key: cord-0256495-999nu6nb
authors: Cui, Junru; Meshesha, Mesfin; Churgulia, Natela; Merlo, Christian; Fuchs, Edward; Breakey, Jennifer; Jones, Joyce; Stivers, James T.
title: Replication-competent HIV-1 in human alveolar macrophages and monocytes despite nucleotide pools with elevated dUTP
date: 2022-05-04
journal: bioRxiv
DOI: 10.1101/2022.05.03.490432
sha: 8f6b723be7cd2b5ca7c3354f667ca935e4f5c605
doc_id: 256495
cord_uid: 999nu6nb

Although CD4+ memory T cells are considered the primary latent reservoir for HIV-1, replication competent HIV has been detected in tissue macrophages in both animal and human studies. During in vitro HIV infection, the depleted nucleotide pool and high dUTP levels in monocyte derived macrophages (MDM) leads to proviruses with high levels of dUMP, which has been implicated in viral restriction or reduced transcription depending on the uracil base excision repair (UBER) competence of the macrophage. Incorporated dUMP has also been detected in viral DNA from circulating monocytes (MC) and alveolar macrophages (AM) of HIV infected patients on antiretroviral therapy (ART), establishing the biological relevance of this phenotype but not the replicative capacity of dUMP-containing proviruses. As compared to in vitro differentiated MDM, AM from normal donors had 6-fold lower levels of dTTP and a 6-fold increased dUTP/dTTP, indicating a highly restrictive dNTP pool for reverse transcription. Expression of uracil DNA glycosylase (UNG) was 8-fold lower in AM compared to the already low levels in MDM. Accordingly, ∼80% of HIV proviruses contained dUMP, which persisted for at least 14-days due to low UNG excision activity. Unlike MDM, AM expression levels of UNG and SAM and HD domain containing deoxynucleoside triphosphate triphosphohydrolase 1 (SAMHD1) increased over 14 days post-HIV infection, while dUTP nucleotidohydrolase expression decreased. These AM-specific effects suggest a restriction response centered on excising uracil from viral DNA copies and increasing relative dUTP levels. Despite the restrictive nucleotide pools, we detected rare replication competent HIV in AM, peripheral MC, and CD4+ T cells from ART-treated donors. These findings indicate that the potential integration block of incorporated dUMP is not realized during in vivo infection of AM and MC due to the near absence of UBER activity. In addition, the increased expression of UNG and SAMHD1 in AM post-infection is too slow to prevent integration. Accordingly, dUMP persists in integrated viruses, which based on in vitro studies, can lead to transcriptional silencing. This possible silencing outcome of persistent dUMP could promote viral latency until the repressive effects of viral dUMP are reversed.

Although CD4 + memory T cells are considered the primary latent reservoir for HIV-1, replication competent HIV has been detected in tissue macrophages in both animal and human studies. During in vitro HIV infection, the depleted nucleotide pool and high dUTP levels in monocyte derived macrophages (MDM) leads to proviruses with high levels of dUMP, which has been implicated in viral restriction or reduced transcription depending on the uracil base excision repair (UBER) competence of the macrophage. Incorporated dUMP has also been detected in viral DNA from circulating monocytes (MC) and alveolar macrophages (AM) of HIV infected patients on antiretroviral therapy (ART), establishing the biological relevance of this phenotype but not the replicative capacity of dUMPcontaining proviruses. As compared to in vitro differentiated MDM, AM from normal donors had 6-fold lower levels of dTTP and a 6-fold increased dUTP/dTTP, indicating a highly restrictive dNTP pool for reverse transcription. Expression of uracil DNA glycosylase (UNG) was 8-fold lower in AM compared to the already low levels in MDM.

Accordingly, ~80% of HIV proviruses contained dUMP, which persisted for at least 14days due to low UNG excision activity. Unlike MDM, AM expression levels of UNG and SAM and HD domain containing deoxynucleoside triphosphate triphosphohydrolase 1 (SAMHD1) increased over 14 days post-HIV infection, while dUTP nucleotidohydrolase expression decreased. These AM-specific effects suggest a restriction response centered on excising uracil from viral DNA copies and increasing relative dUTP levels.

Despite the restrictive nucleotide pools, we detected rare replication competent HIV in AM, peripheral MC, and CD4 + T cells from ART-treated donors. These findings indicate that the potential integration block of incorporated dUMP is not realized during in vivo infection of AM and MC due to the near absence of UBER activity. In addition, the increased expression of UNG and SAMHD1 in AM post-infection is too slow to prevent integration. Accordingly, dUMP persists in integrated viruses, which based on in vitro studies, can lead to transcriptional silencing. This possible silencing outcome of persistent dUMP could promote viral latency until the repressive effects of viral dUMP are reversed. properties that could lead to persistent HIV infection even in the presence of active retroviral therapy (ART) 1, 2 . For instance, macrophages have distinct and malleable metabolic properties compared to T cells that make them intrinsically resistance to HIVinduced cytopathic effects and less susceptible to some antiretroviral drugs 3, 4 . These properties include, but are not limited to, innate immunity pathways that involve viral nucleic acid lethal mutation by enzymatic cytidine deamination 5, 6 , dramatically reduced dNTP pools and composition through the action of SAMHD1 dNTPase 7, 8 , and the ability to transiently access G0 and G1 stages of the cell cycle 4, 9, 10 . In addition, recent reappraisals of macrophage biology indicate that these cells can have multi-year life spans in certain tissue environments and possess self-renewal properties that are similar to those of memory T cells [11] [12] [13] [14] [15] . These considerations have led to the still debated conclusion that myeloid lineage cells serve as a reservoir for HIV, even in the presence of ART 16 .

The experimental evidence supporting the proposal that macrophages serve as a persistent HIV reservoir in humans after initiation of ART is extensive, but less than definitive. In part, the continuing uncertainty arises from the relative difficulty in obtaining tissue macrophages in sufficient numbers to detect and quantify rare viral nucleic acid, excluding the possibility that contamination by infected or phagocytized T cells accounts for any positive result, and the use of different viral detection assays by various researchers, each with different limits-of-detection or specificity (see the complete review by Wong et al) 12, 13 . Nevertheless, reports from different research groups over the past twenty years have consistently (but not uniformly) detected viral RNA, proviral DNA and p24 antigen in isolated monocytes and tissue macrophages from HIV infected patients on ART [17] [18] [19] [20] [21] . In many cases, convincing controls for T cell contamination were performed and genetic analysis of viral progeny indicated distinct genotypes for HIV produced from macrophages as compared to T cells from the same patient 18, 22 . In addition, studies using animal models such as SIV-infected macaques 23, 24 , humanized BLT mice and myeloidonly mice 25, 26 have provided support for the contention that HIV persistence in the presence of ART can be partially attributed to infected myeloid cells, especially in the CNS where ART is less effective 27, 28 .

Three phenotypic traits of in vitro differentiated monocyte-derived macrophages (MDM) that are distinct from CD4 + T cells are their overall low dNTP pool levels 4 This simple picture of non-dividing MDM being repair-deficient has been complicated by recent findings that these cells exist in two populations: (i) a major G0 population where the dNTP pool is depleted, the DNA repair activity is low and HIV proviruses are heavily uracilated 20 , and (ii) a minor pseudo-G1 population characterized by a normal dNTP pool, DNA repair activity, and low levels of repair-associated DNA replication 4, 9, 36 . The pseudo-G1 state can be transiently and reversibly accessed from the G0-state, providing a mechanism for long-lived macrophages to repair genomic DNA, but also an increased chance of viral infection while in the pseudo-G1 state. The presence of two populations with greatly different susceptibilities to infection complicates quantitative evaluation of the effects of depleted and imbalanced dNTP pools 9, 20 . The presence of G0 and pseudo-G1 populations is but one example of the malleable nature of the macrophage phenotype that is determined by the microenvironment of the tissue in which the cell resides 36 .

Importantly, the nucleotide pool and DNA repair attributes of in vitro differentiated MDM have never been compared to in vivo differentiated tissue macrophages. Nevertheless, abundant viral dU/A pairs have been detected in HIV infected monocytes and alveolar macrophages isolated from patients receiving ART suggesting that similar nucleotide pool compositions exist for in vitro and in vivo differentiated macrophages 20 .

Here we report the first evaluation of the dNTP pool status of in vivo differentiated alveolar macrophages (AM) obtained from normal and HIV-infected donors using ART.

Our findings confirm that the phenotype of depleted dNTP pools and elevated dUTP/TTP extends to in vivo differentiated AM and that the expression of the uracil base excision enzyme uracil DNA glycosylase (UNG) is vanishingly low in these cells, which promotes the persistence of dUMP in proviral DNA. Since dU/A base pairs have previously been associated with suppression of RNAPII transcription 20, 31, 34 , uracilated proviral DNA may be a unique transcriptional silencing mechanism in macrophages. We also report using a quantitative viral outgrowth assay (QVOA) that replication competent HIV virus can be isolated from in vivo infected AM and peripheral blood monocytes.

Alveolar macrophages (AM) have very low dNTP pools, high dUTP/TTP and depleted uracil DNA glycosylase (UNG)

Using a highly sensitive ddPCR assay we previously reported the detection of HIV DNA in peripheral MC from 6 out of 6 fully ART-suppressed HIV patients (50-100 pol copies/10 6 cells) 20 . Further, 5/6 of the patient MC were positive for dUMP content, as were two AM samples obtained from a single patient pre-and post-ART 20 . Importantly, the viral uracilation phenotype was unique to viral DNA obtained from MC and AM, but was absent in CD4 T cells 20, 31 . These results established that (i) HIV DNA was detectable in myeloid cells using the most sensitive detection method, and (ii) viral DNA from these cells was uniquely marked by the presence of dUMP. However, these studies did not assess whether the viral DNA was integrated or whether replication competent viruses were present.

To further characterize the dNTP pool and UBER phenotypes of AM, we purified AM obtained by bronchoalveolar lavage from non-infected donors ( Table 1) . The AM were purified as described in Methods, and the absence of contaminating T cells was established using RT-qPCR targeting the TCR mRNA, which is selectively expressed in T cells 39 . From this analysis, we confirmed that T cells were present at <1 T cell/10 2 AM.

Further, ninety-five percent of the isolated AM were CD68 positive and 90% were functional as judged by a positive phagocytosis assay employing pHrodo E. coli cells (Fig.   1a, b) .

We used a modified single-nucleotide primer extension assay to assess the dUTP and dTTP levels in AM for comparison with in vitro differentiated MDM and dividing HAP1 human cells ( Fig. 1c and Supplemental Fig. S1 ) 30, 31 . As compared to the HAP1 cell line, the dTTP levels were reduced by 2,700 and 170-fold in AM and MDM, respectively. Accordingly, the dUTP/dTTP ratio in AM is 32-fold greater than the HAP1 line, and more than 6-fold greater than the already elevated ratio seen in MDM (Fig. 1d) . These measurements of high dUTP levels in AM aligns with our previous Ex-ddPCR measurements of high uracil content in HIV DNA isolated from infected AM (Fig. 1c) 20 .

We then used RT-qPCR to probe the base line expression levels of key UBER enzymes (UNG, APE1, pol , lig III, DUT) and the dNTPase SAMHD1 in AM and MDM prior to infection with HIV (Fig. 1e) . For comparative purposes, we normalized the measured expression levels to those found in the HAP1 dividing cell line. These measurements established that UNG, the first enzyme in the UBER pathway that excises uracil bases from DNA, was present at an 8-fold lower level in AM as compared to MDM and about 50-fold less than the HAP1 line. The other UBER enzymes showed similar expression levels in AM and MDM, which were about 2 to 4-fold less than HAP1 cells. As expected, both AM and MDM showed a 4-fold greater expression of SAMHD1 compared to the HAP1 cells. Taken together, these data show that AM have extraordinarily low dNTP levels, elevated dUTP/dTTP and greatly reduced UNG activity. Thus, dUMP incorporated into viral DNA during infection of AM would be expected to persist at higher levels than MDM due to the severe depletion of UNG and dTTP.

There have been very limited studies of HIV infection of AM or any other primary tissue macrophages. To ascertain the kinetics and relative efficiency of HIV infection of AM and to examine how many viral DNA products contain detectable dUMP, we infected AM and MDM with an equivalent number of R5 tropic HIV Bal virus (MOI = 1) and monitored viral DNA products and viral p24 levels over 21 days (Fig. 2) . Using PCR primers designed to detect early (ERT) and late (LRT) reverse transcription DNA products we found that the ERT levels peaked at 3 copies per cell at 1-day post infection of AM while the LRT copies reached a final level of about 0.5 copies per cell (Fig. 2a) . Thus, only about one-sixth of the ERT produced in AM were converted to more mature viral DNA products, suggesting a restrictive step early in infection of AM. In contrast, the ERT and LRT levels in MDM increased steadily over the course of infection, peaking at 4 copies per cell at 7-days post infection. By day 14, both the ERT and LRT copies in MDM diminished by almost a factor of two, suggesting a late restrictive step not observed with AM (Fig. 2a) .

Using the Alu-gag qPCR method for amplification of integrated viruses, we found that proviral copies of infected AM reached a plateau of about one copy per cell after about 7 days and remained stable out to day 21 (Fig. 2b) . In contrast, proviral copies of infected MDM increased steadily from one copy per cell at 1-day post infection to four copies at 7-days and remained roughly the same out to 21-days post infection (Fig. 2b) . Thus, AM and MDM showed distinct kinetics for viral integration.

We used the uracil-excision qPCR (Ex-qPCR) method to determine the fraction of LTR DNA products that contained dUMP from 1 to 21 days post-infection of AM and MDM ( Fig. 2c) . For AM, viral dUMP was present in ~80% of the LTR products at day one and the level was unchanged over the course of the infection. In contrast, MDM showed no detectable dUMP in LTR DNA products at day 1 and the levels increased between days 3 and 7 up to 60% of viral LTR copies. Unlike AM, the fraction of LTR copies containing dUMP decreased to about 40% in MDM between days 7 and 21 (Fig. 2c) . The absence of detectable dUMP in viral DNA at one day post-infection of MDM has been observed consistently and attributed to rapid infection of the MDM cell population present in the pseudo-G1 state where dUTP is not present 40 . The slow reduction in dUMP over time has been attributed to the slow replacement of proviral dUMP with TMP by base excision repair 40 . The early appearance of dUMP and its persistence in AM suggests that the pseudo-G1 population is a minor contribution, and that uracil excision activity is significantly reduced compared to MDM (consistent with the UNG expression profiling).

The p24 levels in culture supernatants of infected AM and MDM steadily increased over 21 days for both MDM and AM (Fig. 2d) , with the final p24 level at day 21 about three-fold higher for MDM, which is similar in magnitude to the greater proviral copy number for MDM (Fig. 2b) . Taken together, these data indicate slower infection kinetics, decreased infection efficiency, and increased dUMP persistence for AM as compared to MDM.

We measured large changes in the expression levels of UNG and two other enzymes in our panel at various times after infection of AM with R5 tropic HIV Bal virus (Fig. 3) .

Strikingly, AM showed a 20-fold induction of UNG mRNA between days 3 and 14 post infection, while no significant increases in UNG expression were observed with MDM over the same time frame. In addition, the expression level of SAMHD1 was increased ~30fold in AM and dUTPase showed a rapid 16-fold decrease in expression by day 3 (Fig.   3a) . These changes in SAMHD1 and dUTPase expression were not observed for MDM, which instead showed an ~30-fold increase in expression of DNA cytidine deaminase A3A, but not A3G (Fig. 3b) . The post-infection upregulation of UNG and SAMHD1 and the downregulation of dUTPase suggest a coordinated response to the invading virus.

Specifically, these changes would be expected to impede viral replication via increased SAMHD1 dNTPase activity and other SAMHD1-dependent mechanisms 37 , increasing the already high dUTP levels by decreasing dUTPase activity, and promoting excision of viral uracils by increasing UNG activity. However, these changes in expression levels in AM may not be rapid enough to impact the early stages of viral infection in a significant way.

Further, the increased levels of UNG appear to be ineffective at removing proviral uracils over 21 days post-infection (Fig. 2c) . This effect may indicate that integrated viruses with uracil are protected by chromatin compaction or other mechanisms.

An optimized QVOA assay for macrophages and monocytes was used to determine the number of productively infected myeloid and T cells obtained from five patients ( Table   2 ) (Fig. 4a) 23 . A total of 2.6 to 15.0 million myeloid cells (AM or MC) and 3.0 to 20.0 million T cells were examined for each donor, which sets the limit of detection for each cell type ( Table 2) . The QVOA experiments with AM and MC were assessed for the presence of TCR mRNA using RT-qPCR and the percent contamination of T cells in each QVOA was not greater than 0.9% (AM) or 1.7% (MC) (Fig. 4b) . Since replication competent viruses are present at a frequency of less than 1 per million CD4 + T cells (see below), the probability of infected CD4 + T cells harboring replication competent virus in the AM and MC QVOA reactions was 0.02 and 0.04 per million AM and MC, respectively.

Accordingly, this level of T cell contamination cannot account for any of the positive AM and MC QVOA reactions (see below).

Using qPCR, we detected viral RNA in the QVOA supernatants originating from one AM sample out of three that were analyzed and from four of the five MC QVOA experiments ( Table 2 and Fig. 4b ). In addition, viral RNA was detected in four out of five T cell QVOA experiments. In these experiments, QVOA wells that contained at least 50 RNA copies/mL were deemed positive and the limit of detection was 10 HIV RNA copies per reaction. The number of latently infected cells that contained replication-competent virus was then determined using an algorithm for maximum likelihood estimation of IUPM 38 .

For P1, blood monocytes were isolated before the initiation of ART (viral load of 31,800 copies/mL) and four months after initiation of ART (when circulating virus was undetectable) and QVOAs were performed ( Table 2) Table 2 ) (Fig 4b) . As expected, these numbers of productively infected peripheral blood MC and T cells are much less than the mean levels of viral DNA previously detected using ddPCR in these cell types isolated from 6 ART-suppressed donors (~700 copies/10 6 T cells and ~60 copies/10 6 MC) 20 . It is commonly accepted that most viral DNA copies are not replication competent, and further, the QVOA assay is not 100% efficient. Thus, these results provide a minimum estimate of the level of productively infected HIV target cells 39 .

To further establish that the viral particles present in the AM-QVOA, MC-QVOA and T cell QVOA supernatants contained replication competent virus, we spin-inoculated MOLT-4/CCR5 target cells and followed the time course for appearance of nascent viral RNA in the culture supernatants using RT-qPCR (Fig. 5a ). Supernatants were collected at days 0, 4, 10, and 15 after spin-inoculation and viral spread was observed in all MC and T cell QVOAs for P2, P4 and P5 (Fig. 5b, 5d , 5e). However, for P3 viral spread was only observed in MC, not T cells (Fig. 5c) . We estimated that 0.7 % to 2.2 % of the viral particles present in the MC and T cell QVOAs were replication competent. We also calculated an upper limit of 0.2 % for replication competent virus present in the P3 T cell QVOA supernatants based on the absence of outgrowth (Fig. 5f) .

We used targeted amplicon next generation sequencing to perform a sequence analysis of the V3 regions of HIV env using virus isolated from the QVOA wells from CD4 + T cells, AM and MC from each patient (Fig. 6a, 6b) . Comparisons of the predominant viral sequences isolated from myeloid cells (AM or MC) and T cell QVOAs from the same donor showed 100 % (P2), 100 % (P3), 91 % (P4), and 97 % (P5) pairwise identity.

Further, all major variants analyzed were predicated to use the R5 co-receptor (Supplemental Table S1 ). These data, which only reliably report on the major variants present, do not address the question of whether common or distinct myeloid and T cell viral pools exist. The predicted amino acid sequences of each major variant were matched to a single HIV consensus sequence calculated from 2635 HIV clinical sequences archived in the Los Alamos HIV sequence database 40 . Amino acid substitutions were clustered in various regions with the same pattern for P2 and P3 and more variabilities for P4 and P5 (Fig. 6c ). An APOBEC-induced hypermutation analysis using Hypermut 2.0 (https://www.hiv.lanl.gov/content/sequence/HYPERMUT/hypermut.html) did not reveal any definitive A3G or A3F deamination sites (Supplemental Table S2 ). These studies also provide rare (but limited) measurements of the replication capacity of proviruses present in circulating MC and resident alveolar macrophages of the lung. be transcriptionally repressive and may even contribute to a unique mechanism of latency in long-lived macrophages such as microglial cells of the brain.

An important finding in this study is that rare replication competent viruses can be detected in both AM and peripheral MC from HIV-infected individuals that were virally suppressed using ART. Beyond the obvious need to control for contaminating T cells in AM and MC QVOAs, a frequent critique is that phagocytized infected T cells containing integrated virus could give rise to the observed signals. A phagocytic mechanism is difficult to discount with one hundred percent certainty, but is made unlikely given that on 

Board and informed consent was obtained from all participants before study enrollment.

Peripheral blood was obtained from healthy volunteers and ART suppressed patients with viral loads <20 copies HIV-1 RNA/mL. Monocytes and resting CD4 + T cells were obtained from donor PBMCs as described below. Characteristics of study participants are given in Table 1) . To obtain alveolar macrophages (AM), bronchoscopy and lavage were done, as described elsewhere 48 . Briefly, BAL samples were filtered using a sterilized gauze pad and transferred into 50-mL centrifuge tubes. A pellet was obtained using a short spin (250 × g for 5 min) and was resuspended The PCR product is diluted 20-fold and five L of the diluted PCR product is used as an input material for the second PCR reaction, which is performed using LRT forward and reverse primers and probe using the Rotor-Gene Probe PCR kit (Qiagen) as described above. Proviral copy numbers were determined using the J-lat cell integration standard as previously described 20, 30, 31 . Genomic DNA copy numbers were determined using the RPP30 reference gene using the same amount of input DNA used to measure proviral copy numbers.

Uracil content of viral DNA. Uracil content of viral DNA was determined using uracil excision qPCR method (Ex-qPCR) as previously described with some modifications as described below 20, 30, 31 . Ex-qPCR was used to determine uracil-containing fraction of viral DNA. The sample was split into two equal portions and one portion was treated with UDG.

Briefly, 0.2 units of UDG (NEB) was added into the qPCR master mix to excise uracils from viral DNA. The qPCR reaction was modified to include the UDG reaction time and heat-cleavage of the resulting abasic sites. Thermocycler program for this reaction was: Supplemental Table S3 . Primers and probe sets targeting RPP30 were used to calculate Frac U DNA using the Ct method.

undetectable plasma viral loads) were separated from whole blood by a standard Ficoll gradient. CD4 + T cells were obtained using a negative CD4 magnetic selection kit infected patients and AMs were isolated and purified as described above. The cells were then re-suspended in RPMI medium, seeded in successive 1:5 or 1:10 serial dilutions in triplicate wells of a 6-well plate or 24-well plate (starting from 1 × 10 6 cells), and kept for 72 h at 37 °C with T20 entry inhibitor (10 μM) to prevent reinfection. Next, the cells were washed and co-cultured with MOLT-4/CCR5 cells (0.5 × 10 6 cells) for 2 weeks at 37 °C, with fresh medium containing lipopolysaccharide (Sigma-Aldrich, L5293) replenished after 1 week. HIV-1 in the co-culture supernatants was measured using RT-qPCR.

Infection frequencies were determined by maximum likelihood statistics as described previously, using the application available online at http://silicianolab.johnshopkins.edu, and expressed as infectious units per million (IUPM).

Blood samples were collected from HIV-1 infected patients and MCs were purified and differentiated with T20 entry inhibitor (10 μM) to prevent reinfection as described above. The cells were then re-suspended in RPMI medium, seeded in successive 1:5 or 1:10 serial dilutions in triplicate wells of a 6-well plate or 24-well plate (starting from 1 × 10 6 cells) and co-cultured with MOLT-4/CCR5 cells (0.5 × 106 cells) for two weeks at 37 °C, with fresh medium containing lipopolysaccharide (Sigma-Aldrich, L5293) replenished after 1 week. HIV-1 in the co-culture supernatants was measured using RT-qPCR and IUPM was calculated as above. After cDNA synthesis, a 359 bp segment of the env region was amplified using nested PCR (outer primers E90 and Nesty8 and inner primers DLoop and E115, as described previously) 50 . Thermocycler settings were the same for both outer and inner env PCR The amplification product (20 ng/uL, 50 uL total) were sent to GENEWIZ for targeted amplicon sequencing. Sequencing adapters are listed in Supplemental Table S4 . Unique sequence and abundance raw sequence data were demultiplexed using bcl2fastq version 2.17.1.14. Read pairs were trimmed for adapter sequences and low quality basecalls using Trimmomatic version 0.36. Reads were discarded if they were less than 30 bases long. Each read pair was then merged using the bbmerge tool from the bbtools software toolkit. Reads that could not be merged were discarded from further analysis. The target sequence between conserved flanking primers was extracted from each merged pair. The forward and reverse reads were joined and assigned to samples based on barcode and truncated by cutting off the barcode and primer sequence. Quality filtering on joined sequences was performed (Supplemental Table S4 ). The sequences are grouped according to their abundance and identities (single nucleotide differences were considered different). 

Not applicable

The sequence datasets generated and/or analyzed during the current study are available in the GenBank repository: Sequence Read Archive (SRA); BioProject ID PRJNA832816.

Supporting data generated or analyzed during this study are included in this published article and its supplementary information files. The complete data for this study is accessible for download without any restrictions at: 5-03-2022_CUI_AM_Ms_DATA

The authors declare that they have no competing interests.

This work was supported by NIH grants R01 AI124777 and RO1 GM056834. AM data was collected from one healthy donor and the errors are from three technical replicates. The MDM data is the average from three healthy donors. The fold-change in copy number is relative to the levels before infection and the 18S ribosomal RNA expression was used as the calibration standard for all samples. All primers are listed in Table S1 . Errors are means with standard deviations, N = 2 for AM and N = 3 for MDM (where N is number of donors). 

Biology, and Diseases of Tissue Macrophages

Increased phagocytosis in the presence of enhanced M2-like macrophage responses correlates with increased primary and latent HSV-1 infection

Cell Cycle Regulation in Macrophages and Susceptibility to HIV-1

APOBEC3A cytidine deaminase induces RNA editing in monocytes and macrophages

APOBEC3A is a specific inhibitor of the early phases of HIV-1 infection in myeloid cells

dNTP pool modulation dynamics by SAMHD1 protein in monocyte-derived macrophages

Effect of induced dNTP pool imbalance on HIV-1 reverse transcription in macrophages

A G1-like state allows HIV-1 to bypass SAMHD1 restriction in macrophages

TLR4-Mediated Pathway Triggers Interferon-Independent G0 Arrest and Antiviral SAMHD1 Activity in Macrophages

The human heart contains distinct macrophage subsets with divergent origins and functions

Monocyte-derived macrophages and myeloid cell lines as targets of HIV-1 replication and persistence

Tissue macrophages: heterogeneity and functions

Long-term persistence of human donor alveolar macrophages in lung transplant recipients

The Lifespan and Turnover of Microglia in the Human Brain

The Interplay of HIV-1 and Macrophages in Viral Persistence

Monocytes harbour replication-competent, non-latent HIV-1 in patients on highly active antiretroviral therapy

Continued evolution of HIV-1 circulating in blood monocytes with antiretroviral therapy: genetic analysis of HIV-1 in monocytes and CD4+ T cells of patients with discontinued therapy

Blood monocytes harbor HIV type 1 strains with diversified phenotypes including macrophage-specific CCR5 virus

Diverse fates of uracilated HIV-1 DNA during infection of myeloid lineage cells

Reversible Human Immunodeficiency Virus Type-1 Latency in Primary Human Monocyte-Derived Macrophages Induced by Sustained M1 Polarization

Compartmentalization of human immunodeficiency virus type 1 between blood monocytes and CD4+ T cells during infection

Infectious Virus Persists in CD4+ T Cells and Macrophages in Antiretroviral Therapy-Suppressed Simian Immunodeficiency Virus-Infected Macaques

Tissue-resident macrophages can contain replicationcompetent virus in antiretroviral-naive, SIV-infected Asian macaques

Macrophages sustain HIV replication in vivo independently of T cells

HIV persistence in tissue macrophages of humanized myeloidonly mice during antiretroviral therapy

Microglial Cells: The Main HIV-1 Reservoir in the Brain

Lentiviral Infections Persist in Brain despite Effective Antiretroviral Therapy and Neuroimmune Activation

Tight interplay among SAMHD1 protein level, cellular dNTP levels, and HIV-1 proviral DNA synthesis kinetics in human primary monocyte-derived macrophages

Deficient uracil base excision repair leads to persistent dUMP in HIV proviruses during infection of monocytes and macrophages

Uracil DNA glycosylase initiates degradation of HIV-1 cDNA containing misincorporated dUTP and prevents viral integration

Abundant non-canonical dUTP found in primary human macrophages drives its frequent incorporation by HIV-1 reverse transcriptase

HIV-related proteins prolong macrophage survival through induction of Triggering receptor expressed on myeloid cells-1

Deoxyuridine in DNA has an inhibitory and promutagenic effect on RNA transcription by diverse RNA polymerases

Incorporation of uracil into minus strand DNA affects the specificity of plus strand synthesis initiation during lentiviral reverse transcription

IFN-independent G0 arrest and SAMHD1 activation following TLR4 activation in macrophages

SAMHD1-Mediated Negative Regulation of Cellular dNTP Levels: HIV-1, Innate Immunity, and Cancers

Measuring the Frequency of Latent HIV-1 in Resting CD4+ T Cells Using a Limiting Dilution Coculture Assay

Quantitation of Productively Infected Monocytes and Macrophages of Simian Immunodeficiency Virus-Infected Macaques

Peculiarities of recognition of CCA/TGG sequences in DNA by restriction endonucleases MvaI and EcoRII

REBASE-a database for DNA restriction and modification: enzymes, genes and genomes

Endogenous DNA lesions can inhibit the binding of the AP-1 (c-Jun) transcription factor

Defective repair of uracil causes telomere defects in mouse hematopoietic cells

Minimal differentiation of classical monocytes as they survey steady-state tissues and transport antigen to lymph nodes

Cellular pharmacology and potency of HIV-1 nucleoside analogs in primary human macrophages

Antiretroviral therapy in macrophages: implication for HIV eradication

Activation-induced cell death drives profound lung CD4(+) T-cell depletion in HIV-associated chronic obstructive pulmonary disease

Effect of the thymidylate synthase inhibitors on dUTP and TTP pool levels and the activities of DNA repair glycosylases on uracil and 5-fluorouracil in DNA

Sequencing Primers for HIV-1. Human retroviruses and AIDS

Geno2pheno: estimating phenotypic drug resistance from HIV-1 genotypes

Improved coreceptor usage prediction and genotypic monitoring of R5-to-X4 transition by motif analysis of human immunodeficiency virus type 1 env V3 loop sequences

The authors acknowledge the support of the Johns Hopkins Institute for Clinical and Translational Research team and especially Ms. Christina Bunch for recruiting, consenting, and performing blood draws from healthy subjects. The authors thank Dr.Rebecca Veenhuis (Johns Hopkins School of Medicine) for QVOA protocol optimization.