Role and timing of GTP binding and hydrolysis during EF-G-dependent tRNA translocation on the ribosome

Abstract

The translocation of tRNA and mRNA through the ribosome is promoted by elongation factor G (EF-G), a GTPase that hydrolyzes GTP during the reaction. Recently, it was reported that, in contrast to previous observations, the affinity of EF-G was much weaker for GTP than for GDP and that ribosome-catalyzed GDP-GTP exchange would be required for translocation [Zavialov AV, Hauryliuk VV, Ehrenberg M (2005) J Biol 4:9]. We have reinvestigated GTP/GDP binding and show that EF-G binds GTP and GDP with affinities in the 20 to 40 microM range (37 degrees C), in accordance with earlier reports. Furthermore, GDP exchange, which is extremely rapid on unbound EF-G, is retarded, rather than accelerated, on the ribosome, which, therefore, is not a nucleotide-exchange factor for EF-G. The EF-G.GDPNP complex, which is very labile, is stabilized 30,000-fold by binding to the ribosome. These findings, together with earlier kinetic results, reveal that EF-G enters the pretranslocation ribosome in the GTP-bound form and indicate that, upon ribosome-complex formation, the nucleotide-binding pocket of EF-G is closed, presumably in conjunction with GTPase activation. GTP hydrolysis is required for rapid tRNA-mRNA movement, and P(i) release induces further rearrangements of both EF-G and the ribosome that are required for EF-G turnover.

Fig. 1.

GTP/GDP binding to EF-G. (A) Fluorescence titrations with mant-GTP (circles) and mant-GDP (triangles). (B) TLC of mant-GTP and mant-GDP. (C) Chase titrations with GTP (circles) or GDP (triangles) monitoring the fluorescence of Bodipy FL-GDP. (D) Titrations monitoring tryptophan fluorescence. GTP (circles) or GDP (triangles). (Inset) Concentration dependence of GTP binding kinetics. k_on = 0.58 ± 0.04 μM⁻¹·s⁻¹; k_off = 13 ± 1 s⁻¹. (E) Binding of [³H]GDP (triangles) or [³H]GTP (circles) to nitrocellulose filters in the presence (filled symbols) or absence (open symbols) of EF-G. (F) Retention of EF-G·[³H]GDP (triangles) and EF-G·[³H]GTP (circles) on nitrocellulose filters. Shown are difference titration curves from E. In A–D, continuous lines represent fits (see Materials and Methods).

Fig. 2.

Equilibrium binding of GDPNP to free and ribosome-bound EF-G. (A) mant-GDPNP. Free EF-G (open symbols) and ribosome-bound EF-G (filled symbols) were titrated with mant-GDPNP as in Fig. 1A. Control titrations without EF-G showed no signal change. Apparent K_d values: Free EF-G, 120 μM; ribosome-bound EF-G, 0.5 μM. (B) GDPNP. Mant-GDPNP was chased from ribosome-bound EF-G by adding increasing amounts of GDPNP. Apparent K_d = 2 μM. Continuous lines represent fits (see Materials and Methods); standard deviations of K_d values were ±20%.

Fig. 3.

Dissociation of mant-labeled nucleotides from free and ribosome-bound EF-G. The dissociation of mant-GDP (traces 1 and 2), mant-GTP (traces 3 and 4), and mant-GDPNP (traces 5 and 6) from free EF-G (traces 1, 3, and 5) or ribosome-bound EF-G (traces 2, 4, and 6) as induced by rapid mixing with excess unlabeled GDP (traces 1–4) or GDPNP (traces 5 and 6) was monitored by fluorescence. Dissociation rate constants are summarized in Table 1. For details, see Materials and Methods.

Fig. 4.

Translocation with GTP, GDP, and GDPNP. Time courses of translocation were measured by fluorescence stopped-flow, monitoring the fluorescence of proflavin-labeled fMetPhe-tRNA^Phe (traces 1, 4, and 6) or of mRNA(fluorescein + 14) (traces 2, 3, and 5). Time courses were evaluated by single-exponential fitting (see Materials and Methods) to obtain the following values for k_app (SD ±15%): trace 1, GTP, 21 s⁻¹; trace 2, GTP, 18 s⁻¹; trace 3, GDPNP, 0.8 s⁻¹; traces 4 and 5, GDP, 0.9 s⁻¹; trace 6, mant-GDP, 0.6 s⁻¹.