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

Abstract

Understanding the cellular and anatomical sites of latent virus that contribute to human immunodeficiency virus (HIV) rebound is essential for eradication. In HIV-positive patients, CD4⁺ T lymphocytes comprise a well-defined functional latent reservoir, defined as cells containing transcriptionally silent genomes able to produce infectious virus once reactivated. However, the persistence of infectious latent virus in CD4⁺ T cells in compartments other than blood and lymph nodes is unclear. Macrophages (Mϕ) are infected by HIV/simian immunodeficiency virus (SIV) and are likely to carry latent viral genomes during antiretroviral therapy (ART), contributing to the reservoir. Currently, the gold standard assay used to measure reservoirs containing replication-competent virus is the quantitative viral outgrowth assay (QVOA). Using an SIV-macaque model, the CD4⁺ T cell and Mϕ functional latent reservoirs were measured in various tissues using cell-specific QVOAs. Our results showed that blood, spleen, and lung in the majority of suppressed animals contain latently infected Mϕs. Surprisingly, the numbers of CD4⁺ T cells, monocytes, and Mϕs carrying infectious genomes in blood and spleen were at comparable frequencies (∼1 infected cell per million). We also demonstrate that ex vivo viruses produced in the Mϕ QVOA are capable of infecting activated CD4⁺ T cells. These results strongly suggest that latently infected tissue Mϕs can reestablish productive infection upon treatment interruption. This study provides the first comparison of CD4⁺ T cell and Mϕ functional reservoirs in a macaque model. It is the first confirmation of the persistence of latent genomes in monocytes in blood and Mϕs in the spleen and lung of SIV-infected ART-suppressed macaques. Our results demonstrate that transcriptionally silent genomes in Mϕs can contribute to viral rebound after ART interruption and should be considered in future HIV cure strategies.IMPORTANCE This study suggests that CD4⁺ T cells found throughout tissues in the body can contain replication-competent SIV and contribute to rebound of the virus after treatment interruption. In addition, this study demonstrates that macrophages in tissues are another cellular reservoir for SIV and may contribute to viral rebound after treatment interruption. This new insight into the size and location of the SIV reservoir could have great implications for HIV-infected individuals and should be taken into consideration for the development of future HIV cure strategies.

Keywords: CD4+ T cells; latency; macrophages; quantitative viral outgrowth assay; reservoir; simian immunodeficiency virus.

FIG 1

Viral load in plasma and CSF of seven suppressed SIV-infected ART-treated macaques (34). Seven SIV-infected pigtailed macaques were treated with similar ART regimens (tenofovir, darunavir, integrase inhibitor, and ritonavir). Symbols represent the plasma (A) and CSF (B) viral load values for each animal. Analyses of samples with values below the limit of quantitation (LOQ) for the SIV RT-qPCR assay (430 SIV RNA copies/ml; top dotted line) were repeated using RT-ddPCR (LOQ, 5 SIV RNA copies/ml; bottom dotted line).

FIG 2

Quantification of SIV DNA and SIV RNA from tissue samples. Cellular DNA and RNA were extracted from spleen, lung, and PBMCs from seven SIV-infected ART-suppressed macaques. SIV gag DNA (left) and SIV gag RNA (right) were measured by ddPCR, with the limit of quantitation (LOQ, dashed lines) being 1 copy per reaction mixture. Open symbols indicate animals that were treated with LRAs in vivo; closed symbols indicate animals that were not treated with LRAs.

FIG 3

Quantification of SIV DNA and RNA from CD4⁺ T cells and CD11b⁺ cells isolated from tissues. CD4⁺ T cells (A and B) and CD11b⁺ cells (C, D, and E) were isolated from spleen, lung, and PBMCs from seven SIV-infected ART-suppressed macaques and two SIV-infected untreated macaques. Cellular DNA and RNA were then extracted and analyzed for SIV gag DNA (left), SIV gag RNA (middle), and SIV tat/rev RNA (right) by PCR. The limit of quantitation (LOQ) for ddPCR is 1 copy per reaction, and that for qPCR is 10 copies per reaction. The dashed lines represent the LOQ for ddPCR. Open symbols indicate animals that were treated with LRAs in vivo; closed symbols indicate animals that were not treated with LRAs.

FIG 4

SIV DNA and RNA levels were similar between CD4⁺ T cells and macrophages or monocytes isolated from tissues. Comparison of SIV DNA and SIV RNA levels in CD4⁺ T cells and CD11b⁺ cells isolated from spleen (A) and blood (B). The dotted line represents the limit of quantitation for ddPCR. Each symbol represents an animal. Open symbols indicate animals that were treated with LRAs in vivo; closed symbols indicate animals that were not treated with LRAs.

FIG 5

In vivo LRA treatment does not have a measurable effect on the level of SIV gag DNA or RNA in isolated cells. SIV gag DNA (A) and RNA (B) levels measured in cells isolated from SIV-infected macaques treated with LRA were compared to those measured in cells isolated from SIV-infected macaques without LRA treatment. CD4⁺ T cells were isolated from blood and spleen. CD11b⁺ cells (monocytes/macrophages) were isolated from blood, spleen, and lung.

FIG 6

The macrophage population in culture is stable over time. IFN-β gene (A, C, and E) and SIV gag (B, D, and F) DNA copies were quantitated by qPCR in CD11b⁺ cells isolated from spleen, lung, and PBMCs before being plated for Mϕ QVOA and also from cells in the QVOA wells at 14 or 21 days postplating. Values are presented as the number of copies per 10⁶ cells. Open symbols indicate animals that were treated with LRAs in vivo; closed symbols indicate animals that were not treated with LRAs. D0, D14, and D21, days 0, 14, and 21, respectively.

FIG 7

Functional latent reservoirs are detected in both CD4⁺ T cells and monocytes/macrophages isolated from SIV-infected ART-suppressed animal tissues. The frequency of latently infected CD4⁺ T cells (left) and monocytes/macrophages (right) isolated from the spleen, lung, and PBMCs of suppressed ART-treated macaques were quantitated by cell-specific QVOAs. The horizontal black lines represent the median number of infectious units per million cells (IUPM), and each symbol represents an animal. Open symbols indicate animals that were treated with LRAs in vivo; closed symbols indicate animals that were not treated with LRAs. LOD, limit of detection.

FIG 8

In vivo LRA treatment has no measurable effect on IUPM. A comparison of the IUPM values obtained from cells isolated from SIV-infected macaques treated with LRA compared to those obtained from cells isolated from SIV-infected macaques without LRA treatment is shown. (A) Number of IUPM from CD4⁺ T cells isolated from spleen; (B) number of IUPM from CD4⁺ T cells isolated from blood; (C) number of IUPM from macrophages isolated from spleen; (D) number of IUPM from monocyte-derived macrophages isolated from blood; (E) number of IUPM from alveolar macrophages isolated from lung.

FIG 9

Supernatants from Mϕ QVOA cultures contain viruses that establish de novo infection in activated macaque PBMCs. Activated PBMCs from a healthy pigtailed macaque were spinoculated with culture supernatant from positive Mϕ QVOA wells with spleen tissue (A), lung tissue (B), and blood (C) from the seven suppressed macaques. The graphs show the viral kinetics over time postinfection, as measured by an SIV RNA RT-qPCR. Each line represents viral replication from one Mϕ QVOA well.