS-CoV-2 pool peptides derived from structural (S, M, N, E) and non-structural viral proteins (Miltenyi Biotec), respectively. All stimulations included 2 µg/ml of Brefeldin A (Biolegend), 1 µg/ml of Golgi stop (BD Biosciences), 10 µg/ml of anti-CD28/anti-CD49d and 3 µg/ml of CD107a surface degranulation marker. Intracellular cytokines were analysed by multiparametric flow cytometry, as previously described [9]. The specific T-cell response was determined by the sum of studied function combinations (CD107a, TNF-α, IFN-γ, and IL2). T-cell response analyses included background subtraction using an unstimulated condition as a negative control and a positive control stimulated with staphylococcal enterotoxin B (SEB, Sigma Aldrich).

All details for antibody panels are collected in Supplementary Tables 1, and a representative gating strategy for specific T-cell responses can be found in Supplementary Fig. 1E.

SARS-CoV-2’spike specific B-cell response analysis

A SARS-CoV-2 spike B cell analysis kit (Miltenyi Biotech) was used to evaluate B-cell responses. Thawed 107 PBMCs from HIV/SCV2 and HIV/post-SCV2 time-points were resuspended in RPMI media, washed with PBS/BSA 3% and incubated for 30 min at 4ºC with 7-AAD for viability staining and a mix of fluorochrome-conjugated antibodies (CD19, CD27, IgG, IgA and IgM), according to the manufacturer’s instructions. Recombinant biotinylated SARS-CoV-2 Spike protein was conjugated separately with streptavidin PE and PE-Cy7 for 15 min at room temperature and pooled in a 1:2 ratio, and then added to the final staining mix. The schematic gating strategy for B cells can be found in Supplementary Fig. 1F.

Flow cytometry was performed on a CytoFlex S cytometer (Beckman Coulter) and analysed using FlowJo V10 software (TreeStar).

Statistics

Quantitative variables were expressed as median and interquartile range (IQR). Categorical variables were represented as numbers and percentages. Normality of data distribution was determined by frequency histogram normal test and Shapiro- Wilk test (n < 50). Due to the non-normal distribution of most variables, a logarithmic transformation was applied to normalize the data and increase statistical power before fitting a mixed linear model. The evolution of different variables between the HIV/pre-SCV2, HIV/SCV2, and HIV/post-SCV2 time-points was studied using linear mixed models considering each individual as a random effect. Wilcoxon matched pairs signed-rank test was conducted to compare the evolution of SARS-CoV-2-specific T and B cell responses between the HIV/SCV2, and HIV/post-SCV2 time-points. Associations between clinical data, immunophenotyping analysis, and cytokine and chemokine levels were assessed using Spearman’s rank test. P-values are adjusted by Bonferroni correction and statistical significance was set at P < 0.05. Statistical analyses were performed using the Statistical Package for the Social Sciences (SPSS), STATA 18.0 and GraphPad Prism 9.0 software.