Distinct functions of elongation factor G in ribosome recycling and translocation

Abstract

Elongation factor G (EF-G) promotes the translocation step in bacterial protein synthesis and, together with ribosome recycling factor (RRF), the disassembly of the post-termination ribosome. Unlike translocation, ribosome disassembly strictly requires GTP hydrolysis by EF-G. Here we report that ribosome disassembly is strongly inhibited by vanadate, an analog of inorganic phosphate (Pi), indicating that Pi release is required for ribosome disassembly. In contrast, the function of EF-G in single-round translocation is not affected by vanadate, while the turnover reaction is strongly inhibited. We also show that the antibiotic fusidic acid blocks ribosome disassembly by EF-G/RRF at a 1000-fold lower concentration than required for the inhibition of EF-G turnover in vitro and close to the effective inhibitory concentration in vivo, suggesting that the antimicrobial activity of fusidic acid is primarily due to the direct inhibition of ribosome recycling. Our results indicate that conformational coupling between EF-G and the ribosome is principally different in translocation and ribosome disassembly. Pi release is not required for the mechanochemical function of EF-G in translocation, whereas the interactions between RRF and EF-G introduce tight coupling between the conformational change of EF-G induced by Pi release and ribosome disassembly.

FIGURE 1.

Single-round GTP hydrolysis by EF-G in the absence and presence of RRF. Post-termination complex (Materials and Methods; 1 μM final concentration after mixing) was rapidly mixed with IF3 (2 μM), [γ-³²P]GTP (30 μM), and EF-G (3 μM) alone (●) or together with RRF (5 μM) (○). The extent of GTP hydrolysis is presented relative to the input of post-termination complex. Single-exponential fitting yielded a rate of GTP hydrolysis of 75 ± 25 sec⁻¹ (4°C).

FIGURE 2.

Effect of RRF on the release of inorganic phosphate from ribosome-bound EF-G. (A) Effect of RRF on the delay of Pi release. Post-termination complex (Materials and Methods; final concentration after mixing, 0.2 μM) with GTP (200 μM) was rapidly mixed in a stopped-flow apparatus with IF3 (2 μM), GTP (200 μM), and EF-G (2 μM) (1) or EF-G (2 μM) and RRF (5 μM) (2). The release of Pi from ribosome-bound EF-G was monitored by the fluorescence increase of MDCC-labeled PBP (2.5 μM) caused by the binding of Pi. The delays are indicated by vertical dotted lines. (B) Pi release in the presence of FA, short time window. Pi release was measured as in A (traces 1 and 2) or in the presence of FA (200 μM) without (trace 3) or with RRF (trace 4). (C) Pi release in the presence of FA, long time window.

FIGURE 3.

Effect of vanadate on EF-G function on the ribosome. (A) Inhibition by vanadate of post-termination complex disassembly. Post-termination complex was rapidly mixed with EF-G/RRF in the absence (1) or presence (2) of vanadate (2 mM), and complex disassembly was monitored by light scattering (LS) (Materials and Methods). (B) Concentration dependence of the inhibition by vanadate. Post-termination complex disassembly was measured as in A in the presence of increasing concentrations of vanadate. Plotted is the half-life (τ_1/2) of the reaction measured in the absence relative to that measured in the presence of vanadate; τ_1/2 was ∼5 sec in the absence and ∼150 sec in the presence of vanadate. (C) Comparison of post-termination complex disassembly and translocation. Rates of post-termination complex disassembly (black bars) were measured as in A. The rate of single-round translocation (∼15 sec⁻¹, open bars) was measured by fluorescence stopped-flow, multiple-round translocation (hatched bars) by the puromycin assay (Materials and Methods); plotted is the relative initial rate of translocation measured in the absence (2 sec⁻¹, set to 1.0) and presence (0.17 sec⁻¹) of vanadate. The concentration of vanadate was 2 mM.

FIGURE 4.

Effect of fusidic acid (FA) on EF-G function on the ribosome. (A) Inhibition of post-termination complex disassembly. Post-termination complex was rapidly mixed with EF-G/RRF as in Figure 3A in the absence (1) and presence (2) of FA (200 μM), and ribosome disassembly was monitored by light scattering (LS). (B) Dependence of the inhibition on FA concentration. Post-termination complex disassembly by EF-G/RRF was measured as in A at various concentrations of FA (●); 50% inhibition was observed at 0.1 μM FA. For comparison, the inhibition of EF-G turnover in translocation is shown (○); 50% inhibition was observed at ∼200 μM FA. (C) Time course of translocation in the absence and presence of FA. Translocation was monitored by the fluorescence of fluorescence-labeled peptidyl-tRNA (Materials and Methods).

FIGURE 5.

Sequence of steps in translocation (A) and ribosome disassembly (B). The ribosomal subunits are depicted in light blue (50S) and yellow (30S) in the ratcheted (50S tilted) or nonratcheted (50S straight) conformation with the tRNAs (red, P site; green, A site) in hybrid (tilted) or classic (upright) positions. EF-G is colored in blue, RRF in magenta. Different conformations of EF-G are not indicated. In B, it is assumed that ribosome disassembly proceeds through the ratcheted conformation of the ribosome. The inhibition by fusidic acid (FA) of ribosome relocking following tRNA-mRNA movement (A) or of subunit splitting (B) is indicated. The mRNA bound to the 30S subunit is omitted for clarity.