Medroxyprogesterone acetate, unlike norethisterone, increases HIV-1 replication in human peripheral blood mononuclear cells and an indicator cell line, via mechanisms involving the glucocorticoid receptor, increased CD4/CD8 ratios and CCR5 levels

Abstract

High usage of progestin-only injectable contraceptives, which include the intramuscular injectables depo-medroxyprogesterone acetate (DMPA-IM, Depo-Provera) and norethisterone (NET) enanthate (NET-EN or Nur-Isterate), correlates worldwide with areas of high HIV-1 prevalence. Epidemiological data show a significant association between usage of DMPA-IM and increased HIV-1 acquisition but no such association from limited data for NET-EN. Whether MPA and NET have similar effects on HIV-1 acquisition and pathogenesis, and the relationship between these effects and the dose of MPA, are critical issues for women's health and access to suitable and safe contraceptives. We show for the first time that MPA, unlike NET, significantly increases HIV-1 replication in peripheral blood mononuclear cells (PBMCs) and a cervical cell line model. The results provide novel evidence for a biological mechanism whereby MPA, acting via the glucocorticoid receptor (GR), increases HIV-1 replication by at least in part increasing expression of the CCR5 HIV-1 coreceptor on target T-lymphocytes. MPA, unlike NET, also increases activation of T-cells and increases the CD4/CD8 ratio, suggesting that multiple mechanisms are involved in the MPA response. Our data offer strong support for different biological mechanisms for MPA versus NET, due to their differential GR activity. The dose-dependence of the MPA response suggests that significant effects are observed within the range of peak serum levels of progestins in DMPA-IM but not NET-EN users. Dose-response results further suggest that effects of contraceptives containing MPA on HIV-1 acquisition and disease progression may be critically dependent on dose, time after injection and intrinsic factors that affect serum concentrations in women.

Fig 1. MPA, unlike NET, dose-dependently increases R5 HIV-1 replication in PBMCs in the range of peak serum levels of DMPA users, via a GR-dependent mechanism.

(a-d) Non-activated PBMCs were pre-treated with various ligands at the concentrations indicated, or with vehicle control, (0.1% v/v EtOH) for 24 hours. Thereafter, PBMCs were infected with 10 IU/mL HIV-1_BaL-Renilla IMC for 2 hours and maintained in full RPMI supplemented with 30 IU/mL IL2 for 5 days before harvesting for Renilla luciferase or MTT viability. Relative infection for each donor for progestins relative to vehicle, where the only variable was the absence or presence of progestin, was calculated by normalizing RLU against corresponding MTT values, and calibrated to vehicle set to 100% relative infection. Each condition was performed at least in triplicate. The data (a-d) are represented as mean ± SEM. In (a-c), individual experimental means are depicted as black dots. Statistical analysis was performed using (a-c) a non-parametric Kruskal-Wallis one-way ANOVA with a subsequent Wilcoxon non-parametric t- test to compare vehicle and MPA or (b) a Mann Whitney test when comparing MPA to RU486/MPA. (d) A non-linear regression line was generated to calculate EC₅₀ and statistical analysis was performed using a parametric one-way ANOVA with Tukey multiple comparisons post-test comparing all conditions. Statistical significance is indicated with * or ** denoting p<0.05 or p<0.01, respectively. In a) the dexamethasone (Dex) (a synthetic GR agonist) result is from 6 donors, MPA is from 14 donors and NET is from 8 donors, b shows results for 6 donors and c-d show results for 10 donors, where each point on the histogram shows the result from a separate donor.

Fig 2. MPA, unlike NET, increases CCR5 levels and activation of T-cells in PBMCs after 24 hours.

PBMCs were stimulated with 100 nM MPA, 100 nM NET or vehicle for 24 hours after which relative levels of CD25 and CCR5 were determined in CD3+, CD4+ and CD8+ T-cells using flow cytometry. Results are shown as either relative frequency of CD3+, CD4+ and CD8+ T-cells (a), CD4/CD8 ratios (b), relative frequency of cells expressing CD25 (c) or CCR5 (e), or relative MFI of CD25 (d) or CCR5 (f) in CD3+, CD4+ and CD8+ T-cells. (a, c-f) show results from 10 independent donor experiments, while (b) shows CD4/CD8 ratios for 8 independent donor experiments. The data (a-f) are represented as mean ± SEM. Each point on the histogram shows the result from a separate donor, with experiments for each donor performed with vehicle, MPA and NET in parallel, and the value for vehicle set to 1. Statistical significance was determined by using a non-parametric Kruskal-Wallis one-way ANOVA with Dunn’s post-test with * or ** denoting p<0.05 or p<0.01 respectively.

Fig 3. Longer exposure to MPA increases the CD4/CD8 ratio, decreases CD8+ T-cell frequency, and increases the density of the CCR5 coreceptor on CD4+ and CD8+ T-cells.

PBMCs were stimulated with 100 nM MPA vehicle for 7 days, after which relative levels of CD69 and CCR5 were determined in CD3+, CD4+ and CD8+ T-cells and CD14+ monocytes using flow cytometry. Results are shown as either relative frequency of CD3+, CD4+ and CD8+ T-cells and CD14+ monocytes (a), CD4/CD8 ratio (b), relative frequency of cells expressing CD69 (c) or CCR5 (e), or relative MFI of CD69 (d) or CCR5 (f) in CD3+, CD4+ and CD8+ T-cells and CD14+ monocytes. Results are shown from 8 independent donor experiments. The data (a-f) are represented as mean ± SEM. Each point on the histograms shows the result from a separate donor, with experiments for each donor performed with vehicle and MPA in parallel, with the values for vehicle set to 1. Data were analyzed using a parametric unpaired t-test comparing vehicle to MPA. Statistical significance is shown with *, ** or *** denoting p<0.05, p<0.01 or p<0.001, respectively.

Fig 4. MPA, unlike NET, dose-dependently increases R5 HIV-1 replication in the TZM-bl cervical cell line, via a GR-dependent mechanism.

TZM-bl cells were stimulated in parallel for 24 hours with the various ligands, or combinations thereof or vehicle control (0.1% v/v EtOH), at the concentrations indicated, before infection with 20 IU/mL R5 HIV-1_BaL-Renilla. Samples were harvested 48 hours later for Renilla luciferase and MTT viability assays. Each condition was performed at least in triplicate. Relative infection was calculated by normalising RLU against corresponding MTT values. This value for vehicle was set to 1 for (a) and (b), while for (c), data were analysed relative to the maximal response generated by MPA which was set to 100%. Statistical analysis was performed using (a) a parametric one-way ANOVA with Dunnett’s multiple comparisons post test when comparing to vehicle or (b) a non-parametric Kruskal-Wallis one-way ANOVA with a subsequent unpaired t test comparing vehicle and MPA or comparing MPA and MPA/RU486. (c) A non-linear regression line was generated to calculate EC₅₀ and statistical analysis was performed using a parametric one-way ANOVA with Tukey multiple comparisons post-test comparing all conditions. Significant differences are shown by *, ** or *** denoting p<0.05, p<0.01 or p<0.001, respectively. The data (a-c) are represented as mean ± SEM and show the results of nine, five and six independent experiments, respectively, with each point performed in triplicate at least. (d-e) The GR is the predominant steroid receptor protein expressed in the TZM-bl cell line. Cell lysates were prepared and the steroid receptor mRNA and protein levels were detected by qRT-PCR and western blotting, respectively. (d) indicates that TZM-bl cells express detectable GR, AR and MR mRNA while (e) shows that TZM-bl cells express detectable GR and AR protein.

Fig 5. MPA-induced HIV-1 replication requires the GR in TZM-bl cells.

TZM-bl cells transfected with 10 nM non-silencing control (NSC) or siRNA targeting the human GR (siGR) for 48 hours. (a) Cells were re-seeded into 96-well plates for 24 hours, followed by stimulation for 24 hours with 100 nM MPA or vehicle (0.1% v/v EtOH), then infected with 10 IU/mL HIV-1_{BaL_Renilla} (“HIV-BaL”) or equivalent volume of virus control (“Uninfected”). Cells were harvested 72 hours later for luciferase (infection) and for MTT (cell viability). The results were pooled from 3 independent experiments where each point was in quadruplicate and are represented as mean ± SEM. Relative infection was calculated as luciferase (RLU) divided by average absorbance at 595nm (MTT) for the quadruplicates. Infection was plotted relative to HIV-1_{BaL_Renilla} vehicle control set to 1. Statistical comparisons were carried out using a two-way ANOVA with Tukey’s multiple comparisons post-test, with **** denoting p<0.001. (b) Cells seeded and transfected in parallel were harvested 48 hours after transfection in SDS sample buffer. Lysates were analyzed for GR levels by western blotting using GAPDH as a loading control. A representative western blot is shown.

Fig 6. Increased R5 HIV-1 replication with MPA in TZM-bl cells requires the CCR5 coreceptor.

TZM-bl cells were treated for 24 hours with 100 nM MPA, 100 nM NET or vehicle control (EtOH, 0.1% v/v). Cells were infected with 20 IU/mL of HIV-1_BaL-Renilla or control in the presence of progestins in the absence or presence of varying concentrations of MVC or DMSO. Cells were harvested 48 hours later and infection determined with BrightGlo luciferase. Cell viability was measured by MTT assay and read at an absorbance of 595 nm. Luciferase readings were expressed over MTT and infection determined relative to HIV-1_BaL-Renilla vehicle control (EtOH/DMSO, 0.1% v/v) set to 1. Results show MVC dose- response curves for four independent experiments, each performed in triplicate and is represented as mean ± SEM. Non-linear regression was used and generated best fit slopes with R² values of 0.87, 0.81 and 0.63 for vehicle, 100 nM MPA or NET, respectively, to determine IC₅₀ of MVC. Statistical significance was assessed using a two-way ANOVA with a post hoc Tukey tests between MPA and NET to the vehicle control, or MPA compared to NET with *, ** denoting p<0.05, p<0.001 respectively.

Fig 7. MPA, unlike NET, increases CD4 and CCR5 mRNA levels in TZM-bls, via a GR-dependent mechanism.

TZM-bl cells were stimulated for 24 hours with indicated ligands (a-b) 100 nM MPA, 100 nM NET or vehicle control (EtOH) (0.1% v/v ethanol) or (c-d) 10 nM MPA or (e-f) 100 nM MPA, 100 nM RU486 (RU), or combinations thereof (RU/MPA). RNA was isolated, cDNA was synthesised and relative CCR5, CD4 mRNA levels were determined by real time qPCR, normalized to GAPDH and relative fold change in expression was determined by setting vehicle control to 1. Results are shown from 6 independent experiments for each panel and is represented as mean ± SEM. Statistical significance was determined by using (a-b, e-f) one-way ANOVA with a post hoc Tukey tests between conditions or (c-d) using a parametric unpaired t-test comparing the vehicle to 10 nM MPA with * or ** denoting p<0.05 or p<0.01 respectively.