The myxobacterial metabolite ratjadone A inhibits HIV infection by blocking the Rev/CRM1-mediated nuclear export pathway

Abstract

Background: The nuclear export of unspliced and partially spliced HIV-1 mRNA is mediated by the recognition of a leucine-rich nuclear export signal (NES) in the HIV Rev protein by the host protein CRM1/Exportin1. This makes the CRM1-Rev complex an attractive target for the development of new antiviral drugs. Here we tested the anti-HIV efficacy of ratjadone A, a CRM1 inhibitor derived from myxobacteria.

Results: Ratjadone A inhibits HIV infection in vitro in a dose-dependent manner with EC₅₀ values at the nanomolar range. The inhibitory effect of ratjadone A occurs around 12 hours post-infection and is specific for the Rev/CRM1-mediated nuclear export pathway. By using a drug affinity responsive target stability (DARTS) assay we could demonstrate that ratjadone A interferes with the formation of the CRM1-Rev-NES complex by binding to CRM1 but not to Rev.

Conclusion: Ratjadone A exhibits strong anti-HIV activity but low selectivity due to toxic effects. Although this limits its potential use as a therapeutic drug, further studies with derivatives of ratjadones might help to overcome these difficulties in the future.

Figure 1

Schematic representation of Rev-mediated nuclear export of HIV mRNAs and chemical structures of RaTA and LMB. (A) Binding of the Nuclear Localization Signal (NLS) in Rev to importin-β (Imp-β) triggers the nuclear internalization of Rev through the Nuclear Pore Complex (NPC). Once in the nucleus, Rev binds to the Rev-responsive element (RRE) of the HIV mRNA. This interaction exposes the Nuclear Export Signal (NES) of Rev for recognition by CRM1. The CRM1-Rev-mRNA complex is stabilized by the phosphorylated form of Ran (RanGTP) and crosses the nuclear pore into the cytoplasm where Ran is dephosphorylated (RanGDP) and the complex is disassembled making HIV mRNAs available for translation. (B) The specific groups within the chemical structures of RaTA and LMB that are involved in the interaction with CRM1 are circled. A complete description of these interactions can be found in (55). Structures are freely available from: http://www.chemspider.com. Chemspider IDs: 5293127 and 4948244.

Figure 2

Potency of ratjadone A against HIV infection of TZM-bl cells. Cells were seeded in 96-well plates in triplicate and treated with increasing concentrations of RaTA (A) or LMB (B) and infected with HIV_LAI at a multiplicity of infection (MOI) of 0.5. 48 h after infection cells were assayed for luciferase activity and the mean relative light units (RLU) were plotted as % relative to DMSO (vehicle) for both infectivity and cell viability. Effective Concentration 50 (EC₅₀) and Cytotoxic Concentration 50 (CC₅₀) were estimated by non-linear regression of log inhibitor vs. normalized response and used to calculate the Selectivity Index (SI) value (see Materials and methods). Bars: standard error of the mean (SEM). For panels (C) and (D), cells were HIV infected by spinoculation, seeded in 6-well plates and treated with increasing concentrations of RaTA. 48 h after infection, cells were lysed and analysed by Western Blotting (C). HIV p24 bands are shown in the upper row. Lanes 1 to 3: cells incubated with RaTA at different concentrations; Lane 4: DMSO control; Lane 5: uninfected DMSO control. Tubulin was used as a loading control (lower row). Every band in panel C was normalized respect to the loading control and quantified. The relative p24 production was plotted as % relative to DMSO control (D). The drug solvent concentration (0.1% DMSO) in every sample was constant.

Figure 3

Ratjadone A inhibits HIV p24 expression in MT-2-infected cells. MT-2 cells were infected with HIV_LAI and treated with ratjadone A or leptomycin B. 48 h after infection cells were fixed and stained for HIV-p24 protein (green signals) and with DAPI (blue signals). Not infected and infected but untreated cells were used as negative and positive controls respectively. (A) non-infected cells. (B) p24 distribution in infected cells without drugs. (C) p24 distribution in infected cells treated with ratjadone A, and (D) p24 distribution in infected cells treated with leptomycin B.

Figure 4

Ratjadone A blocks HIV infection at a time point corresponding to viral mRNA nuclear export. TZM-bl cells were synchronously infected with HIV_LAI and plated in 96-wells plates in duplicate. Then 10 nM ratjadone A or 1 μM T20 were added at the indicated time points. The drug solvent (0.1% DMSO) was used as control. 72 h after infection luciferase activity was measured for every time point. Values are plotted as relative light units per seconds of exposure and are the mean of the duplicates. Error bars are standard error of the mean (SEM).

Figure 5

Ratjadone A inhibits the Rev-dependent HIV-RNA nuclear export. TZM-bl cells were transfected with a GagPol-RRE plasmid (A to C); co-transfected with GagPol-RRE and Rev expression plasmids (D to F) or with a GagPol-CTE expression plasmid (G to I). Cells were left untreated or incubated with the indicated drugs at 2 nM (RaTA) and 4 nM (LMB) concentrations. 48 h after transfection cells were fixed and stained for HIV-p24 protein (green signal) and DAPI (blue signal). Samples (A) to (C) serve as negative controls, and (D) as positive control. Arrows in (H) indicate p24 positive cells. (J) The percentage of p24 positive cells for every sample is shown. Significant differences with p < 0.01 are marked with asterisks. Error bars are standard error of the mean (SEM) of p24-positive cells counted in 10 microscopic fields.

Figure 6

Ratjadone A protects CRM1 from proteolysis. Lysates from Rev-transfected 293 T cells were mock-treated with DMSO or incubated with 200 nM ratjadone A or leptomycin B and digested with Pronase. Samples were loaded on a SDS-PAGE gel and developed by Western Blot (see Materials and methods). Lane 1: undigested DMSO control; Lane 2: digested DMSO control; Lane 3: digested sample incubated with RaTA; Lane 4: digested sample incubated with LMB. Percentages of protection relative to undigested controls are shown below the protein bands. Both RaTA and LMB protect CRM1 (upper row) but not Rev (middle row) from degradation. Tubulin was used as a loading control (lower row). The drug solvent concentration (0.1% DMSO) in every sample was constant. Abbreviations: RAT = Ratjadone A; LMB = Leptomycin B; DMSO = dimethyl sulfoxide; n.a. = not applicable.