Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2011 Dec 22.
Published in final edited form as: Structure. 2010 May 12;18(5):627–637. doi: 10.1016/j.str.2010.02.011

Recognition of the regulatory nascent chain TnaC by the ribosome

Leonardo G Trabuco a,b, Christopher B Harrison a, Eduard Schreiner a, Klaus Schulten a,c,*
PMCID: PMC3244694  NIHMSID: NIHMS343793  PMID: 20462496

Abstract

Regulatory nascent chains interact with the ribosomal exit tunnel and modulate their own translation. To characterize nascent chain recognition by the ribosome at the atomic level, extensive molecular dynamics simulations of TnaC, the leader peptide of the tryptophanase operon, inside the exit tunnel were performed for an aggregate time of 2.1 μs. The simulations, complemented by quantum chemistry calculations, suggest that the critical TnaC residue W12 is recognized by the ribosome via a cation-π interaction, whereas TnaC's D16 forms salt bridges with ribosomal proteins. The simulations also show that TnaC-mediated translational arrest does not involve a swinging of ribosomal protein L22, as previously proposed. Furthermore, bioinformatic analyses and simulations suggest nascent chain elements which may prevent translational arrest in various organisms. Altogether, the current study unveils atomic-detail interactions that explain the role of elements of TnaC and the ribosome essential for translational arrest.

Introduction

Gene expression is regulated at multiple levels, such as transcription, post-transcriptional modification, mRNA processing, mRNA stability, translation, post-translational modification, and protein degradation. These regulatory levels do not necessarily operate in isolation. For instance, there exist several examples in which translation of certain messages can control transcription of downstream genes via regulation of translation by specific nascent peptides (Tenson and Ehrenberg, 2002). The exit tunnel of the ribosome, once viewed as a passive path, is now known to respond in different ways to signals present in nascent chains. Possible responses include modulation of the rate of translation (Lu and Deutsch, 2008), recruitment of ribosome-binding partners such as the signal recognition particle (Bornemann et al., 2008), and inhibition of peptide bond formation (Nakatogawa and Ito, 2002) or translation termination (Gong and Yanofsky, 2002).

Most known examples of ribosome regulation by nascent chains involve translational stalling. A classical example in bacteria is the SecM protein, translated from an open reading frame (ORF) upstream to the ORF coding for SecA, involved in protein export (Nakatogawa and Ito, 2002). When expression levels of SecA are low, SecM-mediated translational stalling leads to rearrangement of the mRNA, exposing the Shine-Dalgarno sequence of the downstream SecA ORF, thus leading to SecA translation. When SecA levels are high, however, SecA mechanically pulls SecM out of the ribosome, inhibiting its own expression via a negative feedback mechanism (Nakatogawa and Ito, 2002). In contrast to SecM, which can induce stalling without any additional factors, several regulatory nascent chains stall their own translation only in the presence of a small effector. For instance, non-lethal concentrations of certain antibiotics lead to translational stalling of regulatory leader peptides, inducing translation of downstream genes conferring antibiotic resistance (reviewed in (Ramu et al., 2009)). Interestingly, in vivo analysis of ribosome occupancy on yeast mRNAs showed a surprisingly large number of translated regions upstream of known ORFs (Ingolia et al., 2009), which suggests that regulatory nascent chains may be much more prevalent than previously thought.

The regulatory nascent chain with the most biochemical and structural data available is TnaC, the leader peptide of the tryptophanase (tna) operon responsible for tryptophan degradation in E. coli. The tna operon contains two structural genes, tnaA (tryptophanase) and tnaB (a tryptophan permease), downstream of the tnaC ORF. Transcription initiation of the tna operon is regulated by catabolite repression, i.e., at high concentrations of cyclic AMP (which correlates with low concentrations of glucose) the catabolite activating protein binds at the promoter region and induces transcription (Deeley and Yanofsky, 1982). A transcription pause site exists at the end of the tnaC gene, allowing time for the ribosome to bind to the transcript and initiate translation. The ribosome then decreases the life-time of this transcription pause complex, synchronizing transcription and translation (Gong and Yanofsky, 2003). At low concentrations of tryptophan, the ribosome completes translation of the TnaC peptide. Rho factor then binds at a site immediately downstream of the TnaC ORF, moves along the mRNA toward the 3' end, and interacts with the RNA polymerase complex paused at one of the pause sites located prior to the structural genes, inducing transcription termination. At high concentrations of tryptophan, however, termination of TnaC synthesis by the ribosome is inhibited, and the ribosome stalls at the TnaC stop codon, precluding binding of Rho factor, hence preventing transcription termination. Thus, high concentrations of tryptophan lead to expression of the structural genes via transcription antitermination (Stewart and Yanofsky, 1985).

Specific elements of both ribosome and TnaC are required for the formation of a tryptophan binding site at the ribosome, likely at the peptidyl transferase center (PTC) near the A site (Gong and Yanofsky, 2002). Table 1 summarizes the available experimental data on TnaC and ribosome residues regarding tna induction to the best of our knowledge. The structure of the E. coli 70S·TnaC complex was recently determined by cryo-electron microscopy (cryo-EM) single-particle reconstruction at 5.8 Å (Seidelt et al., 2009), which showed that TnaC adopts a distinct conformation in the exit tunnel, with several contact points between TnaC and the ribosome. An atomic model was obtained from the cryo-EM data by applying the molecular dynamics flexible fitting (MDFF) method (Figure 1) (Trabuco et al., 2008), which revealed that universally conserved 23S rRNA nucleotides A2602 and U2585, located at the PTC, adopt conformations incompatible with co-habitation by release factors, explaining how the PTC is inactivated in the stalled complex (Seidelt et al., 2009). Even though the register of the TnaC peptide could be identified from the cryo-EM data, side chain positions in the TnaC structure became only approximately known.

Table 1.

Summary of experimental data on TnaC and ribosome residues regarding tna induction. Δ = deletion; Ins = insertion. All data correspond to E. coli unless otherwise noted.

Molecule Residue Experimental data References
TnaC M1–W12 The following alterations have little or no effect on induction: N2F, I3S, T9I, ΔN2I3L4, ΔC7, ΔN2I3L4H5I6, Ins after T9 (A10), Ins after T9 (G10G11), Ins after L4 (H5H6H7H8H9), Ins after I6 (N7I8L9H10I11), Ins after T9 (H10I11C12V13T14). (Gong and Yanofsky, 2002)
K11 UV cross-links with A750 (23S). (Cruz-Vera et al., 2005)
W12 W12R, W12L, and W12R–F13W eliminate induction. Substitution of 33 unique codons (encoding 16 different amino acids) eliminate induction (Eshoo and Yanofsky, unpublished data). In P. vulgaris, W20L, W20R, or ΔW20 (W12 in E. coli) eliminate induction. (Gollnick and Yanofsky, 1990; Gish and Yanofsky, 1995; Konan and Yanofsky, 1997; Gong and Yanofsky, 2002; Kamath and Yanofsky, 1997; Cruz-Vera et al., 2009)
W12–P24 The following alterations reduce or eliminate induction: substitution of W12 with a stop codon (UAG or UGA), ΔK18, ΔV20D21, Ins after D21 (A22), ΔH22, ΔP24, ΔR23P24. Synonymous mutations at positions 13–17 and 19 do not appreciably affect induction levels, but changes at positions 16 and 19 increase basal activity 2–3 fold. (Gollnick and Yanofsky, 1990; Gong and Yanofsky, 2002; Gish and Yanofsky, 1995)
F13 F13W results in semiconstitutive expression of the tna operon. F13C, F13L, and W12R–F13W eliminate induction. F13I greatly reduces induction. (Gollnick and Yanofsky, 1990; Cruz-Vera et al., 2005; Gish and Yanofsky, 1995)
N14 N14I, N14S, and N14K eliminate induction. (Gish and Yanofsky, 1995)
I15 I15A has no effect on induction. I15F and I15N eliminate induction. (Cruz-Vera and Yanofsky, 2008; Gish and Yanofsky, 1995)
D16 D16A, D16N, D16E, D16W, D16S, D16L, D16C, D16V, and D16K eliminate induction. In P. vulgaris, D24A (D16 in E. coli) eliminate induction. (Cruz-Vera and Yanofsky, 2008; Gish and Yanofsky, 1995; Cruz-Vera et al., 2009)
K18 K18R and K18A have little or no effect on induction. K18Q reduces induction. (Cruz-Vera et al., 2005; Cruz-Vera and Yanofsky, 2008; Gish and Yanofsky, 1995)
I19 I19T and I19N eliminate induction. In P. vulgaris, L27A (I19 in E. coli) slightly reduces induction. (Gish and Yanofsky, 1995; Cruz-Vera et al., 2009)
P24 P24A and P24S eliminate induction. In P. vulgaris, P32A (P24 in E. coli) eliminates induction. (Gong and Yanofsky, 2002; Cruz-Vera and Yanofsky, 2008; Cruz-Vera et al., 2009)
>P24 The following insertions after P24 (replacing P24-UGA stop codon) reduce or eliminate induction: F25-UGA, M25H26T27Q28K29P30T31L32E33L34L35T36-UGA, R25P26-UGA, V25D26H27R28P29-UGA. P. vulgaris contains K33K34 after P32 (P24 in E. coli); the following changes do not reduce induction: K33R, K33I, K33R–K34R, K33I–K34I, ΔK33–34, and TAA35TGA. (Gong and Yanofsky, 2002; Cruz-Vera et al., 2009)
23S A750 UV cross-links with K11 (TnaC). (Cruz-Vera et al., 2005)
A751 Insertion of an A at position 751 eliminates induction and the ability of Trp to compete with puromycin or sparsomycin. See also A2572. (Cruz-Vera et al., 2005, 2007)
A752 A752C and A752T reduce induction. A752C eliminates ability of Trp to compete with puromycin or sparsomycin. See also A2572. (Cruz-Vera et al., 2007)
U754 U754A reduces induction. (Cruz-Vera et al., 2005)
A788 Presence of W12 (TnaC), compared to W12R, protects A788 from methylation, implying A788 displacement. (Cruz-Vera et al., 2005)
A792 A792 is slightly protected from methylation by TnaC. (Cruz-Vera et al., 2005)
A2058–A2059 The antibiotics chloramphenicol and sparsomycin enhance methylation of A2058–A2059 when ribosomes are treated with the methylating agent DMS. Trp does not affect A2059 methylation induced by chloramphenicol, indicating that it does not affect chloramphenicol interaction with the ribosome. A2059 methylation induced by sparsomycin is reduced by Trp, indicating that Trp affects the ability of sparsomycin to interact with the PTC. (Cruz-Vera et al., 2006, 2007)
A2058 A2058G slightly increases induction. (Cruz-Vera et al., 2005)
A2062, A2503 A2062C or A2503G does not prevent stalling. (Vasquez-Laslop and Mankin, personal communication)
A2572 Free Trp blocks methylation of A2572 of wild-type 70S·TnaC complexes but not of 70S·TnaC(W12R) complexes or A752C (23S), +A751 (23S), or U2609C (23S), indicating that Trp is unable to bind effectively to these mutant ribosomes. (Cruz-Vera et al., 2006, 2007)
2583–2587 The PTC mutations G2583A and U2584C reduce the ability of ribosome-bound Trp to inhibit peptidyl transferase activity, but do not affect Trp binding or sparsomycin action. Other changes in the U2585 region (U2586G, U2586C, A2587G, A2587C) do not affect induction. (Yang et al., 2009)
U2585, A2602 Cryo-EM data show that PTC residues U2585 and A2602 adopt conformations incompatible with co-habitation by release factors. (Seidelt et al., 2009)
U2609 U2609A and U2609C reduces and eliminates induction, respectively. U2609C eliminates the ability of Trp to compete with puromycin or sparsomycin. See also A2572. (Cruz-Vera et al., 2005, 2007)
L22 K90 K90W, K90H, K90A, K90A–G91K, and K90W–G91K eliminate induction. K90W eliminates the ability of Trp to compete with puromycin or sparsomycin. (Cruz-Vera et al., 2005, 2007)
G91 G91D, G91W, and G91K reduce induction, but G91A does not significantly affect induction. (Cruz-Vera et al., 2005)
A93 A93S does not significantly affect induction. (Cruz-Vera et al., 2005)
V107 V107M slightly increases induction. (Cruz-Vera et al., 2005)
L4 K57–K58 K57E–K58Q, K57Q–K58Q, K57E–K58E, and K57Q–K58E slightly reduce induction. (Cruz-Vera et al., 2005)
Open in a new tab

Figure 1.

Figure 1

Open in a new tab

Structure of TnaC in the exit tunnel. Shown are selected residues of the 23S rRNA, the portions of ribosomal proteins L4 and L22 forming the exit tunnel's constriction site, and ten models of the nascent chain found to be consistent with cryo-EM data (Seidelt et al., 2009). 23S rRNA residues displayed in black represent bacterial sequence signatures (Roberts et al., 2008). Movie S1 shows a 360-degree view of this figure. See also Figure S1.

In order to characterize the precise atomic interactions through which the ribosome rec ognizes TnaC, we performed extensive molecular dynamics (MD) simulations of TnaC in the exit tunnel. The simulations, supported by quantum chemistry calculations, indicate that the critical TnaC residue W12 is recognized by the ribosome via a cation-π interaction with R92 of ribosomal protein L22. Another critical TnaC residue, D16, is seen in the simulations to be involved in salt bridges with ribosomal proteins, in particular with K90 of L22, another essential element for tna induction. The simulations also confirm that TnaC-mediated stalling does not involve swinging of the tip of L22, as previously proposed (Berisio et al., 2003). Furthermore, bioinformatic analysis of naturally occurring proteins that contain the TnaC consensus suggests that mutations at positions 19 and 20 of TnaC may alleviate translational stalling, for which additional simulations provide further evidence. Our results thus shed light onto the precise atomic-level interactions involved in the recognition of TnaC by the ribosome.

Results

Evolutionary analysis of TnaC and the exit tunnel

Recently, tnaC genes in various bacterial species were identified computationally (Cruz-Vera and Yanofsky, 2008). Starting with the 31 sequences reported by the authors, we selected a non-redundant set using the sequence QR algorithm (Sethi et al., 2005). The obtained non-redundant set contains 13 sequences and is presented in Figure 2A. According to the evolutionary profile, the critical residues W12, D16, and P24 are indeed the most conserved ones. Furthermore, certain mutations at relatively conserved positions 13–15 and 19 reduce or eliminate induction (Table 1). Given the sequence conservation in the alignment and the available biochemical data, it is unlikely that other very specific requirements for TnaC-mediated stalling exist. Interestingly, all organisms other than E. coli presented in Figure 2A contain one or more residues after P24, which poses the question whether stalling would occur at the stop codon or at the conserved proline codon in these organisms. It was recently shown that, in Proteus vulgaris, TnaC stalling occurs at the conserved proline codon, implying in this case inhibition of protein elongation instead of termination (Cruz-Vera et al., 2009).

Figure 2.

Figure 2

Open in a new tab

Non-redundant sequence alignments of TnaC (A) and the regions of ribosomal proteins L4 (B) and L22 (C) comprising the exit tunnel's constriction site. The sequence alignments and corresponding structures are colored by sequence similarity using the BLOSUM 30 matrix and a red-white-blue color scale, with most conserved residues shown in blue. For TnaC (A), complete sequences are shown, i.e., there exists a stop codon at the end of each presented sequence. See also Table S1.

Sequence alignments were also obtained for ribosomal proteins L4 and L22 (Figure 2B and C, respectively), which form the constriction site in the exit tunnel. The regions of L4 and L22 forming the constriction site are visibly very conserved. Residue K90 in the β-hairpin at the tip of L22 is critical for TnaC stalling in E. coli as judged by mutational studies (K90W, K90H, and K90A; see Table 1). Interestingly, K90 is not strictly conserved in bacteria, and Figure 2C shows that other residues occur at this position, which may point to either absence of TnaC stalling in some bacteria or a different recognition mode. It would be worthwhile to investigate the effect of further naturally occurring substitutions of K90 on tna induction.

Non-redundant sequence alignments of 16S and 23S rRNAs were recently obtained for both bacteria and archaea, and sequence and structure signatures distinguishing these two domains of life were identified (Roberts et al., 2008). None of the structure signatures lie in the vicinity of TnaC, but four sequence signatures (23S rRNA residues) do (Figure 1). Using the sequence alignments obtained by the authors, we calculated the base composition of 23S rRNA in proteobacteria (phylum to which E. coli belongs), the remaining bacterial phyla, and archaea (Table S1). Only two nucleotides in the vicinity of TnaC show significantly different base compositions in proteobacteria versus the other bacterial phyla, namely U754 and A792.

Consensus stalling sequence in bacterial genomes

Considering only TnaC elements critical for stalling and highly conserved (W12, D16, P24, and the spacing between W12 and P24), current experimental data suggest that the consensus sequence WxxxDxxxxxxxP is necessary for TnaC-mediated stalling (see Table 1). Furthermore, substitution of TnaC's stop codon with a tryptophan codon leads to constitutive, tryptophan-independent translational arrest (Gong and Yanofsky, 2002). We thus searched all bacterial protein sequences corresponding to available genomes for the consensus sequence WxxxDxxxxxxxPW. After removing redundancy of the underlying sequences, we found a slight evolutionary selection against the consensus in proteobacteria, but not in other bacterial phyla. Specifically, 114 sequences from proteobacteria matched the WxxxDxxxxxxxPW consensus, compared to an expected number of 134 ± 12 by chance (standard deviation given). On the other hand, 125 bacterial sequences from phyla other than proteobacteria matched the same consensus, compared to an expected number of 79 ± 9.

If the searched consensus sequences were sufficient to induce arrest, they should not be found in any protein. However, several proteins do match the consensus, as described above. Three possible, non-mutually exclusive explanations are: (1) there are other TnaC elements required for arrest, (2) proteins that match the consensus have features that prevent arrest, and/or (3) there are differences in ribosomes that make them less susceptible to arrest by the consensus sequence. Intriguingly, there exist proteins from E. coli and close relatives that match the consensus WxxxDxxxxxxxPW. Thus, the fact that these proteins presumably do not induce constitutive stalling cannot be explained by differences in the corresponding ribosomes.

For each protein sequence that matches the consensus WxxxDxxxxxxxPW, the relative frequency of each amino acid was calculated for each position and compared to frequencies obtained from the entire data set. A bias toward presence of an arginine residue at positions 19 and 20 was identified in the sequences matching the consensus, with frequencies equal to 16.3 ± 4.7% and 13.8 ± 4.4%, respectively, compared to 6.0 ± 0.002% for the entire data set (95% confidence intervals given). Similar results were obtained repeating the analysis for proteobacteria and non-proteobacteria separately (data not shown). Note that positions 19 and 20 are relatively conserved according to the TnaC evolutionary profile (Figure 2A). It is possible that presence of these arginine residues helps prevent translational arrest of the identified proteins. In fact, MD simulations show that the TnaC mutations I19R and V20R can displace a 23S rRNA residue critical for stalling (see below). Alterations of TnaC residues 19 and 20 should be investigated experimentally, particularly the mutations I19R and V20R, which may reduce tna induction. Indeed, mutations I19T and I19N have already been shown to eliminate induction (Gish and Yanofsky, 1995).

Recognition of TnaC by the exit tunnel

A 5.8-Å cryo-EM map of the E. coli 70S·TnaC complex was recently obtained (Seidelt et al., 2009), from which an atomic model of the ribosome and ten different TnaC models were generated using the MDFF method (Trabuco et al., 2008; Seidelt et al., 2009) (see Figure 1). MDFF allows for automated generation of atomic models using cryo-EM data which are ultimately checked for consistency by direct comparison with the density maps (Villa et al., 2009; Hsin et al., 2009; Gumbart et al., 2009; Seidelt et al., 2009; Becker et al., 2009). In order to characterize the specific interactions between TnaC and the ribosome, extensive all-atom MD simulations were performed on ten different systems containing TnaC and the exit tunnel. The simulation setup is illustrated in Supplementary Figure 1. Each of the ten systems comprises about 125,000 atoms and the total aggregate simulation time was 1.8 μs. The goal was to sample as much as possible the conformational space of TnaC side chains inside the tunnel, allowing for favorable TnaC-ribosome interactions to be identified.

The TnaC models derived from the cryo-EM reconstruction (Seidelt et al., 2009), as well as the simulations presented here, show that the universally conserved TnaC residue W12, critical for tna induction, is located in the exit tunnel at the constriction site formed by ribosomal proteins L4 and L22 (see Figure 1). In several simulations, W12 exhibited potential interactions with positively charged residues (arginine or lysine) from L4 and L22. Aromatic residues such as tryptophan can form cation-π interactions with positively charged residues, a strong and specific interaction important in different cases of molecular recognition (Gallivan and Dougherty, 1999). Unfortunately, MD force fields do not accurately describe cation-π interactions (Caldwell and Kollman, 1995; Donini and Weaver, 1998; Minoux and Chipot, 1999), in particular due to the lack of polarizability (Cubero et al., 1998). However, it was shown that the electrostatic component of the MD interaction energy between aromatic and positively charged residues correlates well with binding energies computed via quantum-mechanical (QM) calculations (Gallivan and Dougherty, 1999). Thus, we calculated the MD electrostatic interaction energy between the side chains of W12 and of each nearby positively charged residue for each frame of the ten independent simulations, which allowed us to select conformations corresponding to potential cation-π interactions involving W12. The potential cation-π interaction partner with the highest frequency from the MD trajectories was R92 from L22 (Figure S2A).

To confirm that the identified conformations indeed correspond to cation-π interactions, we computed the interaction energy between the relevant side chains using QM calculations (Figure 3A and Figure S 2B). Ammonium and guanidinium ions were used to represent lysine and arginine, respectively, whereas tryptophan was represented as indole (Gallivan and Dougherty, 1999). Using minimum energy structures for these fragments and their relative orientation from the MD trajectory frames, their binding energy was calculated using second-order Møller-Plesset perturbation theory (see Methods). The QM calculations confirm that several of the selected conformations indeed correspond to cation-π interactions between W12 (TnaC) and mostly R92 (L22), with interaction energies of up to around - 13.5 kcal/mol (Table S 2, Figure 3A and Figure S 2B). If the classical force field used in the MD simulations more accurately described cation-π interactions, the frequency of such interactions would be presumably higher than observed in our simulations. The involvement of a cation-π interaction between W12 and the exit tunnel during TnaC stalling may be confirmed experimentally by incorporation of a series of fluorinated tryptophan derivatives, as recently done in the binding site of G protein-coupled receptors (Torrice et al., 2009).

Figure 3.

Figure 3

Open in a new tab

Recognition of key TnaC residues by the exit tunnel. (A) Interaction energies between W12 (TnaC) and R92 (L22) calculated using the MD force field from the ten independent simulations starting from the different TnaC models shown in Figure 1 (labeled “model 0–9”). Several trajectory frames were selected as indicative for cation-π interactions (circles; see Figure S2A), for which the interaction energies were also determined using QM calculations (diamonds; full results given on Table S2). The same procedure was followed for all positively charged residues in the vicinity of W12 (Figure S2B). (B) Snapshot from one of the simulation trajectories showing a cation-π interaction between W12 (TnaC) and R92 (L22) as well as a salt bridge between D16 (TnaC) and K90 (L22) (see Figure 4). See also Figure S2 and Table S2.

The universally conserved TnaC residue D16, required for tna induction, is involved in salt bridges with ribosomal proteins in about half of the simulations (Figure 4). Observed interaction partners of D16 comprise K90 (L22), R92 (L22), R61 (L4), and K18 (TnaC). The other negatively charged TnaC residue D21 is not involved in salt bridges with ribosomal proteins but, like D16, salt bridges with positively charged residues from TnaC itself (Figure S3). Given our finding that W12 (TnaC) is likely recognized by R92 (L22) via a cation-π interaction, and also that K90 (L22) is critical for induction as shown by several mutations (Cruz-Vera et al., 2005), it is conceivable that among the observed salt bridges the most relevant for induction is the one between D16 (TnaC) and K90 (L22). The importance of this salt bridge may be tested experimentally by constructing the double mutant D16K (TnaC) and K90D (L22), which we predict will have little effect on tna induction, even though the D16K (TnaC) mutation eliminates induction (see Table 1). Figure 3B shows a snapshot from a trajectory where a cation-π interaction between W12 (TnaC) and R92 (L22) as well as a salt bridge between D16 (TnaC) and K90 (L22) can be observed.

Figure 4.

Figure 4

Open in a new tab

Salt bridges formed between D16 (TnaC) and the exit tunnel in all simulations. The plots show the distance between the center of mass of the oxygen atoms in the acidic side chain and the center of mass of the nitrogen atoms in the basic side chain. For each interaction pair, only simulations in which a stable salt bridge was observed are shown. The ten independent simulations starting from different TnaC models shown in Figure 1 are labeled “model 0–9.” See also Figure S3.

Since our bioinformatic analysis suggests an important role of TnaC residues I19 and V20 (see above), we characterized the contacts between these residues and the ribosome. The frequencies at which I19 and V20 contacts ribosomal residues was calculated from the last 40 ns of each of the ten simulations (Figure 5). The 23S rRNA residues contacting I19 most often were 2058–2059, 2061–2062, 2503, and 2609–2611, with I19 usually inserting between two RNA bases. Since mutations of U2609 reduced or eliminated tna induction (see Table 1), the frequencies of contacts between U2609 and TnaC residues were also characterized. V20 accounted for the majority of U2609 contacts (data not shown); likewise, the most frequent contact partner of V20 was U2609 (Figure 5). Interestingly, our bioinformatic analysis suggests that the mutations I19R or V20R may prevent translational stalling. Thus, it is conceivable that mutations of I19 or V20 could affect the critical residue U2609 and reduce stalling. Each of the ten simulations were extended for 15 ns introducing either the TnaC mutation I19R or V20R, for a total additional 300 ns of simulation time. The mutant simulations show that either change is able to displace U2609 by up to 2–3 Å in the relatively short simulation time of 15 ns (Table S3). Note that the boundaries of the exit tunnel (10 Å away from TnaC) were restrained in space (see Methods), which limited actually the magnitude of possible displacements of U2609.

Figure 5.

Figure 5

Open in a new tab

Frequency of contacts between TnaC residues I19 and V20 and ribosomal residues. Contacts were defined based on a cut-off distance of 3.5 Å between heavy atoms of side chains, and the last 40 ns of each of the 10 independent simulations were considered in the analysis. See also Table S3 and Figure S4.

Recently, three-dimensional free-energy profiles for different amino-acid side chains inside the exit tunnel have been calculated using MD-based umbrella sampling simulations (Petrone et al., 2008). Even though the free-energy profiles were obtained by these authors using a Haloarcula marismortui 50S model, the structure of the region of the exit tunnel discussed in the current work is very similar to the E. coli one (Figure S4). The authors have described a significant energy barrier for tryptophan binding at the constriction site (ΔG > 6kBT, see Figure 2F in (Petrone et al., 2008)), which is inconsistent with the W12 position derived from cryo-EM data (Seidelt et al., 2009). The discrepancy is likely due to the improper treatment of cation-π interactions by the MD force field used to calculate the free-energy profile. On the other hand, the salt bridges between D16 and either side of the exit tunnel (L4 or L22) observed in our simulations are consistent with local free energy minima (ΔG < −5kBT, see Figure 2C in (Petrone et al., 2008)) in the free-energy profile for aspartate positioning inside the tunnel. Interestingly, the position of TnaC residue I19 is also consistent with a local free energy minimum (ΔG < −3kBT, see Figure 2E in (Petrone et al., 2008)) for isoleucine positioning inside the tunnel. We proposed above that the mutations I19R or V20R may reduce tna induction. Although a free-energy map for arginine is not available, presence of a positive charge (lysine) at positions occupied by I19 and V20 in our simulations is unfavorable (see Figure 2B in (Petrone et al., 2008)). Note that the effect may be less pronounced for arginine due to potential favorable stacking interactions with 23S rRNA bases.

TnaC stalling does not involve a swung conformation of L22

It has been proposed that the conserved β-hairpin at the tip of L22 (Figure 2C), located at the constriction site of the exit tunnel (Figure 1), may act as a tunnel gate controlling nascent chain elongation (Berisio et al., 2003). The suggestion was based on the crystal structure of the Deinococcus radiodurans 50S (D50S) subunit in complex with the antibiotic troleandomycin (TAO), in which the L22 β-hairpin assumes a swung conformation and interacts with L4, blocking the tunnel (Berisio et al., 2003). The authors hypothesized that SecM- and TnaC-mediated translational arrest could involve a swung conformation of the L22 β-hairpin (Berisio et al., 2003). However, the cryo-EM reconstruction of the E. coli 70S·TnaC complex (Seidelt et al., 2009) and the simulations presented here do not support this hypothesis for TnaC-mediated stalling. Figure 6A shows the structure obtained from the cryo-EM data (Seidelt et al., 2009), as well as a model of the swung conformation of L22 for comparison, obtained by aligning L22 from D50S·TAO (PDB 1OND) (Berisio et al., 2003) to L22 from 70S·TnaC. From the figure it is clear that a swung L22 would severely clash with the TnaC density. Moreover, L22 remains in the native conformation during all simulations (data not shown). In the case of SecM-mediated translational arrest, involvement of a swung conformation of L22 is also unlikely, since it was recently shown that mutants in which the tip of L22 was deleted still show significant SecM stalling (Lawrence et al., 2008).

Figure 6.

Figure 6

Open in a new tab

Path of the TnaC peptide inside the exit tunnel. (A) Atomic model of ribosomal proteins L4 and L22 obtained from the 70S·TnaC cryo-EM map (Seidelt et al., 2009), along with the density corresponding to TnaC. For comparison, the swung conformation of the tip of L22 is shown in red (PDB 1OND) (Berisio et al., 2003). (B) Conformations explored by TnaC during the MD simulations according to clustering analysis of the Cα atoms (see Methods). Two possible paths for a poly-alanine nascent chain had been previously proposed, namely “case-over” and “case-under” R92 from L22 (Ishida and Hayward, 2008).

The L22 swinging mechanism of translational arrest had been advocated based on previous computational studies. Rigidity analysis showed that the L22 β-hairpin is flexible outside the ribosome but becomes rigid within the ribosome, whereas those residues that form a hinge, around which the β-hairpin would swing, retain their flexible form (Fulle and Gohlke, 2009). However, MD simulations of a complete ribosome showed that the tip of L22 does not exhibit significant flexibility (Figure S5). Another computational study predicted paths of the nascent chain inside the ribosome, which were classified in two categories: “case-over” and “case-under” L22's R92 (Ishida and Hayward, 2008). Via pulling simulations, the authors hypothesized that the case-under category corresponds to protein elongation, whereas the case-over corresponds to translational arrest, based in part on the proposed L22 swinging mechanism of arrest (Berisio et al., 2003). However, the path of TnaC inside the exit tunnel clearly corresponds to the case-under category (Figure 6B). Altogether, our data show that TnaC-mediated translational stalling does not involve swinging of the tip of L22.

Discussion

It has been increasingly realized that modulation of ribosome function by specific nascent chains is a prevalent regulatory mechanism. Most known examples involve a regulatory nascent chain encoded by an upstream ORF, whose translation controls expression of downstream genes. For instance, nascent chain-mediated translational stalling controls expression of several genes conferring antibiotic resistance in bacteria (Ramu et al., 2009). The recent discovery that a large proportion of mRNAs in yeast contain small upstream ORFs which are translated in vivo point to a central role of regulatory nascent chains in the control of gene expression also in eukaryotes (Ingolia et al., 2009).

In the present work, we employed a range of computational techniques to address how TnaC, the leader peptide of the tryptophanase (tna) operon, is recognized by the ribosome. Using recent atomic models of the E. coli 70S·TnaC complex obtained from a 5.8-Å cryo-EM map (Seidelt et al., 2009), we performed extensive MD simulations of TnaC inside the exit tunnel for a total aggregate time of 2.1 μs (1.8 μs and 0.3 μs of wild-type and mutant simulations, respectively). The MD simulations, complemented by QM calculations, indicate that the critical TnaC residue W12 is recognized by R92 of ribosomal protein L22 via a cation-π interaction (Figure 3 and Table S2). The simulations further show that TnaC residue D16, also critical for stalling, forms salt bridges with ribosomal proteins (Figure 4), in particular with K90 from L22, which, too, is essential for stalling (see Table 1). All findings presented here are experimentally testable. Incorporation of a series of fluorinated tryptophan derivatives may be used to confirm that W12 is recognized by the ribosome via a cation-π interaction, as recently done in the binding site of G protein-coupled receptors (Torrice et al., 2009). In order to test the relevance of the salt bridge between D16 (TnaC) and K90 (L22), the double mutant D16K (TnaC) and K90D (L22) may be constructed and its ability to induce tna expression analyzed.

Bioinformatic analysis suggests that presence of an arginine at TnaC positions 19 or 20 reduces translational arrest. MD simulations of TnaC inside the exit tunnel show that TnaC residues I19 and V20 contact 23S rRNA residues critical for stalling, in particular U2609. Simulations of TnaC mutants I19R and V20R reveal that these changes are able to displace U2609, supporting the hypothesis that these mutations reduce stalling. The effect of these mutations on tna induction remains to be tested experimentally. Finally, the current study provides evidence that TnaC-mediated stalling does not involve swinging of the tip of L22, as previously hypothesized (Berisio et al., 2003). SecM-mediated arrest likely also does not involve L22 swinging, since deletions of the tip of L22 still allow for significant stalling (Lawrence et al., 2008).

Three main open questions regarding TnaC-mediated translational stalling remain. The first question is how does the recognition of TnaC by the exit tunnel (described in this paper) lead to the creation of a tryptophan binding site in the ribosome. Possible communication pathways between critical elements of TnaC and the PTC have been proposed based on cryo-EM data (Seidelt et al., 2009), but further investigations will be necessary to unveil the signaling mechanism. The second open question is the precise location of the tryptophan binding site at the PTC. A previous study showed that substituting the TnaC's stop codon with a tryptophan codon leads to constitutive stalling, indicating that a tRNA bearing a tryptophan residue has the same effect as free tryptophan and induces stalling (Gong and Yanofsky, 2002). Thus, the binding site for free tryptophan is likely formed near the A site, but a high-resolution structure will be required to determine the exact location. Finally, it was recently found that TnaC-mediated stalling in P. vulgaris involves inhibition of elongation instead of termination (Cruz-Vera et al., 2009), but the structural basis for elongation inhibition is not known.

Methods

Bioinformatic analyses

All bioinformatic analyses were performed with the VMD (Humphrey et al., 1996) plugin MultiSeq 3.0 (Roberts et al., 2006). Representative sequences of TnaC, L4, and L22 were selected using the sequence QR algorithm (Sethi et al., 2005), which sorts a sequence alignment by increasing linear dependency. For TnaC, 31 bacterial sequences previously identified computationally (Cruz-Vera and Yanofsky, 2008) were utilized, and 13 sequences were selected using a 50% sequence identity cut-off. For each protein L4 and L22, over 600 bacterial sequences were retrieved from the NCBI nr database using a BLAST search (Altschul et al., 1990), and the sequences aligned with ClustalW (Thompson et al., 1994). Representative sets containing 19 and 14 L4 and L22 sequences were obtained using 43% and 53% sequence identity cut-offs, respectively.

Protein sequences corresponding to bacterial genomes were downloaded from NCBI. To reduce redundancy, only one genome from each species was considered. In total, 513 genomes were analyzed, corresponding to 1,696,002 protein sequences and 545,566,949 amino acids. The relative frequency of each amino acid residue was calculated in different data sets, and a binomial distribution was assumed in each case to calculate 95% confidence intervals. In order to calculate the expected number of matches to a consensus sequence of size n at random, the probability pn of a sequence of size n matching the consensus was calculated using the relative frequencies of each amino acid. Introducing the approximation that overlapping sequences of size n are independent, the expected number of matches in the entire data set is given by pn Σi(Ni − n + 1), where the sum is performed over all sequences i in the data set, Ni is the length of each sequence i, and Ni > n. Standard deviations were calculated assuming a binomial distribution with probability pn and sample size Σi(Ni − n + 1).

MD simulations of TnaC-exit tunnel

An atomic model of the 70S ribosome complexed with TnaC, as well as 10 different models for TnaC itself, were obtained from a 5.8-Å cryo-EM map using MDFF (Trabuco et al., 2008, 2009), as previously described (PDB 2WWL/2WWQ, (Seidelt et al., 2009)). For each of the ten 70S·TnaC models, a subsystem was defined by selecting all residues within 20 Å of TnaC. Each system was then neutralized by adding difusively-bound Mg2+ ions (minimum ion-solute distance of 5 Å) using cionize, a GPU-accelerated program that iteratively places ions at electrostatic potential minima (Stone et al., 2007), and the solvation shell of each Mg2+ ion was completed (Eargle et al., 2008; Trabuco et al., 2009). Subsequently, DOWSER (Zhang and Hermans, 1996) was used to place water molecules into internal cavities; the VMD plugin Dowser (Gumbart et al., 2009), which extends DOWSER to support systems containing RNA, was used in this step. Each system was then placed in a water box using the VMD plugin solvate, after which neutralization was completed by placing an additional Na+ ion with the VMD plugin autoionize. The total size of each of the ten systems was about 125,000 atoms.

MD simulations were conducted using NAMD 2.7 (Phillips et al., 2005) with the CHARMM27 force field (MacKerell Jr. et al., 1998; Foloppe and MacKerell Jr., 2000) and CMAP correction (MacKerell Jr. et al., 2004) and the TIP3P water model. The equations of motion were integrated using a 2-fs time step, with all interactions calculated every 2 fs, except for long-range electrostatic interactions, which were calculated every 6 fs using the particle mesh Ewald (PME) method with pencil decomposition. The PME grid density was never less than 1/Å3, and periodic boundary conditions were applied. Short-range, nonbonded interactions were calculated using a distance cut-off of 8 Å. Each system was equilibrated in the NpT ensemble (T =300 K, p=1 atm) using the following protocol: water and ions were equilibrated for 420 ps, keeping the rest of the atoms restrained; side chains were then free to move and equilibrated for 3 ns; all restraints on TnaC were subsequently removed, but ribosome backbone restraints were kept, and each system was equilibrated for 60 ns; finally, only ribosome backbone atoms farther than 10 Å from TnaC were restrained, and each system was equilibrated for an additional 120 ns.

QM calculations of cation-π interactions

For each frame from the MD simulation trajectories, the electrostatic interaction energy between W12 (TnaC) and all nearby positively charged amino acid residues from ribosomal proteins was calculated using the VMD plugin NAMDEnergy with a 10 Å cut-off (see Figure S2A). Only fragments representing the side chains were considered, namely indole (tryptophan), ammonium (lysine), and guanidinium (arginine) (Gallivan and Dougherty, 1999). For each interaction pair, several frames corresponding to local energy minima were manually chosen (Figure S 2A), from which the interaction energies were estimated using QM calculations.

QM energies for the selected MD geometries were calculated with the Gaussian03 revision E.01 package (Frisch et al., 2003) using second-order Møller-Plesset perturbation theory and a triple-ζ Pople basis set including double diffuse functions and polarization functions added to d and p functions (6–311++G(d,p)). Similar interaction energy values were obtained with the GAMESS package (Schmidt et al., 1993). Estimates of the basis-set superposition error (BSSE) were determined using the Counterpoise correction method (Boys and Bernardi, 1970). Interaction energy values from the MD potential were also calculated using NAMDEnergy for comparison purposes.

Clustering analysis

Trajectory frames from the last 40 ns of each MD simulation were concatenated and the conformation of TnaC was analyzed using the GROMOS clustering method (Daura et al., 1999; van der Spoel et al., 2005). RMSDs were calculated using all heavy atoms of the protein backbone, with cut-offs chosen from the pairwise frame-frame RMSD distributions of each selected set of atoms between clustering calculations.

Supplementary Material

01
02

Acknowledgments

The authors thank Roland Beckmann and colleagues for fruitful collaborations, Elijah Roberts, John Eargle, and Zan Luthey-Schulten for rRNA sequence alignments and help with MultiSeq and NAMDEnergy scripting, and Christophe Chipot for useful discussions. This work was supported by the National Institutes of Health (P41-RR005969) and the National Science Foundation (PHY0822613). Computer time was provided through the National Resources Allocation Committee (MCA93S028).

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local alignment search tool. J. Mol. Biol. 1990;215:403–410. doi: 10.1016/S0022-2836(05)80360-2. [DOI] [PubMed] [Google Scholar]
  2. Becker T, Mandon E, Bhushan S, Jarasch A, Armache JP, Funes S, Jossinet F, Gumbart J, Mielke T, Berninghausen O, et al. Structure of monomeric yeast and mammalian Sec61 complexes interacting with the translating ribosome. Science. 2009;326:1369–1373. doi: 10.1126/science.1178535. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Berisio R, Schluenzen F, Harms J, Bashan A, Auerbach T, Baram D, Yonath A. Structural insight into the role of the ribosomal tunnel in cellular regulation. Nat. Struct. Biol. 2003;10:366–370. doi: 10.1038/nsb915. [DOI] [PubMed] [Google Scholar]
  4. Bornemann T, Jöckel J, Rodnina MV, Wintermeyer W. Signal sequence-independent membrane targeting of ribosomes containing short nascent peptides within the exit tunnel. Nat. Struct. Mol. Biol. 2008;15:494–499. doi: 10.1038/nsmb.1402. [DOI] [PubMed] [Google Scholar]
  5. Boys SF, Bernardi F. The calculation of small molecular interactions by the differences of separate total energies. Some procedures with reduced errors. Mol. Phys. 1970;19:553–566. [Google Scholar]
  6. Caldwell JW, Kollman PA. Cation-π interactions: nonadditive effects are critical in their accurate representation. J. Am. Chem. Soc. 1995;117:4177–4178. [Google Scholar]
  7. Cruz-Vera LR, Gong M, Yanofsky C. Changes produced by bound tryptophan in the ribosome peptidyl transferase center in response to TnaC, a nascent leader peptide. Proc. Natl. Acad. Sci. USA. 2006;103:3598–3603. doi: 10.1073/pnas.0600082103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cruz-Vera LR, New A, Squires C, Yanofsky C. Ribosomal features essential for tna operon induction: Tryptophan binding at the peptidyl transferase center. J. Bacteriol. 2007;189:3140–3146. doi: 10.1128/JB.01869-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Cruz-Vera LR, Rajagopal S, Squires C, Yanofsky C. Features of the ribosome-peptidyl-tRNA interactions essential for tryptophan induction of tna operon expression. Mol. Cell. 2005;19:333–343. doi: 10.1016/j.molcel.2005.06.013. [DOI] [PubMed] [Google Scholar]
  10. Cruz-Vera LR, Yang R, Yanofsky C. Tryptophan inhibits Proteus vulgaris TnaC leader peptide elongation, activation tna operon expression. J. Bacteriol. 2009;191:7001–7006. doi: 10.1128/JB.01002-09. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Cruz-Vera LR, Yanofsky C. Conserved residues Asp16 and Pro24 of TnaC-tRNAPro participate in tryptophan induction of Tna operon expression. J. Bacteriol. 2008;190:4791–4797. doi: 10.1128/JB.00290-08. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Cubero E, Luque FJ, Orozco M. Is polarization important in cation-π interactions? Proc. Natl. Acad. Sci. USA. 1998;95:5976–5980. doi: 10.1073/pnas.95.11.5976. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Daura X, Gademann K, Jaun B, Seebach D, van Gunsteren WF, Mark AE. Peptide folding: When simulation meets experiment. Angewandte Chemie International Edition. 1999;38:236–240. [Google Scholar]
  14. Deeley MC, Yanofsky C. Transcription initiation at the tryptophanase promoter of Escherichia coli K-12. J. Bacteriol. 1982;151:942–951. doi: 10.1128/jb.151.2.942-951.1982. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Donini O, Weaver DF. Development of modified force field for cation-amino acid interactions: ab initio-derived empirical correction terms with comments on cation-π interactions. J. Comp. Chem. 1998;19:1515–1525. [Google Scholar]
  16. Eargle J, Black AA, Sethi A, Trabuco LG, Luthey-Schulten Z. Dynamics of recognition between tRNA and elongation factor Tu. J. Mol. Biol. 2008;377:1382–1405. doi: 10.1016/j.jmb.2008.01.073. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Foloppe N, MacKerell AD., Jr. All-atom empirical force field for nucleic acids: I. Parameter optimization based on small molecule and condensed phase macromolecular target data. J. Comp. Chem. 2000;21:86–104. [Google Scholar]
  18. Frisch MJ, et al. Gaussian 03 (Revision B.05) Gaussian, Inc.; Pittsburgh, PA: 2003. [Google Scholar]
  19. Fulle S, Gohlke H. Statics of the ribosomal exit tunnel: Implications for cotranslational peptide folding, elongation regulation, and antibiotics binding. J. Mol. Biol. 2009;387:502–517. doi: 10.1016/j.jmb.2009.01.037. [DOI] [PubMed] [Google Scholar]
  20. Gallivan JP, Dougherty DA. Cation-π interactions in structural biology. Proc. Natl. Acad. Sci. USA. 1999;96:9459–9464. doi: 10.1073/pnas.96.17.9459. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Gish K, Yanofsky C. Evidence suggesting cis action by the TnaC leader peptide in regulating transcription attenuation in the tryptophanase operon of Escherichia coli. J. Bacteriol. 1995;177:7245–7254. doi: 10.1128/jb.177.24.7245-7254.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Gollnick P, Yanofsky C. tRNATrp translation of leader peptide codon 12 and other factors that regulate expression of the tryptophanase operon. J. Bacteriol. 1990;172:3100–3107. doi: 10.1128/jb.172.6.3100-3107.1990. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Gong F, Yanofsky C. Instruction of translating ribosome by nascent peptide. Science. 2002;297:1864–1867. doi: 10.1126/science.1073997. [DOI] [PubMed] [Google Scholar]
  24. Gong F, Yanofsky C. A transcriptional pause synchronizes translation with transcription in the tryptophanase operon leader region. J. Bacteriol. 2003;185:6472–6476. doi: 10.1128/JB.185.21.6472-6476.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Gumbart J, Trabuco LG, Schreiner E, Villa E, Schulten K. Regulation of the protein-conducting channel by a bound ribosome. Structure. 2009;17:1453–1464. doi: 10.1016/j.str.2009.09.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Hsin J, Gumbart J, Trabuco LG, Villa E, Qian P, Hunter CN, Schulten K. Protein-induced membrane curvature investigated through molecular dynamics flexible fitting. Biophys. J. 2009;97:321–329. doi: 10.1016/j.bpj.2009.04.031. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Humphrey W, Dalke A, Schulten K. VMD – Visual Molecular Dynamics. J. Mol. Graphics. 1996;14:33–38. doi: 10.1016/0263-7855(96)00018-5. [DOI] [PubMed] [Google Scholar]
  28. Ingolia NT, Ghaemmaghami S, Newman JRS, Weissman JS. Genome-wide analysis in vivo of translation with nucleotide resolution using ribosome profiling. Science. 2009;324:218–223. doi: 10.1126/science.1168978. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Ishida H, Hayward S. Path of nascent polypeptide in exit tunnel revealed by molecular dynamics simulation of ribosome. Biophys. J. 2008;95:5962–5973. doi: 10.1529/biophysj.108.134890. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Kamath AV, Yanofsky C. Roles of the tnaC-tnaA spacer region and Rho factor in regulating expression of the tryptophanase operon of Proteus vulgaris. J. Bacteriol. 1997;179:1780–1786. doi: 10.1128/jb.179.5.1780-1786.1997. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Konan KV, Yanofsky C. Regulation of the Escherichia colitna operon: nascent leader peptide control at the tnaC stop codon. J. Bacteriol. 1997;179:1774–1779. doi: 10.1128/jb.179.5.1774-1779.1997. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Lawrence MG, Lindahl L, Zengel JM. Effects on translation pausing of alterations in protein and RNA components of the ribosome exit tunnel. J. Bacteriol. 2008;190:5862–5869. doi: 10.1128/JB.00632-08. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Lu J, Deutsch C. Electrostatics in the ribosomal tunnel modulate chain elongation rates. J. Mol. Biol. 2008;384:73–86. doi: 10.1016/j.jmb.2008.08.089. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. MacKerell AD, Jr., Bashford D, Bellott M, Dunbrack RL, Jr., Evanseck J, Field MJ, Fischer S, Gao J, Guo H, Ha S, et al. All-atom empirical potential for molecular modeling and dynamics studies of proteins. J. Phys. Chem. B. 1998;102:3586–3616. doi: 10.1021/jp973084f. [DOI] [PubMed] [Google Scholar]
  35. MacKerell AD, Jr., Feig M, Brooks CL., III Extending the treatment of backbone energetics in protein force fields: Limitations of gas-phase quantum mechanics in reproducing protein conformational distributions in molecular dynamics simulations. J. Comp. Chem. 2004;25:1400–1415. doi: 10.1002/jcc.20065. [DOI] [PubMed] [Google Scholar]
  36. Minoux H, Chipot C. Cation-π interactions in proteins: Can simple models provide an accurate description? J. Am. Chem. Soc. 1999;121:10366–10372. [Google Scholar]
  37. Nakatogawa H, Ito K. The ribosomal exit tunnel functions as a discriminating gate. Cell. 2002;108:629–636. doi: 10.1016/s0092-8674(02)00649-9. [DOI] [PubMed] [Google Scholar]
  38. Petrone PM, Snow CD, Lucent D, Pande VS. Side-chain recognition and gating in the ribosome exit tunnel. Proc. Natl. Acad. Sci. USA. 2008;105:16549–16554. doi: 10.1073/pnas.0801795105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Phillips JC, Braun R, Wang W, Gumbart J, Tajkhorshid E, Villa E, Chipot C, Skeel RD, Kale L, Schulten K. Scalable molecular dynamics with NAMD. J. Comp. Chem. 2005;26:1781–1802. doi: 10.1002/jcc.20289. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Ramu H, Mankin A, Vazquez-Laslop N. Programmed drug-dependent ribosome stalling. Mol. Microbiol. 2009;71:811–824. doi: 10.1111/j.1365-2958.2008.06576.x. [DOI] [PubMed] [Google Scholar]
  41. Roberts E, Eargle J, Wright D, Luthey-Schulten Z. MultiSeq: Unifying sequence and structure data for evolutionary analysis. BMC Bioinformatics. 2006;7:382. doi: 10.1186/1471-2105-7-382. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Roberts E, Sethi A, Montoya J, Woese CR, Luthey-Schulten Z. Molecular signatures of ribosomal evolution. Proc. Natl. Acad. Sci. USA. 2008;105:13953–13958. doi: 10.1073/pnas.0804861105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Schmidt MW, Baldridge KK, Boatz JA, Elbert ST, Gordon MS, Jensen JJ, Koseki S, Matsunaga N, Nguyen KA, Su S, et al. General atomic and molecular electronic structure system. J. Comp. Chem. 1993;14:1347–1363. [Google Scholar]
  44. Seidelt B, Innis CA, Wilson DN, Gartmann M, Armache JP, Villa E, Trabuco LG, Becker T, Mielke T, Schulten K, et al. Structural insight into nascent polypeptide chain-mediated translational stalling. Science. 2009;326:1412–1415. doi: 10.1126/science.1177662. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Sethi A, O'Donoghue P, Luthey-Schulten Z. Evolutionary profiles from the QR factorization of multiple sequence alignments. Proc. Natl. Acad. Sci. USA. 2005;102:4045–4050. doi: 10.1073/pnas.0409715102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Stewart V, Yanofsky C. Evidence for transcription antitermination control of tryptophanase operon in Escherichia coli K-12. J. Bacteriol. 1985;164:731–740. doi: 10.1128/jb.164.2.731-740.1985. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Stone JE, Phillips JC, Freddolino PL, Hardy DJ, Trabuco LG, Schulten K. Accelerating molecular modeling applications with graphics processors. J. Comp. Chem. 2007;28:2618–2640. doi: 10.1002/jcc.20829. [DOI] [PubMed] [Google Scholar]
  48. Tenson T, Ehrenberg M. Regulatory nascent peptides in the ribosomal tunnel. Cell. 2002;108:591–594. doi: 10.1016/s0092-8674(02)00669-4. [DOI] [PubMed] [Google Scholar]
  49. Thompson JD, Higgins DG, Gibson TJ. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucl. Acids Res. 1994;22:4673–4680. doi: 10.1093/nar/22.22.4673. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Torrice MM, Bower KS, Lester HA, Dougherty DA. Probing the role of the cation-π interaction in the binding sites of GPCRs using unnatural amino acids. Proc. Natl. Acad. Sci. USA. 2009;106:11919–11924. doi: 10.1073/pnas.0903260106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Trabuco LG, Villa E, Mitra K, Frank J, Schulten K. Flexible fitting of atomic structures into electron microscopy maps using molecular dynamics. Structure. 2008;16:673–683. doi: 10.1016/j.str.2008.03.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Trabuco LG, Villa E, Schreiner E, Harrison CB, Schulten K. Molecular Dynamics Flexible Fitting: A practical guide to combine cryo-electron microscopy and X-ray crystallography. Methods. 2009;49:174–180. doi: 10.1016/j.ymeth.2009.04.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. van der Spoel D, Lindahl E, Hess B, Groenhof G, Mark AE, Berendsen HJC. Gromacs: Fast, flexible, and free. J. Comp. Chem. 2005;26:1701–1718. doi: 10.1002/jcc.20291. [DOI] [PubMed] [Google Scholar]
  54. Villa E, Sengupta J, Trabuco LG, LeBarron J, Baxter WT, Shaikh TR, Grassucci RA, Nissen P, Ehrenberg M, Schulten K, et al. Ribosome-induced changes in elongation factor Tu conformation control GTP hydrolysis. Proc. Natl. Acad. Sci. USA. 2009;106:1063–1068. doi: 10.1073/pnas.0811370106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Yang R, Cruz-Vera L, Yanofsky C. 23S rRNA nucleotides in the peptidyl transferase center are essential for tryptophan operon induction. J. Bacteriol. 2009;191:3445–3450. doi: 10.1128/JB.00096-09. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Zhang L, Hermans J. Hydrophilicity of cavities in proteins. Proteins: Struct., Func., Gen. 1996;24:433–438. doi: 10.1002/(SICI)1097-0134(199604)24:4<433::AID-PROT3>3.0.CO;2-F. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

01
NIHMS343793-supplement-01.pdf (2.4MB, pdf)
02
NIHMS343793-supplement-02.gif (8.7MB, gif)

RESOURCES