Sequential treatment with 5-aza-2'-deoxycytidine and deacetylase inhibitors reactivates HIV-1

Abstract

Reactivation of HIV gene expression in latently infected cells together with an efficient cART has been proposed as an adjuvant therapy aimed at eliminating/decreasing the reservoir size. Results from HIV clinical trials using deacetylase inhibitors (HDACIs) question the efficiency of these latency-reversing agents (LRAs) used alone and underline the need to evaluate other LRAs in combination with HDACIs. Here, we evaluated the therapeutic potential of a demethylating agent (5-AzadC) in combination with clinically tolerable HDACIs in reactivating HIV-1 from latency first in vitro and next ex vivo. We showed that a sequential treatment with 5-AzadC and HDACIs was more effective than the corresponding simultaneous treatment both in vitro and ex vivo. Interestingly, only two of the sequential LRA combinatory treatments tested induced HIV-1 particle recovery in a higher manner than the drugs alone ex vivo and at concentrations lower than the human tolerable plasmatic concentrations. Taken together, our data reveal the benefit of using combinations of 5-AzadC with an HDACI and, for the first time, the importance of treatment time schedule for LRA combinations in order to reactivate HIV.

Keywords: HDACIs; HIV latency; HIV reservoir; demethylating agent; epigenetics.

Figure 1. The DNA methylation inhibitor 5‐AzadC induces HIV‐1 expression in latently infected T cells

1. A–D

   J‐Lat 8.4 cells were mock‐treated or treated with increasing concentrations of 5‐AzadC or 5‐AzaC. At 72 h post‐treatment, viral production was measured by quantifying p24 antigen production in culture supernatants (A); metabolic activity was assessed by a WST‐1 assay (B); viral protein expression was analyzed by FACS (C); and initiated (primers TAR) or elongated (primers tat) transcripts were quantified by RT–qPCR (D). The selected dose was indicated by an arrow.

2. E

   J‐Lat 8.4 cells were mock‐treated or treated with 5‐AzadC (400 nM) or TNF‐α (10 ng/ml) as a positive control. At 24, 48 or 72 h post‐treatment, initiated (primers TAR) or elongated (primers tat) transcripts were quantified by RT–qPCR.

Data information: For (D, E), results were normalized using β‐actin gene primers and are presented as histograms indicating the fold inductions compared to mock‐treated condition for each time period. For (A–E), means and standard errors of the means from three independent biological duplicates (n = 6) are indicated. The result obtained with mock‐treated cells was arbitrarily set at a value of 1 (D, E) and 100% (B).

Figure 2. Determination of an optimal concentration for each HDACI to induce HIV‐1 production in latently infected cells

1. A–D

   J‐Lat 8.4 cells were mock‐treated or treated with increasing concentrations of HDACIs. At 24 h post‐treatment, viral production was measured by quantifying p24 antigen production in culture supernatants (A, B) and metabolic activity was assessed by a WST‐1 assay (C, D). Means and standard errors of the means from three independent biological duplicates (n = 6) are indicated. The result obtained with mock‐treated cells was arbitrarily set at a value of 100% (C, D). The selected doses are indicated by an arrow. Plasmatic concentrations after usual dosage (C _max) of each drug in human therapy are indicated below the LRA names and by a box in the graph.

Figure 3. Determination of 5‐AzadC + SAHA treatment schedule in vitro and ex vivo

1. A–D

   J‐Lat 8.4 cell line was mock‐treated or treated with 5‐AzadC and/or SAHA for different periods of time as indicated. Samples were harvested at the indicated times. Viral production was measured by quantifying p24 antigen production in culture supernatants (A, B) and metabolic activity was assessed by a WST‐1 assay (C, D). Means and standard errors of the means from three independent biological duplicates (n = 6) are indicated. The result obtained with mock‐treated cells was arbitrarily set at a value of 100% (C, D).

2. E

   From data of ex vivo cultures of CD8⁺‐depleted PBMCs isolated from 24 HIV ⁺ patients presented in Appendix Table S2, the extracellular HIV‐1 genomic RNA levels for each LRA treatment are represented. One night after cell purification, cells were mock‐treated or simultaneously treated with 5‐AzadC (1 μM) and/or SAHA (1 μM). Six days after treatment, the concentration of viral RNA in culture supernatants was determined (in copies/ml). The results were reported as the actual HIV RNA copy numbers/ml or as an estimated value calculated as 50% of the smallest value when HIV RNA was not detected in order to assign a log value. Means are represented. Nonparametric one‐way ANOVA for independent samples (Kruskal–Wallis) followed by paired comparisons between each treated condition and the mock‐treated condition (Mann–Whitney test) are performed.

Figure 4. Sequential 5‐AzadC + HDACI treatments synergistically activate HIV‐1 gene expression and production in latently infected cells

1. A–F

   J‐Lat 8.4 (A, C, E) and 15.4 (B, D, F) cell lines were mock‐treated or treated with 5‐AzadC. At 48 h post‐treatment, HDACIs were then added for 24 h. At 72 h 5‐AzadC post‐treatment, samples were harvested and analyzed as follows: viral p24 production in the cell supernatant was measured (A, B); FACS analyses were performed and the percentages of GFP ⁺ cells are presented as histograms (C, D); initiated (primers TAR) or elongated (primers tat) transcripts were quantified by RT–qPCR, and results were normalized using the β‐actin gene primers and are presented as histograms representing fold inductions compared to the mock‐treated condition (E, F). Means and standard errors of the means from three independent biological duplicates (n = 6) are indicated. The result obtained with mock‐treated cells was arbitrarily set at a value of 1 (E, F).

Figure 5. Sequential 5‐AzadC + HDACI treatments induced a variation of cellular metabolic activity depending on the treatment in CD8⁺‐depleted PBMCs from HIV‐negative donors

One night after cell purification, CD8⁺‐depleted PBMCs from 12 HIV‐negative donors were mock‐treated or treated with 5‐AzadC. Three days post‐treatment, 1/3 of medium was replaced and HDACIs were added to the cultures. Six days after 5‐AzadC treatment, WST‐1 assay was performed. Mock value was arbitrarily fixed at 100% for each individual. Means are indicated by a line.

Figure 6. Representation of reactivation status of ex vivo cultures of CD8⁺‐depleted PBMCs isolated from HIV ⁺ patients

1. A–C

   From data of ex vivo cultures of CD8⁺‐depleted PBMCs isolated from HIV ⁺ patients (Table 2), the extracellular HIV‐1 genomic RNA levels for each LRA treatment are represented from all patient cell cultures (A), from patient cell cultures presenting no viral reactivation in mock condition (C), and from patient cell cultures exhibiting reactivation in mock condition (B). One night after cell purification, cells were mock‐treated or treated with 5‐AzadC. Three days post‐treatment, 1/3 of medium was replaced and HDACIs were added to the cultures. Six days after 5‐AzadC treatment, the concentration of viral RNA in culture supernatants was determined (in copies/ml). The results were reported as the actual HIV RNA copy numbers/ml or as an estimated value calculated as 50% of the smallest value when HIV RNA was not detected in order to assign a log value. Means are represented. Nonparametric one‐way ANOVA for independent samples (Kruskal–Wallis) followed by paired comparisons between each treated condition and the mock‐treated condition (Mann–Whitney test) are performed.

Figure 7. Comparison of simultaneous 5‐AzadC + SAHA combined treatment with its corresponding sequential treatment in ex vivo cultures of CD8⁺‐depleted PBMCs isolated from aviremic cART‐treated HIV ⁺ patients

1. Reactivation status of patient cell cultures presenting no viral reactivation in mock condition isolated from aviremic cART‐treated HIV ⁺ patients (shown in Appendix Table S2).

2. Reactivation status of patient cell cultures presenting no viral reactivation in mock condition isolated from aviremic cART‐treated HIV ⁺ patients (shown in Table 2).

Data information: The results were reported as the actual HIV RNA copy numbers/ml or as an estimated value calculated as 50% of the smallest value when HIV RNA was not detected in order to assign a log value. Means are represented by a line. Nonparametric one‐way ANOVA for independent samples (Kruskal–Wallis) followed by paired comparisons between each treated condition and the mock‐treated condition (Mann–Whitney test) are performed.

Figure 8. Representation of reactivation status of ex vivo cultures of resting CD4⁺ T cells

1. A–C

   From data of ex vivo cultures of resting CD4⁺ T cells isolated from HIV ⁺ patients (Table 3), the extracellular HIV‐1 genomic RNA levels for each LRA treatment are represented from all patient cell cultures (A), from patient cell cultures presenting no viral reactivation in mock condition (C), and from patient cell cultures exhibiting reactivation in mock condition (B). One night after cell purification, cells were mock‐treated or treated with 5‐AzadC. Three days post‐treatment, 1/3 of medium was replaced and HDACIs were added in the cultures. Six days after 5‐AzadC treatment, the concentration of viral RNA in culture supernatants was determined (in copies/ml). The results were reported as the actual HIV RNA copy numbers/ml or as an estimated value calculated as 50% of the smallest value when HIV RNA was not detected in order to assign a log value. Means are represented. Nonparametric one‐way ANOVA for independent samples (Kruskal–Wallis) followed by paired comparisons between each treated condition and the mock‐treated condition (Mann–Whitney test) are performed.

Figure 9. 5‐AzadC + panobinostat and 5‐AzadC + romidepsin treatments do not induce global T‐cell activation

1. A–H

   One night after cell purification, CD8⁺‐depleted PBMCs (E–H) or HLA DR ⁻ CD69⁻ CD25⁻ CD4⁺ T cells (A–D) from 4 HIV‐negative donors were mock‐treated or treated with 5‐AzadC. Three days post‐treatment, 1/3 of medium was replaced and HDACIs were added to the cultures. The activation status of CD4⁺ T‐cell subset was assessed 6 days after 5‐AzadC treatment by flow cytometry analysis of cellular activation markers relative to mock treatment before 5‐AzadC stimulation corresponding to day 0. Means are represented. Nonparametric one‐way ANOVA for independent samples (Kruskal–Wallis) followed by paired comparisons between each treated condition and the mock‐treated condition (Mann–Whitney test) are performed.

Figure 10. 5‐AzadC + panobinostat and 5‐AzadC + romidepsin treatments induce a significant decrease in the cell surface CD4 receptor expression

One night after cell purification, HLA DR ⁻ CD69⁻ CD25⁻ CD4⁺ T cells from 4 HIV‐negative donors were mock‐treated or treated with 5‐AzadC. Three days post‐treatment, 1/3 of medium was replaced and HDACIs were added to the cultures. The median fluorescence intensity of CD4 receptor of viable CD4⁺ T‐cell subset was assessed 6 days after 5‐AzadC treatment by flow cytometry analysis relative to mock treatment before 5‐AzadC stimulation corresponding to day 0. Means are represented. Nonparametric one‐way ANOVA for independent samples (Kruskal–Wallis) followed by paired comparisons between each treated condition and the mock‐treated condition (Mann–Whitney test) are performed.