Effect of tRNA on the Maturation of HIV-1 Reverse Transcriptase

Abstract

The mature HIV-1 reverse transcriptase is a heterodimer that comprises 66 kDa (p66) and 51 kDa (p51) subunits. The latter is formed by HIV-1 protease-catalyzed removal of a C-terminal ribonuclease H domain from a p66 subunit. This proteolytic processing is a critical step in virus maturation and essential for viral infectivity. Here, we report that tRNA significantly enhances in vitro processing even at a substoichiometric tRNA:p66/p66 ratio. Other double-stranded RNAs have considerably less pronounced effect. Our data support a model where interaction of p66/p66 with tRNA introduces conformational asymmetry in the two subunits, permitting specific proteolytic processing of one p66 to provide the mature RT p66/p51 heterodimer.

Keywords: HIV-1; RNase H; maturation; proteolysis; reverse transcriptase; tRNA.

Figure 1

(a) Concerted and Sequential Models for RT maturation ^{; –}; (b) location of the RNH domain in the known p66/p51 RT structure, and (c) an expanded view of the RNH domain. In (a)–(c), the p51 and the RNH domains in the p66 subunit are shown in cyan and orange, respectively, while the p51 subunit is shown in lime color. In (b) and (c) the p51-RNH cleavage site (F440-Y441) is shown as a red ribbon. The images were generated using PDB code 1DLO.

Figure 2

Time dependence of p66 processing by HIV-1 PR in 20 mM sodium acetate buffer at pH 5.2 and 37 °C, monitored by SDS-PAGE (a and b) at a high concentration of p66 (8 μM as a p66 monomer, 4 μM as a dimer) proteolytically processed by 1 μM HIV-1 PR in the presence of 0, 0.5, 8 μM tRNA concentrations, (c and d) at a low concentration of p66 (1 μM p66 monomer concentration, 0.5 μM as a dimer) processed by 0.25 μM HIV-1 PR in the presence of 0, 0.1, and 2 μM tRNA concentrations, and (e and f) at a high concentration of p66 (8 μM as a p66 monomer, 4 μM as a dimer) proteolytically processed by 1 μM HIV-1 PR in the presence of 0.5 μM ds-RNA (40 nt/22 nt), ss-RNA (40 nt), and ss-RNA (22 nt). In (b, d, and f), p66/p51 fractions, determined from the gel images in (a, c, and e), are shown, respectively. Average data points were used to fit a curve with one standard deviation as an uncertainty of each data point. Since p66/p51 production occurs in parallel with p66 degradation by PR, the build-up curves do not necessarily reach a plateau. Because of these multiple factors, normalized χ², χ_n², values for the curve fits were between 0.2 and 10.7.

Figure 3

Effects of varying amounts of (a) tRNA, (b) NaCl, (c) tRNA in the presence of 100 mM NaCl, and (d) heparin on p66 processing by HIV-1 PR for 20 min at 37 ºC, and (e) the time dependence of p66 processing by HIV-1 PR in the presence of 100 mM NaCl at 20 ºC. In all the experiments, reactions were initiated by the addition of 1 μM HIV-1 PR to p66 at a high concentration of p66 (8 μM as a monomer, 4 μM as a dimer) and monitored by SDS-PAGE, and the buffer contains 20 mM sodium acetate at pH 5.2. In (e), experiments were carried out in the presence of 0, 0.5, and 8 μM tRNA. (f) Plots of intensity changes and the fit curves, χ_n² values 2.7 and 4.5, are shown for (a) and (c), respectively. (g) Plots of intensity changes and the fit curves, χ_n² values from 0.88 to 2.2, are shown for (f). In (b), aberrant non-specific cleavage products are shown by arrows.

Figure 4

tRNA binding to p66 protein, monitored by (a – c) SEC elution profiles of (a) p66 protein solution, (b) tRNA solution, and (c) that mixed with tRNA, and (d, e) change of fluorescence emission of Cy3-labeled tRNA at varying p66 concentrations obtained by spectrofluorometry. In (a – c), UV absorbance at 254 nm (gray line) and 280 nm (black line), and fluorescence emission at 560 nm (dashed line) are shown. In (d), changes in fluorescence of total 1 μM tRNA, containing 40 nM Cy3-labeled tRNA, as a function of p66 concentration were recorded. In (e), fluorescence intensity changes at 560 nm (d) are plotted. The dash line is a fit-curve calculated with a single binding mode model (χ_n² = 6.5), and the solid line indicates a fit-curve calculated with a two-binding mode model (χ_n² = 0.8). The null hypothesis was rejected with p = 0.00036.