Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Feb 4.
Published in final edited form as: Mol Cell. 2016 Feb 4;61(3):341–351. doi: 10.1016/j.molcel.2016.01.008

SYNONYMOUS CODONS DIRECT CO-TRANSLATIONAL FOLDING TOWARDS DIFFERENT PROTEIN CONFORMATIONS

Florian Buhr 1,#, Sujata Jha 2,#, Michael Thommen 3,#, Joerg Mittelstaet 3, Felicitas Kutz 1, Harald Schwalbe 1,*, Marina V Rodnina 3,*, Anton A Komar 2,4,5,*
PMCID: PMC4745992  NIHMSID: NIHMS750905  PMID: 26849192

SUMMARY

In all genomes, most amino acids are encoded by more than one codon. Synonymous codons can modulate protein production and folding, but the mechanism connecting codon usage to protein homeostasis is not known. Here we show that synonymous codon variants in the gene encoding gamma-B crystallin, a mammalian eye lens protein, modulate the rates of translation and co-translational folding of protein domains monitored in real time by Förster resonance energy transfer and fluorescence intensity changes. Gamma-B crystallins produced from mRNAs with changed codon bias have the same amino acid sequence, but attain different conformations as indicated by altered in vivo stability and in vitro protease resistance. 2D NMR spectroscopic data suggest that structural differences are associated with different cysteine oxidation states of the purified proteins, providing a link between translation, folding, and the structures of isolated proteins. Thus, synonymous codons provide a secondary code for protein folding in the cell.

graphic file with name nihms-750905-f0001.jpg

INTRODUCTION

The genetic code is degenerate, with up to six synonymous codons encoding a given amino acid. The occurrence of synonymous codons in protein-coding open reading frames (ORFs) of genes is not random, thus revealing the existence of evolutionary pressure on codon choice (Hershberg and Petrov, 2008; Sharp et al., 2010; Plotkin and Kudla, 2011; Pechmann and Frydman, 2011; Chaney and Clark, 2015). The occurrence of synonymous codons and the abundance of the corresponding isoacceptor tRNAs are a major cause of non-uniformity in mRNA translation (Ikemura, 1985; Buchan and Stansfield, 2007; Komar, 2009; Pop et al., 2014; Dana and Tuller 2014; Gardin et al., 2014), although several other factors, such as mRNA structure or the presence of anti-Shine-Dalgarno-like sequences in bacterial ORFs (Li et al., 2012), may also contribute (Pop et al., 2014; Ingolia, 2014). In a given organism, frequently used codons are translated more rapidly than infrequently used ones (Ikemura, 1985; Buchan and Stansfield, 2007; Ingolia et al., 2009; Komar, 2009; Ingolia, 2014; Dana and Tuller 2014; Gardin et al., 2014). Non-random and non-uniform distribution of codons in gene ORFs provides a unique organism-specific pattern that modulates local translation elongation rates (Clarke and Clark 2008; Komar, 2009; Pechmann and Frydman, 2011; Chaney and Clark, 2015) and contributes to mRNA stability (Pedersen et al., 2011; Presnyak et al., 2015).

In vivo, protein folding begins co-translationally as nascent peptide chains emerge from the ribosome exit tunnel (Hartl and Hayer-Hartl, 2009; Komar, 2009; Kramer et al., 2009; Cabrita et al., 2010; Waudby et al., 2013; Pechmann et al, 2013; Gloge et al, 2014; Chaney and Clark, 2015). Variations in local translation rates may facilitate protein folding by allowing ordered, sequential structuring of the nascent polypeptide chains emerging from the ribosome (Tsai et al., 2008; Kramer et al., 2009; Komar, 2009; Zhang and Ignatova, 2011; Waudby et al., 2013; O’Brien et al., 2014; Gloge et al, 2014; Chaney and Clark, 2015). The significance of synonymous codon usage on protein folding is highlighted by findings showing that synonymous mutations and naturally occurring synonymous single nucleotide polymorphisms (sSNPs) can affect proteins’ activity (Komar et al., 1999; Kimchi-Sarfaty et al., 2007; Yu et al., 2015), interactions with drugs and inhibitors (Kimchi-Sarfaty et al., 2007), phosphorylation profiles (Zhou et al., 2013), sensitivity to limited proteolysis (Kimchi-Sarfaty et al., 2007; Zhang et al., 2009; Zhou et al., 2013), spectroscopic properties (Sander et al. 2014), and aggregation propensity (Hu et al., 2013; Sander et al. 2014; Kim et al., 2015), which ultimately can cause diseases (Sauna and Kimchi-Sarfaty, 2011; Hunt et al., 2014). Synonymous codon choice has been also suggested to affect efficient interaction of nascent polypeptides with the signal recognition particle (Pechmann et al., 2014). Changes in codon context caused by synonymous mutations may also induce mistranslation leading to protein misfolding (Drummond and Wilke, 2008). An effect of synonymous codon usage on co-translational folding properties of several proteins has been suggested based on computational work or experiments that utilized stalled ribosome-nascent chain complexes (for reviews see Tsai et al., 2008; Komar, 2009; Zhang and Ignatova, 2011; O’Brien et al., 2014; Gloge et al, 2014). However, little attention has been paid to the correct amino acid composition of the synonymous protein variants under study and the direct link between the synonymous mutations, the kinetics of translation, as well as real-time co-translational folding, and the structure of the released protein has not been shown to date.

To investigate how differential usage of synonymous codons affects translation kinetics, co- and post-translational folding, and protein stability, we analyzed in vivo expression of the recombinant bovine eye lens protein gamma-B crystallin in Escherichia coli cells and in vitro in a completely reconstituted high-performance translation system from E. coli. Gamma-B crystallin is a globular protein consisting of two homologous all-beta Greek-key domains connected by a six-residue flexible linker (Bloemendal et al., 2004). We have chosen gamma-B-crystallin, because translation of this two-domain protein in a mammalian cell-free system is a non-uniform process (Komar and Jaenicke, 1995). We suggested that the codon usage and translation rates in gamma-B-crystallin are optimized to tune the synthesis and folding of this protein in the cell (Komar and Jaenicke, 1995); however the direct experimental evidence in support of this hypothesis was lacking. The two domains of gamma-B crystallin comprise a fold that is very common in globular proteins and its folding pathway may be representative for many beta-rich folds in general (Bloemendal et al., 2004). Here, we monitored the kinetics of synthesis and co-translational folding of gamma-B crystallin synonymous variants in real time by fluorescence and Förster resonance energy transfer (FRET). The final protein conformation was assessed by 2D NMR. The stability of the protein variants was also probed in vivo by protein expression analysis and in vitro by limited proteolysis with Proteinase K (PK). Taken together, our data show how synonymous codon usage regulates translation elongation, co-translational folding, and protein quality in the cell.

RESULTS

Synonymous codon replacements improve the stability and solubility of gamma-B crystallin

The codon distribution in the native gamma-B crystallin mRNA sequence is non-optimal for translation in E. coli due to the presence of a number of codons that are more commonly used in the native host, Bos taurus, than in E. coli (Table S1). To study the role of synonymous mutations, we sought to compare the expression of this mRNA with a variant that would have a codon choice more optimal for E. coli. Common approaches to optimize heterologous gene expression include substitution of the majority of infrequently used codons with synonymous frequently used ones often combined with the elimination of extreme GC content that could contribute to formation of stable mRNA secondary structures thereby reducing translation efficiency (Gustafsson et al., 2004). However, maximizing the global speed of translation through these strategies often yields biologically inactive products that form insoluble aggregates and have to be re-folded in order to recover their native structure and biological activity (Gustafsson et al., 2004; Rosano and Ceccarelli, 2009, Komar, 2009). Moreover, even when proteins expressed in heterologous hosts remain soluble, they are not necessarily natively folded (de Marco et al., 2005). We chose to use an alternative optimization strategy and compare expression of the original B. taurus sequence (denoted U for un-optimized) with that of a so-called “harmonized” (Angov et al., 2008) (H) gene variant. Harmonization aims to optimize translation by introducing synonymous codons that have the most similar usage frequencies in the native and target host organisms (Angov et al., 2008; Komar, 2009) (Figure 1A; Table S1). Rather than simply replacing rare codons with frequent ones, this strategy involves alteration of both rare and frequent codons and is expected to establish mRNA translation kinetics in a heterologous host that mimic those observed in the native host organism (Angov et al., 2008; Komar, 2009). Harmonization may lead to increased GC content and a decreased Codon Adaptation Index (CAI) (a measure for synonymous codon usage bias) (Sharp et al., 1987), which is also widely used as a measure for the likelihood of success of heterologous gene expression (Sharp et al., 1987; Gustafsson et al., 2004). Nevertheless, the translation yields obtained with harmonized mRNA sequences may be higher than those with mRNA sequences optimized with frequent codons (Angov et al., 2008). This suggests that harmonization may facilitate proper protein folding and production of proteins with structures closely similar to their native analogues (Komar, 2009).

Figure 1. Effect of synonymous codon choice on the expression and stability of gamma-B crystallin in E. coli.

Figure 1

Open in a new tab

(A) Top - codon frequency profiles of gamma-B crystallin variants in: B. taurus (green) and E. coli U (red), H (blue). Bottom - relative differences in usage frequencies.

(B) Expression of U and H variants of gamma-B crystallin (γB) in E. coli. Top panel – SDS PAGE: total, soluble (S) and pellet (P) fractions and pET15b empty vector control. Proteins were visualized by Coomassie Brilliant Blue (CBB) staining. Bottom panel – quantification of full-length gamma-B crystallin (γB) and its distribution between S and P fractions. The total protein expression level (set to 100%) is the sum of S and P fractions.

(C) Western blotting using polyclonal anti-γB antibodies. The parenthesis indicates truncated products of the U variant expression that are not present or less abundant with H variant.

(D) Expression of U and H variants of gamma-B crystallin (γB) in E. coli based on quantitation of all bands detected by Western blotting in Figure 1C (dark grey bars) or by ELISA (light grey bars). The amounts of protein in soluble and pellet fractions are represented as a fraction of total protein. Error bars (B-D) represent the standard error of the mean (SEM); *p<0.05, **p<0.01 by Student’s t-test.

(E) Detection of expression products of U and H variants that contain a C-terminal 6×His-tag in total, soluble (S) and pellet (P) fractions of E. coli extracts by Western blotting using monoclonal anti-poly-histidine antibody.

See also Figure S1 and Tables S1-S3.

Harmonization of the gamma-B crystallin mRNA (H variant) resulted in a codon usage profile more similar to that of the native host organism (B. taurus) than the U variant (Figure 1A). It also led to an overall increase in the number of codons frequently used in E. coli, as indicated by an increase in the CAI score for expression in E. coli from 0.64 for the U variant to 0.78 for the H variant, closer to the 0.85 score for the native (U) mRNA translated in B. taurus. In E. coli, expression of the H variant yielded about 1.5- to 1.6-fold more full-length protein than expression of the U variant (Figure 1B and 1C), despite identical mRNA levels (Figure S1A). Compared to the U variant, expression of the H variant yielded more soluble protein (Figure 1D; middle and right panels) and fewer truncated protein products (Figure 1C; upper panel). Western blotting with anti-His antibodies recognizing the C-terminal 6×His-tag of the recombinant gamma-B-crystallin protein and MS analysis indicated that the observed “truncated” products were, at least in part, fragments of the full-length proteins (Figure 1E; S1B and Table S2). Because both the C-terminal His-tag and C-terminal peptides were detected by these methods, these fragments most likely arise from degradation rather than premature termination of translation. The different degradation patterns of the two protein variants suggested that they adopt different conformations. Importantly, MS analysis and sequencing of the full-length recombinant protein products (Figure S1C and Table S3) confirmed that the amino acid sequences of the U and H variants were identical.

To explore the effects of local synonymous codon substitutions at certain gene regions, to further reduce intracellular protein degradation and to increase protein solubility, we created a number of additional synonymous variants of gamma-B crystallin (Figure S2A). While protein solubility was improved up to ~70% (Figure S2B), the expression of some of these variants led to substantial levels of miscoding, in particular resulting in frequent replacement of Arg92 with Lys (Figure S2C and S2D). Thus, it could not be excluded that reduced degradation and increased solubility in some cases resulted from an altered protein sequence. We thus focused our further analysis on the U and H variants only, as the expression of these mRNAs did not give rise to any miscoding events. Together, these results indicate that synonymous codon content can lead to differences in the intracellular protein conformation and stability and in certain cases to increased levels of miscoding.

Purified gamma-B crystallin synonymous variants attain different conformations as measured by 2D NMR

To investigate whether differences in protein conformations observed during intracellular U and H protein expression were preserved in the final structure of the isolated soluble proteins, we conducted NMR analyses of the U and H variants expressed in E. coli. Preparative ion exchange chromatography (IEX) revealed one major peak for the U variant and two peaks for the H variant (H-P1 and H-P2, respectively; Figure 2A). Because all of the protein variants were of identical mass, the shift in the elution position between H-P1 and H-P2 suggested a lower isoelectric point (i.e., a different surface charge distribution) in H-P2. Further analysis by reversed-phase high-performance liquid chromatography (RP-HPLC) showed predominantly one peak for U, two peaks for H-P1 and a higher degree of heterogeneity for H-P2 (Figure 2B). Gamma-B crystallin contains seven cysteine residues, six of which are localized close to one another in the N-terminal domain (NTD) (Figure 2C). It cannot be excluded that these cysteines may become oxidized in cytosol of E. coli. Although disulfide bonds are rarely formed in E. coli cytoplasmic proteins, exceptions to this general rule do exist (Aslund et al., 1999; Jakob et al., 1999; Seras-Franzoso et al., 2012). Our NMR analysis of U, H-P1 and H-P2 revealed chemical shift differences in particular for the six cysteine residues at positions 15, 18, 22, 32, 41, and 78 (in NTD) (Figure 2C; inset and Table 1), both in 1D spectra (Figure 2D) and in 2D-1H-15N correlated experiments (Figure 2E and 2F). Upon dithioerythritol (DTT) treatment, or refolding under reducing conditions (Figure S3A), all protein species converged to a single state identical to U, strongly suggesting that U represents the fully reduced state of the protein (Figure 2D and S3A). This notion is supported by the Cα/Cβ chemical shift values of U variant cysteine residues measured in 3D-NMR experiments, which were characteristic for reduced cysteines. Therefore, purified U and H variants of gamma-B-crystallin expressed in E. coli differ structurally with respect to the oxidation state of cysteine residues within the NTD. This conclusion is further supported by the results of analysis of U, H-P1 and H-P2 protein spectra in which only cysteine residues were 15N-labeled (Figure 2C and Table 1). Importantly, peak doubling and chemical shift changes were also observed for other residues in 2D-1H-15N correlated experiments (Figure 2E and 2F). This suggests that the oxidation of at least two cysteine residues within the NTD is a major cause of overall structural changes which affect the conformation of the respective cysteines as well as other residues in the purified protein. These and previous data (Bloemendal et al., 2004), including a crystal structure of gamma-B crystallin (Najmudin et al., 1993), suggest the existence of two alternative conformations for Cys18 and Cys22 in the native protein purified from bovine eye lenses (Najmudin et al., 1993). Therefore, in terms of its oxidation state, the H variant produced in E. coli appears to be closer than the U variant produced in E. coli to the natural (N) protein variant purified from bovine eye lenses. Overnight incubation of U with catalytic amounts of Cu(II) under air supply (known to catalyze cysteine oxidation to cystine (Cavallini et al., 1969)) resulted in a 15N backbone spectrum similar to H-P1, albeit with a different peak intensity distribution, suggesting that the respective peaks in the NMR spectra of the H protein reflect the differences in the oxidation state (Figure S3B).

Figure 2. Physico-chemical and structural properties of gamma-B crystallin U and H variants expressed in E. coli as determined by RP-HPLC and 2D-1H-15N-correlation NMR.

Figure 2

Open in a new tab

(A) Preparative pH-gradient ion exchange chromatography: U (red), H (blue).

(B) Analytical RP-HPLC of ion exchange fractions U (red), H-P1 (blue), H-P2 (black).

(C) Overlay of 2D-1H-15N correlated NMR backbone spectra of 15N-cysteine-labeled U, H-P1 and H-P2. Insets show 1D rows to visualize differential cysteine peak intensities, which are normalized against C109. Full list of peak integrals is presented in Table 1. Lower right quadrant: Crystal structure of bovine gamma-B crystallin (PDB-ID 4GCR); cysteine residues highlighted.

(D) 1D summation of rows extracted from 2D-1H-15N backbone datasets recorded for U, H-P1 and H-P2 expressed in 15N-labeled rich medium. Upper panel: Spectra comparison. Lower panel: Spectra after treatment of samples with 10 mM DTT.

(E,F) Overlay of 2D-1H-15N correlated NMR spectra for U, H-P1 and U, H-P2. Addition of DTT resulted in full convergence to U-like spectra (Figure S3A). See also Table 1, Figure S3B and S4.

Table 1. Differential cysteine peak intensities of 2D-1H-15N correlated NMR backbone spectra of 15N-cysteine-labeled U, H-P1 and H-P2. See also Figure 2.

Residue U H-P1 H-P2
Cys 109 (Reference) 1.0 1.0 1.0
Cys 15 (red) 1.05 0.53 0.30
Cys 15 (ox) 0.01 0.39 0.59
Cys 18 (red) 0.83 0.45 0.21
Cys 18 (ox) N/A N/A N/A
Cys 22 (red) 0.87 0.40 0.22
Cys 22 (ox) 0.01 0.38 0.70
Cys 32 (red) 0.86 0.45 0.33
Cys 32 (ox) 0.03 0.30 0.50
Cys 41 (red) 0.97 0.55 0.41
Cys 41 (ox) 0.02 0.36 0.50
Cys 78 (red) 0.80 0.43 0.28
Cys 78 (ox) 0.01 0.42 0.80
Open in a new tab

Expression of U and H variants in the SHuffle T7 E. coli (NEB) strain (which constitutively expresses a chromosomal copy of the disulfide bond isomerase DsbC lacking its signal sequence, hence retaining the protein in the cytoplasm and thus facilitating disulfide bond formation in the E. coli cytoplasm (Lobstein et al., 2012), revealed overall similar differences for the two variants compared to the expression in BL21 strain (Figure S4). However, use of this strain when grown in minimal medium resulted in extensive protein misfolding (especially for the U variant (Figure S4A)), possibly due to excessive oxidation and/or early formation of mixed disulfides leading to aggregation and precluding accurate NMR analysis under these conditions. Expression in SHuffle T7 E. coli grown in 15N-rich medium resulted in yet another oxidized variant of U (with chemical shifts partially overlapping with, but still different from, those of the U variant expressed in the BL21 strain (Figure S4C)). In contrast, the spectra of H-P1 were remarkably similar after expression in either strain of bacteria (Figure S4B and 4D). This suggests that the intracellular folding of H is more robust and less dependent on environmental conditions than the folding of U, possibly due to improved co-translational folding kinetics resulting from harmonization of the codon usage profile in the H variant. We thus believe that optimized translation kinetics can help to produce recombinant proteins more similar to their natural analogs.

Figure 4. Co-translational folding of the NTD monitored in real-time by FRET.

Figure 4

Open in a new tab

Left panel, positions of the donor (BOF) and acceptor (BOP) dyes in the structure of gamma-B crystalline. Middle and right panels, Time-resolved folding of U (red) and H (blue) peptides monitored by FRET between BOP-Met at position 1 and BOF-Cys at position 88 in the stopped-flow apparatus. DA, both donor and acceptor dyes were present; A, control in the absence of the donor. Middle panel: direct comparison of FRET due to folding for the U and H variants. Right panel: FRET signal vs. acceptor direct excitation at donor wavelength; for better comparison, the traces with and without the donor for the U or H variants, respectively, were adjusted to the same starting level; the H traces are arbitrarily shifted from the U traces for visual clarity. Time courses were evaluated using a two-step model comprising a delay phase which shows no change in fluorescence and an exponential phase corresponding to a monomolecular folding reaction using GraphPad Prism. Delay times are 35 ± 0.3 s for H and 50 ± 0.5 s for U. The folding times determined by exponential fitting after delay are 39 ± 1 s for H and 59 ± 1 s for U. See also Figure S5.

Synonymous mRNA variants are translated with different kinetics

The translational kinetics of the U and H variants was compared in a completely reconstituted high-performance cell-free in-vitro translation system from E. coli (Mittelstaet et al, 2013). In this system, proteins were N-terminally labeled during translation by using BodipyFL-Met-tRNAfMet as the initiator tRNA (BOF-Met-tRNAfMet). BOF, as well as BOP (Bodipy 576/589) used in the experiments below, were previously utilized to study co-translational helix formation within the ribosome (Woolhead et al., 2004) and have been validated as reporters for translation and co-translational folding (Johnson, 2005). The ribosomes were synchronized at the translation initiation step and translation elongation was initiated by the addition of aminoacyl-tRNAs, EF-Tu and EF-G. After translating the mRNA to the 3’-end, the ribosomes remained bound to the mRNA and the nascent peptide was not released from the ribosome due to the lack of a stop codon in the mRNA and absence of termination factors from the translation system. In this reconstituted system, the H mRNA was translated more rapidly that than the U mRNA (Figure 3A and 3B). Full-length gamma-B crystallin protein was first detected after 43 s of H mRNA translation compared to 57 s for U mRNA (Figure 3B). The average protein synthesis rate, calculated from the sum of the delay and synthesis times, was 2.2 amino acids (aa)/s for the H variant, compared to 1.8 aa/s for the U variant.

Figure 3. Different translation kinetics of U and H variants in a fully reconstituted E. coli cell-free translation system.

Figure 3

Open in a new tab

(A) Accumulation and size distribution of U (upper panel) and H (lower panel) in vitro translation products. Peptides were separated by SDS-PAGE and visualized by the fluorescence of the BOP label attached to the N-terminus of the peptides. N and C indicate peptides arising during translation of the NTD and CTD, respectively; γBN is the fragment corresponding to the NTD.

(B) Kinetic analysis of accumulation of full-length gamma-B crystallin. “Delay” is the time before appearance of the full-length product. “Rate” is the average translation rate (amino acids per second). Error bars show standard deviations (SD) calculated from n=3 replicates. We further tested whether the difference in delay times for H and U is statistically equivalent to a single shared parameter for the delay time in synthesis of γB-crystallin from the U and H mRNA (null hypothesis) or alternatively if two independent parameters for the delay for U or H are justified. According to the extra sum-of-squares F test, the null hypothesis of an identical delay time for U/H is rejected with a significance of p<0.0001 (****).

(C) Lifetime of translation intermediates corresponding to the NTD upon translation of U (red) and H (blue) mRNAs. The total intensity of all bands indicated as N in Figure 3A (U and H, respectively) after 10 s of translation was set to 1.

(D) Stopped-flow kinetics of synthesis and movement through the ribosome exit tunnel of U (red) and H (blue) nascent chains monitored by a fluorescence reporter (BOP) at the N-terminus of the nascent peptides. The maximum delay in translation of the U sequence relative to H is indicated. Apparent variations in the height of fluorescence changes between U and H are due to differences in the accumulation of the respective nascent peptides resulting from the altered translation rates.

(E) Same as D, except using BOF fluorescence as the reporter.

See also Figure S5.

At any given time during translation, individual ribosomes in the reaction carry nascent peptides of different lengths (Figure 3A). At the beginning of translation, short peptides are prevalent, but at later time points, the full-size protein becomes predominant. Analysis of the transient accumulation of nascent peptides indicated that the U and H variants of gamma-B crystallin differ not only in global translation rate, but also in local translation kinetics (Figure 3A). Specifically, more translational pausing (and hence greater prevalence of shorter nascent chains <70 aa) was observed during synthesis of the NTD of the U protein, compared to the H protein (Figure 3A and 3C). The cumulative lifetime (τ) of nascent NTD chains before they were converted into longer peptides was 48 s for U, compared to 25 s for H. In addition, a smaller proportion of ribosomes continued translation past the NTD on the U mRNA than on the H mRNA (Figure 3C). Slow synthesis of the NTD is consistent with the abundance of rare codons in the corresponding part of the U mRNA (Figure 1A; codons 16-24) and explains the observed delayed synthesis of the full-length U protein. In contrast, accumulation of nascent chains with lengths of ~130-160 aa was detected during translation of the H variant, but not the U variant, suggesting ribosome pausing during synthesis of the C-terminal domain (CTD) of the H variant.

To improve the time resolution of translation, we performed an experiment similar to that shown in Figure 3A in which we monitored fluorescence changes of the BOP reporter attached to the N-terminus of the nascent peptide moving through the ribosomal exit tunnel in real time (Figure 3D). This approach also demonstrated that the U variant was translated more slowly than the H variant due to an approximately 20 s delay between the H and U variants (Figure 3D). The difference in kinetics appeared early (detected after 1 s of translation) and became more pronounced through the duration of NTD synthesis (compare Figure 3C and 3D), consistent with translational pausing due to rare codons in the NTD of the U variant.

Finally, when BOF at the N-terminus of the growing nascent peptide was monitored, the fluorescence changes for the U and H variants were identical initially, but started to deviate after about 1-2 s of translation, again coinciding with the presence of rare codons at the beginning of the U mRNA (Figure 3E). The 6 s difference in synthesis between the U and H accounts for a delay in the production of the initial parts of the NTD; differences in translation of the C-terminal portions of the protein are not reported by BOF fluorescence. Taken together, these data indicate that the synonymous codon substitutions that distinguish the U and H mRNA variants affected the global and local kinetics of gamma-B crystallin translation. It cannot be excluded that the observed alterations in translation kinetics are not due to differences in decoding rates of tRNA selection, but rather are related to possible differences in mRNA secondary structure or other factors unknown so far. However, regardless of the underlying molecular mechanism(s), it is clear that the choice between synonymous codons can affect translation kinetics.

Synonymous codon usage modulates the kinetics of co-translational folding

Next we examined whether the observed differences in translation kinetics led to differences in co-translational folding of the U and H nascent polypeptides by performing real-time FRET measurements (Figure 4). The FRET acceptor BOP was attached at the N-terminus of nascent peptides by priming translation with BOP-Met-tRNAfMet. To incorporate the fluorescence donor, BOF, we introduced an UAG stop codon at position 88 of the U and H mRNAs, which was decoded by a modified amber suppressor BOF-Cys-tRNA. Thus, the fluorescence donor was introduced co-translationally. Importantly, decoding of the UAG stop codon by BOF-Cys-tRNA did not affect the translation kinetics of either mRNA (Figure S5). In the unfolded protein, donor and acceptor fluorophores are too far apart for FRET to occur. Upon protein compaction during folding, BOP and BOF come into proximity, resulting in FRET (Figure 4). The onset of folding was significantly delayed for the U variant, compared to the H variant (50 s versus 35 s, respectively), consistent with the observed differences in H and U translation kinetics (Figure 3A; 3B and 3C). Given that the synthesis of the NTD was completed after 20 s (H) or 30 s (U) (Figure 3A and S5), the delay between NTD synthesis and folding was 15 s and 20 s for H and U, respectively. At the respective translation rates of 2.2 aa/s and 1.8 aa/s (Figure 3B), this delay should allow the ribosome to synthesize 33-36 amino acids past the NTD which would fill the polypeptide exit tunnel of the ribosome, such that the NTD itself is extruded from the tunnel. This suggests that global folding of the NTD (H or U), as reported by FRET, occurred co-translationally, shortly after the corresponding portion of the polypeptide emerged from the exit tunnel. Formation of locally restricted structural elements within the exit tunnel cannot be excluded, however. Thereafter, the average time of NTD folding was 59 s for U and 39 s for H. Notably, the 20 s difference represents the divergence in the H and U folding rates after the corresponding portion of the protein has emerged from the ribosome. This reveals the specific effect of synonymous codon replacement on the kinetics of co-translational protein folding, beyond the differences in production times for nascent chains of the same length. The overall effect, taking into account also the additional delay time for the NTD and linker synthesis, gives a total folding time of 109 s for U and 74 s for H.

Gamma-B crystallin synonymous variants display differential protease sensitivity in vitro

To further distinguish between folded and unfolded states of the protein, we probed the conformation of ribosome-bound nascent chains using pulse proteolysis with proteinase K (PK) (Figure 5A; 5B and 5C). Isolated recombinant or natural gamma-B crystallin is completely resistant to PK digestion (Figure S6A; S6B and S6C), while partial denaturation with 3M guanidinium chloride (GuHCl) results in complete gamma-B crystallin degradation by PK (Figure S6A; S6B and S6C). PK treatment of ribosome-bound nascent chains produced upon in vitro translation of U or H mRNAs resulted in two major products: (i) full-length protein and (ii) a fragment of about 80-90 residues in length, corresponding in size to the NTD (Figure 5A; S6D). While NTD folding was too rapid to evaluate potential differences in the folding rates of U and H variants by proteolysis, the approach did reveal that formation of a PK-resistant full-length nascent polypeptide occurred on a significantly shorter time-scale with H than with U (Figure 5B). Because the synthesis of the nascent chains is essentially complete after 2 min, protection against PK digestion must represent variations in folding dynamics of the emerging nascent U and H proteins. Notably, because the ribosome tunnel occludes about ~30-40 aa of the protein from the C terminus (Wilson and Beckmann, 2011), folding of the CTD is not completed until the entire nascent chain is released from the ribosome. Therefore, the appearance of the full-length protein after PK digestion is an indication of the initial stages of compaction of the emerging CTD. Overall, these results show that synonymous codon replacements in gamma-B crystallin resulted in altered conformational dynamics of ribosome-bound nascent chains.

Figure 5. Sensitivity of U and H polypeptide chains to pulse proteolysis.

Figure 5

Open in a new tab

(A) Proteolysis of ribosome-bound gamma-B crystallin chains assessed by SDS-PAGE. At different time points of in vitro translation, chains were digested with 5.4 pmol PK for 2 min at 37°C. γB (full-length protein) and γBN (N-terminal domain) protease-resistant products are indicated by arrows.

(B) Time courses of accumulation of PK-resistant N-terminal (open circles) and full-length (closed circles) products relative to the undigested protein. The last data point (60 min) was excluded from the exponential fitting, as it represents a decrease in the portion of PK-resistant NTD due to accumulation of the full-length protein. Error bars show the SEM for n=7 replicates.

(C) Left panel: PK proteolysis of puromycin-released chains after 3 min of translation. Right panel: quantification of PK resistance of released U (black symbols) and H (gray symbols) products from the data shown in the left panel. Error bars show the SD for n=9 replicates. *p<0.05, **p<0.01, ***p<0.005, ****p<0.0001 by Student’s two-tailed unpaired t-test. See also Figure S6.

To examine whether the folding states of the U and H variants are also different after their release from the ribosome, we tested the PK resistance of gamma-B crystallin chains released from the ribosome by puromycin (Figure 5C). Upon release, both U and H proteins rapidly fold into a PK-resistant conformation similar to that of the purified recombinant protein (Figure S6). The yield of folded U protein did not change significantly with time; however, with H, another 16% of the protein became more PK-resistant within 30 minutes after release from the ribosome. This suggests that, after release from the ribosome, the conformational ensemble of H is sufficiently pre-organized to allow more efficient post-translational folding. The final yield of PK-resistant protein was greater for the H variant (86%) than for the U variant (65%). The peptides that did not fold properly were PK-sensitive and likely represent part of the protease-sensitive protein found in the insoluble fraction in vivo (Figure 1). Thus, differences in velocities of co-translational folding resulted in different conformational states of U and H variants on the ribosome and after release from the ribosome.

DISCUSSION

Our analysis of the effects of synonymous codon choice on the translation of gamma-B crystallin mRNAs shows that codon bias alters local and global translation rates and results in the formation of alternative conformations of the nascent protein on the ribosome and in solution after the release of the completed protein (Figure 6). Rare codons at the beginning of the U mRNA slow down the synthesis of the NTD and lead to significantly slower global NTD folding of in the U variant compared to the H variant. It is possible that slow translation allows for partitioning of elemental folding events to alternative pathways (O’Brien et al., 2014). The spectrum of conformations attained by the nascent chains on the ribosome may determine different trajectories of their folding in the vestibule region (Tu et al., 2014; Lu and Deutsch, 2014; Holtkamp et al., 2015) of the polypeptide exit tunnel (Wilson and Beckmann, 2011) and/or after the chain emerges from the tunnel and is released from the ribosome. This, in turn, may change the local environment and orientation of cysteine residues leading to distinct patterns of disulfide bridges and affecting the resulting protein’s stability and propensity to form aggregates.

Figure 6. Synonymous codon usage directs co-translational folding towards different protein conformations.

Figure 6

Open in a new tab

Alternatively, the slower folding of U compared to H may be due to the decreased rate of CTD synthesis, which is manifested e.g. by the different rates of translation past the NTD for U and H (Figure 3A and 3C). In this case, folding of the NTD may be affected by interaction(s) with the ribosome (Kaiser et al., 2011; Holtkamp et al., 2015) and the rate with which the NTD moves away from the ribosome surface and/or altered interactions with the emerging CTD and the dynamics of the sampling between NTD and CTD. While such mechanistic details remain to be clarified, our results clearly demonstrate that different synonymous codon usage influences the kinetics of translation, resulting in different co-translational folding trajectories and ultimately different final conformations of the protein. Thus, synonymous codon usage serves as a secondary code that guides in vivo protein folding and constitutes an additional source of conformational variability of proteins.

EXPERIMENTAL PROCEDURES

Plasmids, E. coli cells, and protein expression

Bovine gamma-B crystallin cDNA (Hay et al., 1987) was a gift of Dr. J. Mark Petrash (University of Colorado, Denver, CO, USA). For expression in E. coli, the un-optimized (U) gamma-B crystallin cDNA was PCR amplified, fused to a C-terminal 6×His-tag and cloned into the pET15b vector (Novagen) via NcoI and XhoI sites. Variants H of the gamma-B crystallin gene were chemically synthesized by GeneArt AG (Life Technologies) and cloned into pET15b to obtain a construct with 5’ and 3’ untranslated regions identical to those of the U variant. Analytical-scale protein expression was performed in E. coli cells as described in Supplemental Experimental Procedures.

Analysis of protein expression

For analysis of the relative expression levels of the U and H variants and comparison of soluble (supernatant) and insoluble (pellet) protein levels, Coomassie Brilliant Blue (CBB) staining of proteins separated by Tris-Tricine SDS-PAGE (Schägger and von Jagow, 1987), Western blotting (WB) and Enzyme-Linked Immunosorbent Assay (ELISA) were used. See Supplemental Experimental Procedures for detailed descriptions.

In vitro translation

Single-turnover in vitro translation was carried out in a reconstituted minimum translation system (Doerfel et al., 2013; Mittelstaet et al., 2013). Components of the reconstituted in vitro translation system were prepared as described (Mittelstaet et al., 2013), see Supplemental Experimental Procedures for details.

Pulse proteolysis of in vitro translation products

Pulse proteolysis of translation products was carried out using either ribosome-bound nascent chains or the released chains obtained after puromycin (1.3 mM final concentration) treatment. Routinely, 25 μL translation reaction mix was subjected to digestion with proteinase K (PK) as indicated in Figures 4 and S6 for 2 min at 37°C. Following digestion, PMSF was added to a final concentration of 10 mM to inactivate the protease and samples were immediately flash frozen in liquid nitrogen. Samples were thawed and subjected to RNaseA treatment (5 μL, 2 mg/mL RNase A, 30 min at 37°C). After RNase digestion, 2× Tris-tricine loading buffer was added and peptides were resolved by Tris-tricine SDS PAGE as described above.

Protein expression and purification for NMR spectroscopy

Protein expression was performed in BL21(DE3) and Shuffle T7 E. coli strains. For uniform 15N-labeling, a commercial rich medium was used. 15N/13C–labeled protein for backbone assignment was expressed in M9 minimal medium containing [15N]NH4Cl and [U-13C]glucose. For selective 15N-cysteine labeling, a modified M9 medium containing [15N]cysteine was used, as described (Muchmore et al., 1989). Cells were harvested by centrifugation and lysates were purified by Nickel affinity chromatography (HisTrap HP Ni-NTA), size exclusion chromatography (Superdex 75 pg), and ion exchange chromatography (HiTrap Q XL). NMR buffer: 50 mM Tris, pH 8.0, 200 mM NaCl, 10% D2O, 0.1% DSS. Samples for C-4 reverse-phase HPLC were taken before and after ion exchange chromatography. See Supplemental Experimental Procedures for details.

NMR spectroscopy

NMR experiments were conducted on Bruker spectrometers (AV600-AV900). Processing and analysis of the data were performed using the software programs TopSpin 2.1-3.2 (Bruker BioSpin, Rheinstetten) and Cara 1.9.0 (Rochus L. J. Keller). 2D 1H-15N backbone spectra were recorded using the BEST-TROSY experiment. Backbone resonance assignment was performed using standard 3D pulse sequences. Sample concentrations were 250-350 μM after expression in 15N-labeled rich media, 0.8-1.2 mM after expression in minimal media. For further details see Supplemental Experimental Procedures (Tables S4-S7).

Supplementary Material

1
2

Highlights.

  • Quality of protein folding in cells is guided by synonymous codon usage

  • NMR reveals multiple conformational and oxidation states of synonymous variants

  • Synonymous codon usage in mRNA alters translation kinetics

  • Real-time co-translational folding is guided by synonymous codon usage

Acknowledgements

We thank Drs. Jun Qin and Xianqing Song for many helpful discussions, Patricia Stanhope Baker for help with manuscript preparation, Shrutee Jakhanwal for help in preparing functionally-active tRNACys, and Olaf Geintzer, Christina Kothe, Sandra Kappler, Anna Pfeifer, Theresia Uhlendorf, Tanja Wiles, and Franziska Hummel for expert technical assistance. This work was supported by grants of the Human Frontier Science Program (grant #RGP0024/2010 to A.A.K., M.V.R. and H.S.), 13GRNT17070025 (AHA) and 1R15HL121779-01A1 (NIH) (to A.A.K.) and of the Deutsche Forschungsgemeinschaft (FOR1805 to M.V.R.).

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.

Author contributions

H.S., M.V.R. and A.A.K. supervised the research. F.B., S.J., M.T., J.M. and F.K. performed the research. F.B., S.J., M.T., J.M, F.K., H.S., M.V.R. and A.A.K. analyzed data. H.S., M.V.R. and A.A.K. wrote the manuscript and all authors contributed to the final version of the manuscript.

Competing financial interests

The authors declare no competing financial interests.

REFERENCES

  1. Angov E, Hillier CJ, Kincaid RL, Lyon JA. Heterologous protein expression is enhanced by harmonizing the codon usage frequencies of the target gene with those of the expression host. PLoS ONE. 2008;3:e2189. doi: 10.1371/journal.pone.0002189. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Aslund F, Zheng M, Beckwith J, Storz G. Regulation of the OxyR transcription factor by hydrogen peroxide and the cellular thiol-disulfide status. Proc. Natl. Acad. Sci. U. S. A. 1999;96:6161–6165. doi: 10.1073/pnas.96.11.6161. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bloemendal H, de Jong W, Jaenicke R, Lubsen NH, Slingsby C, Tardieu A. Ageing and vision: structure, stability and function of lens crystallins. Prog. Biophys. Mol. Biol. 2004;86:407–485. doi: 10.1016/j.pbiomolbio.2003.11.012. [DOI] [PubMed] [Google Scholar]
  4. Buchan JR, Stansfield I. Halting a cellular production line: responses to ribosomal pausing during translation. Biol. Cell. 2007;99:475–487. doi: 10.1042/BC20070037. [DOI] [PubMed] [Google Scholar]
  5. Cabrita LD, Dobson CM, Christodoulou J. Protein folding on the ribosome. Curr. Opin. Struct. Biol. 2010;20:33–45. doi: 10.1016/j.sbi.2010.01.005. [DOI] [PubMed] [Google Scholar]
  6. Cavallini D, De Marco C, Duprè S, Rotilio G. The copper catalyzed oxidation of cysteine to cystine. Arch. Biochem. Biophys. 1969;130:354–361. doi: 10.1016/0003-9861(69)90044-7. [DOI] [PubMed] [Google Scholar]
  7. Chaney JL, Clark PL. Roles for Synonymous Codon Usage in Protein Biogenesis. Annu Rev Biophys. 2015;44:143–166. doi: 10.1146/annurev-biophys-060414-034333. [DOI] [PubMed] [Google Scholar]
  8. Clarke TF, 4th, Clark PL. Rare codons cluster. PLoS One. 2008;3:e3412. doi: 10.1371/journal.pone.0003412. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. de Marco A, Vigh L, Diamant S, Goloubinoff P. Native folding of aggregation-prone recombinant proteins in Escherichia coli by osmolytes, plasmid- or benzyl alcohol-overexpressed molecular chaperones. Cell Stress Chaperones. 2005;10:329–339. doi: 10.1379/CSC-139R.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Dana A, Tuller T. The effect of tRNA levels on decoding times of mRNA codons. Nucleic Acids Res. 2014;42:9171–9181. doi: 10.1093/nar/gku646. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Doerfel LK, Wohlgemuth I, Kothe C, Peske F, Urlaub H, Rodnina MV. EF-P is essential for rapid synthesis of proteins containing consecutive proline residues. Science. 2013;339:85–88. doi: 10.1126/science.1229017. [DOI] [PubMed] [Google Scholar]
  12. Drummond DA, Wilke CO. Mistranslation-induced protein misfolding as a dominant constraint on coding-sequence evolution. Cell. 2008;134:341–352. doi: 10.1016/j.cell.2008.05.042. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Gardin J, Yeasmin R, Yurovsky A, Cai Y, Skiena S, Futcher B. Measurement of average decoding rates of the 61 sense codons in vivo. Elife. 2014;3 doi: 10.7554/eLife.03735. eLife.03735. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Gloge F, Becker AH, Kramer G, Bukau B. Co-translational mechanisms of protein maturation. Curr. Opin. Struct. Biol. 2014;24:24–33. doi: 10.1016/j.sbi.2013.11.004. [DOI] [PubMed] [Google Scholar]
  15. Gubbens J, Kim SJ, Yang Z, Johnson AE, Skach WR. In vitro incorporation of nonnatural amino acids into protein using tRNA(Cys)-derived opal, ochre, and amber suppressor tRNAs. RNA. 2010;16:1660–1672. doi: 10.1261/rna.2024810. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Gustafsson C, Govindarajan S, Minshull J. Codon bias and heterologous protein expression. Trends Biotechnol. 2004;22:346–353. doi: 10.1016/j.tibtech.2004.04.006. [DOI] [PubMed] [Google Scholar]
  17. Hartl FU, Hayer-Hartl M. Converging concepts of protein folding in vitro and in vivo. Nat. Struct. Mol. Biol. 2009;16:574–581. doi: 10.1038/nsmb.1591. [DOI] [PubMed] [Google Scholar]
  18. Hay RE, Woods WD, Church RL, Petrash JM. cDNA clones encoding bovine gamma-crystallins. Biochem. Biophys. Res. Commun. 1987;146:332–338. doi: 10.1016/0006-291x(87)90729-7. [DOI] [PubMed] [Google Scholar]
  19. Hershberg R, Petrov DA. Selection on codon bias. Annu. Rev. Genet. 2008;42:287–299. doi: 10.1146/annurev.genet.42.110807.091442. [DOI] [PubMed] [Google Scholar]
  20. Holtkamp W, Kokic G, Jäger M, Mittelstaet J, Komar AA, Rodnina MV. Cotranslational protein folding on the ribosome monitored in real time. Science. 2015;350:1104–1107. doi: 10.1126/science.aad0344. [DOI] [PubMed] [Google Scholar]
  21. Hu S, Wang M, Cai G, He M. Genetic code-guided protein synthesis and folding in Escherichia coli. J. Biol. Chem. 2013;288:30855–30861. doi: 10.1074/jbc.M113.467977. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Hunt RC, Simhadri VL, Iandoli M, Sauna ZE, Kimchi-Sarfaty C. Exposing synonymous mutations. Trends Genet. 2014;30:308–321. doi: 10.1016/j.tig.2014.04.006. [DOI] [PubMed] [Google Scholar]
  23. Ingolia NT. Ribosome profiling: new views of translation, from single codons to genome scale. Nat. Rev. Genet. 2014;15:205–213. doi: 10.1038/nrg3645. [DOI] [PubMed] [Google Scholar]
  24. 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]
  25. Ikemura T. Codon usage and tRNA content in unicellular and multicellular organisms. Mol. Biol. Evol. 1985;2:13–34. doi: 10.1093/oxfordjournals.molbev.a040335. [DOI] [PubMed] [Google Scholar]
  26. Jakob U, Muse W, Eser M, Bardwell JC. Chaperone activity with a redox switch. Cell. 1999;96:341–352. doi: 10.1016/s0092-8674(00)80547-4. [DOI] [PubMed] [Google Scholar]
  27. Johnson AE. The co-translational folding and interactions of nascent protein chains: a new approach using fluorescence resonance energy transfer. FEBS Lett. 2005;579:916–920. doi: 10.1016/j.febslet.2004.11.046. [DOI] [PubMed] [Google Scholar]
  28. Kaiser CM, Goldman DH, Chodera JD, Tinoco I, Jr., Bustamante C. The ribosome modulates nascent protein folding. Science. 2011;334:1723–1727. doi: 10.1126/science.1209740. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kim SJ, Yoon JS, Shishido H, Yang Z, Rooney LA, Barral JM, Skach WR. Protein folding. Translational tuning optimizes nascent protein folding in cells. Science. 2015;348:444–448. doi: 10.1126/science.aaa3974. [DOI] [PubMed] [Google Scholar]
  30. Kimchi-Sarfaty C, Oh JM, Kim IW, Sauna ZE, Calcagno AM, Ambudkar SV, Gottesman MM. A “silent” polymorphism in the MDR1 gene changes substrate specificity. Science. 2007;315:525–528. doi: 10.1126/science.1135308. [DOI] [PubMed] [Google Scholar]
  31. Komar AA. A pause for thought along the co-translational folding pathway. Trends Biochem. Sci. 2009;34:16–24. doi: 10.1016/j.tibs.2008.10.002. [DOI] [PubMed] [Google Scholar]
  32. Komar AA, Jaenicke R. Kinetics of translation of gamma B crystallin and its circularly permutated variant in an in vitro cell-free system: possible relations to codon distribution and protein folding. FEBS Lett. 1995;376:195–198. doi: 10.1016/0014-5793(95)01275-0. [DOI] [PubMed] [Google Scholar]
  33. Komar AA, Lesnik T, Reiss C. Synonymous codon substitutions affect ribosome traffic and protein folding during in vitro translation. FEBS Lett. 1999;462:387–391. doi: 10.1016/s0014-5793(99)01566-5. [DOI] [PubMed] [Google Scholar]
  34. Kramer G, Boehringer D, Ban N, Bukau B. The ribosome as a platform for co-translational processing, folding and targeting of newly synthesized proteins. Nat. Struct. Mol. Biol. 2009;16:589–597. doi: 10.1038/nsmb.1614. [DOI] [PubMed] [Google Scholar]
  35. Lu J, Deutsch C. Regional discrimination and propagation of local rearrangements along the ribosomal exit tunnel. J. Mol. Biol. 2014;426:4061–4073. doi: 10.1016/j.jmb.2014.10.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Li GW, Oh E, Weissman JS. The anti-Shine-Dalgarno sequence drives translational pausing and codon choice in bacteria. Nature. 2012;484:538–541. doi: 10.1038/nature10965. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Lobstein J, Emrich CA, Jeans C, Faulkner M, Riggs P, Berkmen M. SHuffle, a novel Escherichia coli protein expression strain capable of correctly folding disulfide bonded proteins in its cytoplasm. Microb. Cell Fact. 2012;11:56. doi: 10.1186/1475-2859-11-56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Mittelstaet J, Konevega AL, Rodnina MV. A kinetic safety gate controlling the delivery of unnatural amino acids to the ribosome. J. Am. Chem. Soc. 2013;135:17031–17038. doi: 10.1021/ja407511q. [DOI] [PubMed] [Google Scholar]
  39. Muchmore DC, McIntosh LP, Russell CB, Anderson DE, Dahlquist FW. Expression and nitrogen-15 labeling of proteins for proton and nitrogen-15 nuclear magnetic resonance. Methods Enzymol. 1989;177:44–73. doi: 10.1016/0076-6879(89)77005-1. [DOI] [PubMed] [Google Scholar]
  40. Najmudin S, Nalini V, Driessen HP, Slingsby C, Blundell TL, Moss DS, Lindley PF. Structure of the bovine eye lens protein gammaB(gammaII)-crystallin at 1.47 A. Acta Crystallogr. D Biol. Crystallogr. 1993;49:223–233. doi: 10.1107/S0907444992007601. [DOI] [PubMed] [Google Scholar]
  41. O’Brien EP, Ciryam P, Vendruscolo M, Dobson CM. Understanding the influence of codon translation rates on cotranslational protein folding. Acc. Chem. Res. 2014;47:1536–1544. doi: 10.1021/ar5000117. [DOI] [PubMed] [Google Scholar]
  42. Pechmann S, Chartron JW, Frydman J. Local slowdown of translation by nonoptimal codons promotes nascent-chain recognition by SRP in vivo. Nat. Struct. Mol. Biol. 2014;21:1100–1105. doi: 10.1038/nsmb.2919. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Pechmann S, Frydman J. Evolutionary conservation of codon optimality reveals hidden signatures of cotranslational folding. Nat. Struct. Mol. Biol. 2011;20:237–243. doi: 10.1038/nsmb.2466. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Pechmann S, Willmund F, Frydman J. The ribosome as a hub for protein quality control. Mol Cell. 2013;49:411–421. doi: 10.1016/j.molcel.2013.01.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Pedersen M, Nissen S, Mitarai N, Lo Svenningsen S, Sneppen K, Pedersen S. The functional half-life of an mRNA depends on the ribosome spacing in an early coding region. J Mol Biol. 2011;407:35–44. doi: 10.1016/j.jmb.2011.01.025. [DOI] [PubMed] [Google Scholar]
  46. Plotkin JB, Kudla G. Synonymous but not the same: the causes and consequences of codon bias. Nat. Rev. Genet. 2011;12:32–42. doi: 10.1038/nrg2899. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Pop C, Rouskin S, Ingolia NT, Han L, Phizicky EM, Weissman JS, Koller D. Causal signals between codon bias, mRNA structure, and the efficiency of translation and elongation. Mol. Syst. Biol. 2014;10:770. doi: 10.15252/msb.20145524. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Presnyak V, Alhusaini N, Chen YH, Martin S, Morris N, Kline N, Olson S, Weinberg D, Baker KE, Graveley BR, Coller J. Codon optimality is a major determinant of mRNA stability. Cell. 2015;160:1111–1124. doi: 10.1016/j.cell.2015.02.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Rodnina MV, Wintermeyer W. GTP consumption of elongation factor Tu during translation of heteropolymeric mRNAs. Proc. Natl. Acad. Sci U. S. A. 1995;92:1945–1949. doi: 10.1073/pnas.92.6.1945. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Rosano GL, Ceccarelli EA. Rare codon content affects the solubility of recombinant proteins in a codon bias-adjusted Escherichia coli strain. Microb. Cell Fact. 2009;8:41–50. doi: 10.1186/1475-2859-8-41. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Sander IM, Chaney JL, Clark PL. Expanding Anfinsen’s principle: contributions of synonymous codon selection to rational protein design. J. Am. Chem. 2014;136:858–861. doi: 10.1021/ja411302m. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Sauna ZE, Kimchi-Sarfaty C. Understanding the contribution of synonymous mutations to human disease. Nat. Rev. Genet. 2011;12:683–691. doi: 10.1038/nrg3051. [DOI] [PubMed] [Google Scholar]
  53. Schägger H, von Jagow G. Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal. Biochem. 1987;166:368–379. doi: 10.1016/0003-2697(87)90587-2. [DOI] [PubMed] [Google Scholar]
  54. Seras-Franzoso J, Affentranger R, Ferrer-Navarro M, Daura X, Villaverde A, García-Fruitós E. Disulfide bond formation and activation of Escherichia coli β-galactosidase under oxidizing conditions. Appl. Environ. Microbiol. 2012;78:2376–2385. doi: 10.1128/AEM.06923-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Sharp PM, Emery LR, Zeng K. Forces that influence the evolution of codon bias. Philos. Trans. R. Soc. Lond. B Biol. Sci. 2010;365:1203–1212. doi: 10.1098/rstb.2009.0305. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Sharp PM, Li WH. The codon Adaptation Index - a measure of directional synonymous codon usage bias, and its potential applications. Nucleic Acids Res. 1987;15:1281–1295. doi: 10.1093/nar/15.3.1281. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Tsai CJ, Sauna ZE, Kimchi-Sarfaty C, Ambudkar SV, Gottesman MM, Nussinov R. Synonymous mutations and ribosome stalling can lead to altered folding pathways and distinct minima. J. Mol. Biol. 2008;383:281–291. doi: 10.1016/j.jmb.2008.08.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Tu L, Khanna P, Deutsch C. Transmembrane segments form tertiary hairpins in the folding vestibule of the ribosome. J. Mol. Biol. 2014;426:185–198. doi: 10.1016/j.jmb.2013.09.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Waudby CA, Launay H, Cabrita LD, Christodoulou J. Protein folding on the ribosome studied using NMR spectroscopy. Prog. Nucl. Magn. Reson. Spectrosc. 2013;74:57–75. doi: 10.1016/j.pnmrs.2013.07.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Wilson DN, Beckmann R. The ribosomal tunnel as a functional environment for nascent polypeptide folding and translational stalling. Curr. Opin. Struct. Biol. 2011;21:274–282. doi: 10.1016/j.sbi.2011.01.007. [DOI] [PubMed] [Google Scholar]
  61. Woolhead CA, McCormick PJ, Johnson AE. Nascent membrane and secretory proteins differ in FRET-detected folding far inside the ribosome and in their exposure to ribosomal proteins. Cell. 2004;116:725–736. doi: 10.1016/s0092-8674(04)00169-2. [DOI] [PubMed] [Google Scholar]
  62. Yu CH, Dang Y, Zhou Z, Wu C, Zhao F, Sachs MS, Liu Y. Codon usage influences the local rate of translation elongation to regulate co-translational protein folding. Mol Cell. 2015;59:744–754. doi: 10.1016/j.molcel.2015.07.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Zhang G, Ignatova Z. Folding at the birth of the nascent chain: coordinating translation with co-translational folding. Curr. Opin. Struct. Biol. 2011;21:25–31. doi: 10.1016/j.sbi.2010.10.008. [DOI] [PubMed] [Google Scholar]
  64. Zhang G, Hubalewska M, Ignatova Z. Transient ribosomal attenuation coordinates protein synthesis and co-translational folding. Nat. Struct. Mol. Biol. 2009;16:274–280. doi: 10.1038/nsmb.1554. [DOI] [PubMed] [Google Scholar]
  65. Zhou M, Guo J, Cha J, Chae M, Chen S, Barral JM, Sachs MS, Liu Y. Non-optimal codon usage affects expression, structure and function of clock protein FRQ. Nature. 2013;495:111–115. doi: 10.1038/nature11833. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

1
NIHMS750905-supplement-1.pdf (1.7MB, pdf)
2
NIHMS750905-supplement-2.xls (38KB, xls)

RESOURCES