Mechanism of inhibition of translation termination by blasticidin S

Abstract

Graphical abstract

Revealing the structural mechanisms of translation termination inhibitors may inform development of new antibacterials [1] or drugs to treat genetic diseases caused by mutations that introduce premature termination codons [2]. Recent studies uncovered molecules that inhibit peptide release [3] or trap a post-release ribosome state [4]. We have shown that the nucleoside analog blasticidin S (BlaS; Fig. 1a), which binds the peptidyl transferase center of bacterial 70S [3] and eukaryotic 80S [5] ribosomes, inhibits a release stage of translation termination [3]. The structural explanation for termination inhibition by BlaS, however, remains unsolved because ribosome•BlaS structures lacked release factors. Release factors (RF1 and RF2 in bacteria) recognize a stop codon in the ribosomal A site and catalyze the hydrolysis of peptidyl-tRNA bound in the P site, resulting in peptide release (reviewed in ref [6]). In this work, we report a 3.4-Å resolution crystal structure of a bacterial 70S ribosome bound to release factor 1 (RF1) and BlaS (Fig. 1b-c, see also Supplementary Data).

Open in a separate window

Figure 1

Crystal structure of RF1 bound to the 70S ribosome in the presence of blasticidin S

(a) Chemical structure of blasticidin S. (b) Structure of RF1 bound to the 70S ribosome in the presence of BlaS. (c) Structure and unbiased F_o-F_c density for BlaS and RF1 with bent domain 3 (see also Fig. S1). (d, e) BlaS induces rearrangement of the peptidyl transferase center. Movement of U2585 toward the catalytic backbone NH group of Gln 235 in response to BlaS is shown by arrow (d). Structure of the 70S•RF1 complex in the absence of BlaS [8] (PDB 5J4D) is shown in panel (e). See also Figures S2-S5.

BlaS induces conformational changes in the peptidyl transferase center that explain how it inhibits termination by RF1. As in the structure without RF1 [3], BlaS intercalates between C74 and A76 at the ⁷⁴CCA⁷⁶ 3′ end of the P-site tRNA, displacing C75 by ~7 Å toward the A site. The displacement of P-site tRNA nucleotide C75 shifts the position of RF1’s universally conserved ²³³GGQ²³⁵ catalytic motif by ~2 Å (more than twice the coordinate error; see Fig. S1, Table S1 in Supplementary Data), compared to its position in 70S•RF1 structures [7]. A2602 of 23S rRNA, which is also critical for termination, normally packs in a pocket within domain 3 of RF1 to stabilize the GGQ motif [8]. In the presence of BlaS, however, A2602 shifts by ~10 Å (measured at the N6 atom) away from its position in the 70S•RF1 complex and thus disengages from the GGQ loop of RF1 (Fig. S2a-b). Moreover, the displacement of C75 by BlaS shifts the phosphate backbone of A76. U2585, which normally packs on A76 backbone, moves toward BlaS by ~3 Å (measured at the O4 atom; Fig. 1d-e; Fig. S3). As a result, the O4 atom of U2585 is placed within hydrogen-bonding distance of the backbone NH group of Gln 235 from RF1’s GGQ motif (Fig. 1d, Fig. S4).

The backbone NH of Gln 235 is essential for peptidyl-tRNA hydrolysis. In crystal structures of RF1- and RF2-bound 70S termination complexes without BlaS, the backbone NH group of Gln 235 hydrogen bonds with the substrate [9] and leaving group (i.e., the A76 ribose of the P-site tRNA), and may stabilize the transition-state intermediate [7]. Furthermore, mutation of Gln 235 to proline—but not other amino acids—eliminates peptide release activity without disrupting the conformation of release factor on the ribosome [10]. Our structure therefore suggests that BlaS inhibits peptidyl-tRNA hydrolysis by shifting the GGQ motif and preventing the catalytic group from optimally interacting with the substrate ([9], Fig. S5), product, or reaction intermediates, and/or by disrupting the coordination of the nucleophile. Translation inhibitors that induce a similar P-site tRNA distortion (e.g., bactobolin A [11]) may have a similar mechanism of action in termination and ribosome rescue pathways [12–15].

Our structure also provides new insight into termination accuracy. BlaS distorts domain 3 of RF1, which carries the catalytic GGQ motif. Helix α7 is curved (Fig. 1c), with His 279 at its center shifted by ~3 Å relative to its position in the normally straight helix (Fig. S1). The position of RF1’s codon-recognition domain 2 and its interactions with the stop codon, however, are nearly identical to those in termination complexes without BlaS [7,8]. Thus, our structure demonstrates that accommodation of the codon-recognition domain on the stop codon can occur independently of full accommodation of the catalytic domain. Inter-domain flexibility has been proposed to underlie termination accuracy, in which peptide release is strictly coordinated with stop-codon recognition (reviewed in [6]. A stop-codon in the ribosomal A site is thought to induce a large-scale rearrangement of RF1, from a compact to an extended conformation, that inserts the catalytic domain into the peptidyl transferase center [16]. Recent cryo-EM structures of ArfA•RF2-bound ribosome rescue complexes revealed that RF2 can adopt a compact state when bound to the rescue factor ArfA in the absence of an A-site codon [17, 18]. In every structure of a termination complex bound to a stop codon, however, release factor is fully extended, with domain 3 in the catalytically engaged conformation. Although the conformation of RF1 in our structure could represent an “off-pathway” intermediate caused by BlaS, it might also coincide with an intermediate state of RF1 opening. Regardless, displacement of the catalytic domain relative to the codon-recognition domain in this structure is consistent with the mechanistic model of termination accuracy due to inter-domain movement.

Supplementary Material

Acknowledgments

Footnotes

References