Screening of an FDA-approved compound library identifies levosimendan as a novel anti-HIV-1 agent that inhibits viral transcription

Abstract

Combination antiretroviral therapy (cART) has been proven to efficiently inhibit ongoing replication of human immunodeficiency virus type 1 (HIV-1), and significantly improve the health outcome in patients of acquired immune deficiency syndrome (AIDS). However, cART is unable to cure HIV-1/AIDS. Even in presence of cART there exists a residual viremia, contributed from the viral reservoirs of latently infected HIV-1 proviruses; this constitutes a major hurdle. Currently, there are multiple strategies aimed at eliminating or permanently silence these HIV-1 latent reservoirs being intensely explored. One such strategy, a recently emerged "block and lock" approach is appealing. For this approach, so-called HIV-1 latency-promoting agents (LPAs) are used to reinforce viral latency and to prevent the low-level or sporadic transcription of integrated HIV-1 proviruses. Although several LPAs have been reported, there is still a question of their suitability to be further developed as a safe and valid therapeutic agent for the clinical use. In this study, we aimed to identify new potential LPAs through the screening an FDA-approved compound library. A new and promising anti-HIV-1 inhibitor, levosimendan, was identified from these screens. Levosimendan is currently used to treat heart failure in clinics, but it demonstrates strong inhibition of TNFα-induced HIV-1 reactivation in multiple cell lines of HIV-1 latency through affecting the HIV-1 Tat-LTR transcriptional axis. Furthermore, we confirmed that in primary CD4⁺ T cells levosimendan inhibits both the acute HIV-1 replication and the reactivation of latent HIV-1 proviruses. As a summary, our studies successfully identify levosimendan as a novel and promising anti-HIV-1 inhibitor, which should be immediately investigated in vivo given that it is already an FDA-approved drug.

Keywords: CD4(+) T cell; FDA-approved compound; HIV cure; Levosimendan; Transcriptional suppression.

Figure 1. Screening for FDA-approved compounds that inhibit reactivation of latent HIV-1

(A) Schematic representation of the compound screening. J-LAT A2 cells were co-treated with compounds (5 μM) and TNFα (10 ng/ml) for 24 hr. Cells were then stained with Hoechst 33342 followed by imaging analysis. The assay was performed in quadruplicate. (B) Fluorescence images showing the reactivation (GFP, green) and cell nucleus (blue) in J-LAT A2 cells treated with TNFα for the indicated compounds. Values represent the mean ± s.d. of the % GFP-positive cells (n ≥ 4). (C) Scatter plot of the % GFP-positive cells vs % cell numbers for all tested compounds. Results were normalized to DMSO control. LSM, levosimendan; SPR, spironolactone; 9AA, 9-aminoacridine.

Figure 2. Effect of anti-HIV-1 inhibitor candidates on viral reactivation in HIV-1 latency cell lines

(A) J-LAT A2 cells were co-treated with compounds at the indicated concentrations and TNFα (10 ng/ml) for 24 hr. The GFP-positive cells were analyzed by flow cytometry. Cell viability was measured using CellTiter-Glo assay. The % of GFP-positive cells and % cell viability were plotted for each compound. Results were normalized to DMSO control. (B–E) The % of GFP-positive cells upon compound treatment were determined for the indicated HIV-1 latency cell lines (B. J-LAT 6.3; C. J-LAT 10.6; D. CA5; E. EF7). Results were normalized to DMSO control. Values represent the mean ± s.d. (n = 4–6). * p < 0.05, one-way ANOVA followed by Tukey’s multiple-comparison test.

Figure 3. Levosimendan inhibits HIV-1 Tat-LTR mediated transcription

(A, B) Indicated HIV-1 latency cells (A. J-LAT 10.6; B. CA5) were treated with levosimendan or spironolactone (5, 10, or 20 μM) in combination with TNFα (10 ng/ml). After 24 hr, mRNA was extracted, and reverse transcribed to cDNA for measurement of HIV-1 initiated or elongated (proximal [Pro], intermediate [Int], and distal [Dis]) transcripts by qPCR. Results were normalized to DMSO control. Values represent the mean ± s.d. (n = 4–6). (C) JLTRG cells were pre-treated with indicated compounds for 6 hr and transduced with either pQCXIP empty or pQCXIP-Tat for 48 hr. The % of GFP-positive cells was determined by flow cytometry. Results were normalized to DMSO control (Tat+). Values represent the mean ± s.d. (n = 4–6). (D) Jurkat cells were pre-treated with indicated compounds (10 μM) for 6 hr, then infected with HIV-1 IIIB viruses for 18 hr. HIV-1 reverse-transcriptase inhibitor (AZT, 25 μM) and integrase inhibitor (raltegravir [RAL], 1 μM) were also used. The DNA was extracted for measurement of late RT products by qPCR. Results were normalized to DMSO control. Values represent the mean ± s.d. (n = 9). * p < 0.05, one-way ANOVA followed by Tukey’s multiple-comparison test. NS, not significant (p > 0.05).

Figure 4. Levosimendan inhibits HIV-1 replication and reactivation in primary CD4+ T cells

(A) Primary CD4⁺ T cells were isolated from three healthy donors and activated using anti-CD3/CD28 antibodies. Cells were then infected with HIV-IIIB in the presence of indicated compound (10 μM). At each time point, supernatant was harvested and subjected to HIV-1 p24 ELISA assay. (B) Isolated primary CD4⁺ T cells from donors were infected with HIV-1 Luc followed by long-term culture to allow for the establishment of HIV-1 latency (the Vicente Planelles’ model). HIV-1 latently infected cells were then stimulated with anti-CD3/CD28 antibodies in the presence of indicated compound (10 μM) for 3 days. Luciferase activity was measured and normalized to the DMSO control. (C) CD4⁺ T cells isolated from 3 cART-treated, HIV-infected aviremic patients were stimulated with anti-CD3/CD28 antibodies in the presence of levosimendan or spironolactone (10 μM) for 6 days. cART (100 nM efavirenz [EFV], 180 nM zidovudine [AZT], 200 nM raltegravir [RAL]) was used as a positive control. HIV-1 RNAs in the supernatant were quantified by ultra-sensitive qPCR. (D) The viability of cells in (C) was measured using the LIVE/DEAD™ Fixable Green Dead Cell Stain Kit. Results of (C) and (D) were normalized to DMSO control. *p < 0.05, one-way ANOVA followed by Tukey’s multiple-comparison test).

Figure 5. Levosimendan does not reduce Tat protein level or Tat-P-TEFb interaction but PI3K inhibiton rescues its effect

(A, B) HeLa cells stably expressing Tat-Flag were treated with levosimendan (10 μM) for 24 hr. (A) Total cell lysates were subjected to western blot to detect Tat-Flag or endogenous GAPDH protein level using Flag antibody or GAPDH antibody, respectively. (B) Total cell lysates were immunoprecipitated using Flag antibody followed by western blot to detect endogenous cyclin T1 or Tat-flag protein level using Cyclin T1 or Flag antibody, respectively. (C) J-LAT 10.6 cells were pre-treated with 3-MA at indicated concentrations for 6 hr, then co-treated with levosimendan (10 μM) and TNFα (10 ng/ml) for 24 hr. The GFP-positive cells were analyzed by flow cytometry. Results were normalized to DMSO control. Values represent the mean ± s.d. (n = 3). * p < 0.05, one-way ANOVA followed by Tukey’s multiple-comparison test.