Molecular Cell
Article
Molecular Mechanism
of Drug-Dependent Ribosome Stalling
Nora Vazquez-Laslop,1,* Celine Thum,1 and Alexander S. Mankin1,*
1Center for Pharmaceutical Biotechnology, University of Illinois at Chicago, 900 South Ashland Avenue, m/c 870, Chicago, IL 60607, USA
*Correspondence: [email protected] (A.S.M.), [email protected] (N.V.-L.)
DOI 10.1016/j.molcel.2008.02.026
SUMMARY
Inducible expression of the erm erythromycin resistance genes relies on drug-dependent ribosome stalling. The molecular mechanisms underlying stalling are
unknown. We used a cell-free translation system to
elucidate the contribution of the nascent peptide, the
drug, and the ribosome toward formation of the stalled
complex during translation of the ermC leader cistron.
Toe-printing mapping, selective amino acid labeling,
and mutational analyses revealed the peptidyl transferase center (PTC) as the focal point of the stalling
mechanism. In the ribosome exit tunnel, the C-terminal
sequence of the nascent peptide, critical for stalling, is
in the immediate vicinity of the universally conserved
A2062 of 23S rRNA. Mutations of this nucleotide eliminate stalling. Because A2062 is located in the tunnel, it
may trigger a conformational change in the PTC, responding to the presence of a specific nascent peptide. The cladinose-containing macrolide antibiotic in
the tunnel positions the nascent peptide for interaction
with the tunnel sensory elements.
INTRODUCTION
The ribosome assembles amino acids into polypeptides in the
peptidyl transferase center (PTC) located at the interface side
of the large ribosomal subunit. The nascent peptide is threaded
through the exit tunnel, which spans the entire body of the subunit and opens at its solvent side. On its way through the tunnel,
whose walls are built primarily of rRNA, the nascent peptide can
interact with a number of RNA residues. At the same time, segments of several ribosomal proteins are exposed in the tunnel lumen. In particular, loops of two proteins, L4 and L22, contribute
to formation of a constriction in the tunnel located 30 Å away
from the PTC (Berisio et al., 2003; Gabashvili et al., 2001; Harms
et al., 2001; Jenni and Ban, 2003; Mitra et al., 2006; Nissen et al.,
2000).
Most proteins are thought to smoothly pass through the tunnel
on their way out. However, accumulating evidence shows that
some nascent polypeptides establish specific interactions with
tunnel elements, resulting in ribosome stalling. In several characterized cases, stalling regulates gene expression (Cao and Geballe, 1996; Gong and Yanofsky, 2002; Nakatogawa and Ito,
190 Molecular Cell 30, 190–202, April 25, 2008 ª2008 Elsevier Inc.
2002; Onouchi et al., 2005; Wang and Sachs, 1997). Ribosome
stalling during translation of the secretion monitor gene secM
in Escherichia coli controls the expression of the downstream
secA cistron, which encodes a component of the secretion apparatus. When a specific sequence of the SecM nascent peptide
is synthesized and properly placed in the exit tunnel, it induces
a conformational change in the ribosome, resulting in ribosome
stalling (Mitra et al., 2006; Muto et al., 2006; Nakatogawa and
Ito, 2002). If the cellular secretion is lethargic, the stalling persists
for a long time, leading to activation of secA expression (Nakatogawa and Ito, 2001). In another well-characterized example in
bacteria, ribosome stalling is used for regulation of expression
of tryptophan catabolizing enzymes (Gong and Yanofsky,
2002). In the presence of high concentrations of tryptophan,
the ribosome stalls at the last sense codon of the tnaC cistron,
occluding the transcription terminator and activating the expression of the downstream genes (Gong et al., 2001; Gong and Yanofsky, 2002). Specific interactions of the TnaC nascent peptide
with the tunnel mediate formation of a long-lived stalled complex
and inhibition of peptide release.
Though ribosome stalling has received a surge of attention in
the past few years, the first example of nascent peptide-dependent translation arrest was described more than a quarter century ago when Weisblum and Dubnau laboratories characterized
inducible expression of the erythromycin resistance genes
(Gryczan et al., 1980; Horinouchi and Weisblum, 1980). Erythromycin and other macrolide antibiotics bind in the nascent peptide tunnel in the vicinity of the PTC and are thought to block
translation by inhibiting progression of the nascent peptide
when it reaches the size of 3–9 amino acids long (Tenson
et al., 2003; Vazquez, 1975). The inducible ermC cassette, which
is often found in resistant pathogens, consists of the resistance
methyltransferase gene (ermC) preceded by a 19-codon open
reading frame (ORF) encoding the regulatory leader peptide
ErmCL (Figures 1A and 1B). In the absence of an inducing antibiotic, ermC expression is repressed because its ribosome binding site is sequestered in the mRNA secondary structure. Subinhibitory concentrations of an inducer (erythromycin or similar
macrolide antibiotics) cause the ribosome to stall at the ermCL
ORF, triggering a conformational change in the mRNA that releases the ermC ribosome binding site, and thus activates the
expression of the methylase gene (Horinouchi and Weisblum,
1980; Narayanan and Dubnau, 1985). Because nonsense mutations beyond the 10th codon of the leader ORF do not alleviate
induction, it has been suggested that only a portion of the leader
peptide is required for stalling (Mayford and Weisblum, 1989).
Molecular Cell
Drug-Dependent Ribosome Stalling
Mutational analyses have also shown that changes in the first
few amino acids of the leader had little effect on inducibility,
but alterations of residues Ile-6, Phe-7, and Val-8 severely impair
it (Mayford and Weisblum, 1989). Though the precise stalling
point has not been determined, the results of the mutational
studies implied that in the stalled complex the ermCL nascent
peptide would be only a few amino acids long (Mayford and
Weisblum, 1989; Weisblum, 1995). Therefore, drug-dependent
ribosome stalling on the ermC leader seemed to be principally
different from the stalling at SecM or TnaC nascent peptides,
which span the entire length of the tunnel and do not require
the presence of antibiotic.
In spite of the time passed since the discovery of inducible erm
genes, very little has been learned about the molecular mechanisms of ribosome stalling, which is at the heart of the translation
attenuation model of antibiotic induction. Several key questions
remained essentially unanswered: why do some macrolide antibiotics efficiently induce erm expression, while other similar
drugs act as much poorer inducers? If macrolide antibiotics
block translation of all polypeptides, why does the sequence of
the nascent peptide matter for stalling? Which elements of the
ribosome are involved in stalling? Which step in translation is
blocked in the stalled complex? How does stalling on the ermCL
ORF compare with stalling on the tnaC or secM cistrons?
In order to get insights into the molecular details of drug-dependent ribosome stalling, we took a biochemical approach to
study stalling of the ribosome during translation of the ermC
leader peptide. By analyzing the nature and placement of peptidyl-tRNA in the stalled ribosome, and detecting the presence
and exact position of the stalled complex on ermCL mRNA, we
gained important information on how the stalling mechanism
may operate. Combining biochemical data with the knowledge
of the structure of the ribosome tunnel, we were able to identify
some of the key molecular players involved in the translation
arrest.
RESULTS
Erythromycin-Dependent Induction of ermC
in the E. coli Cell-Free Translation System
Inducible expression of ermC was originally characterized
in Gram-positive bacteria (Horinouchi and Weisblum, 1980;
Shivakumar et al., 1980), and subsequent studies have shown
that ermC can be inducibly expressed in E. coli cells (Bailey
et al., 2008; Hardy and Haefeli, 1982). Because our experimental strategy required the use of an E. coli cell-free system, it
was important to test whether erythromycin-dependent inducible expression of ermC could be reproduced in vitro (Shimizu
et al., 2001). When a bicistronic DNA template containing the
leader peptide ermCL and methylase ermC ORFs was expressed in the in vitro transcription-translation system in the
absence of erythromycin, the main translation product corresponded to the 2.2 kDa leader peptide, whereas no ErmC
was synthesized (Figure 1C). Only when erythromycin was
present did translation of a 29 kDa polypeptide representing
a full-size ErmC became evident. In agreement with in vivo
data (Mayford and Weisblum, 1990; Narayanan and Dubnau,
1985), induction of ermC expression in vitro was observed
only within a specific range of erythromycin concentrations
(0.05–2 mM) (Figure 1D). At very low drug concentrations ribosome stalling at the leader ORF is apparently inefficient,
whereas at high concentrations erythromycin inhibits translation altogether.
These results show that inducible expression of ermC can be
faithfully reproduced in vitro, therefore demonstrating that all the
elements required for ribosome stalling at the ermCL ORF are
present in the E. coli cell-free system.
The Site of Ribosome Stalling on the ermC Leader mRNA
The key element of the ermC induction model is the erythromycin-dependent ribosome stalling on the leader ORF. However,
not only has the exact stalling site remained unknown, but
the existence of the stalled ribosome complex on the leader
mRNA has never been directly shown. Here, we used the toeprinting technique (Hartz et al., 1988; Muto et al., 2006) to test
whether the stalled complex exists and is stable enough to block
reverse transcriptase progression; toe-printing could also help
to determine the exact position of the stalled ribosome on
mRNA. The ermCL ORF controlled by a T7 promoter was transcribed and translated in the cell-free system in the absence or
presence of erythromycin. Subsequently, a radiolabeled primer
was annealed to the 30 -proximal region of the leader ORF
transcript and extended by reverse transcriptase. The reaction
products were resolved on a denaturing polyacrylamide gel (Figure 1E). When erythromycin was present at sufficiently high concentrations, a strong reverse transcriptase stop was observed at
positions +40 and +41 relative to the 50 end of the ermCL ORF
(Figure 1B). A clear toe-printing signal was observed already at
0.2 mM erythromycin, and its strength increased with increasing
concentration of the drug. This result revealed the erythromycindependent formation of a stable stalled complex on the leader
mRNA. Because in the assay reverse transcriptase stops 16 to
17 residues away from the first nucleotide of the P site codon
(Hartz et al., 1988), the observed signal places the Ile9 codon
in the P site and the Ser10 codon at the A site of the stalled ribosome.
Placement of Peptidyl-tRNA
in the Stalled Complex
The toe-printing data described above are compatible with two
versions of the stalled complex. If the ribosome stalls after the
translocation step, then peptidyl-tRNAIle, carrying a 9 amino
acid-long nascent peptide with a C-terminal isoleucine, is expected to be in the classic P/P state. Alternatively, if the stalled
ribosome is trapped in the pretranslocation state, the 10 amino
acid-long peptidyl-tRNASer (having serine as its C-terminal
amino acid) will reside either in the A/A site or in the A/P hybrid
state (Moazed and Noller, 1989). In order to distinguish between
these possibilities, we used amino acid-specific labeling to define the nature of the C-terminal peptide residue of the peptidyl-tRNA in the stalled complex.
Ribosome stalling in the cell-free translation system should
lead to the accumulation of a specific peptidyl-tRNA trapped in
the stalled complex. Indeed, when products of the translation
reaction (directed by the ermCL ORF and carried out in the presence of [35S]methionine) were separated by gel electrophoresis in
Molecular Cell 30, 190–202, April 25, 2008 ª2008 Elsevier Inc. 191
Molecular Cell
Drug-Dependent Ribosome Stalling
192 Molecular Cell 30, 190–202, April 25, 2008 ª2008 Elsevier Inc.
Molecular Cell
Drug-Dependent Ribosome Stalling
a buffer system that preserved the integrity of the peptidyl-tRNA
ester bond, a slow-migrating band was observed (Figure 2A). The
appearance of this radioactive band depended on the presence
of erythromycin, suggesting that it represented a protein-containing component of the stalled complex. At the same time,
the material in the band was RNase sensitive, showing that the
protein was linked to RNA. Upon phenol extraction this material
partitioned into the aqueous phase—an indication that its protein
component was a short peptide or an amino acid (Figure 2A). Finally, the electrophoretic mobility of this band was slower than
that of [35S]fMet-tRNA (Figure 2B), which led to the conclusion
that this material represented peptidyl-tRNA from the stalled ribosome complex. The accumulating peptidyl-tRNA was largely
resistant to the action of peptidyl-tRNA hydrolase, which argued
that most of it remained tightly associated with the ribosome
rather than being a drop-off product (N.V.-L. and A.S.M., unpublished data).
The ermCL ORF contains three isoleucine codons (positions 3,
6, and 9). In order to obtain a leader peptide ORF variant with Ile9
as the sole isoleucine codon, the sequence of codons 3 and 6
was changed from ATT (Ile) to CTT (Leu). These mutations (I3L/
I6L) did not affect formation of the erythromycin-induced stalled
complex (Figure 2C). Next, the wild-type and the I3L/I6L mutant
ermCL ORFs were translated in the cell-free system in the presence of [14C]Ile. In both cases, the accumulation of 14C-labeled
peptidyl-tRNA was readily observed (Figure 2D), confirming
that peptidyl-tRNA in the stalled complex carries at least a 9
amino acid-long nascent peptide. A similar approach was used
to determine whether Ser10 is incorporated into the peptidyltRNA in the stalled complex. The ermCL sequence contains
two serine codons, Ser5 and Ser10. The Ser5 codon (AGT)
was mutated to an Ala codon (GCT), and toe-printing analysis
was again used to verify that this mutation allowed formation
of the stalled complex (Figure 2C). In agreement with this result,
when expression of S5A leader template was carried out in the
presence of [35S]Met, erythromycin-dependent accumulation
of peptidyl-tRNA was observed (Figure 2E). However, when the
same reaction was carried out in the presence of [14C]Ser, radioactive peptidyl-tRNA was seen only with the wild-type template
(reflecting incorporation of Ser5), but not with the S5A mutant.
This result shows that in the stalled complex, peptide bond
formation between Ile9 and Ser10 does not take place and Ile9
remains the last amino acid residue incorporated in the nascent
peptide before translation is arrested. Combining these data with
the results of the toe-printing analysis, we concluded that peptidyl-tRNAIle, carrying a 9 amino acid-long peptide MGIFSIFVI,
is positioned in the P/P site of the arrested ribosome.
The Stalled Ribosome Is Unable to Catalyze
Peptide Bond Formation
The presence of peptidyl-tRNA in the P site argued that the
stalled ribosome might be impaired in peptide bond formation.
Indeed, treatment of the stalled complex with puromycin, which
is expected to release the nascent peptide upon peptide bond
formation, had no effect on the toe-printing signal, indicating
that the stalled ribosome is refractory to the action of this antibiotic (data not shown). To verify this observation, we prepared
stalled complex on ermCL mRNA truncated immediately after
the Ile9 codon. Because of the lack of tmRNA in the system,
the ribosomes that reach the Ile9 codon at the 30 end of the truncated mRNA should remain associated with it and with peptidyltRNA, even in the absence of erythromycin. In the presence of
erythromycin, an authentic stalled complex is expected to assemble on truncated ermCL mRNA. Both types of complexes
were treated with 1 mM puromycin, and peptidyl-tRNA was
analyzed by gel electrophoresis (Figure 2F). Puromycin easily
released the nascent peptide from the peptidyl-tRNA in the complex formed in the absence of erythromycin. In contrast, the
stalled complex with erythromycin was refractory to puromycin
action. The most straightforward explanation of these results is
that the stalled ribosome is unable to catalyze peptide bond
formation.
Structural Elements of the Nascent Peptide
Critical for Stalling
Mutational studies have shown the importance of amino acid
residues 6–8 of the leader peptide for induction of ermC expression (Mayford and Weisblum, 1989). Although illuminating, these
results could not readily distinguish between the effects of mutations on mRNA conformation versus ribosome stalling. The toeprinting assay provides a more precise tool to investigate the
properties of the nascent peptide that directly affect formation
of the stalled complex. In order to identify the amino acid residues important for stalling, we performed alanine-scanning
Figure 1. The Site of Ribosome Stalling on the ermC Leader ORF
(A) Structure of the regulatory region of the inducible ermC gene in the noninduced conformation. Shine-Dalgarno regions and initiator codons of the leader peptide ORF (ermCL) and the resistance methylase cistron (ermC) are shown in bold. The amino acid sequences of the entire ErmCL peptide and the N-terminal
sequence of ErmC are indicated above the corresponding codons.
(B) Structure of the regulatory region in the induced conformation. The location of the toe-printing signal is shown by arrows, and the codon positioned in the
ribosomal P site in the stalled complex is underlined.
(C) Gel electrophoresis analysis of [35S] radiolabeled protein products accumulating in the cell-free translation system programmed with the monocistronic ermC
template (lane 1) or ermCL-ermC bicistronic template (lanes 2 and 3) in the absence () or presence (+) of 2 mM erythromycin. Positions of the full-size ErmC
protein and the ErmCL leader peptide are indicated by arrows.
(D) Erythromycin concentration dependence of ermC expression in the cell-free system. The data points were obtained by quantifying the amount of radioactivity
in the gel band corresponding to the full-size ErmC and normalizing versus the amount of ErmC expressed in the absence of erythromycin. Shown are the average
values of two experiments with the dispersion of actual values represented by error bars.
(E) The stalled ribosome complex on ermCL mRNA. Cell-free translation was carried out in the presence of increasing concentrations of erythromycin, and the
position of the stalled ribosome was determined by toe-printing analysis. The reaction products were analyzed in a sequencing gel alongside sequencing reactions (C, U, A, G) obtained from the same primer used for toe-printing. Positions of the toe-printing signals are shown by arrows, and the position of the full-size
reverse transcriptase product corresponding to the 50 end of ermCL mRNA transcript is indicated by an open arrow. The sequence of a portion of the ermCL ORF
and the encoded amino acids are expanded next to the sequencing lanes, with the Ile9 codon positioned in the P site of the stalled ribosome boxed.
Molecular Cell 30, 190–202, April 25, 2008 ª2008 Elsevier Inc. 193
Molecular Cell
Drug-Dependent Ribosome Stalling
Figure 2. Characterization of Peptidyl-tRNA from
the Stalled Complex
(A) Gel electrophoresis analysis of [35S] radiolabeled
products accumulating during in vitro translation of ermCL
mRNA. Reaction products were analyzed by gel electrophoresis in the Tricine buffer system (see the Supplemental Experimental Procedures). The position of the RNasesensitive material that accumulates in the presence of
erythromycin is indicated by a filled triangle.
(B) Gel electrophoresis analysis in the Bis-Tris buffer system (see the Supplemental Experimental Procedures) of
[35S]fMet-tRNA (indicated by an open triangle) and
[35S]peptidyl-tRNA (filled triangle) accumulating during
cell-free translation of ermCL mRNA in the presence of
erythromycin.
(C) Toe-printing analysis of stalled complex formation on
I3L/I6L or S5A ermCL mutant mRNAs. Bands corresponding to the stalled complex-specific toe-printing signals are
indicated by arrows.
(D) Accumulation of [14C]Ile peptidyl-tRNA during cell-free
translation of wild-type (WT) or mutant (I3L/I6L) ermCL
mRNA in the presence of erythromycin. The position of
peptidyl-tRNA is indicated by a filled triangle. Open triangles (in [D]–[F]) indicate a faster-migrating band, whose
nature is unclear, detected in some experiments. However, its presence was not erythromycin-dependent, suggesting that it did not originate from the stalled complex.
(E) [35S]Met or [14C]Ser peptidyl-tRNA accumulating
during translation of wild-type (WT) or mutant (S5A) ermCL
mRNA.
(F) Puromycin sensitivity of [35S]peptidyl-tRNA in the
stalled complex. Leader mRNA templates truncated immediately after Ile9 codon (ermCLD) or full-length ermCL
were expressed for 15 min in the cell-free system in the absence or presence of erythromycin and then treated for
90 s at 37 C with 1 mM puromycin. Reaction products
were phenol extracted and analyzed by gel electrophoresis.
mutagenesis of the Gly2-Ser10 segment of the leader peptide
and used toe-printing to monitor the ability of mutant nascent
peptides to stall the ribosome (Figure 3A). Alanine substitutions
at residues 2–5 or the mutation of Ser10 to alanine did not prevent erythromycin-dependent translation arrest. On the other
hand, substitutions of residues Ile6, Phe7, Val8, or Ile9 for Ala
abolished stalling. These results show that the C-terminal segment of the leader peptide is the most critical for erythromycindependent ribosome stalling.
The importance of the C-terminal amino acid of the nascent
peptide, located within the PTC active site, confirmed our conclusion that the PTC is likely the target point of the stalling mech-
194 Molecular Cell 30, 190–202, April 25, 2008 ª2008 Elsevier Inc.
anism. Because substitution of Ala for Ile9 introduces a rather dramatic change in the amino
acid structure, we tested whether a more conservative change, Ile9 to Leu, would also affect
formation of the stalled complex. The toe-printing result shows that even a relatively minor alteration in the structure of the P site amino
acid residue prevents stalling (Figure 3A, right
panel) re-emphasizing the notion of the PTC importance in stalling.
Mutations in codon 9 of the mRNA affect not only the P site
amino acid, but also the tRNA moiety of the peptidyl-tRNA,
which hypothetically may also affect stalling. However, substitution of the AUC9 codon, decoded by tRNA1Ile, with the AUA codon, decoded by tRNA2Ile, did not eliminate stalling (data not
shown). Therefore, the structure of the tRNA does not appear
to be critical, whereas the nature of the C-terminal amino acid
of the nascent peptide constitutes a key element in the mechanism of stalling.
The length of the ErmCL nascent peptide in the stalled complex is much shorter than that of the SecM and TnaC stalling
peptides. Even though the alanine scanning showed that only
Molecular Cell
Drug-Dependent Ribosome Stalling
Figure 3. Dependence of Stalled Complex
Formation on the Size and Sequence of
the Nascent Peptide
(A) Formation of the stalled complex on ermCL
mRNAs that carry mutations of individual ermCL
codons to alanine codons (left panel) and the
mutation of the Ile9 codon AUC to a Leu codon
CUC (right panel).
(B) Stalled complex formation on mutant ermCL
mRNA with addition or deletion of codons. The
following templates were used: 1, deletion of
the Gly2 codon; 2, deletion of Gly2 and Ile3 codons; 3, deletion of Gly2, Ile3, and Phe4 codons;
+1, +2, and +3 indicate addition of one, two, or
three GCG (Ala) codons after the initiator AUG codon. In (A) and (B), the toe-printing signal corresponding to the stalled complex is indicated by arrows. (B) Efficiency of stalled complex formation
on ermCL mRNA with addition or deletion of
codons.
(C) Quantification of the bands on the gel shown in
(B). The plot was obtained by calculating relative intensity of the two stalled complex-specific toeprinting bands versus the sum of these bands and
the full (ribosome-free) mRNA band in the same gel.
the sequence of the four C-terminal amino acids of the ErmCL is
critical for stalling, we wondered whether the N terminus of the
ErmCL nascent peptide plays any role in stalling. Site-directed
mutagenesis was used to delete one (Gly2), two (Gly2-Ile3), or
three (Gly2-Ile3-Phe4) codons in the leader ORF. Similarly, another set of templates was prepared by introducing one, two,
or three extra alanine codons GCG after the initiator codon of
the ermCL ORF. Toe-printing analysis showed that, regardless
of the codon deletions or additions, the site of stalling did not
change: translation was always arrested at the same Ile codon
corresponding to Ile9 in the wild-type leader sequence. Thus,
the IFVI C-terminal sequence of the nascent peptide defines
the stalling site and constitutes a strict prerequisite for the formation of the stalled complex (Figure 3B). However, changing the
length of the nascent peptide dramatically reduced the efficiency
of stalling (Figure 3C). This result leads to the conclusion that, although the N-terminal sequence of the ErmCL peptide is not critical, the N terminus has to reach a particular place in the tunnel to
promote stalling.
Specific Features of the Antibiotic Required for Stalling
The initial studies of ermC induction have demonstrated the importance of the structure of the inducing macrolide antibiotic
(Mayford and Weisblum, 1990). More recently, the newest generation of clinical macrolides, the ketolides, have been praised for
their lower erm-inducing activity compared to macrolides of previous generations (Bonnefoy et al., 1997; Bryskier, 2000). The inability of ketolides to induce erm was tentatively linked to the lack
of a C3 cladinose sugar, which in ketolides is replaced with
a keto group (Clarebout and Leclercq, 2002) (Figure 4, top).
The dissimilar effect of ketolides and macrolides on erm induction may have several different explanations, including the possibility that ketolides might stall the ribosome at a ‘‘wrong’’ place
on the ermCL ORF, which is not conducive to ermC induction.
Alternatively, ketolides may not induce stalling altogether.
To distinguish between these scenarios, we examined the formation and location of the stalled complex on ermCL mRNA using several different macrolide and ketolide antibiotics (Figure 4).
Substitution of erythromycin with telithromycin eliminated the
toe-printing signal, suggesting that some of the structural differences between these two drugs were critical for formation of the
stalled complex. Replacement of telithromycin with RU69847,
which is isostructural to telithromycin except for the presence
of the C3 cladinose, restored stalling, indicating that the presence of the C3 cladinose is critical for the formation of the stalled
complex. This result was confirmed using another pair of drugs,
HMR3004 and RU66252, which differed only in the presence of
cladinose (RU66252) or keto group (HMR3004) at the C3 position
of the lactone ring. Here again only the cladinose-containing
drug was able to induce stalling.
Ribosomal Elements Critical for Stalling
The cladinose sugar of the ribosome-bound macrolide antibiotics, which, as shown above, is critical for stalling, projects into
the tunnel and narrows its opening (Hansen et al., 2002; Schlünzen et al., 2001, 2003; Tu et al., 2005). On the opposite wall, another protrusion is formed by the base of a universally conserved
nucleotide, A2062 of 23S rRNA (Figure 6A). A2062 is optimally
positioned to directly interact with the critical C-terminal IFVI
sequence of the ErmCL nascent peptide (Figure 6B). Therefore,
this nucleotide was mutated to test its importance for stalling.
Ribosomes with all three possible A2062 mutations could support cell growth when expressed in the E. coli strain that lacked
wild-type ribosomes (Asai et al., 1999; Cruz-Vera et al., 2005).
However, in agreement with a previous report (Cochella and
Green, 2004), the A2062G mutant grew very slowly, which prevented using this mutation in our experiments. Importantly, introduction of either A2062U or A2062C mutations in 23S rRNA did
not notably affect sensitivity of the cells to erythromycin, indicating
that the drug can readily bind to the mutant ribosomes. This
Molecular Cell 30, 190–202, April 25, 2008 ª2008 Elsevier Inc. 195
Molecular Cell
Drug-Dependent Ribosome Stalling
Figure 4. Dependence of Stalled Complex Formation on the Nature of the Macrolide Antibiotic
Toe-printing analysis of the stalled complex on the ermCL mRNA upon in vitro translation in the presence of 50 mM cladinose-containing macrolide antibiotics (C)
or ketolides (K). Lane marked ‘‘’’ corresponds to the reaction carried in the absence of antibiotics. The sequence of a portion of the ermCL ORF and the encoded
amino acids are indicated, and the Ile9 codon positioned in the P site of the stalled ribosome is boxed.
conclusion was confirmed by RNA footprinting (data not shown).
We then tested the ability of A2062U and A2062C mutant ribosomes to translate proteins and induce stalling in the cell-free system. Mutant ribosomes were as active as wild-type ribosomes in
translating a control DHFR template (data not shown). However,
in contrast to wild-type ribosomes, both mutants, A2062C and
A2062U, failed to produce a stalled ErmCL complex in the presence of erythromycin (Figure 5A). These results revealed A2062
as a tunnel RNA element critical for drug-dependent ribosome
stalling.
196 Molecular Cell 30, 190–202, April 25, 2008 ª2008 Elsevier Inc.
In the TnaC and SecM nascent peptides, the identities of
amino acid residues located some distance from the PTC
were found to be important for stalling (Cruz-Vera et al., 2005;
Nakatogawa and Ito, 2002). In the extended conformation,
these segments of the nascent peptides would be close to
the b hairpin of protein L22 exposed in the tunnel lumen. Mutational and crystallographic studies led to the proposal that L22
may play an important role in ribosome stalling on secM and
tnaC cistrons (Berisio et al., 2003; Cruz-Vera et al., 2005; Nakatogawa and Ito, 2002; Tu et al., 2005; Woolhead et al., 2006).
Molecular Cell
Drug-Dependent Ribosome Stalling
DISCUSSION
Figure 5. Ribosomal Elements Involved in Drug-Dependent Stalling
(A) The effect of A2062 mutations on stalled complex formation.
(B) Effect of a 3 amino acid deletion in the b loop of protein L22 on ribosome
stalling. In (A) and (B), translation reactions were carried out in the cell-free
translation system devoid of endogenous wild-type ribosomes. Filled arrows
indicate the toe-printing signal corresponding to the stalled complex, and
open arrows indicate the 50 end of the mRNA transcript.
The notion that the ErmCL nascent peptide should be of a certain length for drug-dependent stalling to occur (Figures 3B and
3C) prompted us to test whether mutations in the L22 b hairpin
would also affect ribosome stalling at the ermCL ORF. We employed mutant ribosomes that lack three amino acids (Met82Arg84) in the L22 b hairpin (Chittum and Champney, 1994).
This deletion does not significantly alter erythromycin binding
(Chittum and Champney, 1994; Tu et al., 2005) and thus was
suitable for our purposes. Furthermore, L22 mutant ribosomes
were as active in translating a control protein template as wildtype ribosomes (data not shown). Toe-printing showed that the
3 amino acid deletion in the L22 b hairpin considerably (3-fold)
reduced the relative amount of the stalled complex formed in
the cell-free translation system, therefore implicating involvement of protein L22 in the mechanism of drug-dependent ribosome stalling.
The use of the cell-free transcription-translation system provided
important insights into the structure of the stalled complex and
the molecular mechanism of antibiotic-dependent stalling. Our
study made it possible to detect the presence of the stalled
ribosome on ermCL mRNA and assess the contribution of the
peptide, the drug, and the ribosome on formation of the stalled
complex.
Toe-printing placed the Ile9 codon of the ermCL mRNA in the
P site of the stalled ribosome, and selective amino acid labeling
showed that the C-terminal residue of the nascent peptide is isoleucine. These data revealed the nature of the peptide and the
position of the peptidyl-tRNA in the stalled complex and showed
that a 9 amino acid-long nascent peptide is esterified to tRNAIle
positioned in the classic P/P site of the ribosome. The inability of
puromycin to release the peptide and disassemble the stalled
complex hints that the stalled ribosome is unable to catalyze
peptide bond formation and that the PTC may be the focal point
of the stalling mechanisms. This conclusion is reinforced by the
observation that the nature of the amino acid residue that occupies the PTC P site in the stalled complex is critically important—even a conservative change of the C-terminal isoleucine
to leucine drastically reduces stalling.
Alanine scanning identified the C-terminal sequence IFVI9 of
the nascent chain as crucial for stalling. Changes in the nascent
peptide length, in the structure of the inducing macrolide antibiotic, or even in the ribosome itself can diminish or even completely abolish stalling, but never ‘‘shift’’ the stalling site. Thus,
the display of this critical 4 amino acid-long C-terminal sequence
of the nascent peptide in the PTC-proximal tunnel segment
appears to be a necessary prerequisite for the erythromycindependent stalling.
It has long been known that the antibiotic structure is important for erm induction, although the molecular reasons for this
were obscure (Mayford and Weisblum, 1990). Our results show
that the cladinose sugar at the C3 position of the macrolide lactone is the specific drug element essential for stalling. The placement of the C3 cladinose in the ribosome-bound antibiotic is
such that its direct contact with the IFVI sequence of the nascent
peptide is practically unavoidable (Figures 6A and 6B). This specific drug-peptide interaction likely plays a key role in the mechanism of stalling. Most of the inducible erm genes are preceded
by short ORFs that often show only marginal similarity to the
ermC leader peptide and are induced by different types of macrolide drugs. Thus the nature of the stalling nascent peptide sequence might be dictated by the chemical structure of the inducing antibiotic. Our finding that the cladinose-lacking antibiotics
fail to promote ribosome stalling provides the molecular basis
for the clinically relevant observation that ketolides are poor
inducers of erm (Bailey et al., 2008; Bonnefoy et al., 1997; Clarebout and Leclercq, 2002; Rosato et al., 1998).
The narrowing of the tunnel by the cladinose-containing macrolides brings the nascent peptide into a close proximity with
a universally conserved 23S rRNA residue, A2062, exposed on
the wall of the tunnel precisely at the level with the IFVI sequence
(Figures 6A and 6B). Our mutational studies showed that the
identity of this universally conserved residue is crucial for
Molecular Cell 30, 190–202, April 25, 2008 ª2008 Elsevier Inc. 197
Molecular Cell
Drug-Dependent Ribosome Stalling
Figure 6. A Model of the Molecular Mechanism of Antibiotic-Dependent Ribosome Stalling
(A) A model of the 9 amino acid-long ErmCL nascent peptide in the exit tunnel of the erythromycin-bound ribosome. The peptide is shown in ball-and-stick
representation. The five N-terminal amino acid residues are colored pale green, and the four C-terminal residues, critical for stalling, are shown in bright green.
Peptide sequence is indicated. The CCA 30 end of the P site-bound peptidyl-tRNA is colored wheat. Erythromycin is shown as salmon-colored sticks with the
cladinose sugar highlighted in red. The van der Waals surface of the drug is represented by mesh. A2062 of 23S rRNA is shown in hot pink, C2063 and G2061 are
pale pink, and A2451 is magenta. The b loop of protein L22 is colored cyan.
(B) Relative location of the nascent peptide, erythromycin, and A2062 in the exit tunnel (viewed from the PTC down the tunnel). Color scheme as in (A).
(C) Relative location of A2062, exposed in the tunnel, linked to nucleotides in the PTC active site of the E. coli ribosome (Schuwirth et al., 2005).
(D) A general model of the drug- and nascent peptide-dependent stalling. The color scheme matches that of (A). Solid arrow marks communication of a signal
from A2062 to the PTC active site. Dashed arrows indicate possible contribution of L22 (or other tunnel elements) to establishing inactive conformation of the PTC
or stabilizing peptidyl-tRNA in the ribosome.
198 Molecular Cell 30, 190–202, April 25, 2008 ª2008 Elsevier Inc.
Molecular Cell
Drug-Dependent Ribosome Stalling
stalling. Crystallographic studies revealed the mobile nature of
A2062 (Hansen et al., 2002). Two residues that flank A2062,
G2061 and C2063, project directly into the PTC active site (Figures 6A and 6C). G2061 is hydrogen bonded with A2451, one of
the key residues in the PTC, and C2063 is hydrogen bonded with
its immediate neighbor, A2450. Mutations of either G2061,
C2063, A2450, or A2451 are deleterious (Bayfield et al., 2004;
Hesslein et al., 2004; Muth et al., 2000). Because of its link to
the PTC residues, A2062 is optimally positioned to sense the nature of the nascent peptide in the exit tunnel and to communicate
signals to the PTC.
Formation of the stalled complex at the ermCL mRNA requires
a nascent peptide of a specific length. A possible interpretation of
this finding is that the nascent peptide has to engage more distal
elements of the tunnel for a stable complex to form. The loop of
protein L22, which is exposed in the tunnel lumen near the N terminus of the extended 9 amino acid-long ErmCL nascent peptide, is an immediate suspect. Not only was the L22 b loop shown
to be important for the formation of stalled complexes in SecM
and TnaC systems, but alteration of this loop decreased the
amount of stalled complex formed on the ermCL template
(Figure 5B). The flexibility of the L22 b loop and its proposed function in gating the tunnel (Berisio et al., 2003; Tu et al., 2005) are in
line with its possible contribution to drug-dependent stalling.
Whereas placement of the IFVI sequence too close to the N
terminus is not conducive for stalling, moving this sequence
too far from the terminus also prevents formation of the stalled
complex. This observation is in line with the results of recent
studies that showed that erythromycin efficiently promotes
drop-off of peptidyl-tRNAs with a nascent peptide exceeding
9 amino acids in length (Lovmar et al., 2004; Tenson et al.,
2003). With the IFVI sequence moved farther away from the N
terminus, the ribosome is likely to lose the peptidyl-tRNA before
the critical stalling segment of the nascent peptide is synthesized. Thus, in the course of evolution of the ermC induction
mechanism, the placement of the IFVI sequence within the
ErmCL nascent peptide structure has been carefully optimized:
its position within the nascent peptide allows for the peptide N
terminus to reach and engage distant elements of the tunnel
(possibly protein L22), but it is close enough to the N terminus
of the nascent peptide to avoid macrolide-induced peptidyltRNA drop-off prior to synthesis of the stalling sequence.
Combined, our experimental data suggest how the stalling
mechanism may operate (Figure 6D). Three main players involved in formation of the stalled complex during translation
of the ermCL ORF are the drug, the nascent peptide, and the ribosome. A macrolide antibiotic narrows the nascent peptide
exit tunnel in the vicinity of the PTC; the presence of a cladinose
sugar at the C3 position of the lactone ring is essential to sufficiently reduce the tunnel aperture. Once the IFVI sequence is
synthesized during translation of the ermCL ORF, its specific interactions with the drug bring it into the proximity of A2062 and
promote reorientation of this trigger residue. The conformational change of A2062 is communicated, likely via its immediate neighbors, G2061 and C2063, to the peptidyl transferase
active site (Figure 6C). Alternatively, the conformational change
can be transmitted to the PTC via the nascent peptide itself.
Changes in the PTC conformation prevent formation of the
next peptide bond. The side chain of the nascent peptide C-terminal isoleucine located in the P site helps to switch the PTC to
an inactive conformation. Engagement of a more distant segment of the tunnel, possibly involving the b loop of protein
L22, is essential for either communicating the stalling signal to
the PTC or, possibly, for stabilizing the association of peptidyl-tRNA with the stalled complex and preventing rapid
drop-off.
Drug-induced ribosome stalling during translation of the ermC
leader, though idiosyncratic in many of its features, shares important similarities with the ribosome stalling at secM and tnaC
regulatory ORFs (Gong et al., 2001; Nakatogawa and Ito, 2002).
In all three systems the catalytic activity of the PTC is inhibited:
the stalled ribosome fails to catalyze peptide bond formation
(ermCL and secM) or peptide release (tnaC). The identity of the
amino acid residues present in the P site of the PTC (Ile9 for
ErmCL, Pro24 for TnaC, Gly165 for SecM) is critical for stalling
and may help shape the inactive conformation of the catalytic
center. Thus the mechanism of stalling takes advantage of the
known ability of the PTC active site to sense the nature of the
amino acid residue (Dale and Uhlenbeck, 2005; Krayevsky and
Kukhanova, 1979).
Ribosome stalling at tnaC, secM, or ermCL ORFs requires the
presence of an effector that directly or indirectly mediates stalling. Binding of free tryptophan near the PTC A site is thought to
drive formation of the stalled complex at the tnaC mRNA (CruzVera et al., 2006; Gong and Yanofsky, 2002). The presence of
Pro-tRNAPro in the A site of the SecM stalled complex is critical
for stalling the ribosome with SecM1–165-tRNAGly bound in the P
site (Muto et al., 2006). In the ErmCL stalled complex, the A site
seems to play a less important role. The identity or even the mere
presence of aminoacyl-tRNA in the A site is not consequential:
missense mutations at the Ser10 codon do not affect stalled
complex formation (Figure 3C), and nonsense mutations do not
alleviate ermC induction (Mayford and Weisblum, 1989). Instead
of an effector in the A site, the ErmCL stalled complex requires
the presence of a macrolide molecule in the tunnel. While the A
site-bound effectors may help sculpt the inactive conformation
of the PTC in the SecM and TnaC stalled complexes, the inducing antibiotic may promote specific interactions of the nascent
peptide with the sensory elements of the tunnel during stalling
at the ermC leader ORF.
The presence of a specific amino acid sequence at the C terminus of the nascent peptide is required for ribosome stalling at the
secM and ermCL ORFs. However, there is no obvious similarity
between the SecM C-terminal stalling sequence (GIRAG165) (Nakatogawa and Ito, 2002) and the ErmCL stalling sequence IFVI9.
This is not surprising given that ErmCL and SecM nascent peptides may use different strategies to engage tunnel sensors.
The identity of the IFVI sequence of ErmCL is likely dictated by
the antibiotic with which this sequence interacts. A particular
C-terminal sequence of the SecM nascent peptide appears to
promote a specific compact conformation of the nascent peptide
in the tunnel (Woolhead et al., 2006). In such a conformation, the
GIRAG sequence of the SecM nascent peptide might come into
contact with the same elements of the tunnel (including possibly
A2062) that are contacted by the ErmCL peptide IFVI sequence in
the presence of antibiotic.
Molecular Cell 30, 190–202, April 25, 2008 ª2008 Elsevier Inc. 199
Molecular Cell
Drug-Dependent Ribosome Stalling
Several rRNA residues of the PTC-proximal segment of the
exit tunnel are likely essential in the mechanism of stalling. Two
nucleotide residues, A2058 and U2609, were found to be important for stalling at the SecM and TnaC peptides, respectively
(Cruz-Vera et al., 2005; Nakatogawa and Ito, 2002). Interestingly,
both of these residues are involved in macrolide binding (Ettayebi et al., 1985; Garza-Ramos et al., 2001; Schlünzen et al.,
2001; Tu et al., 2005) and thus are also involved in the drugdependent stalling at the ermCL ORF. We found here that mutations of A2062 completely eliminate stalling in the ErmCL system. It would be interesting to examine the role of this nucleotide
in other stalling bacterial systems. Furthermore, given the nearuniversal conservation of A2062 this residue may be essential
even for stalling of eukaryotic ribosomes (Cao and Geballe,
1996; Onouchi et al., 2005; Wang and Sachs, 1997).
The 9 amino acid ErmCL stalling peptide is much shorter than
the nascent peptides in the TnaC and SecM stalled complexes
(the latter span the entire length of the tunnel). In spite of this difference, all these nascent peptides appear to form critical interactions at the tunnel constriction. In TnaC and SecM peptides,
amino acid residues critical for stalling are located 9–11 residues
away from the C terminus (Gong and Yanofsky, 2002; Nakatogawa and Ito, 2002). The optimal length of the ErmCL peptide
in the stalled complex is 9 amino acids, and thus the N-terminal
fMet is expected to be placed close to the tunnel segment occupied by the critical residues of SecM and TnaC peptides. The
tunnel element that might be universally involved in interaction
with these distal sequences of the stalling peptides is the
b loop of protein L22 (Figures 5B and 6B). Indeed, the L22 mutations affect stalling in all of the three systems.
The picture that emerges from comparison of SecM, TnaC,
and ErmCL stalling peptides shows that the key components
of the stalling mechanism are similar and involve specific elements of the ribosome tunnel, the nascent peptide, and the
PTC. Therefore, it appears that the ability to stall or pause in response to a specific nascent peptide sequence is an intrinsic
property of the ribosome. Yet the strategies that SecM, TnaC,
and ErmCL peptides use to activate stalling are substantially
different. This notion hints that nascent peptide-dependent
stalling may not be an ‘‘exotic’’ trait used for regulation of expression of very few specific genes, but a general mechanism
of translation that adjusts the rate of polypeptide synthesis to
the specific requirements of expression, folding, and targeting
of proteins.
Cell-free Translation
Construction of DNA templates is described in the Supplemental Experimental
Procedures available online. The ‘‘Classic’’ PURE cell-free transcription-translation system (Post Genome Institute, Inc., Japan) (Shimizu et al., 2001) was
used for in vitro protein synthesis (Muto et al., 2006). In a typical experiment,
0.5–1 pmol of the purified template DNA (plasmid or PCR product) was combined on ice with 6.25 ml of the manufacturer’s solution A and 1.25 ml solution B
(containing 5–10 pmol ribosomes). When necessary, antibiotics were added
from concentrated stocks. The final concentration of erythromycin in most reactions was 50 mM. In experiments when radiolabeling of the translation products was required, 5 mCi of [35S]Met or 100 mCi of [14C]Ile or Ser was added to
the reaction. The reaction was initiated by placing tubes in 37 C water bath.
Reactions were typically carried out for 1 hr.
In experiments with mutant ribosomes, the ‘‘D ribosome’’ version of the
PURE system was supplemented with ribosomes (25 pmol per reaction).
Translation products were analyzed as described in the Supplemental
Experimental Procedures.
Model of the Nascent Peptide in the Ribosome
The model of an 8 amino acid-long ErmCL nascent peptide in the ribosome
tunnel presented in Tu et al. (2005) was used as a starting point. The ninth
amino acid (Ile) and formyl group at the N-terminal methionine were added
computationally, and the peptide was translocated down the tunnel to provide
the best backbone overlap of the amino acid residues 2–8 of the new peptide
with the peptide in the model of Tu et al. The orientation of the side chains and
of the N-terminal fMet was manually adjusted to avoid clashes with the tunnel
walls and with the antibiotic.
SUPPLEMENTAL DATA
Supplemental Data include Supplemental Experimental Procedures and can
be found with this article online at http://www.molecule.org/cgi/content/full/
30/2/190/DC1/.
ACKNOWLEDGMENTS
We wish to thank Liqun Xiong for help with footprinting experiments, Beatriz
Llano-Sotelo for guidance in some experimental procedures, Bernard Santarsiero and Debbie Mulhearn for help with modeling the nascent peptide in the
tunnel, Scott Champney for the gift of the strain N281, Catherine Squires for
making the strain SQ171 available to us and advising on its use, Gregor Blaha
for sharing the PDB file of the 8 amino acid-long peptide in the tunnel, Takuya
Ueda for advice on using PURESYSTEM, Bernard Weisblum for discussion,
and Lisa Smith for proofreading the manuscript. This work was supported
by a grant from the National Science Foundation (MCB-0515934).
Received: December 20, 2007
Revised: January 31, 2008
Accepted: February 27, 2008
Published: April 24, 2008
REFERENCES
EXPERIMENTAL PROCEDURES
Mutagenesis of rRNA and Preparation of Ribosomes
The pLK35 plasmid (Douthwaite et al., 1989) was used for introduction of mutations at A2062 of the 23S rRNA gene. Three possible mutations were introduced using three mutagenizing primers 50 -CGGCAAGACGGAAAGXCCC
CGTGAACCTTT-30 where X was C, U, or G. Mutated plasmids were transformed in POP2136 cells and propagated at 30 C. After the presence of the
mutation was verified by sequencing, the mutant plasmids were transformed
into SQ171 cells that lack chromosomal copies of rRNA operons (Asai et al.,
1999) (S. Quan and C.L. Squires, personal communication). Plasmid exchange
was done as in Zaporojets et al. (2003).
Mutant cells were grown in 1 liter Luria-Bertani broth to A600 of 0.6
and harvested, and ribosomes were prepared by the method of Ohashi
et al. (2007).
200 Molecular Cell 30, 190–202, April 25, 2008 ª2008 Elsevier Inc.
Asai, T., Zaporojets, D., Squires, C., and Squires, C.L. (1999). An Escherichia
coli strain with all chromosomal rRNA operons inactivated: complete
exchange of rRNA genes between bacteria. Proc. Natl. Acad. Sci. USA 96,
1971–1976.
Bailey, M., Chettiath, T., and Mankin, A.S. (2008). Induction of erm(C) expression by non-inducing antibiotics. Antimicrob. Agents Chemother. 52, 866–874.
Bayfield, M.A., Thompson, J., and Dahlberg, A.E. (2004). The A2453-C2499
wobble base pair in Escherichia coli 23S ribosomal RNA is responsible for
pH sensitivity of the peptidyltransferase active site conformation. Nucleic
Acids Res. 32, 5512–5518.
Berisio, R., Schluenzen, F., Harms, J., Bashan, A., Auerbach, T., Baram, D.,
and Yonath, A. (2003). Structural insight into the role of the ribosomal tunnel
in cellular regulation. Nat. Struct. Biol. 10, 366–370.
Molecular Cell
Drug-Dependent Ribosome Stalling
Bonnefoy, A., Girard, A.M., Agouridas, C., and Chantot, J.F. (1997). Ketolides
lack inducibility properties of MLS(B) resistance phenotype. J. Antimicrob.
Chemother. 40, 85–90.
Horinouchi, S., and Weisblum, B. (1980). Posttranscriptional modification of
mRNA conformation: mechanism that regulates erythromycin-induced resistance. Proc. Natl. Acad. Sci. USA 77, 7079–7083.
Bryskier, A. (2000). Ketolides-telithromycin, an example of a new class of
antibacterial agents. Clin. Microbiol. Infect. 6, 661–669.
Jenni, S., and Ban, N. (2003). The chemistry of protein synthesis and voyage
through the ribosomal tunnel. Curr. Opin. Struct. Biol. 13, 212–219.
Cao, J., and Geballe, A.P. (1996). Inhibition of nascent-peptide release at
translation termination. Mol. Cell. Biol. 16, 7109–7114.
Krayevsky, A.A., and Kukhanova, M.K. (1979). The peptidyltransferase center
of ribosomes. Prog. Nucleic Acid Res. Mol. Biol. 23, 1–51.
Chittum, H.S., and Champney, W.S. (1994). Ribosomal protein gene sequence
changes in erythromycin-resistant mutants of Escherichia coli. J. Bacteriol.
176, 6192–6198.
Lovmar, M., Tenson, T., and Ehrenberg, M. (2004). Kinetics of macrolide action: the josamycin and erythromycin cases. J. Biol. Chem. 279, 53506–53515.
Clarebout, G., and Leclercq, R. (2002). Fluorescence assay for studying the
ability of macrolides to induce production of ribosomal methylase. Antimicrob.
Agents Chemother. 46, 2269–2272.
Cochella, L., and Green, R. (2004). Isolation of antibiotic resistance mutations
in the rRNA by using an in vitro selection system. Proc. Natl. Acad. Sci. USA
101, 3786–3791.
Cruz-Vera, L.R., Rajagopal, S., Squires, C., and Yanofsky, C. (2005). Features
of ribosome-peptidyl-tRNA interactions essential for tryptophan induction of
tna operon expression. Mol. Cell 19, 333–343.
Cruz-Vera, L.R., Gong, M., and Yanofsky, C. (2006). Changes produced by
bound tryptophan in the ribosome peptidyl transferase center in response to
TnaC, a nascent leader peptide. Proc. Natl. Acad. Sci. USA 103, 3598–3603.
Dale, T., and Uhlenbeck, O.C. (2005). Amino acid specificity in translation.
Trends Biochem. Sci. 30, 659–665.
Douthwaite, S., Powers, T., Lee, J.Y., and Noller, H.F. (1989). Defining the
structural requirements for a helix in 23S ribosomal RNA that confers erythromycin resistance. J. Mol. Biol. 209, 655–665.
Ettayebi, M., Prasad, S.M., and Morgan, E.A. (1985). Chloramphenicol-erythromycin resistance mutations in a 23S rRNA gene of Escherichia coli. J. Bacteriol. 162, 551–557.
Gabashvili, I.S., Gregory, S.T., Valle, M., Grassucci, R., Worbs, M., Wahl, M.C.,
Dahlberg, A.E., and Frank, J. (2001). The polypeptide tunnel system in the ribosome and its gating in erythromycin resistance mutants of L4 and L22. Mol.
Cell 8, 181–188.
Garza-Ramos, G., Xiong, L., Zhong, P., and Mankin, A. (2001). Binding site of
macrolide antibiotics on the ribosome: new resistance mutation identifies
a specific interaction of ketolides with rRNA. J. Bacteriol. 183, 6898–6907.
Mayford, M., and Weisblum, B. (1989). ermC leader peptide. Amino acid sequence critical for induction by translational attenuation. J. Mol. Biol. 206,
69–79.
Mayford, M., and Weisblum, B. (1990). The ermC leader peptide: amino acid
alterations leading to differential efficiency of induction by macrolide-lincosamide-streptogramin B antibiotics. J. Bacteriol. 172, 3772–3779.
Mitra, K., Schaffitzel, C., Fabiola, F., Chapman, M.S., Ban, N., and Frank, J.
(2006). Elongation arrest by SecM via a cascade of ribosomal RNA rearrangements. Mol. Cell 22, 533–543.
Moazed, D., and Noller, H.F. (1989). Intermediate states in the movement of
transfer RNA in the ribosome. Nature 342, 142–148.
Muth, G.W., Ortoleva-Donnelly, L., and Strobel, S.A. (2000). A single adenosine with a neutral pKa in the ribosomal peptidyl transferase center. Science
289, 947–950.
Muto, H., Nakatogawa, H., and Ito, K. (2006). Genetically encoded but nonpolypeptide prolyl-tRNA functions in the A site for SecM-mediated ribosomal
stall. Mol. Cell 22, 545–552.
Nakatogawa, H., and Ito, K. (2001). Secretion monitor, SecM, undergoes selftranslation arrest in the cytosol. Mol. Cell 7, 185–192.
Nakatogawa, H., and Ito, K. (2002). The ribosomal exit tunnel functions as a
discriminating gate. Cell 108, 629–636.
Narayanan, C.S., and Dubnau, D. (1985). Evidence for the translational attenuation model: ribosome-binding studies and structural analysis with an in vitro
run-off transcript of ermC. Nucleic Acids Res. 13, 7307–7326.
Nissen, P., Hansen, J., Ban, N., Moore, P.B., and Steitz, T.A. (2000). The structural basis of ribosome activity in peptide bond synthesis. Science 289,
920–930.
Gong, F., and Yanofsky, C. (2002). Instruction of translating ribosome by
nascent peptide. Science 297, 1864–1867.
Ohashi, H., Shimizu, Y., Ying, B.W., and Ueda, T. (2007). Efficient protein
selection based on ribosome display system with purified components.
Biochem. Biophys. Res. Commun. 352, 270–276.
Gong, F., Ito, K., Nakamura, Y., and Yanofsky, C. (2001). The mechanism of
tryptophan induction of tryptophanase operon expression: tryptophan inhibits
release factor-mediated cleavage of TnaC-peptidyl-tRNA(Pro). Proc. Natl.
Acad. Sci. USA 98, 8997–9001.
Onouchi, H., Nagami, Y., Haraguchi, Y., Nakamoto, M., Nishimura, Y., Sakurai,
R., Nagao, N., Kawasaki, D., Kadokura, Y., and Naito, S. (2005). Nascent peptide-mediated translation elongation arrest coupled with mRNA degradation in
the CGS1 gene of Arabidopsis. Genes Dev. 19, 1799–1810.
Gryczan, T.J., Grandi, G., Hahn, J., Grandi, R., and Dubnau, D. (1980). Conformational alteration of mRNA structure and the posttranscriptional regulation of
erythromycin-induced drug resistance. Nucleic Acids Res. 8, 6081–6097.
Rosato, A., Vicarini, H., Bonnefoy, A., Chantot, J.F., and Leclercq, R. (1998). A
new ketolide, HMR 3004, active against streptococci inducibly resistant to
erythromycin. Antimicrob. Agents Chemother. 42, 1392–1396.
Hansen, J.L., Ippolito, J.A., Ban, N., Nissen, P., Moore, P.B., and Steitz, T.A.
(2002). The structures of four macrolide antibiotics bound to the large
ribosomal subunit. Mol. Cell 10, 117–128.
Schlünzen, F., Zarivach, R., Harms, J., Bashan, A., Tocilj, A., Albrecht, R., Yonath, A., and Franceschi, F. (2001). Structural basis for the interaction of antibiotics with the peptidyl transferase centre in eubacteria. Nature 413, 814–821.
Hardy, K., and Haefeli, C. (1982). Expression in Escherichia coli of a staphylococcal gene for resistance to macrolide, lincosamide, and streptogramin type
B antibiotics. J. Bacteriol. 152, 524–526.
Schlünzen, F., Harms, J.M., Franceschi, F., Hansen, H.A., Bartels, H., Zarivach, R., and Yonath, A. (2003). Structural basis for the antibiotic activity of
ketolides and azalides. Structure 11, 329–338.
Harms, J., Schluenzen, F., Zarivach, R., Bashan, A., Gat, S., Agmon, I., Bartels,
H., Franceschi, F., and Yonath, A. (2001). High resolution structure of the large
ribosomal subunit from a mesophilic eubacterium. Cell 107, 679–688.
Schuwirth, B.S., Borovinskaya, M.A., Hau, C.W., Zhang, W., Vila-Sanjurjo, A.,
Holton, J.M., and Cate, J.H. (2005). Structures of the bacterial ribosome at
3.5 Å resolution. Science 310, 827–834.
Hartz, D., McPheeters, D.S., Traut, R., and Gold, L. (1988). Extension inhibition
analysis of translation initiation complexes. Methods Enzymol. 164, 419–425.
Shimizu, Y., Inoue, A., Tomari, Y., Suzuki, T., Yokogawa, T., Nishikawa, K., and
Ueda, T. (2001). Cell-free translation reconstituted with purified components.
Nat. Biotechnol. 19, 751–755.
Hesslein, A.E., Katunin, V.I., Beringer, M., Kosek, A.B., Rodnina, M.V., and
Strobel, S.A. (2004). Exploration of the conserved A+C wobble pair within
the ribosomal peptidyl transferase center using affinity purified mutant ribosomes. Nucleic Acids Res. 32, 3760–3770.
Shivakumar, A.G., Hahn, J., Grandi, G., Kozlov, Y., and Dubnau, D. (1980).
Posttranscriptional regulation of an erythromycin resistance protein specified
by plasmic pE194. Proc. Natl. Acad. Sci. USA 77, 3903–3907.
Molecular Cell 30, 190–202, April 25, 2008 ª2008 Elsevier Inc. 201
Molecular Cell
Drug-Dependent Ribosome Stalling
Tenson, T., Lovmar, M., and Ehrenberg, M. (2003). The mechanism of action of
macrolides, lincosamides and streptogramin B reveals the nascent peptide
exit path in the ribosome. J. Mol. Biol. 330, 1005–1014.
Tu, D., Blaha, G., Moore, P.B., and Steitz, T.A. (2005). Structures of MLSBK
antibiotics bound to mutated large ribosomal subunits provide a structural
explanation for resistance. Cell 121, 257–270.
Vazquez, D. (1975). The Macrolide Antibiotics. In Antibiotics III. Mechanism of
action of antimicrobial and antitumor agents, J.W. Corcoran and F.E. Hahn,
eds. (New York: Springer-Verlag), pp. 459–479.
202 Molecular Cell 30, 190–202, April 25, 2008 ª2008 Elsevier Inc.
Wang, Z., and Sachs, M.S. (1997). Ribosome stalling is responsible for arginine-specific translational attenuation in Neurospora crassa. Mol. Cell. Biol.
17, 4904–4913.
Weisblum, B. (1995). Insights into erythromycin action from studies of its activity as inducer of resistance. Antimicrob. Agents Chemother. 39, 797–805.
Woolhead, C.A., Johnson, A.E., and Bernstein, H.D. (2006). Translation arrest
requires two-way communication between a nascent polypeptide and the
ribosome. Mol. Cell 22, 587–598.
Zaporojets, D., French, S., and Squires, C.L. (2003). Products transcribed from
rearranged rrn genes of Escherichia coli can assemble to form functional
ribosomes. J. Bacteriol. 185, 6921–6927.