JBC Connect with Cosmo for Collagen Detection

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M701395200 on April 5, 2007

J. Biol. Chem., Vol. 282, Issue 26, 18929-18936, June 29, 2007
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
282/26/18929    most recent
M701395200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bernstein, J. R.
Right arrow Articles by Liao, J. C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bernstein, J. R.
Right arrow Articles by Liao, J. C.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Directed Evolution of Ribosomal Protein S1 for Enhanced Translational Efficiency of High GC Rhodopseudomonas palustris DNA in Escherichia coli*Formula

Jeffrey R. Bernstein{ddagger}§, Thomas Bulter{ddagger}, Claire R. Shen{ddagger}, and James C. Liao{ddagger}§1

From the {ddagger}Department of Chemical and Biomolecular Engineering and §Biomedical Engineering Interdepartmental Program, UCLA, Los Angeles, California 90095

Received for publication, February 16, 2007 , and in revised form, March 16, 2007.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
The expression of foreign DNA in Escherichia coli is important in biotechnological applications. However, the translation of genes from GC-rich organisms is inefficient in E. coli.To overcome this problem, we applied directed evolution to E. coli ribosomal protein S1. Two selected mutants enabled 12- and 8-fold higher expression levels from GC-rich DNA targets. General improvements in translation efficiency over a range of genes from Rhodopseudomonas palustris and E. coli was achieved using an S1 mutant selected against multiple genes from R. palustris. This method opens new opportunities for the expression of GC-rich genes in E. coli.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
The recombinant expression of DNA in Escherichia coli is important for biotechnological applications, enabling the extension of pathways for the production of non-native metabolites with a growing economic and social impact (1–5). Heterologous expression of GC-rich genes in E. coli poses a unique set of challenges, in comparison with other DNA. Overcoming these problems will generate increased flexibility in the production of proteins and metabolites in this host. Although most efforts for recombinant expression in this host are achieved using E. coli promoters and optimal ribosome-binding sites (RBS),2 it would be advantageous to express large pathways, consisting of many genes, without needing to change each promoter and RBS. Such a strategy would enable hybrid functionalities from multiple organisms using technology such as the fusing of two genomes (6) or transformation-associated recombination (7). Here we assess the ability of E. coli to express native DNA, including transcription and translation initiation regions, from the high GC {alpha}-bacterium Rhodopseudomonas palustris (8), and we propose a novel method for enhancing expression of foreign transcripts, altering E. coli ribosomal protein S1 by directed evolution.

A number of studies have examined the ability of E. coli to express high GC DNA. One in vitro study directly linked inefficient translation to the GC content from various organisms (9). Others have discussed implications of secondary structure and its ability to regulate and control the strength of protein expression (10, 11). The destruction of stable stem-loops by site-directed mutagenesis in the region of translation initiation of high GC genes has been shown to enhance protein expression (12). Another method introduced a short leader ORF preceding a high GC gene to improve its expression, and suggested that it is beneficial to include an AT-rich translation termination region in the leader ORF that overlaps with the start codon of the high GC target gene (13). Finally, replacing GC-rich codons from one bacterium with those more commonly used by E. coli enabled a 10-fold increase in expression of a recombinant protein for crystallization purposes (14).

To improve expression from high GC heterologous mRNAs in E. coli, it is important to consider features involved in translation. Upstream of the start codon is the 5'-untranslated region (UTR), which contains the Shine-Dalgarno (SD) sequence that is able to base pair with a complementary sequence on the 3' end of the 16S rRNA (15). However, the SD interaction was found to be nonessential for correct initiation (16, 17), indicating that prokaryotic ribosomes have other capacities for start site selection. Ribosomal protein S1, encoded by rpsA, has been shown to play critical roles in translation initiation and elongation (18, 19), and it has been shown to promote transcriptional cycling in vitro (20). S1 interacts with mRNA upstream of the SD sequence (21), and translation efficiency has been positively correlated to the presence of AU-rich sequences at this position of the UTR (22, 23). The requirement for S1 in translating mRNAs with a weak or no SD sequence has also been detailed (24). S1 is an essential protein required for the translation of most transcripts within E. coli (25), and it contains six repeats of the oligonucleotide-binding (OB) fold, contained in other RNA-binding proteins, which raise the nonspecific affinity of the ribosome to single-stranded RNA (26). In this regard, ribosomal protein S1 is an attractive target for improving the expression of GC-rich genes that cannot be efficiently translated in E. coli. We therefore employed directed evolution methods to S1 to generate S1 variants that can recognize GC-rich sequences.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Media and Chemicals—Luria-Bertani (LB) broth (Difco) was prepared according to the manufacturer's instructions and was supplemented with antibiotics as appropriate. Ampicillin was at 100 µg/µl, kanamycin was at 75 µg/µl, and chloramphenicol (Cm) was at 75 µg/µl. Oligonucleotides were from Invitrogen or from Alpha DNA (Montreal, Canada) and are listed in supplemental Table 1. o-Nitrophenyl-beta-D-galactopyranoside and isopropyl beta-D-thiogalactopyranoside were purchased from Sigma.

Plasmid Constructs—To create plasmid pJRB320, E. coli rpsA under the control of its native promoter was amplified from the genome using KOD Hot Start Polymerase (Novagen). The resulting PCR fragment was cloned pZS*24MCS-1, a low copy plasmid (27). A mutation -2 bp upstream of the rpsA ATG start codon installed a unique BspHI restriction site upstream of the coding sequence. All constructs described herein were confirmed by DNA sequencing (MCLab).

To generate pTB120, part of the lactose repressor gene lacI and the Ptac promoter was removed from pJF118EH (28) and replaced with the E. coli rrnB terminator. The truncated lacZ gene missing its first 9 amino acids from pRS414 (29) was inserted downstream from the terminator. This plasmid allows for protein fusions to the BamHI site at the start of lacZ.To generate pTB130, the CAT gene was amplified from pACYC184 (New England Biolabs) by changing the start codon to CAG (30). The PCR product was inserted into pTB120 replacing the truncated lacZ gene to generate pTB130, which allows for protein fusions to CAT. Protein fusions listed in Table 2 were inserted into pTB120, and fusions to the CAT gene used for selection were inserted into pTB130. pJRB202 was created to study transcriptional fusions to lacZ. The plasmid is derived from pJF118EH (28) and contains an rrnB terminator followed by the PlacO1 promoter and UTR upstream of the complete lacZ gene. The PlacO1 promoter can be removed using XmaI and EcoRI restriction sites and replaced with the promoter and UTR sequence from another gene, maintaining 27 bp of sequence from expression vector UTR upstream of the ATG start codon, including the SD region. Transcriptional fusions listed in Table 2 were cloned into this vector, and the pJRB202 UTR is listed in supplemental Table 2.


View this table:
[in this window]
[in a new window]

  		TABLE 2
beta-Galactosidase expression from E. coli cells containing plasmids with R. palustris, P. aeruginosa, and E. coli leader sequences fused to the lacZ gene

 
To create pJRB40xba used for the low copy protein fusions of R. palustris and E. coli UTRs and badD sequence, the PlacO1 promoter was placed into the pZS*24 vector (27), and the truncated lacZ from pTB120 was amplified by PCR and cloned downstream of PlacO1 to create pJRB40. An XbaI cut site was added to this construct upstream of the transcription start site using oligonucleotides annealed and cloned into Xho and EcoRI restriction sites.

beta-Galactosidase Measurements—beta-Galactosidase assays were performed in triplicate for each sample using flat bottom 96-well polypropylene plates (Grenier). 20 µl of overnight culture was diluted in 280 µl of Z-Buffer, and subsequently 4 µl of 0.1% SDS and 8 µl of chloroform were added. Plates were covered with aluminum seal film (Corning Glass) and shaken at 1400 rpm for 90 s in a plate incubator/shaker (Eppendorf). After centrifugation, 50 µl of 4 mg/ml o-nitrophenyl-beta-D-galactopyranoside (dissolved in Z-buffer) was added to each well, and the beta-galactosidase activity was measured in a plate reader (Bio-Tek) at 420 nm for ~5 min at 30 °C. Slopes from the linear portion of the curve were obtained using KC Junior software. Activity was determined as described previously (31).

badD UTR Comparison—To compare various features of the BadD-LacZ protein fusion, we used either the 47 bp preceding the badD gene (Rpal UTR) or the 47 bp of the PlacO1 expression vector (E. coli UTR). Additionally, we compared either the first 19 native R. palustris badD codons fused to lacZ or the same amino acid sequence optimized for the most commonly used E. coli codons (32). Finally, to test the differences in the SD sequence between the R. palustris and PlacO1 expression vector UTRs, we replaced the 14 bp preceding the ATG start from the badD UTR with the corresponding sequence from the PlacO1 expression vector (Ec RBS). Constructs were generated by annealing complementary oligonucleotides with restriction site overhangs in 1x annealing buffer (10 mM Tris, pH 8.0, 50 mM NaCl, 1 mM EDTA) and slowly cooling to room temperature from 95 °C. Annealed oligonucleotides were then diluted in ligation buffer (Roche Applied Science) and incubated with T4 polynucleotide kinase (New England Biolabs) and were ligated into the appropriate vector generated from pJRB40xba. To test the expression from these clones, constructs were transformed to BW25113-Z1, and two independent colonies per construct were grown overnight in LB plus 0.2% glucose. The following day, 15 µl were used to inoculate 3 ml of LB plus 0.2% glucose with 1 mM isopropyl beta-D-thiogalactopyranoside. Cells were grown at 37 °C for 4 h until the A600 was 0.4 to 0.9, and cells were then placed on ice and used in a beta-galactosidase assay as described above.

Error-prone PCR—Mutagenic PCR was performed using the Gene Morph II mutagenesis kit (Stratagene). 350 ng of plasmid pJRB320 or mutant variant was used as template, and the reaction was performed according to the manufacturer's protocol to obtain medium mutagenesis rates of ~20-fold amplification of the target sequence.

Cloning and Mutant Library Generation—The mutant PCR product was purified using agarose gel electrophoresis and QIAQuick PCR purification kit (Qiagen) and digested with appropriate restriction enzymes (New England Biolabs). Remaining parent template was removed using DpnI enzyme, and the purified digested product was ligated into an open vector of pJRB320 using Rapid DNA ligation kit (Roche Applied Science). 10 µl of the ligation reaction was used to transform 100 µl of XL10 Gold Ultracompetent cells (Stratagene) according to the manufacturer's protocol. The resultant library size was calculated (~8,000–12,000 colonies), and clones were washed from each plate using 2.8 ml of LB medium and a sterile hockey stick. The washed library was vortexed miniprepped using QIAprep Spin Miniprep columns (Qiagen) and eluted using 1/10 concentration elution buffer. The library was subsequently concentrated under vacuum (Labconco).

Mutant Selection and Screening—Approximately 125 ng of both rpsA library plasmid and selection plasmid containing the R. palustris DNA sequence fused to CAT (derived from pTB130) were electroporated into E. coli BW25113. Cells were plated onto LB containing appropriate antibiotics and variable Cm concentrations to restrict background growth of the library. The transformation efficiency was estimated using LB plates lacking Cm, with a target of ~10,000 clones tested per selection plate. Approximately 20–24 potential positive clones were identified 48 h post-transformation. Cultures were miniprepped (Qiagen) and eluted in 1/10 elution buffer and dried to ~5 µl. 2 µl of the potential positive plasmid preparation were digested using NotI (New England Biolabs), which cuts the CAT fusion selection plasmid without digesting the rpsA mutant plasmid. The reaction was heat-inactivated, and 2 µl was used to transform BW25113 with addition of 1 µl of the reporter plasmid (~100 ng/µl) containing the R. palustris protein fusion to lacZ. beta-Galactosidase assays were performed on these cells as stated above. The best new mutant identified by beta-galactosidase assay was miniprepped and transformed into E. coli Top 10 (Invitrogen) using kanamycin selection to remove the reporter plasmid (confirmed by ampicillin susceptibility), and template was isolated for subsequent rounds of mutagenesis and DNA sequencing. Two independent colonies of each positive mutant with the appropriate protein fusion were tested in triplicate using the beta-galactosidase assay, and the results reported here are the average of these data.

Minimum Inhibitory Concentration (MIC) Assay—Overnight 3 ml of LB cultures containing wild type or mutant rpsA and BadR-CAT protein fusion were diluted 20-fold in fresh LB medium and grown to an A600 of 2.0. Cells were chilled on ice and then diluted in LB to a concentration of 107 cells/ml, and 1 µl of this stock was then spotted onto plates with increasing Cm concentrations ranging from 5 to 45 µg/µl. Increments of 1 µg/µl were used below 10 µg/µl, then increments of 2.5 µg/µl were used to 25 µg/µl, and finally 5 µg/µl increases in antibiotic concentration were used for the remaining plates. The MIC reported is the lowest concentration of Cm that restricted growth of a given construct after 16 h at 37 °C.

Time Course Study and Quantitative PCR—60 ml of LB cultures were inoculated (1%) with overnight BW25113 transformed with the BadR-LacZ protein fusion along with either wild type, 4R, or 5R evolved rpsA constructs and grown in baffled Erlenmeyer shake flasks at 37 °C and at 200 rpm. Replicate cultures of each clone were performed. Samples were removed and chilled on ice prior to measuring the A600 or beta-galactosidase activity. Samples removed during mid-exponential phase (A600 = 0.49–0.75) were pelleted and frozen in a dry ice/ethanol bath and stored at -80 °C until RNA extraction by RNeasy mini kit (Qiagen) following the manufacturer's protocol. Samples were diluted in 1x DNase I buffer (Ambion) and treated with 0.5 µl of RNase OUT (Invitrogen) and 0.8 µl of DNase I (139 units/µl; Invitrogen). Samples were purified using acid/phenol extraction and ethanol precipitation. First strand cDNA synthesis was performed according to protocol (Invitrogen) with 150 ng of random hexamers and 150 ng of total E. coli RNA and 1 µl of SSIII reverse transcriptase. Reactions were diluted 3-fold with water upon completion. Real time PCR was performed with gene-specific PCR primers designed using MyPROBES software (33) for lacZ and chaA, a housekeeping gene involved in potassium, sodium, and calcium ion regulation within the cell. The reactions were conducted on a Smart Cycler (Cepheid) with 2 µl of cDNA added per 25 µl of real time PCR sample using the QuantiTect SYBR green PCR kit (Qiagen). A four-step program consisting of denaturation, annealing, extension, and data acquisition was used, and melt curves verified that only one species of DNA product was amplified using the primer pairs. Calibration curves were performed for each primer pair using genomic DNA template amounts over 4 orders of magnitude, and curves with linearity R2 values of at least 0.99 were used to determine the amount of lacZ transcript relative to the chaA control.


Figure 1
View larger version (31K):
[in this window]
[in a new window]

  		FIGURE 1.
Expression determinants of badD mRNA in E. coli. a, expression of badD mRNA transcribed from the PlacO1 promoter using either the 47 bp preceding the R. palustris badD gene (Rpal UTR), the 47 bp UTR from the PlacO1 construct (Ec UTR), the first 19 codons of badD sequence from the native R. palustris gene (Rpal badD), or the same 19 amino acids optimized for E. coli codon usage (Ec badD). *, the last construct replaces the 14 bp preceding the R. palustris ATG start codon with the corresponding 14 bp from the PlacO1 expression vector UTR (Ec RBS). Red is used for the R. palustris sequence, and tan is used for E. coli. ND denotes the lacZ data that was not detected. The badD codons are fused to a LacZ variant missing its first nine amino acids (29). beta-gal, beta-galactosidase. b, sequences of the 47 bp preceding the badD gene from R. palustris and the 47-bp UTR from the PlacO1 expression construct (27) are displayed, including the ATG start codon (boldface type).

 

   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
The 5'-UTR of a GC-rich Gene Is Responsible for Translational Deficiency—To understand the elements of the high GC R. palustris mRNAs that are responsible for reduced translational efficiency in E. coli, we selected the badD gene from the well characterized benzoate degradation pathway for further study (34). Constructs were generated that used the 47-bp DNA sequence preceding the start codon of badD and compared it with the UTR of equal length from the transcriptional start site of the PlacO1 expression vector (Fig. 1). Additionally, we compared differences in the beginning coding sequence of badD fused to lacZ, by using either native or codon-optimized DNA sequence (32). Finally, we designed a construct with R. palustris UTR and badD codons, but we replaced the 14 bp preceding the ATG with the corresponding 14 bp from the PlacO1 expression vector UTR to determine the effect of an optimized SD sequence on activity. The results of this experiment confirm that the R. palustris UTR causes a translational deficiency such that no activity can be measured regardless of whether using the native coding sequence or one optimized for E. coli codons (Fig. 1). These results complement an earlier study indicating that AT-rich sequences derived from the antiterminator of the E. coli rrn operons enhanced translation up to ~3.6-fold when placed upstream of a strong SD (22). Optimizing the beginning coding sequence from R. palustris to E. coli codons enables an ~2-fold increase in expression with the constructs containing the PlacO1 expression vector UTR. Finally, the effect of switching the 14 bp preceding the ATG start site from the R. palustris UTR with the corresponding region containing the SD sequence from the PlacO1 expression construct allows for a very small amount of beta-galactosidase expression, showing that the presence of a strong RBS alone is not enough to rescue the translational deficiency of R. palustris transcripts in E. coli.


Figure 2
View larger version (23K):
[in this window]
[in a new window]

  		FIGURE 2.
Evolutionary scheme and amino acid mutations accumulated during the directed evolution of ribosomal protein S1 to R. palustris constructs. The beta-galactosidase (beta-gal) activity of wild type and rpsA mutants toward LacZ protein fusions of the badR (a) and crtI (b) genes is shown. The starting point for crtI evolution was chosen from the 4th generation mutant of badR evolution (4R). Amino acid mutations accumulated at each round of directed evolution are listed above the activity bar. c, the evolutionary path and accumulated mutations of an S1 variant that enhances expression from multiple DNA targets is shown. Protein fusions of the indicated genes include their native promoter, UTR, and first ~20 codons fused to CAT. The BadK-LacZ protein fusion contains the promoter, UTR, and each of the indicated ORFs, with the CAT protein fusion to the final badK gene. *, indicates the DNA sequence of the indicated gene is truncated after ~20 codons and fused to LacZ.

 
Directed Evolution of rpsA—Once the translational deficiency of foreign transcripts was determined, we sought to alleviate this problem by directed evolution. We employed a pool of modified rpsA constructs in which the sequences of rpsA were mutagenized by error-prone PCR. These rpsA variants were expressed from their native promoter on a low copy plasmid to avoid growth deficiency because of overexpression (35), and the chromosomal copy of rpsA was maintained. An in vivo selection system was designed in which mutants of the rpsA gene were co-expressed in the presence of a protein fusion of the R. palustris promoter UTR, and beginning coding sequence to the CAT gene conferring resistance to this antibiotic. The growth rate of E. coli cells harboring these two plasmids incubated in media containing Cm reflects the level of CAT expression that in turn depends on the efficacy of expression from the GC-rich UTR and gene leader sequence. The selection pressure was increased after each round of directed evolution by increasing the concentration of Cm. After selection toward the badR fusion construct, two S1 mutants, 4R and 5R, enabled growth to MIC values of 25 and 35 µg/µl Cm, whereas the wild type S1 construct allowed growth to an MIC Cm value of 7.5 µg/µl, indicating our successful isolation of rpsA variants that enhanced expression from the badR construct.

To further improve our selection of rpsA variants with enhanced activity toward high GC templates, we applied a beta-galactosidase screen to our selection method. Multiple positive clones were chosen after selection on Cm plates, and the plasmids harboring the mutant rpsA was then employed in a screen using a fusion of the R. palustris gene sequence to lacZ. Positive clones were screened in parallel for beta-galactosidase activity alongside the parent clone to identify the mutant with the highest activity toward its target gene. This mutant was then subjected to a subsequent round of mutagenesis, selection, and screening to continue improving translation efficiency of the R. palustris DNA sequence. Five rounds of selection and screening were performed, resulting in a mutant (5R) that enabled 12.0-fold higher beta-galactosidase expression than the wild type S1 (Fig. 2a). To further investigate the ability to enhance expression of a different R. palustris gene, we used the fourth generation rpsA mutant (4R), evolved toward badR, as a starting point for increasing expression of the CrtI protein fusion. This 4R clone exhibited 2-fold higher activity toward the crtI construct than the wild type S1. After three rounds of selection, a mutant (3C) was isolated, which increased activity from the LacZ fusion by 8.0-fold (Fig. 2b). We attempted a fourth round of selection toward the crtI construct, but we were unable to isolate further enhanced variants, suggesting that the protein may be saturated for mutations that enhance CrtI-LacZ expression or that it had reached a local maximum because of previously accumulated mutations.

To remove the possibility that selected variants are specialized to translate only one DNA sequence, we adopted a strategy to evolve rpsA against multiple gene targets (Fig. 2c). We used an S1 mutant (4R) as the starting point for evolution toward the BadD protein fusion for one round of selection and screening. This clone (5D) was then selected and screened four additional rounds toward a protein fusion of BadK, the final gene in the badH–K operon, including the four genes upstream of badK expressed from their native promoter and UTR (34).

Sequence Analysis of the Selected Variants—After each round of selection and screening, we sequenced the highest activity clone to determine newly acquired mutations. On average, we observed 1–3 new coding mutations per round of directed evolution (Fig. 2). The 5th round mutant evolved toward badR accumulated 10 mutations, of which 5 mutations map to the 3rd and 4th repeats of the S1 OB fold domain, whereas other mutations occurred in the 1st, 5th, or 6th domains. After three rounds of evolution toward the CrtI protein fusion, an additional four unique mutations were accumulated, three of which map to the 3rd and 4th repeats of the OB fold domain. The final mutant evolved towards multiple R. palustris constructs (9K) collected an additional 10 amino acid mutations unique from the parent clone (4R). Four mutations were lost during the creation of S1 mutant 9K, including a Q261H mutation occurring in the first round of evolution toward badK, which was again mutated in the 3rd round of selection toward this construct (H261L). Additionally, three mutations (A471V, Y477D, and A530T) occurring in the 6th repeat of the S1 domain were lost during the second round of mutation toward badK in which a frameshift mutation occurred at 1,351 bp, creating an S1 variant missing the final repeat of its S1 domain.

The Truncated S1 Mutant Improves Expression from Its RNA Target—One S1 clone (7K) that was selected in the 2nd round of selection toward the badK operon contained a frameshift mutation (Fig. 2c). To establish whether this frameshift mutation was truly beneficial, we repaired the mutation adding on the 101 amino acid tail from the parent clone of 7K to both the 7K and 9K mutants from the 2nd and 4th round of selection toward the badK construct. In both cases the expression from the BadK-LacZ fusion was higher when expressed with the mutant rpsA containing the frameshift compared with the corresponding mutant with the repaired C terminus, indicating that this mutation was beneficial for enhanced expression from the BadK operon fusion.

beta-Galactosidase Rate and mRNA Levels Indicate S1 Mutants Enhance Both Transcript Abundance and Protein Expression—To further understand the ability of rpsA mutants to enhance expression from various R. palustris DNA sequences, we monitored beta-galactosidase expression over time for the wild type S1 and two mutants isolated in the evolution toward badR (Fig. 3). Although the rpsA mutants display somewhat reduced growth rates (exponential growth doubling times of 44.6 and 49.3 min for 4R and 5R versus 38.4 min for wild type), the mutant S1 constructs allow greater expression from the foreign DNA throughout all stages of growth. We also measured transcript levels by quantitative PCR. The results (Table 1) show that there is approximately a 2.8- and 5.6-fold difference in lacZ transcript abundance levels between cells expressing the two mutants versus the wild type rpsA. These results are consistent with previous work that showed a positive correlation between enhanced translation because of changes in the UTR and increased lacZ mRNA stability (23).


View this table:
[in this window]
[in a new window]

  		TABLE 1
Normalized lacZ mRNA abundance and beta-galactosidase activity from cells expressing the badR::lacZ protein fusion and rpsA constructs Numbers in parentheses indicate fold change versus wild type for the quantitative PCR assay.

 
"Generalist" Mutant Improves Expression from Multiple Constructs—The resulting mutant from our multiple gene evolutionary strategy (9K) was tested versus the wild type rpsA construct in combination with a variety of LacZ protein fusions to genes from R. palustris, Pseudomonas aeruginosa, and E. coli (Table 2). The 9K clone enhanced expression from multiple constructs, including those toward which it was evolved (badD, badR, aliB, and badK) and from constructs to which it had not been exposed during the selection process. Interestingly, this clone enhanced translation from the high GC P. aeruginosa omlA construct by ~7-fold and also improved the translation efficiency of a number of E. coli constructs, indicating that the mutations accumulated during evolution toward R. palustris DNA allow for varying improvements in translation efficiency from a variety of DNA sequences.


Figure 3
View larger version (12K):
[in this window]
[in a new window]

  		FIGURE 3.
beta-Galactosidase activity from the BadR::LacZ protein fusion in the presence of various rpsA constructs. Expression was monitored over time for the wild type (WT)(blue diamonds) and mutant rpsA constructs 4R (orange squares) and 5R (purple circles).

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
Ribosomal protein S1 consists of six repeats of a 70-residue domain known as the OB fold for its ability to bind single-stranded nucleic acids (26). The structure of this S1 RNA-binding domain present in the PNPase of E. coli was determined by NMR (36) and has been used to shed light on the RNA binding capacity of S1 (26). Alignment of multiple S1 OB fold domains revealed seven highly conserved amino acids proposed to be involved in mRNA binding (26). When analyzing the location of mutations in the 5R and 3C rpsA variants evolved toward badR and crtI, 8 of 14 unique mutations map to the 3rd and 4th repeats of the OB fold (Fig. 4). It is possible that these two domains represent a hot spot of mutation enhancing expression from these foreign constructs. Multiple mutations occur at a region of the 4th repeat on or within four amino acids of the conserved His-34 residue believed to be involved in RNA binding. Similarly, three mutations occurring in the 3rd OB fold repeat are adjacent to either the Asp-64, Arg-68, or Lys-74 implicated in nucleic acid binding, whereas another residue five amino acids downstream from Lys-74 was twice mutated (residue numbering is from the PNPase S1 domain structure paper (36)).


Figure 4
View larger version (34K):
[in this window]
[in a new window]

  		FIGURE 4.
Location of mutations accumulated during the directed evolution of rpsA. a, diagram of the S1 protein and the six repeats of the OB fold (blue boxes with numbers denote the start and finish of each domain). Locations of the unique mutations in three mutants (5R, 3C, and 9K) are marked with black arrows, and those carried over from a previous template are marked with gray arrows. b, alignment of the 3rd and 4th repeats of the OB fold within S1 to the same domain sequenced from PNPase (36). Highly conserved residues believed to be involved in RNA binding (26) are in red are highlighted gray in the PNPase sequence. Unique mutations from the badR evolution (5R) are highlighted in blue boxes, and those from the crtI evolution (3C) are orange, and those from generalist mutant 9K are green. The new amino acid appears above each highlighted mutation.

 
The frameshift mutation present in the generalist clone collected in the second round of selection toward the BadK protein fusion results in a truncated S1 protein similar in length to another mutant reported in the literature (37). It was shown that the missing sixth repeat of the OB fold domain from this truncated mutant participated in the translational auto-regulation of S1 protein synthesis but was not required for translation initiation of other proteins. Another group has shown that S1 variants truncated in the final two C-terminal OB fold domains are incapable of stimulating transcriptional cycling in vitro, but a mutant with a disruption in its final C-terminal domain was only slightly impaired in this property (20). Additionally, it has also been demonstrated that the N-terminal portion of S1, consisting of the first two repeats of the OB fold, is responsible for binding to the 30S subunit of the ribosome (38), and the first three repeats together bind a model mRNA with high efficiency (35). The fact that many of the mutations resulting in enhanced expression of GC-rich DNA reside in repeats 3 and 4 of the OB fold further implicates the involvement of these domains in mRNA binding in vivo. The S1 mutant evolved toward multiple high GC DNA targets improved expression from other GC-rich constructs that were not part of its evolutionary scheme and also enhanced expression from a number of E. coli constructs with more balanced GC content. Future experiments will enable the characterization of binding efficiencies of S1 mutants to different mRNA targets.

In the comparison of BadD protein fusions with either UTRs from E. coli or R. palustris, and native or optimized codon sequence, we show that in this case codon usage of the beginning ORF sequence is unable to account for the low expression of R. palustris genes in E. coli (Fig. 1). This was also the case when we co-expressed a plasmid containing seven rare tRNAs alongside multiple R. palustris protein fusions (data not shown). The 2-fold increase in expression between the construct with native R. palustris badD codons and those optimized for expression in E. coli is potentially because of the reduction in GC content of this stretch of DNA from 64.9 to 61.4%. The R. palustris UTR for badD has a particularly high GC content (83% in the 47 bp preceding the start codon), whereas the UTR from the PlacO1 construct we used as a comparison has 46.8% GC content. It is possible that the high GC content of the R. palustris genome presents a significant challenge to E. coli when initiating translation from its mRNAs.

Although the directed evolution of ribosomal protein S1 improves the translational efficiency of expression from these foreign DNA sequences, we recognize that other proteins and perhaps also the RNA core of the ribosome are likely to influence the expression of foreign DNA in the cell. Other proteins such as translation initiation factors IF1, IF2, and IF3 certainly play a role in translation initiation (11, 39, 40), and we cannot rule out the possibility of accessory proteins in E. coli or from a donor organism that may be used to stabilize mRNAs or prepare them for efficient translation. We have considered that many high GC bacteria have optimal growth rates near 30 °C, yet they have evolved to express their high GC mRNAs at temperatures that would allow for extensive formation of secondary structures within transcripts.

The inability of E. coli to efficiently translate high GC mRNAs has evolutionary implications pertaining to horizontal gene transfer. Recently, an example of genetic silencing of foreign DNA was detailed in Salmonella where the H-NS protein was shown to transcriptionally repress genes with lower GC content than that of the genome average (41). The authors showed that a low GC gene integrated to the genome of Salmonella was expressed >15-fold higher in an hns mutant compared with the wild type strain, and propose that the bacterium has evolved a mechanism to defend against the expression of foreign DNA. In this case, the necessity to guard against invading DNA is explicitly directed toward the transcription of low GC DNA as a means of distinguishing "self" from "foreign." Our data show that impediments to the expression of GC-rich DNA in E. coli are likely to reside at the translational level, obviating the need for transcriptional repression.

A group recently reported the stable cloning of the Synechocystis PCC6803 genome into Bacillus subtilis (6); however, the resultant strain had growth limitations and was unable to stably maintain the Synechocystis ribosomal RNA operon, possibly because of the increased expression of proteins toxic to the cell. Their work highlights both the growing ease at which large clusters of DNA can be transferred among bacteria and the difficulty of expressing DNA from distantly related species using unaltered transcription and translation signals. R. palustris exhibits an array of metabolically desirable reactions such as photosynthesis, nitrogen fixation, carbon fixation, and aromatic degradation, which are often clustered in discrete chromosomal loci (8). The directed evolution of rpsA for enhancing the expression of GC-rich DNA in E. coli is a tool that when combined with other techniques will enable the transfer of new metabolic phenotypes from GC-rich organisms, allowing hybrid functionalities for biotechnological applications.


   FOOTNOTES
 
* This work was supported by NIGMS Grants 1R01GM077625-01 and DOE DE-FG02-01ER63241 from the National Institutes of Health and UCLA-DOE Institute of Genomics and Proteomics. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

Formula The on-line version of this article (available at http://www.jbc.org) contains supplemental Tables 1 and 2. Back

1 To whom correspondence should be addressed: 5531 Boelter Hall, 420 Westwood Plaza, Los Angeles, CA 90095. Tel.: 310-825-1656; Fax: 310-206-4107; E-mail: liaoj{at}ucla.edu.

2 The abbreviations used are: RBS, ribosome-binding site; Km, kanamycin; Cm, chloramphenicol; SD, Shine-Dalgarno; MIC, minimum inhibitory concentration; OB fold, oligonucleotide-binding fold; ORF, open reading frame; PNPase, polynucleotide phosphorylase; CAT, chloramphenicol acetyltransferase; UTR, untranslated region. Back


   ACKNOWLEDGMENTS
 
We thank Shota Atsumi for advice during preparation of the manuscript and for contributing to the expression study of BadD protein fusion variants.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Bulter, T., Bernstein, J. R., and Liao, J. C. (2003) Biotechnol. Bioeng. 84, 815-821[CrossRef][Medline] [Order article via Infotrieve]
  2. Tong, I. T., Liao, H. H., and Cameron, D. C. (1991) Appl. Environ. Microbiol. 57, 3541-3546[Abstract/Free Full Text]
  3. Misawa, N., Satomi, Y., Kondo, K., Yokoyama, A., Kajiwara, S., Saito, T., Ohtani, T., and Miki, W. (1995) J. Bacteriol. 177, 6575-6584[Abstract/Free Full Text]
  4. Martin, V. J., Pitera, D. J., Withers, S. T., Newman, J. D., and Keasling, J. D. (2003) Nat. Biotechnol. 21, 796-802[CrossRef][Medline] [Order article via Infotrieve]
  5. Wang, C., Oh, M. K., and Liao, J. C. (2000) Biotechnol. Prog. 16, 922-926[CrossRef][Medline] [Order article via Infotrieve]
  6. Itaya, M., Tsuge, K., Koizumi, M., and Fujita, K. (2005) Proc. Natl. Acad. Sci. U. S. A. 102, 15971-15976[Abstract/Free Full Text]
  7. Leem, S. H., Noskov, V. N., Park, J. E., Kim, S. I., Larionov, V., and Kouprina, N. (2003) Nucleic Acids Res. 31, e29[Abstract/Free Full Text]
  8. Larimer, F. W., Chain, P., Hauser, L., Lamerdin, J., Malfatti, S., Do, L., Land, M. L., Pelletier, D. A., Beatty, J. T., Lang, A. S., Tabita, F. R., Gibson, J. L., Hanson, T. E., Bobst, C., Torres, J. L., Peres, C., Harrison, F. H., Gibson, J., and Harwood, C. S. (2004) Nat. Biotechnol. 22, 55-61[CrossRef][Medline] [Order article via Infotrieve]
  9. Voges, D., Watzele, M., Nemetz, C., Wizemann, S., and Buchberger, B. (2004) Biochem. Biophys. Res. Commun. 318, 601-614[CrossRef][Medline] [Order article via Infotrieve]
  10. Kozak, M. (2005) Gene (Amst.) 361, 13-37[CrossRef][Medline] [Order article via Infotrieve]
  11. Laursen, B. S., Sorensen, H. P., Mortensen, K. K., and Sperling-Petersen, H. U. (2005) Microbiol. Mol. Biol. Rev. 69, 101-123[Abstract/Free Full Text]
  12. Ishida, M., and Oshima, T. (1994) J. Bacteriol. 176, 2767-2770[Abstract/Free Full Text]
  13. Ishida, M., and Oshima, T. (2002) J. Biochem. (Tokyo) 132, 63-70[Abstract/Free Full Text]
  14. Heinz, C., Karosi, S., and Niederweis, M. (2003) J. Chromatogr. B. Analyt. Technol. Biomed. Life Sci. 790, 337-348[Medline] [Order article via Infotrieve]
  15. Shine, J., and Dalgarno, L. (1975) Nature 254, 34-38[CrossRef][Medline] [Order article via Infotrieve]
  16. Calogero, R. A., Pon, C. L., Canonaco, M. A., and Gualerzi, C. O. (1988) Proc. Natl. Acad. Sci. U. S. A. 85, 6427-6431[Abstract/Free Full Text]
  17. Melancon, P., Leclerc, D., Destroismaisons, N., and Brakier-Gingras, L. (1990) Biochemistry 29, 3402-3407[CrossRef][Medline] [Order article via Infotrieve]
  18. Tzareva, N. V., Makhno, V. I., and Boni, I. V. (1994) FEBS Lett. 337, 189-194[CrossRef][Medline] [Order article via Infotrieve]
  19. Potapov, A. P., and Subramanian, A. R. (1992) Biochem. Int. 27, 745-753[Medline] [Order article via Infotrieve]
  20. Sukhodolets, M. V., Garges, S., and Adhya, S. (2006) RNA (Cold Spring Harbor) 12, 1505-1513
  21. Sengupta, J., Agrawal, R. K., and Frank, J. (2001) Proc. Natl. Acad. Sci. U. S. A. 98, 11991-11996[Abstract/Free Full Text]
  22. Komarova, A. V., Tchufistova, L. S., Supina, E. V., and Boni, I. V. (2002) RNA (Cold Spring Harbor) 8, 1137-1147
  23. Komarova, A. V., Tchufistova, L. S., Dreyfus, M., and Boni, I. V. (2005) J. Bacteriol. 187, 1344-1349[Abstract/Free Full Text]
  24. Farwell, M. A., Roberts, M. W., and Rabinowitz, J. C. (1992) Mol. Microbiol. 6, 3375-3383[CrossRef][Medline] [Order article via Infotrieve]
  25. Sorensen, M. A., Fricke, J., and Pedersen, S. (1998) J. Mol. Biol. 280, 561-569[CrossRef][Medline] [Order article via Infotrieve]
  26. Draper, D. E., and Reynaldo, L. P. (1999) Nucleic Acids Res. 27, 381-388[Abstract/Free Full Text]
  27. Lutz, R., and Bujard, H. (1997) Nucleic Acids Res. 25, 1203-1210[Abstract/Free Full Text]
  28. Furste, J. P., Pansegrau, W., Frank, R., Blocker, H., Scholz, P., Bagdasarian, M., and Lanka, E. (1986) Gene (Amst.) 48, 119-131[CrossRef][Medline] [Order article via Infotrieve]
  29. Simons, R. W., Houman, F., and Kleckner, N. (1987) Gene (Amst.) 53, 85-96[CrossRef][Medline] [Order article via Infotrieve]
  30. Prats, A. C., Vagner, S., Prats, H., and Amalric, F. (1992) Mol. Cell. Biol. 12, 4796-4805[Abstract/Free Full Text]
  31. Arvidson, D. N., Youderian, P., Schneider, T. D., and Stormo, G. D. (1991) BioTechniques 11, 733-738[Medline] [Order article via Infotrieve]
  32. Grote, A., Hiller, K., Scheer, M., Munch, R., Nortemann, B., Hempel, D. C., and Jahn, D. (2005) Nucleic Acids Res. 33, 526-531[CrossRef]
  33. Rohlin, L., Trent, J. D., Salmon, K., Kim, U., Gunsalus, R. P., and Liao, J. C. (2005) J. Bacteriol. 187, 6046-6057[Abstract/Free Full Text]
  34. Egland, P. G., and Harwood, C. S. (1999) J. Bacteriol. 181, 2102-2109[Abstract/Free Full Text]
  35. McGinness, K. E., and Sauer, R. T. (2004) Proc. Natl. Acad. Sci. U. S. A. 101, 13454-13459[Abstract/Free Full Text]
  36. Bycroft, M., Hubbard, T. J., Proctor, M., Freund, S. M., and Murzin, A. G. (1997) Cell 88, 235-242[CrossRef][Medline] [Order article via Infotrieve]
  37. Boni, I. V., Artamonova, V. S., and Dreyfus, M. (2000) J. Bacteriol. 182, 5872-5879[Abstract/Free Full Text]
  38. Subramanian, A. R. (1983) Prog. Nucleic Acids Res. Mol. Biol. 28, 101-142[Medline] [Order article via Infotrieve]
  39. Grill, S., Moll, I., Hasenohrl, D., Gualerzi, C. O., and Blasi, U. (2001) FEBS Lett. 495, 167-171[CrossRef][Medline] [Order article via Infotrieve]
  40. Brodersen, D. E., and Nissen, P. (2005) FEBS J. 272, 2098-2108[CrossRef][Medline] [Order article via Infotrieve]
  41. Navarre, W. W., Porwollik, S., Wang, Y., McClelland, M., Rosen, H., Libby, S. J., and Fang, F. C. (2006) Science 313, 236-238[Abstract/Free Full Text]
  42. Lessard, I. A., Pratt, S. D., McCafferty, D. G., Bussiere, D. E., Hutchins, C., Wanner, B. L., Katz, L., and Walsh, C. T. (1998) Chem. Biol. 5, 489-504[CrossRef][Medline] [Order article via Infotrieve]
  43. Datsenko, K. A., and Wanner, B. L. (2000) Proc. Natl. Acad. Sci. U. S. A. 97, 6640-6645[Abstract/Free Full Text]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 J. Biol. Chem.Home page
P. Aliprandi, C. Sizun, J. Perez, F. Mareuil, S. Caputo, J.-L. Leroy, B. Odaert, S. Laalami, M. Uzan, and F. Bontems
S1 Ribosomal Protein Functions in Translation Initiation and Ribonuclease RegB Activation Are Mediated by Similar RNA-Protein Interactions: AN NMR AND SAXS ANALYSIS
J. Biol. Chem., May 9, 2008; 283(19): 13289 - 13301.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
282/26/18929    most recent
M701395200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bernstein, J. R.
Right arrow Articles by Liao, J. C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bernstein, J. R.
Right arrow Articles by Liao, J. C.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS