Materials availability

This study used novel deimmunized aIL-10 (JES3.12G8) and aPD-1 mAbs (1B8 LC3/HC1), which are proprietary reagents developed by Merck & Co. Both aIL-10 and aPD-1 mAbs were the subject of material transfer agreements restrictions as these biologics are currently under investigation in human clinical trials.

Study approval

This study was approved by the Emory University Institutional Animal Care and Use Committee via permit no. 201800047. Experiments were conducted following guidelines set forth by the National Institutes of Health (NIH) and Animal Welfare Act in regard to the housing and welfare of laboratory RMs. All possible efforts were taken to minimize the pain and discomfort experienced by RMs.

Experimental model

A total of 28 Indian-origin, specific-pathogen-free RMs (Macaca mulatta) (25 males and 3 females) were sourced from the Emory National Primate Research Center (ENPRC) colony and single-housed in an animal BSL-2 facility at ENPRC, as previously described¹. RMs were between 38 and 50 months old at the time of infection and Mamu-B*07⁻ and Mamu-B*17⁻. Some RMs were Mamu-A^*01⁺. RMs were infected, treated and followed up for virological and immunological outcomes, as previously described¹⁸. The analysis presented herein was focused on immune responses and virological readouts in plasma, PBMCs and LNMCs pre-ATI and at 24 weeks post-ATI. Sample collection and processing, formulation of the aIL-10 and aPD-1 mAbs, pharmacokinetics (PK) and pharmacodynamics (PD) of aPD-1 and aIL-10 mAbs, levels of antidrug antibodies (ADAs), plasma IL-10 levels, aPD-1 receptor occupancy, complete blood counts, blood chemistry, SIV VL, measurement of CA-vRNA and DNA levels⁶, and measurement of 2-LTR circles were all performed as previously described¹⁸.

Bulk RNA-seq

A total of 100,000 PBMCs or LNMCs from RMs were lysed directly into 700 μl of QIAzol reagent. RNA was isolated using the RNeasy Micro kit (QIAGEN) with on-column DNase digestion. RNA quality was assessed using an Agilent Bioanalyzer and total RNA was used as input for cDNA synthesis using the Clontech SMART-Seq v4 Ultra Low Input RNA kit (Takara Bio) according to the manufacturer’s instructions. Amplified cDNA was fragmented and appended with dual-indexed barcodes using the Nextera XT DNA Library Preparation kit (Illumina). Libraries were validated by capillary electrophoresis on an Agilent 4200 Tape Station, pooled at equimolar concentrations and sequenced on an Illumina NovaSeq6000 at 100SR, yielding 20–25 million reads per sample. The quality of reads was evaluated using Fast QC (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/). Reads were aligned using STAR (v.2.7.3)⁶². The STAR index was built by combining genome sequences for M. mulatta (Mmul_10 Ensembl release 100). Transcript abundance estimates were calculated internal to the STAR aligner using the htseq-count algorithm⁶³. The ReadsPerGene files were used to generate counts in the htseq format using a custom script that also converted the Ensembl ID to gene names using the gtf file. These files were imported in DESeq2 using the DESeqDataSet- FromHTSeqCount function. DESeq2 was used for normalization⁶⁴, producing both a normalized read count table and a regularized log expression table. Regularized log expression values were obtained using the rlog function with the parameters blind = FALSE and filtType = parametric. Thresholds of Padj  1.5 and log-fold-change standard error 65. GSEA was performed using a compiled set of pathways from public databases, including MSigDB v.5.1 (http://software.broadinstitute.org/gsea/msigdb/) and blood cell marker signatures. The GSEA Java desktop program was downloaded from the Broad Institute (http://www.broadinstitute.org/gsea/index.jsp) and used with GSEA preranked module parameters (number of permutations = 1,000; enrichment statistic = weighted; seed for permutation = 111; and 10 ≤ gene set size ≤ 5,000). We used the Dynet Analyzer application implemented in Cytoscape v.3.6.0 to generate gene interacting networks that highlight overlapping genes between the different enriched modules. SLEA⁶⁶ was used to investigate the enrichment of pathways in individual RMs upon the different interventions. In brief, the expression of all genes in a specific pathway was averaged across samples and compared to the average expression of 1,000 randomly generated gene sets of the same size. The resulting z-score was then used to reflect the overall perturbation of each pathway in each individual sample. Data were visualized using ggplot2 (v.3.3.2) in RStudio (v.1.4.1103) with custom code.

Construction of single-cell multiome (GEX and ATAC) libraries

Samples from CA-vDNA^hi and CA-vDNA^lo combo-treated RMs were used as available: CA-vDNA^hi: RBf17, RFl17, Rym17; CA-vDNA^lo: RBv17, RNy16, RWs17, RRh17. Single-cell multiome ATAC and gene expression (GEX) libraries were prepared using the Chromium Single Cell Multiome ATAC+ Gene Expression platform (10x Genomics) with a target nuclei recovery of 8,000 nuclei. Isolated nuclei were transposed and partitioned into Gel Beads-in-emulsion (GEMs) using the 10x Chromium Controller and Next GEM Chip J. ATAC and GEX libraries were generated from the same pool of preamplified transposed DNA/cDNA. Representative traces and quantitation of both libraries were determined using Bioanalyzer High Sensitivity DNA Analysis (Agilent). ATAC libraries were sequenced on an Illumina NovaSeq S2 with a sequencing depth of 25,000 read pairs per nucleus and GEX libraries were sequenced on an Illumina NovaSeq S4 with a sequencing depth of 20,000 read pairs per nucleus. The 10x Barcodes in each library type were used to associate individual reads back to the individual partitions and thereby to each single nucleus.

Analysis of single-cell multiome data (scRNA-seq and scATAC-seq)

Raw reads from 10x Chromium Single Cell Multiome ATAC and Gene Expression libraries were aligned to the reference M. mulatta genome (Mmul_10) using the Cell Ranger ARC Count v.2.0.2 pipeline. The aggregated counts file was then imported into Seurat v.5.01. Quality control parameters were applied to define high-quality cells. These parameters included: RNA counts between 500 and 25,000, detected genes > 300, mitochondrial reads  2 and fragments in peaks > 20%. Following the removal of low-quality cells, DoubletFinder was used to identify and exclude doublet cells. 56,193 high-quality singlets (RBf17: 7,805, RBv17: 9,561, RFl17: 7,146, RNy16: 6,191, RRh17: 8,820, RWs17: 9,618, RYm17: 7,052) were used in downstream analysis. Peak calling was performed using the CallPeaks function. Data normalization and dimensional reduction were performed independently on the gene expression (RNA) and chromatin accessibility (ATAC) datasets and then integrated using the weighted-nearest-neighbor (WNN) method. Clustering was performed on the integrated UMAP dimension reduction and manual annotation was performed using cell type-specific gene markers. CD4⁺ and CD8⁺ T cells were further re-clustered to improve the identification of specific subsets of T cells. DEGs between conditions were identified per cell type using the two-sided Wilcoxon rank-sum test, requiring a minimum cluster cell frequency of 1%, and an adjusted P value (FDR) 2 fold change > 0.25 and FDR P value 42 were included in the analysis. Gene sets derived from these analysis were labeled ‘EC_vs_nonHIC’. Similarly, the Tirosh et al. gene set refers to genes described in Tirosh et al.⁶¹. For ATAC data, a per-cell motif-binding site activity score was calculated using chromVAR, utilizing a collection of 746 TFs from the JASPAR database⁶⁷. Differences in motif activity per time point and group were identified using the FindMarkers function with an adjusted P value (FDR) 

Flow cytometry

Cryopreserved PBMCs and LNMCs (10⁶ per test) were thawed and stained with anti-rhesus or anti-human mAbs that are known to be cross-reactive with RMs cells and were validated in databases maintained by the Nonhuman Primate Reagent Resource. An antibody panel to evaluate antiviral responses was used to stain PBMCs and LNMCs pre-ATI and included CD95 BV605, BioLegend cat. no. 305628, Clone DX2; CD3 BUV395, Becton-Dickinson (BD) cat. no. 564117, Clone SP34-2; Live/Dead BV510, BD cat. no. 564406; CD8 BUV496, BD cat. no. 612942, Clone RPA-T8; CD28 BUV737, BD cat. no. 612815, Clone CD28.2; CD14 BV786, BD cat. no. 563698, Clone M5E2; CD16 BUV661, BD cat. no. custom 3G8; CD4 PerCP/Cy5.5, BD cat. no. 552838, Clone L200; pSTAT3 PE, BD cat. no. 558557, Clone 4/PSTAT3; pIRF7 (pS477/pS479) A647, BD cat. no. 558630, Clone K47-671; p38 (pT180/pY182) PE-Cy7, BD cat. no. 560241, Clone 36/p38; pStat5 PE-CF594, BD cat. no. 562501, Clone 47/Stat5(pY694); pIRF3, Abcam cat. no. ab138449, polyclonal; pStat1 AF488, BD cat. no. 612596, Clone K51-856; and donkey anti-rabbit BV421, BioLegend cat. no. 406410, polyclonal Ig. Cell surface staining was performed at 37 °C for 20 min. Staining was stopped by adding 50 μl of cold 1× TFP Fix/Perm Buffer to the cells (BD cat. no. 565575) and mixing gently. Cells were then fixed and permeabilized for 40 min at 2–8 °C. Cells were washed and rehydrated twice (500g for 5 min) in 200 μl of PBS + 2% fetal bovine serum (FBS) buffer. Then, 50 μl of ice-cold Perm Buffer III was added, cells were mixed gently and incubated on ice for 15 min. Cells were then washed twice with 200 μl of 1× TFP Perm/Wash Buffer and all residual buffer was removed. Intracellular staining was performed by adding mAbs to 25 μl of master mix diluted in 1× TFP Perm/Wash Buffer, followed by gentle mixing. Cells were incubated for 30 min at 2–8 °C and washed with 200 μl 1× TFP Perm/Wash Buffer by centrifugation at 500g for 5 min at 2–8 °C. Finally, cells were resuspended in 100 μl of PBS 2% FBS. Acquisition was performed on a minimum of 100,000 live cells on an A5 Symphony flow cytometer driven by BD FACSDiva software (BD Biosciences).

Assays for validation

IFN signaling protects bystander CD4⁺ T cells from HIV infection in vitro

Memory CD4⁺ T cells were isolated from healthy human donors (n = 6). Cryopreserved PBMCs were thawed and rested overnight in AIMV medium + 10% Serum Replacement (Corning Life Sciences) + 10 mM HEPES. Memory CD4⁺ T cells were isolated using the EasySep Human CD4⁺ T Cell Enrichment kit (STEMCELL Technologies). One half of memory CD4⁺ T cells was stimulated for up to 16 h with 5 ng ml⁻¹ of IFN-β (Peprotech) while the other half remained unstimulated. These cells were further divided into two pools: unlabeled CD4⁺ T cells and CD4⁺ T cells labeled with CTV. Unlabeled CD4⁺ T cells were infected with HIV-1 (p89.6 dual-tropic strain). After infection, cells were cultured in the presence of saquinavir and mixed with autologous CTV-labeled CD4⁺ T cells in a 1:1 ratio. Staining for HIV-1 p24 was performed on day 4 post-infection using BD Phosflow Fix Buffer I (BD) and BD Phosflow Perm Buffer III (BD) according to the manufacturer’s instructions. The staining panel comprised Live/dead BV510, Life Technologies cat. no. L34957; CD3 BUV615, BD cat. no. 612992, Clone UCHT1; CD4 BV605, BioLegend cat. no. 317438, Clone OKT4; CD8 BUV737, BD cat. no. 564629, Clone SK1; CD45RA BV650, BioLegend cat. no. 304136, Clone HI100; CD27 APC-eFluor 780, eBioscience cat. no. 47027942, Clone O323; CCR7 BUV563, BD cat. no. 741317, Clone 3D12; MX2 AF488, Santa Cruz Biotechnology cat. no. 271527, Clone H-7; IFIT1 APC, Novus Biologicals cat. no. NBP2-71005APC, Clone OTI3G8; APOBEC3G AF700, Novus Biologicals cat. no. NBP1-77206AF700, polyclonal; pSTAT1 (p701) PE-CF594, BD cat. no. AB_2737715, Clone 4a; pIRF3 (pSer396) PerCP, Bioss cat. no. bs-3195R-PerCP, polyclonal; pIRF7 (pSer471+pSer472) AF350, Bioss cat. no. bs-3196R-A350, polyclonal; HIV-1 core antigen FITC, Beckman Coulter cat. no. 6604665, Clone KC57 and CTV BV421, Thermo Fisher Scientific cat. no. C34571. Cell surface staining was performed at 37 °C for 20 min in 25 μl of PBS + 2% FBS buffer. Cells were washed with 150 μl of PBS + 2% FBS and were fixed in 50 μl of Phosflow Fix Buffer I (BD) at 4 °C for 30 min. Cells were washed again with 150 μl of PBS + 2% FBS and then permeabilized in 50 μl of Phosflow Perm Buffer III (BD) on ice for 15 min. Cells were washed twice with 150 μl of PBS + 2% FBS at 500g for 5 min at room temperature. Intracellular staining was performed by adding the intracellular antibody mix in 25 μl of master mix diluted in 1× permeabilization buffer and mixing gently. Cells were incubated for 60 min at 4 °C and washed with 150 μl of 1× permeabilization buffer. Cells were then resuspended in 100 μl of PBS 2% FBS. Acquisition was performed on a minimum of 100,000 live cells on an A5 Symphony flow cytometer driven by FACSDiva. Acquired data were initially analyzed using FlowJo v.10.8.1.

Plasma from CA-vDNA^hi RMs induces HDAC11 expression and leads to increased HIV infection in vitro

Plasma samples from healthy control (HC) RMs (TGF-β concentration of 1,260 pg ml⁻¹), CA-vDNA^hi RMs (RFl17, TGF-β concentration of 11,873 pg ml⁻¹) and CA-vDNA^lo RMs (RRh17, TGF-β concentration of 2,084 pg ml⁻¹) were heat-inactivated at 56 °C for 35 min for complement deactivation and used in this experiment at 25% dilution. This plasma concentration was used as cell viability was maintained above 95%. The final concentrations per well were healthy control TGF-β (315 pg ml⁻¹); CA-vDNA^hi TGF-β (2,968 pg ml⁻¹); and CA-vDNA^lo TGF-β (521 pg ml⁻¹). After complement deactivation, plasma samples were cleaned using a 75-μm filter to remove clots. Recombinant TGF-β (rTGF-β; 2,000 pg ml⁻¹) was used as positive control. TGF-β signaling blockade was performed by using anti-TGF-β (10 μg ml⁻¹). AIMV medium was used. Fresh PBMCs were isolated from healthy human donors (n = 5) as described. Memory CD4⁺ T cells were isolated by using the EasySep Human CD4⁺ T Cell Enrichment kit (Stemcell Technologies) and rested overnight in AIMV medium at a concentration of 2.5 × 10⁶ CD4⁺ T cells per ml. After resting, cells were subjected to the following conditions: (1) no plasma; (2) HC plasma; (3) CA-vDNA^hi plasma; (4) CA-vDNA^lo plasma; and (5) rTGF-β. In all cases, cells were cultured in the presence or absence of anti-TGF-β. At 20 min following the addition of the respective conditions, 50,000 cells were removed for the evaluation of SMAD2/3 phosphorylation. Remaining cells were cultured for an additional 16 h, when another 50,000 cells were removed for measuring HDAC11 expression. The remaining cells were spinoculated for HIV infection in vitro as described above. These cells were cultured in AIMV medium supplemented with IL-2 and saquinavir to limit infection to one round, as previously described. HIV protein expression was measured by flow cytometry 4 days post-infection. The staining panel comprised Live/dead BV510, Life Technologies cat. no. L34957; CD3 BUV615, BD cat. no. 612992, Clone UCHT1; CD4 BV605, BioLegend cat. no. 317438, Clone OKT4; CD8 BUV737, BD cat. no. 564629, Clone SK1; CD45RA BV650, BioLegend cat. no. 304136, Clone HI100; CD27 APC-eFluor 780, eBioscience cat. no. 47027942, Clone O323; CCR7 BUV563, BD cat. no. 741317, Clone 3D12; pSMAD2(pS465/pS467)/pSMAD3 (pS423/pS425) PE-CF594, BD cat. no. 562697, Clone O72-670; HDAC11 PE, Santa Cruz cat. no. sc-390737PE, Clone C-5; and HIV-1 (p24) APC, Medimabs cat. no. MM-0289-APC, Clone 28B7. Cell surface staining was performed at 37 °C for 20 min in 25 ml of PBS + 2% FBS buffer. Cells were washed with 150 ml of PBS + 2% FBS and were fixed in 50 ml of Phosflow Fix Buffer I (BD) at 4 °C for 30 min. Cells were washed again with 150 ml of PBS + 2% FBS and then permeabilized in 50 ml of Phosflow Perm Buffer III (BD) on ice for 15 min. Cells were washed twice with 150 ml of PBS + 2% FBS at 500g for 5 min at room temperature. Intracellular staining was performed by adding the intracellular antibody mix in 25 ml of master mix diluted in 1× permeabilization buffer and mixing gently. Cells were incubated for 60 min at 4 °C and washed with 150 ml of 1× permeabilization buffer. Cells were then resuspended in 100 ml of PBS 2% FBS. Acquisition was performed on a minimum of 100,000 live cells on an A5 Symphony flow cytometer driven by FACSDiva.

TGF-β abrogates IFN-induced antiviral signatures and makes cells susceptible to HIV infection

Memory CD4⁺ T cells were isolated from healthy human donors (n = 6) as described above and were then subjected to the following conditions: (1) unstimulated; (2) stimulated with 2 ng ml⁻¹ TGF-β (Peprotech, cat. no. 100-21) for 24 h; (3) stimulated with 0.2 or 2 ng ml⁻¹ IFN-β; and (4) stimulated with 2 ng ml⁻¹ TGF-β in presence or absence of anti-TGF-β antibody (Merck & Co.) for 24 h and then with 0.2 or 2 ng ml⁻¹ IFN-β for 24 h. Stimulated CD4⁺ T cells were then infected with HIV (p89.6 dual-tropic strain) by spinoculation in the presence of saquinavir. Activation of TGF-β and IFN-β signaling pathways (pSMAD2/3 and pSTAT1, respectively) was assessed at 20 min post-stimulation, induction or suppression of ISGs was assessed at 16 h post IFN-β and TGF-β stimulation, and HIV p24 levels were assessed 4 days post-infection. Cell surface and intracellular staining were performed as above. The staining panel comprised Live/dead BV510, Life Technologies cat. no. L34957; CD4 BV605, BioLegend cat. no. 317438, Clone OKT4; CD8 BUV737, BD cat. no. 564629, Clone SK1; CD45RA BV650, BioLegend cat. no. 304136, Clone HI100; CD27 BUV496, BD cat. no. 751678, Clone O323; CCR7 BUV563, BD cat. no. 741317, Clone 3D12; IFIT1 APC, Novus Biologicals cat. no. NBP2-71005APC, Clone OTI3G8; APOBEC3G AF700, Novus Biologicals cat. no. NBP1-77206AF700, polyclonal; pSTAT1 (p701) AF488, BD cat. no. AB_2737715, Clone 4a; pSMAD2(pS465/pS467)/pSMAD3 (pS423/pS425) PE-CF594, BD cat. no. 562697, Clone O72-670; p24 RD1, Beckman Coulter cat. no. 6604667, Clone KC57; PD-1 BV711, BD cat. no. 564017, Clone EH12.1; and BCL-2 BUV395, cat. no. custom, Clone Bcl-2/100. Acquisition and analysis of flow cytometry data were performed as above.

TGF-β induces expression of HDACs

To validate our ex vivo findings showing that TGF-β upregulate HDAC1 and HDAC11 expression. PBMCs were isolated from fresh blood from healthy human donors (n = 5). Cells were cultured in AIMV + 10% Serum Replacement (Corning cat. no. 355500) + 10 mM HEPES overnight. Cells were then subjected to the following conditions: unstimulated, stimulated with: TGF-β (0.2 ng, 2 ng and 20 ng ml⁻¹; Gibco, cat. no. 100-21) for 16 h in absence or presence of anti-TGF-β (Merck, cat. no. customized) or anti-IL-10 antibodies (Merck, cat. no. customized). Cells were first stained with a viability dye, washed and fixed according to standard protocols. Staining was performed as per the BD staining protocol with BD Phosflow Fix Buffer I (BD, cat. no. 557870) and BD Phosflow Perm Buffer III (BD, cat. no. 558050). The staining panel included: Live/dead APC-Cy7, BD cat. no. 565388; CD3 BUV805, BD cat. no. 612895, Clone UCHT1; CD4 BV605, BioLegend cat. no. 317438, Clone OKT4; CD8 BUV737, BD cat. no. 564629, Clone SK1; CD45RA BV650, BioLegend cat. no. 304136, Clone HI100; CD27 BUV615, BD cat. no. 751685, Clone O323; CCR7 PE-CF594, BD cat. no. 562381, Clone 2-L1-A; HDAC11 PE, Santa Cruz cat. no. sc-390737PE, Clone C-5; HDAC1 FITC, Santa Cruz cat. no. sc-81598FITC, Clone 10E-2; HLA-DR BV786, BD Biosciences cat. no. 564041, Clone G46-6; CD19 BUV395, BD Biosciences cat. no. 563549, Clone SJ25C1; CD14 BV570, BioLegend cat. no. 301832, Clone M5E2; CD16 BUV661, BD Biosciences cat. no. 741693, Clone B73.1; CD56 PE-Cy5, BD Biosciences cat. no. 555517, Clone B159. Surface staining was performed at 37 °C for 20 min in 25 μl of PBS + 2% FBS buffer. Cells were washed with 150 μl of PBS + 2% FBS buffer. Cells were fixed in 50 μl of Phosflow Fix Buffer I (BD, cat. no. 557870) at 4 °C for 30 min. Cells were washed again with 150 μl of PBS + 2% FBS buffer and then permeabilized in 50 μl of cold Phosflow Perm Buffer III (BD, cat. no. 558050) on ice for 15 min. Cells were washed twice with 150 μl of 1× permeabilization buffer (Invitrogen cat. no. 00-8333), and all residual buffer was removed. All centrifugation steps for staining were performed at 500g for 5 min at room temperature. Intracellular staining was performed by adding the intracellular antibody mix in 25 μl of master mix (diluted in 1× permeabilization buffer (Invitrogen cat. no. 00-8333)) and mixed gently. Cells were incubated for 60 min at 4^oC and washed with 150 μl of 1× permeabilization buffer (Invitrogen cat. no. 00-8333). Finally, cells were resuspended in 100 μl of PBS for acquisition. Acquisition was performed on a minimum of 100,000 live cells on a A5 Symphony (BD Biosciences) driven by BD FACSDiva software. Acquired data were initially analyzed employing FlowJo software (v.10.8.1). Representative staining is shown for each panel in the respective figures. Data analysis was performed by using manual gating frequencies and MFIs are shown.

TGF-β suppresses IFN-dependent antiviral state through HDACs and promotes HIV infection

To validate our ex vivo findings, PBMCs were isolated from fresh blood from healthy human donors (n = 5). Cells were cultured in AIMV + 10% Serum Replacement (Corning cat. no. 355500) + 10 mM HEPES overnight. Cells were then subjected to the following conditions: (1) unstimulated; or stimulated with HDAC inhibitors (HDACi), such as (2) panobinostat (100 nM), (3) SAHA (10 nM), (4) romidepsin (10 nM), and (5) SIS17 (100 nM) individually for 24 h only; (6) TGF-β only (2 ng ml⁻¹) for 24 h; (7) IFN- β (1 ng ml⁻¹) only for 24 h; (8) IFN-γ (1 ng ml⁻¹) only for 24 h; (9) TGF-β (2 ng ml⁻¹) for 24 h and then IFN-β (1 ng ml⁻¹) for 24 h; (10) TGF-β (2 ng ml⁻¹) for 24 h and then IFN-γ (1 ng ml⁻¹) for 24 h; (11) individual HDACi for 2 h, then TGF-β (2 ng ml⁻¹) for 24 h, and then, IFN- β (1 ng ml⁻¹) for 24 h; and individual, (12) HDACi for 2 h, then TGF-β (2 ng ml⁻¹) for 24 h, and then, IFN-γ (1 ng ml⁻¹) for 24 h. HDAC inhibitors and TGF-β were removed after 24 h of stimulation, just before the start of IFN-β or IFN-γ stimulation. Activation of the TGF-β and IFN-β signaling pathways (pSMAD2/3 and pSTAT1, respectively) was assessed at 20 min post-stimulation. Induction or suppression of ISGs was assessed at 16 h post IFNs and TGF-β stimulation. Staining was performed as per BD staining protocol with BD Phosflow Fix Buffer I (BD, cat. no. 557870) and BD Phosflow Perm Buffer III (BD, cat. no. 558050). The staining panel included Live/dead BV510, Life Technologies cat. no. L34957; CD3 BUV805, BD cat. no. 612895, Clone UCHT1; CD4 BV605, BioLegend cat. no. 317438, Clone OKT4; CD8 BUV737, BD cat. no. 564629, Clone SK1; CD45RA BV650, BioLegend cat. no. 304136, Clone HI100; CD27 BUV615, BD cat. no. 751685, Clone O323; CCR7 PE-CF594, BD cat. no. 562381, Clone 2-L1-A; IRF7 AF488, Novus biologicals cat. no. NBP306987AF488, Clone 3D9; IFIT1 APC, Novus biologicals cat. no. NBP2-71005APC, Clone OTI3G8; pSTAT1 (p701) RB780, BD cat. no. 569144, Clone 4a; pSmad2 (pS465/pS467)/Smad3 (pS423/pS425) R718, BD cat. no. 567080, Clone O72-670; H3K27ac Pacific Blue, Cell signaling cat. no. 23349, Clone D5E4; IRF1 PE, BD cat. no. 566322, Clone 20/IRF1; PD-1 BV711, BD Biosciences cat. no. 564017, Clone EH12.1; HLA-DR BV786, BD Biosciences cat. no. 564041, Clone G46-6; CD19 BUV395, BD Biosciences cat. no. 563549, Clone SJ25C1; CD14 BV570, BioLegend cat. no. 301832, Clone M5E2; CD16 BUV661, BD Biosciences cat. no. 741693, Clone B73.1; and CD56 PE-Cy5, BD Biosciences cat. no. 555517, Clone B159. Surface staining was performed at 37 °C for 20 min in 25 μl of PBS + 2% FBS buffer. Cells were washed with 150 μl of PBS + 2% FBS buffer. Cells were fixed in 50 μl of Phosflow Fix Buffer I (BD, cat. no. 557870) at 4 °C for 30 min. Cells were washed again with 150 μl of PBS + 2% FBS buffer and then permeabilized in 50 μl of cold Phosflow Perm Buffer III (BD, cat. no. 558050) on ice for 15 min. Cells were washed twice with 150 μl of 1× permeabilization buffer (Invitrogen cat. no. 00-8333), and all residual buffer was removed. All centrifugation steps for staining were performed at 500g for 5 min at room temperature. Intracellular Staining was performed by adding the intracellular antibody mix in 25 μl of master mix diluted in 1× permeabilization buffer (Invitrogen cat. no. 00-8333) and mixed gently. Cells were incubated for 60 min at 4 °C and washed with 150 μl of 1× permeabilization buffer. Finally, cells were resuspended in 100 μl of PBS for acquisition. Acquisition was performed on a minimum of 100,000 live cells on a A5 Symphony (BD Biosciences) driven by BD FACSDiva software.

All flow cytometry data were analyzed using FlowJo v.10.8.1. Data analysis was performed by using manual gating frequencies and MFIs are shown or by using UMAP followed by cluster analysis. Representative cytograms are shown for each panel in the respective figures.

Measurement of cytokines

Meso Scale MULTI-ARRAY Technology (Meso Scale Discovery) was used for evaluation of cytokine levels. A cytokine panel containing the following analytes was screened: CTAK, ITAC, IL-10, IL-16, IL-17A, IL-18, IL-4, IP-10, MCP-1, MCP-2, MIP-3β, GROA, IL-22, IL-7, IL-8, TGF-β1, TGF-β2, TGF-β3, Fractalkine, IFN-α, IFN-β, IL-15, IL-2, IL-6, IL-9, MIP-1α and TNF using 25 μl of serum from each donor in duplicates and following the manufacturer’s instructions. Samples were randomized to avoid batch effects. Results were extrapolated from standard curves for each analyte and plotted in pg ml⁻¹ using the DISCOVERY WORKBENCH v.4.0 software (Meso Scale Discovery).

Statistical analysis

Statistical analyses and plots of data generated in the assays for experimental validation were performed with GraphPad Prism v.9.4.0 (GraphPad Software). The statistical test, symbol representation of statistical significance and sample sizes are listed within the figure legends for each analysis.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.