Genetic identification of nascent peptides that induce ribosome stalling

Abstract

Several nascent peptides stall ribosomes during their own translation in both prokaryotes and eukaryotes. Leader peptides that induce stalling can regulate downstream gene expression. Interestingly, stalling peptides show little sequence similarity and interact with the ribosome through distinct mechanisms. To explore the scope of regulation by stalling peptides and to better understand the mechanism of stalling, we identified and characterized new examples from random libraries. We created a genetic selection that ties the life of Escherichia coli cells to stalling at a specific site. This selection relies on the natural bacterial system that rescues arrested ribosomes. We altered transfer-messenger RNA, a key component of this rescue system, to direct the completion of a necessary protein if and only if stalling occurs. We identified three classes of stalling peptides: C-terminal Pro residues, SecM-like peptides, and the novel stalling sequence FXXYXIWPP. Like the leader peptides SecM and TnaC, the FXXYXIWPP peptide induces stalling efficiently by inhibiting peptidyl transfer. The nascent peptide exit tunnel and peptidyltransferase center are implicated in this stalling event, although mutations in the ribosome affect stalling on SecM and FXXYXIWPP differently. We conclude that ribosome stalling can be caused by numerous sequences and is more common than previously believed.

FIGURE 1.

Genetic selection for sequences that induce ribosome stalling. A, structure of the Aph(3′)-IIa protein, homologous to KanR (30). Kanamycin in shown in black. The C-terminal helix (14 residues), shown in red, is essential for KanR activity. A four-residue surface-exposed loop prior to this helix is highlighted in blue. B, 18 random nucleotides (6 codons) were introduced at the C terminus of a KanR protein lacking its last 18 amino acids (yellow). If the random sequence induces stalling, tmRNA-K1 rescues the stalled ribosome and directs the synthesis of the remaining KanR residues (red). Cellular survival on kanamycin plates is therefore tied to stalling on the C terminus of KanR.

FIGURE 2.

Essential elements of a stalling peptide. A, WPPPSI and 12 upstream residues were divided into four groups for analysis. B, N-terminal deletions of the 18-mer stalling peptide were fused to the C terminus of GST. The fusions were analyzed by immunoblot for tagging by a modified tmRNA (tmRNA-H) that encodes an His₆ epitope. The +1FS vector contains the same 18-mer sequence shifted into the +1 frame to test if translation of WPPPSI is necessary for stalling. C, each residue of the 18-mer was mutated individually to Ala and assayed as above. A GST-Stop construct served as a negative control and the intact full-length WPPPSI 18-mer as a positive control.

FIGURE 3.

Stalling during termination at WPPDV*. Tagging of the GST-WPPDV* fusion by tmRNA-H was monitored by anti-His₆ antibodies. Mutations in bold were introduced to determine the role of spacing, stop codons, and the DV residues in this stalling event.

FIGURE 4.

The effect of tRNA levels on stalling and tagging. A, one or more Pro codons in the WPPPSI sequence was switched to CCC (lowercase p) so that it would only be recognized by Pro2 tRNA. An uppercase P represents a Pro codon not recognized by Pro2 tRNA. Tagging of the GST-WPPPSI fusion was monitored in the presence or absence of a plasmid (pRARE) overexpressing the Pro2 tRNA. pRARE also expresses several other rare tRNAs. B, stalling sequences containing one or three rare Arg codons (AGG) were fused to the C terminus of GST. Tagging by tmRNA-H was monitored with anti-His₆ antibodies in the presence or absence of a plasmid (pRARE) overexpressing the cognate tRNA, Arg5.

FIGURE 5.

Direct detection of stalled ribosome complexes. A, stalled ribosome complexes were formed by cell-free translation of a template encoding the FXXYXIWPPP sequence in a larger (64-mer) peptide. The non-stalling Ala mutant FXXYXIWPAP served as a negative control. The position of the ribosome was determined by reverse transcription of the mRNA template, and C and G sequencing lanes were run alongside. Thiostrepton was added in the fourth and sixth lanes to trap the ribosome in the initiation stage, demonstrating the observed toe-print signal in the third lane (marked by arrows) requires translation of the stalling site. The nucleotide and peptide sequence of the stalling site is shown at left. B, the [³⁵S]Met-labeled products of the cell-free translation of the FXXYXIWPPP peptide or the non-stalling FXXYXIWPAP control were analyzed by Tricine-SDS-PAGE. Under these conditions, the peptidyl-tRNA linkage is not hydrolyzed during electrophoresis. The stalled peptidyl-tRNA disappears when treated with RNase (lane 3) but remains in the aqueous layer upon extraction with phenol (lane 5).

FIGURE 6.

Effects of ribosome mutations on stalling on SecM and FXXYXIWPPP. The SecM stalling sequence or the WPPPSI 18-mer (see Table 2) were inserted after residue nine of the full-length lacZ gene. β-Galactosidase activity was measured for the resulting SecM (white) and FXXYXIWPPP (gray) lacZ fusions in a strain overexpressing mutant 23 S ribosomal RNA. The activity is reported in Miller units. The data represent at least three independent experiments. Error bars represent the standard deviation.