Monocyte-derived macrophages contain persistent latent HIV reservoirs

Participants

Blood samples from healthy and HIV-positive donors were obtained with written informed consent and subsequently handled in accordance with protocols approved by the Johns Hopkins University Institutional Review Board. Cohort characteristics are reported in Extended Data Table 1.

Flow cytometry analysis

Whole-blood samples were stained with pre-titrated antibodies using 100 μl of whole blood at room temperature for 20 min. The antibody panel and dilutions are listed in Supplementary Table 1. Whole-blood samples were then lysed and fixed in 2 ml of FACS lysing solution (BD Biosciences) for 10 min at room temperature. Samples were collected in a centrifuge at 400 × g for 5 min, washed in 2 ml of 1× phosphate-buffered saline (PBS) and then resuspended in 0.5 ml of PBS for analysis. Purity was assessed following selection using flow cytometry. PBMCs were stained before isolation and following pan-monocyte selection with pre-titrated monoclonal antibodies and a viability indicator. The antibody panel and dilutions are listed in Supplementary Table 1. TLR2 was used as a general monocyte cell marker as previously published²³. PBMCs were stained, acquired and analysed as described above. Selection purities are reported in Extended Data Table 3. In select instances, purity assessments of MDM 7 d post differentiation were also completed by flow cytometry. MDMs differentiated from heathy donors were removed from the plate with TrypLE (Gibco). The antibody panel and dilutions are listed in Supplementary Table 1. In brief, cells were stained with anti-CD3 and LIVE/DEAD for 30 min at 4 °C. Cells were then permeabilized using Biolegend PermFast and stained with anti-CD68 or matched IgG control. Flow cytometry was performed on a BD LSRFortessa (BD Biosciences). Voltage settings were standardized to daily CS&T Research Bead (BD Biosciences) controls using predetermined application settings in FACSDiva 6.2 to ensure that fluorescent intensity was consistent longitudinally. Data were analysed using FlowJo 10.0.8 software (FlowJo). Representative gating strategy is shown in Extended Data Fig. 6a–d.

Cell lines

Three lymphocyte cell lines were tested during the development of the MDM-QVOA: MT-4 cell line obtained through the NIH HIV Reagent Program, Division of AIDS (NIAID, NIH: MT-4 cells, ARP-120, contributed by Dr Douglas Richman (cat. no. 120)^46,47,48); MOLT-4-CCR5 kindly donated by Dr Robert F. Siliciano from Johns Hopkins Medical School; and CEMX174 purchased from ATCC. All cell lines were propagated and maintained in R10 (RPMI 1640 medium (Gibco) supplemented with 10% heat-inactivated fetal bovine serum, 100 U of penicillin per ml and 100 µg of streptomycin per ml). The MOLT-4-CCR5 were cultured in the presence of G418 (1 mg ml⁻¹) to maintain CCR5 expression. All cell lines were assessed for the necessary receptors and co-receptors for HIV entry by flow cytometry. The antibody panel and dilutions are listed in Supplementary Table 1. Antibody staining was completed as described above.

Development of MDM-QVOA assay

Whole-blood from viremic (v) and virally suppressed people with HIV (vsPWH) was obtained for PBMCs isolation by Ficoll gradient centrifugation. The PBMCs were then used in VOAs to determine the appropriate conditions for QVOA development. All VOAs were completed on MDMs derived from fresh never-frozen PBMCs or negatively isolated monocytes (Pan Monocyte isolation kit, human; Miltenyi Biotec). PBMCs or isolated monocytes were plated at a density of 2–5 × 10⁶ cells per well and cultured in MDM10 + ARV (Dulbecco modified Eagle medium (Life Technologies) supplemented with 10% heat-inactivated human type AB serum (Gemini Bio Products)), 100 U ml⁻¹ penicillin-streptomycin (Life Technologies), 20 μg ml⁻¹ gentamicin (Life Technologies), 2 mM l-glutamine (Life Technologies), 2 mM sodium pyruvate (Sigma), 10 mM HEPES buffer (Life Technologies) and 50 ng ml⁻¹ recombinant human macrophage colony-stimulating factor (R&D) containing anti-retroviral drugs (10 μM zidovudine (Sigma), 25 nM darunavir (Janssen) and 5 nM raltegravir (Merck)). These are considered to be M0 conditions and result in macrophage differentiation without polarization¹⁶. Every 3 d post plating, half of the media was removed and the cultures were replenished with fresh MDM10 + ARV. Monocytes were allowed to differentiate in these conditions for 7 d. Once MDMs were differentiated, the cells were washed twice with sterile PBS, treated with 0.025% trypsin (5 min at room temperature) and then washed twice with sterile PBS again to ensure all contaminating cells were removed and only adherent cells remained in culture. The cells were then activated with MDM10 containing one of the following activating agents: 20 ng ml⁻¹ tumour necrosis factor (TNF𝛼, ProSpec), 0.5 µM ml⁻¹ PMA (Sigma) and 10 ng ml⁻¹ Interleukin-4 (IL-4, Prospec). Lymphocyte cell lines were added in culture to expand the virus released from infected cells. The cell lines tested in the VOAs were MT-4, MOLT-4-CCR5 and CEMX174 at a density 1 × 10⁶ per well. Assay conditions included MDM10 plus one activation reagent (PMA, TNA or IL-4) and cell line (MT-4, MOLT-4-CCR5 or CEMx174), MDM10 plus one activation reagent alone, MDM10 plus one cell line alone and MDM10 only. Supernatant (1 ml) was collected on days 2, 4, 6, 8, 10 and 12 and replaced with fresh MDM10 containing respective activation reagent or media only. Viral RNA was isolated from 1 ml of VOA supernatant at each time point using the QIAamp MinElute virus vacuum kit (Qiagen) according to the manufacturer’s recommendations, and the samples were assessed for the expression of HIV gag RNA by RT–qPCR as described below.

MDM-QVOA assay

Reported here is the final assay used throughout the manuscript. Human PBMCs from vsPWH were isolated as described above. Two-thirds of isolated PBMCs were used for negative monocyte isolation (Pan Monocyte isolation kit, human; Miltenyi Biotec) and the remaining PBMCs reserved for CD4-QVOA assay (see CD4-QVOA assay below for details). Following pan monocyte isolation, 5 × 10⁵ cells were set aside for a purity assessment by flow cytometry, 2 × 10⁶ cells were plated for T cell control wells (1 × 10⁶ per well) and the remaining cells were plated in duplicate at 5-fold limiting dilution (Fig. 1b). All plated cells were cultured in MDM10 + ARV (described above). Monocytes were cultured for 7 d to allow for differentiation to MDMs. MDM10 + ARV was changed every 3 d to prevent viral spread in culture. On day 7, MDMs were washed 2 times with sterile PBS, 1 time with 0.025% trypsin (5 min at room temperature) and 2 more times with PBS to ensure all contaminating cells were removed and only adherent MDMs remained. MDMs were then activated with 0.5 µM ml⁻¹ PMA and 1 × 10⁴–10⁶ MT-4 expander cells were added per well, excluding the T cell control wells. Supernatants were collected and replenished with newly made MDM10 + PMA every 3 d and assessed for HIV gag RNA by RT–qPCR. Supernatants from early activation time points (days 10 and 13) and supernatants from later time points (days 16 and 19) were pooled and assessed for viral RNA as described below. Cells were collected at day 19 and lysed in AllPrep buffer (RLT plus and 1% beta-mercaptoethanol, βME) for RNA and DNA isolation (see below). The frequency of cells harbouring replication-competent virus was determined using the IUPMStats v1.0 infection frequency calculator and expressed as IUPM⁴⁹. Wells were considered positive if either the early or late time point had a cycle threshold (Ct) value less than or equal to 35 as measured by RT–qPCR. All MDM-QVOAs were assessed for CD3+ T cell contamination using RT–qPCR for TCRβ (see below).

Purity checks to assess CD4 T cell contamination

All selected monocyte samples were analysed by flow cytometry to determine the percentage of contaminating T cells before plating. Once plated, the cells were cultured in the presence of ART for 7 d for further purification by adherence. Once the macrophages were differentiated, they were washed extensively with PBS and a low percentage of trypsin to remove any non-adherent cells. Two wells, with a minimum of 1 × 10⁶ monocytes per well, were kept as T cell controls and no MT-4s were added. At the end of the assay (day 19), the control wells were lysed and assessed for T cell contamination by qPCR for T cell receptor beta (TCRβ) RNA. The purpose of assessing TCRβ after MDM activation is to allow contaminating T cells to expand and become easier to detect. During assay development, we also assessed MDMs with and without activation for T cell contamination by flow cytometry and observed no contaminating CD3+ cells (Extended Data Fig. 6e). Additionally, CD4-QVOAs were completed on the same blood draw for all participants to act as a positive control. These measurements were then used to mathematically calculate the percent chance of HIV+ CD4 T cell contamination in the assay contributing to our positive signal (described below and Extended Data Table 4).

Quantitation of CD3+ T cells in MDM-QVOA wells

T cell control wells without MT-4 cells were used for TCRβ RNA analyses. During assay development, we tested two methods to detect CD3+ T cells in the MDM wells: CD3ε and TCRβ. CD3ε and TCRβ RNA expression were quantified using primers, probes and reaction conditions listed in Supplementary Tables 1 and 2. All samples were quantified using target-specific RNA standard curves. In the final MDM-QVOA assay, TCRβ was used to estimate the absolute number of CD3+ T cells in MDM-QVOA (see Extended Data Table 4 for examples of how we calculated the number of CD3+ cells in the T cell control wells). In brief, we assessed the number of TCRβ RNA copies and cell number (IFNβ) in the same sample (RNA and DNA isolated via AllPrep, see below). The median number TCRβ copies per cell was determined to be 174 using CD4 T cells isolated from 10 healthy donors. Therefore, we divided the total TCRβ signal by 174 to equal CD3+ cells in the MDM well. We then used the cell number, calculated by the IFNβ signal divided by 2 (2 copies per cell), to determine the number of CD3+ cells per million. Next, we multiplied the CD3+ cells per million by the number of cells present in the largest MDM-QVOA well, as this is where we found our positive signal majority of the time. Once we had the CD3+ cells in the largest MDM-QVOA well, we multiplied this number by the percentage of CD4 T cells in whole blood at the time of draw to determine how many CD3+ cells were also CD4+. Using this number, we calculated the probability that this number of CD4 T cells could have contributed to our positive signal using the CD4 IUPM value from the same individual. The probability was then multiplied by 100 to estimate the percent chance our signal was from an HIV+ CD4 T cell.

Control experiment to assess HIV+ CD4 transfer of viral nucleic acids

PBMCs were isolated from healthy donor whole blood and monocytes were isolated using the pan monocyte selection kit as described above. Monocytes were then plated at 500,000 cells per well and differentiated in M0 conditions for 7 d. On day 7, MDMs were washed as described above, and CD4 T cells isolated from two HIV+ donors (CP11 and 21) were added in triplicate to MDM wells (range 1 × 10⁴–10¹cells). The MDM + HIV+ CD4 co-cultures were maintained for 12 d with and without PMA activation. On day 12, MDMs were washed and lysed, and assessed for cell-associated HIV RNA and DNA as described below.

CD4-QVOA assay

CD4-QVOA assays were performed as previously described¹⁷. In brief, CD4 T cells were isolated from remaining PBMCs using a negative CD4 selection kit (Neg CD4 Kit, Miltenyi Biotec), plated at 5-fold limiting dilution and cultured in super T cell media. Cells were activated with 0.5 μg ml⁻¹ phytohemagglutinin (Remel) and 10–2.5 × 10⁶ irradiated PBMCs from a heathy donor (feeders) for a minimum of 16 h. Phytohemagglutinin was then removed and 1–0.5 × 10⁶ MT-4s were added to each well. Supernatants and cells were collected on day 7. Supernatants were assessed for HIV gag RNA and cells were lysed in AllPrep buffer (RLT plus+βME) for RNA and DNA isolation (see below).

Quantitation of HIV gag RNA in QVOA supernatants

Viral RNA was isolated from 1 ml of MDM-QVOA supernatant from each serial dilution in duplicate using the QIAamp MinElute virus vacuum kit (Qiagen) according to the manufacturer’s recommendations. Viral RNA was isolated from 0.2 ml of CD4-QVOA supernatant using the QIAamp MinElute virus spin kit (Qiagen) according to the manufacturer’s recommendations. An on-column DNase digestion was performed for all QVOA samples using the RNase-free DNase kit (Qiagen) and 3 U of RQ1 DNase (Promega), and the columns were incubated at room temperature for 20 min. Viral RNA isolated from MDM-QVOA and CD4-QVOA supernatants was assessed by RT–qPCR using the QuantiTect virus kit (Qiagen). Primers, probes and reaction conditions are listed in Supplementary Tables 2 and 3. To control for DNA contamination, one reaction was analysed without reverse transcriptase. The samples were quantified using HIV gag RNA standard curve.

Quantification of cellular HIV gag and tat/rev RNA

HIV RNA cellular gag and tat/rev RNA genes were isolated from cells using AllPrep DNA/RNA mini kit (Qiagen) according to the manufacturer’s recommendations. Primers, probes and reaction conditions are listed in Supplementary Tables 2 and 3. The samples were quantified using target-specific RNA standard curves.

Quantitation of HIV gag DNA

DNA samples were isolated from cells using the AllPrep DNA/RNA mini kit according to the manufacturer’s recommendations. Viral DNA was measured in the cells using the multiplex qPCR with the MP kit (Qiagen). Primers, probes and reaction conditions are listed in Supplementary Tables 2 and 3. For sample normalization and cellular quantitation, we assessed a single-copy gene, human interferon-beta (IFN-β), using primers, probes and reaction conditions listed in Supplementary Tables 1 and 2. The samples were quantified using target-specific DNA standard curves and normalized by cell number input.

IPDA

We performed IPDA as described²¹ to separately measure genetically intact and defective (3’ deleted/hypermutated and 5’ deleted) proviral DNA, with minor modifications made for monocyte assessment. In brief, TLR2+ monocytes were isolated from participant PBMCs using the anti-biotin microbeads kit (Miltenyi Biotec) and a biotinylated TLR2 antibody (1 μg per 10⁷ cells of clone TL2.1, Invitrogen). CD4 T cells were then isolated from the remaining TLR2 negative cells using a negative CD4 selection kit (Neg CD4 kit, Miltenyi Biotec). Selected cells were then assessed for purity by flow cytometry (see above for details) and lysed in AllPrep buffer (RLT plus+βME) for DNA isolation (see above). All primers, probes and reaction conditions used for IPDA are listed in Supplementary Tables 1 and 2. Samples were run in triplicate, or if there was no signal observed, until a minimum of 1 × 10⁶ cells were acquired as determined by measuring the cellular gene RPP30. To estimate the CD4 signal that might have contributed to the results observed in the monocyte IPDA, we utilized the values assessed in the CD4 IPDA and %CD3+/CD4+ determined by flow cytometry. We mathematically calculated the number CD4 T cells present in one million monocytes, the potential intact, 3’ del or 5’ del signal in those cells and subtracted that signal from the monocyte IPDA signal. For example, sample 1 had 2% CD4 T cells in the selected monocytes and 10 intact genomes per million CD4 T cells. We would estimate that there were 20,000 CD4 T cells in 1 × 10⁶ monocytes (2 x (1 × 10⁶ cells) / 100) and 0.2 intact copies were from contaminating CD4s (10 intact / (1 × 10⁶ cells) x 20,000 CD4s), and we would then remove the latter value from the monocyte IPDA signal. The monocyte IPDA data pre and post CD4 adjustment can be found in Extended Data Table 2. All data reported in this manuscript are adjusted for CD4 but not adjusted for DNA shearing to prevent artificial increases in the intact values reported.

In vitro infection of MT-4 with QVOA supernatants

MT-4s (2 × 10⁶) were spinoculated (2 h at 1,200 × g, room temperature) with 500 μl of supernatant from positive MDM or CD4 QVOA wells with available sample. Viral input was normalized to 800 copies of HIV gag for each sample. Post spinoculation, cells were washed once with sterile PBS and resuspended in 2 ml of R10, plated in a 24-well plate and incubated at 37 °C. Supernatants were collected on days 0, 3, 6, 9, 12, 15, 18 and 21 post spinoculation and fresh medium was replaced at each time point. On days 6 and 12 post spinoculation, all cultures were supplemented with an additional 1 × 10⁶ MT-4 and the spinoculation was repeated. RNA was isolated from 1 ml of sample using a QIAamp MinElute virus vacuum kit (Qiagen), and HIV gag RNA was quantitated by RT–qPCR as described above.

Limiting dilution nef sequencing of QVOA virus

DNA was extracted from QVOA cells (MDM and CD4) using the AllPrep kit following the manufacturer’s recommendations. Limiting dilution PCRs to obtain clones of nef were performed as previously described^50,51. In brief, DNA was used in a nested limiting dilution PCR protocol using Platinum Taq HiFi (Life Technologies). The outer PCRs were diluted 1:3 with deionised water, and 10 μl outer PCR DNA was used for nested amplification of full-length nef (661 bp). Primer sets and conditions are listed in Supplementary Tables 1 and 2 and are previously published⁵². Clonality was determined using Poisson statistics, and 2 positives per 10 wells amplified was considered clonal. PCR products were visualized using 1% agarose gels and isolated using the QIAquick gel extraction kit (Qiagen). The products were sent for Sanger sequencing. Contig sequences were generated using CodonCode aligner (v9), alignments done via Bioedit Clustal W method (v7.2) and maximum likelihood phylogenetic trees constructed using the bootstrap method to test phylogeny at 1,000 replications via MEGA software (vX). Bootstrap values greater than or equal to 80 were considered significant.

Statistics and reproducibility

All data were analysed and graphically represented using Excel (v16.61) and/or GraphPad Prism (v9.4.1). All statistical analyses were performed using GraphPad Prism and were either unpaired t-tests, paired t-tests or one-way analyses of variance (ANOVA) with Tukey’s multiple comparisons test. Correlations were performed using simple linear regression. P ≤ 0.05 was considered significant. No statistical method was used to predetermine samples size, no data were excluded from analysis and data distribution was assumed to be normal, but this was not formally tested. Finally, the investigators were not blinded to allocation during experiments and outcome assessment.

Reporting summary

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