Novel postentry inhibitor of human immunodeficiency virus type 1 replication screened by yeast membrane-associated two-hybrid system

Abstract

Human immunodeficiency virus (HIV) Gag protein targets to the plasma membrane and assembles into viral particles. In the next round of infection, the mature Gag capsids disassemble during viral entry. Thus, Gag plays a central role in the HIV life cycle. Using a yeast membrane-associated two-hybrid assay based on the SOS-RAS signaling system, we developed a system to measure the Gag-Gag interaction and isolated 6 candidates for Gag assembly inhibitors from a chemical library composed of 20,000 small molecules. When tested in the human MT-4 cell line and primary peripheral blood mononuclear cells, one of the candidates, 2-(benzothiazol-2-ylmethylthio)-4-methylpyrimidine (BMMP), displayed an inhibitory effect on HIV replication, although a considerably high dose was required. Unexpectedly, neither particle production nor maturation was inhibited by BMMP. Confocal microscopy confirmed that BMMP did not block Gag plasma membrane targeting. Single-round infection assays with envelope-pseudotyped and luciferase-expressing viruses revealed that BMMP inhibited HIV replication postentry but not simian immunodeficiency virus (SIV) or murine leukemia virus infection. Studies with HIV/SIV Gag chimeras indicated that the Gag capsid (CA) domain was responsible for the BMMP-mediated HIV postentry block. In vitro studies indicated that BMMP accelerated disassembly of HIV cores and, conversely, inhibited assembly of purified CA protein in a dose-dependent manner. Collectively, our data suggest that BMMP primarily targets the HIV CA domain and disrupts viral infection postentry, possibly through inducing premature disassembly of HIV cores. We suggest that BMMP is a potential lead compound to develop antiretroviral drugs bearing novel mechanisms of action.

Fig. 1.

Yeast membrane-associated two-hybrid screening for inhibitors of Gag-Gag interaction. (A) Schematic representation of Gag expression plasmids used for yeast SOS recruitment system. The full-length gag gene of HIV-1 (HXB2 strain) was expressed by yeast expression plasmids pMyr and pSOS: pMyr contains the yeast inducible promoter for the GAL1 gene and the URA3 gene as a selective marker and pSOS contains the constitutive promoter for the yeast ADH gene and the LEU2 gene as a selective marker. (B) Principle of yeast membrane-associated two-hybrid assay based on SOS recruitment system. The schematic illustration was adapted with permission from the manuals for the CytoTrap yeast system (Agilent Technologies, Inc.; http://www.genomics.agilent.com/CollectionSubpage.aspx?PageType=Product&SubPageType=ProductDetail&PageID=1311).

Fig. 2.

Screening for inhibitors of Gag assembly by yeast membrane-associated two-hybrid assays. (A) The yeast cdc25Ha strain was transformed with the pSOS and pMyr plasmids. The yeast culture was diluted to an OD₆₀₀ of 0.1 and incubated at 37°C in galactose-plus-raffinose medium (restrictive conditions) with a chemical library (a final concentration of 10 μM). After growth at 37°C for 5 days, cell density was measured at OD₆₀₀. As a control, the OD₆₀₀ of yeast incubated in the presence of DMSO was set to 100%. The yeast transformed with the pSOS and pMyr plasmids, both of which contained the HIV-1 gag gene, was shown as black columns, and the yeast was transformed with the pSOS plasmid containing MAFB and the pMyr plasmid containing the cDNA of SB (as a positive-control combination) as white columns. Data were shown as means with standard deviations from 5 independent experiments. (B) The yeast cdc25Ha strain transformed with the pSOS and pMyr plasmids containing the HIV-1 gag gene was grown at 25°C in glucose medium. The yeast culture was diluted to an OD₆₀₀ of 0.01 and incubated at 25°C in glucose medium with a chemical library (a final concentration of 10 μM). After growth at 25°C for 2 days, cell density was measured at OD₆₀₀. As a control, the OD₆₀₀ of yeast incubated in the presence of DMSO was set to 100%. Data were shown as means with standard deviations from 3 independent experiments. (C) Structures of compounds screened from a chemical library by yeast membrane-associated two-hybrid assays.

Fig. 3.

Inhibition of HIV-1 replication and cell toxicity. (A) MT-4 Luc cells were infected with HIV-1 (HXB2 strain) and incubated at 37°C in the presence of increasing doses of compounds. On day 7, MT-4 Luc cells were subjected to luciferase assay. Data were shown as means from triplicate cultures. (B) 293T cells were incubated with various doses of compounds at 37°C for 2 days and subjected to Alamar Blue assays. Data were shown as means from 3 independent experiments. (C) PBMC stimulated with IL-2 and anti-CD3 antibody were infected with HIV-1 (HXB2 strain) and incubated at 37°C in the presence of 5 and 25 μM 172A6. HIV-1 production in culture medium was temporally quantified by p24CA antigen capture ELISA (top). Uninfected PBMC were cultured with 172A6 at 37°C and were subjected to MTS assay on days 7 and 14. Data were shown as means with standard deviations from triplicate cultures, in which DMSO was used as control. (D) 293FT, HeLa, and MT-4 cells and PBMC were incubated at 37°C with various doses of 172A6 and were subjected to MTS assay on days 4, 3, 4, and 6, respectively. Cell viability was shown as OD₄₉₀. Data were shown as means with standard deviations from triplicate cultures.

Fig. 4.

HIV-1 particle production and Gag plasma membrane targeting. (A) Effects on HIV-1 particle production. 293FT cells were transfected with pHXB2 and incubated at 37°C with 0 to 50 μM BMMP. Two days posttransfection, cells were collected and culture media were subjected to purification of viral particles by ultracentrifugation. Equivalent volumes of samples were subjected to SDS-PAGE followed by Western blotting using anti-HIV-1 p24CA antibody. Representative blots were shown. HIV-1 particle yields in culture media were quantified by p24CA antigen capture ELISA. (B) Intracellular localization of Gag. HeLa cells were transfected with a pNL43 derivative in which Gag was fused with GFP and incubated at 37°C for 1.5 days with 30 μM BMMP. Nuclei were stained with TO-PRO-3, and cells were observed by confocal microscopy. Representative images were shown at the same magnification. Bar = 10 μm.

Fig. 5.

Inhibition of HIV-1 replication postentry. (A) Single-round infection assays in MT-4 cells. 293FT cells were cotransfected with pHIVgag-pol (derived from HXB2 strain), pLenti-luciferase, pRevpac, and either a plasmid expressing HIV-1 Env (middle) or a plasmid expressing VSV-G (top). HIV-1 (NL43 strain) expressing luciferase was similarly pseudotyped with VSV-G protein (bottom). Following incubation, the culture supernatants were recovered and were subsequently inoculated into MT-4 cells with increasing doses of BMMP. (B) Single-round infection assays in 293FT and 293FT-CD4 cells. HIV-1 vectors containing the luciferase gene were similarly pseudotyped with HIV-1 Env (middle) or VSV-G protein (top). Luciferase-expressing SIVmac was pseudotyped with VSV-G protein by cotransfection (bottom). Viruses produced were subsequently inoculated into 293FT (top and bottom) or 293FT-CD4 (middle) cells with 5 and 25 μM BMMP. Viral infectivity was assessed by luciferase reporter assays. Data were shown as means with standard deviations from 3 to 6 independent experiments in panels A and B. *, P < 0.05; **, P < 0.01. (C) Quantitative PCR. MT-4 cells were infected with HIV-1 (HXB2 strain) and incubated in the presence of BMMP. DNA was isolated at 4 or 24 h postinfection and subjected to quantitative PCR for HIV cDNA. Reverse transcripts generated at the early and late phases of HIV reverse transcription were amplified from the DNA isolated at 4 and 24 h postinfection, respectively. The integrated viral genome was amplified as Alu-LTR transcripts from the DNA isolated at 24 h postinfection. EFV (200 nM) was used as positive controls. Three independent infections were performed, and quantitative PCR was carried out in triplicate with each infection sample. Representative data were shown with the means and standard deviations. *, P < 0.05; **, P < 0.01.

Fig. 6.

Mapping of Gag domain responsible for inhibition. (A) Intracellular Gag expression and particle production. 293FT cells were cotransfected with pHIVgag-pol expressing chimeric Gag, pLenti-luciferase, pRevpac, and a plasmid expressing VSV-G, and virus particles produced were purified by ultracentrifugation. Cells and particles were analyzed by Western blotting using anti-HIV-1 p24CA and p17MA and anti-SIVmac p27CA antibodies. (B) Single-round infection assays with the Gag domain chimeras and Gag mutants with amino acid substitutions. The corresponding domain of SIVmac Gag (black) was introduced into HIV-1 Gag background (gray), and the resultant chimera was referred to as “s” plus the name of the corresponding domain of SIVmac. WT, wild type; sMA, HIV containing replacement of MA with SIV MA; sMACA, HIV containing replacement of MACA with SIV MACA; sCA, HIV containing replacement of CA with SIV CA. CypA mt represents HIV with amino acid substitutions G89A and P90A in the CypA-binding loop of CA NTD (denoted by asterisks). Following infection, the cell culture was incubated in the presence of 5 and 25 μM BMMP. Viral infectivity was monitored by luciferase reporter assays. Data were shown as means with standard deviations from 4 to 6 independent experiments. *, P < 0.05; **, P < 0.01.

Fig. 7.

HIV-1 mature capsid disassembly and in vitro assembly assays. (A) In vitro assembly reaction with purified HIV-1 CA. Purified HIV-1 CA protein (100 μM) was incubated with various doses of BMMP at 37°C for 60 min in buffer at high salt concentration. All samples included 1% DMSO. Assembly products were recovered by centrifugation and subjected to SDS-PAGE followed by Coomassie brilliant blue staining. The band intensities were semiquantified by ImageJ software. For quantification, the pelleted products were subjected to p24CA antigen capture ELISA. Data were shown as means with standard deviations from 3 independent experiments. *, P < 0.05. Assembly products were also negatively stained and examined by electron microscopy. Micrographs were shown at the same magnification. Bar = 100 nm. (B) Cell-free assays for HIV-1 uncoating. HIV-1 mature capsids were isolated through a 1% Triton X-100 layer as described previously (3). The HIV-1 cores were incubated with various doses of BMMP at 37°C up to 120 min. For comparison, the cores were similarly incubated with 50 μM AZT. All samples included 1% DMSO. Residual intact cores were recovered by ultracentrifugation and quantified by p24CA antigen capture ELISA. Data were shown as means with standard deviations from 4 independent experiments. *, P < 0.05; **, P < 0.01.