An encoded N-terminal extension results in low levels of heterologous protein production in Escherichia coli
© Orchard and Goodrich-Blair; licensee BioMed Central Ltd. 2005
Received: 14 June 2005
Accepted: 21 July 2005
Published: 21 July 2005
The tdk gene (encoding deoxythymidine kinase) of the gamma-proteobacterium Xenorhabdus nematophila has two potential translation start sites. The promoter-distal start site was predicted to be functional based on amino acid sequence alignment with closely related Tdk proteins. However, to experimentally determine if either of the two possible start codons allows production of a functional Tdk, we expressed the "long-form" (using the promoter-proximal start codon) and "short-form" (using the promoter-distal start codon) X. nematophila tdk genes from the T7 promoter of the pET-28a(+) vector. We assessed Tdk production and activity using a functional assay in an Escherichia coli tdk mutant, which, since it lacks functional Tdk, is able to grow in 5-fluorodeoxyuridine (FUdR)-containing medium.
Short-form Tdk complemented the E. coli tdk mutant strain, resulting in FUdR sensitivity of the strain. However, the E. coli tdk mutant expressing the long form of tdk remained FUdR resistant, indicating it did not have a functional deoxythymidine kinase enzyme. We report that long-form Tdk is at least 13-fold less abundant than short-form Tdk, the limited protein produced was as stable as short-form Tdk and the long-form transcript was 1.7-fold less abundant than short-form transcript. Additionally, we report that the long-form extension was sufficient to decrease heterologous production of a different X. nematophila protein, NilC.
We conclude that the difference in the FUdR growth phenotype between the E. coli tdk mutant carrying the long-or short-form X. nematophila tdk is due to a difference in Tdk levels. The lower long-form protein level does not result from protein instability, but instead from reduced transcript levels possibly combined with reduced translation efficiency. Because the observed effect of the encoded N-terminal extension is not specific to Tdk production and can be overcome with induction of gene expression, these results may have particular relevance to researchers attempting to limit production of toxic proteins under non-inducing conditions.
Proteins from one organism are often expressed in a different species for the purpose of protein purification or complementation studies. When such efforts fail due to non-production of the protein, the underlying cause of failure is often unclear . Protein overproduction is known to induce a heat shock-like response, which results in increased proteolysis in the cell and therefore possible degradation of the desired protein . Other factors such as degradation of the RNA transcript, efficiency of translation and toxic nature of the desired protein may also influence the level of protein production. Thus, common Escherichia coli strains used for protein overproduction include protease mutants (e.g. BL21; lon), RNase mutants (e.g. BL21 Star™; RNaseE mutant strain; Invitrogen, Carlsbad, CA) and those that provide tRNA synthetases corresponding to infrequently used codons (e.g. Rosetta™; Novagen, Madison, WI) to increase translational efficiency. To reduce production of potentially toxic proteins at inappropriate points in the growth of the host strain, heterologous genes are often fused to engineered promoters that limit gene expression under non-inducing conditions. However, all promoters have a degree of "leakiness" and allow some protein production even under non-inducing conditions.
The predicted Xenorhabdus nematophila Tdk protein is 70% identical to E. coli Tdk and has been shown to have deoxythymidine kinase activity , converting salvaged deoxythymidine to deoxythymidine monophosphate . A translational start site was predicted for X. nematophila tdk based on alignment with tdk sequences from other organisms. However, X. nematophila tdk has an additional potential start codon 12 bp 5' from the predicted start site. As part of an effort to establish which start codon is used for native X. nematophila Tdk synthesis, the two forms were expressed from a heterologous promoter in E. coli. These studies revealed that, in contrast to short-form, long-form Tdk is not expressed. Furthermore, the additional four codons present in long-form Tdk are sufficient to decrease production of another unrelated protein, NilC.
Results and discussion
Long-form Tdk does not confer FUdR sensitivity to an E. coli tdk strain
Long-form Tdk is not detected by immunoblot analysis
Long-form NilC-trunc is not detected by immunoblot analysis
To determine if the 5' extension of long-form tdk would reduce production of a different protein, the 12-nt extension was added to an X. nematophila gene, nilC, encoding an outer membrane-associated protein  and for which there were readily available antibodies. To create a cytoplasmic version of NilC, the DNA encoding the 21 amino acid long NilC N-terminal signal sequence was removed creating NilC-trunc (short form), to which the MDGP extension was added (long form). Long-form NilC-trunc was not detected by immunoblot analysis with polyclonal anti-NilC serum while short-form NilC-trunc was (Fig. 2B). Thus, the MDGP-encoding extension is sufficient to reduce production of at least two unrelated proteins.
Long-form tdk transcript is less abundant than short-form transcript but shares the same 5' end
Long-form Tdk protein is stable
Long-form Tdk is expressed in the presence of IPTG and arabinose
Arabinose was used to induce T7 RNA polymerase for the 35S-methionine labeling study, resulting in detectable long-form Tdk protein levels by autoradiography (Fig. 4) but not immunoblot analysis (Fig. 2F). Neither arabinose nor IPTG (de-repression of the pET-28a(+) T7 promoter) were used for the other protein production studies reported herein. When both IPTG and arabinose were added to cultures of KY895 carrying pTara and pETLTdk, long-form Tdk was detected by immunoblot analysis at levels similar to short-form Tdk in comparably induced cultures (Fig. 2G). Thus, increased transcription of long-form tdk is sufficient to promote high long-form protein levels. These data support the hypothesis that the low level of protein expressed from genes with the MDGP-encoding extension results at least in part from a reduction in mRNA levels. These results further indicate that, in non-inducing conditions in cases when overproduction of a toxic protein is desired, the encoded N-terminal extension described here might be beneficial in reducing "leaky" protein production.
Changing the nucleotide and/or amino acid identity of the long-form extension results in variable protein production
X. nematophila tdk plasmid design and resulting Tdk protein levels
Oligonucleotide sequence (5'->3') a
Resulting plasmid name
N-terminus of engineered proteinb
Relative amount of protein expressed (% of short-form) c
A version of X. nematophila Tdk protein with four extra amino acids (long-form Tdk) was not detected by immunoblot analysis of cells carrying plasmid pETLTdk even though the cells could express Tdk lacking these extra amino acids (short-form Tdk encoded on pETSTdk) under the same non-inducing growth conditions. Even with arabinose induction of transcription, long-form Tdk protein was present at a level ~13-fold lower than the short-form. The same extension reduced production of another protein, NilC, suggesting the effect of the extension is not specific to Tdk. Long-form Tdk protein was as stable as short-form Tdk protein, and its levels were not affected by the absence of Lon or ClpP proteases. Therefore the difference in long versus short-Tdk protein levels is not likely due to differential proteolysis of the two protein forms. The long-form tdk mRNA transcript was present at a 1.7-fold lower level than short-form tdk transcript and increased levels of transcript are sufficient to overcome low protein levels, supporting the hypothesis that the low level of long-form Tdk production is due in part to low long-form tdk transcript levels. The bacterial N-end rule  states that proteins containing N-terminal arginine, lysine, leucine, phenylalanine, tyrosine and tryptophan residues are particularly sensitive to Clp-dependent degradation, reducing their in vivo half-lives. The phenomenon reported herein is distinct from the N-end rule as none of the known destabilizing residues occur in the MDGP N-terminal extension, nor does the extension result in protein degradation by Clp (Fig. 2D). Since the same pET-28a(+)-based expression system was used for both Tdk and NilC-trunc production, the observed phenotype may result from a combination of the MDGP extension and the pET-28a(+) expression system. These results may be relevant to researchers attempting to limit production of potentially toxic proteins from the pET-28a(+) expression plasmid under non-inducing conditions in an E. coli host, while still allowing protein overproduction under inducing conditions.
Oligonucleotide primers used in this studya
3' of tdk
5' of E. coli clpP/Tn10 (tetr)
3' of E. coli clpP/ Tn10 (tetr)
E. coli recA
E. coli recA
E. coli tdk
E. coli tdk
Cultures were grown in a tube roller at 30°C in Luria-Bertani (LB) broth  except for the FUdR growth assays, which were performed in semi-defined medium containing 5-fluorodeoxyuridine (; FUdR obtained from Fisher Scientific, Pittsburgh, PA) at 37°C with shaking in a microplate reader (Molecular Devices, Sunnyvale, CA) as described previously . LB agar (20 g l-1) plates and all liquid media were supplemented when appropriate with kanamycin (kan; 20 μg ml-1), chloramphenicol (cam; 20 μg ml-1), isopropyl-β-D-thiogalactopyranoside (IPTG; 0.2 mM) or arabinose (0.2%). M9 medium was prepared as described elsewhere . Permanent stocks of cultures were stored at -80°C in dark-stored LB broth supplemented with 10% dimethylsulfoxide.
Production of recombinant X. nematophila Tdk forms in E. coli
DNA manipulations and transformation of E. coli were performed using standard protocols ( and product literature). To construct plasmids expressing various forms of X. nematophila ATCC19061 tdk [GenBank:AY363171] , the BamHItdk primer (Table 1) was combined with each primer listed in Table 2 in a PCR using either X. nematophila (HGB007; laboratory stock of ATCC19061) chromosomal DNA or a cloned copy of X. nematophila tdk as template DNA. Where a proline residue was desired ahead of other tested residues, an extra alanine codon was engineered between the methionine and proline codons to allow the NcoI restriction site (which includes the translational start site) to be used for cloning. The amplified products from each primer set were cloned into plasmid pET-28a(+) (Novagen, Madison, WI) using the engineered NcoI and BamHI sites. The resulting plasmids, listed in Table 2, in addition to plasmid pTara (arabinose-inducible expression of the T7 RNA polymerase; ), were transformed into E. coli KY895, BL21 Star™ (DE3), KY895 clpP and BL21 (DE3). As a control, pET-28a(+) with no insert was transformed with plasmid pTara into these strain backgrounds.
Production of truncated X. nematophila NilC (NilC-trunc) forms in E. coli
A portion of X. nematophila nilC was PCR-amplified from X. nematophila genomic DNA using primers NcoINilC or NcoIMDGPNilC and BamHINilC (Table 1), resulting in nilC production starting 3', at codon 22, of its predicted signal sequence-encoding region and with an additional 5' alanine codon (GCU). NcoIMDGPNilC encodes an additional four codons, for the amino acid series MDGP, before the added alanine codon, to create long-form NilC. Primers NcoIMAGPNilC, NcoIMDAPNilC and NcoIMDGANilC (Table 1) were used to encode the additional amino acid series MAGP, MDAP, and MDGA, respectively, before the added alanine codon. Products from the amplification reactions were cloned into pET-28a(+) as described above. The resulting plasmids, pETnilC, pETMDGPNilC, pETMAGPNilC, pETMDAPNilC and pETMDGANilC, were transformed with pTara into KY895 for protein production analysis.
Immunoblot detection of Tdk and NilC
For immunoblot detections, samples from overnight cultures of KY895 carrying pTara and the appropriate pET-28a(+)-derived vector were electrophoresed and transferred to 0.2 μm PVDF membrane (Bio-Rad) using standard protocols. Immunoblots were performed using the ECL Plus Western Blotting Kit (Amersham Biosciences, Piscataway, NJ) and a goat anti-rabbit IgG horse radish peroxidase-conjugated secondary antibody (Pierce, Rockford, IL). Anti-Tdk serum was obtained from a bleed of a New Zealand White rabbit at the University of Wisconsin Laboratory Animal Resources polyclonal antibody facility following a series of injections of purified 6× his-tagged X. nematophila Tdk . Anti-Tdk and anti-NilC sera  were used at a final dilution of 1:5,000. Fluorescence was detected on a Storm860 Phosphorimager (Amersham Biosciences, Piscataway, NJ).
Primer extension mapping of short-and long-form transcripts
Total-cell RNA was isolated from KY895 carrying plasmids pTara and pETSTdk or pETLTdk following overnight growth in LB/kan/cam broth to OD600 = 2.0. A PAGE-purified primer, tdkprext2 (Table 1), was end-labeled using T4 Polynucleotide Kinase and [γ32-P]ATP (Perkin-Elmer Biosciences, Wellesley, MA) for 10 min at 37°C and the labeled primer used in cycle sequencing reactions with components of the fmol® DNA Cycle Sequencing System kit (Promega, Madison, WI) and either plasmid pETSTdk or pETLTdk, as indicated in Figure 3. Labeled primer was also hybridized to 5 μg of total-cell RNA by dissociation at 80°C for 10 min, followed by a slow cooling to 37°C. The primer was extended by avian myeloblastosis virus reverse transcriptase enzyme (AMV-RT, Promega, Madison, WI) at 37°C and the extension products and completed sequencing reactions were resolved on a 12% SDS-polyacrylamide gel containing 8 M urea. The resulting gel was dried and visualized by exposure to a phosphor screen overnight, which was scanned on a Storm 860 phoshorimager (Amersham Biosciences, Piscataway, NJ), and the data analyzed with ImageQuant software (Amersham Biosciences, Piscataway, NJ).
Quantitative PCR to measure relative transcript levels
E. coli KY895 carrying plasmids pTara and either pET-28a(+), pETSTdk or pETLTdk was subcultured from an overnight culture and grown to OD600~0.85 in LB broth with 0.2% glucose. Two independent cultures started from individual colonies were used for each strain. Total RNA was isolated using TRIzol (Invitrogen, Carlsbad, CA) and the RNA was DNase treated and used to make cDNA with random hexamer primers (Integrated DNA Technologies, Coralville, IA) and AMV-RT. As a control to detect DNA contamination of the DNased RNA, samples with no added AMV-RT were analyzed by PCR for amplification of E. coli tdk using primers EctdkQPfor and EctdkQPrev (Table 1) and, as expected, none showed product. Reactions for real-time PCR were performed in duplicate in a total volume of 25 μl with iQ™ SYBR® Green Supermix (Bio-Rad, Hercules, CA), cDNA template, appropriate primers and a three-step cycling protocol on a Bio-Rad iCycler and analyzed with Bio-Rad iCycler iQ™ software. The amount of X. nematophila tdk transcript was measured by amplification with XntdkQPfor2 and XntdkQPrev2 primers (Table 1). As a negative control, water was used in place of cDNA template. Cycle threshold results for each sample were adjusted according to E. coli recA levels (amplified with Ecrecaminfor and Ecrecaminrev primers, Table 1, designed from published E. coli recA sequence) and then converted to arbitrary units factoring in a two-fold change in PCR product per cycle.
Pulse labeling and immunoprecipitation
Cells were grown to OD600 = 0.3 in LB/kan/cam, washed and resuspended in M9 containing 0.4% arabinose, 1 μg ml-1 thiamine, kan and cam and incubated for 1 h before being pulse labeled with 20 mCi/ml [35S]-L-methionine (PerkinElmer Life Sciences, Boston, MA) for 3 min at 37°C. Unlabeled L-methionine was added to 300 μM and 50 μl samples were removed and added to 50 μl of a 125 mM Tris-Cl, 4% SDS, pH 6.8 buffer at 20 s, 2 min and 5 min after addition of unlabeled methionine. After immediate freezing in dry ice/ethanol, samples were boiled for 4 min and to each was added 1 ml of immunoprecipitation buffer (50 mM Tris-Cl, 150 mM NaCl, 5 mM EDTA, 1% Triton X-100, pH7.5). The samples were centrifuged and supernatants incubated with 2 μl rabbit anti-Tdk serum with rocking for 1 h at 4°C. The samples were then incubated with Protein A immobilized on sepharose CL-4B (Sigma, St. Louis, MO) for 1 h at 4°C, with rocking. The pelleted beads were washed 3 times in immunoprecipitation buffer containing 0.1% SDS, separated on a 12% acrylamide denaturing gel and the gel was dried and exposed to a phosphor screen and analyzed as for the primer extension reactions.
The authors gratefully acknowledge Charles E. Cowles for providing anti-NilC antibody. This work was supported by National Institutes of Health RO1 grant GM59776, the Investigators in Pathogenesis of Infectious Disease Award from the Burroughs Wellcome Foundation and by USDA/CREES grant CRHF-0-6055 (awarded to HGB and used to support SSO in part). SSO received additional support through the National Institutes of Health Predoctoral Training Grant T32 GM07215 in Molecular Biosciences and through the National Science Foundation Graduate Teaching Fellows in K-12 Education award DUE-9979628 to the K-Through-Infinity program.
- Sati SP, Singh SK, Kumar N, Sharma A: Extra terminal residues have a profound effect on the folding and solubility of a Plasmodium falciparum sexual stage-specific protein over-expressed in Escherichia coli. Eur J Biochem. 2002, 269: 5259-5263. 10.1046/j.1432-1033.2002.03237.x.View ArticleGoogle Scholar
- Gottesman S: Minimizing proteolysis in Escherichia coli: genetic solutions. Methods Enzymol. 1990, 185: 119-129.View ArticleGoogle Scholar
- Orchard SS, Goodrich-Blair H: Pyrimidine nucleoside salvage confers an advantage to Xenorhabdus nematophila in its host interactions. Appl and Environ Microbiol.Google Scholar
- Neuhard J: Utilization of preformed pyrimidine bases and nucleosides. Metabolism of Nucleotides, Nucleosides and Nucleobases in Microorganisms. Edited by: Munch-Peterson A. 1983, London, Academic press, 95-148.Google Scholar
- Miller JH: Experiments in Molecular Genetics. 1972, Cold Spring Harbor, NY, Cold Spring Harbor Laboratory Press, 466-Google Scholar
- Cowles CE, Goodrich-Blair H: Characterization of a lipoprotein, NilC, required by Xenorhabdus nematophila for mutualism with its nematode host. Mol Microbiol. 2004, 54: 464-477. 10.1111/j.1365-2958.2004.04271.x.View ArticleGoogle Scholar
- Grunberg-Manago M: Messenger RNA stability and its role in control of gene expression in bacteria and phages. Annu Rev Gen. 1999, 33: 193-227. 10.1146/annurev.genet.33.1.193.View ArticleGoogle Scholar
- Sherman F, Stewart JW, Tsunasawa S: Methionine or not methionine at the beginning of a protein. BioEssays. 1985, 3: 27-31. 10.1002/bies.950030108.View ArticleGoogle Scholar
- Tobias JW, Shrader TE, Rocap G, Varshavsky A: The N-end rule in bacteria. Science. 1991, 254: 1374-1377.View ArticleGoogle Scholar
- Hiraga S, Igarashi K, Yura T: A deoxythymidine kinase-deficient mutant of Escherichia coli. I. Isolation and some properties. Biochim Biophys Acta. 1967, 145: 41-51.View ArticleGoogle Scholar
- Yu D, Ellis HM, Lee EC, Jenkins NA, Copeland NG, Court DL: An efficient recombination system for chromosome engineering in Escherichia coli. Proc Natl Acad Sci USA. 2000, 97: 5978-5983. 10.1073/pnas.100127597.View ArticleGoogle Scholar
- Summers WC, Raksin P: A method for selection of mutations at the tdk locus in Escherichia coli. J Bacteriol. 1993, 175: 6049-6051.Google Scholar
- Sambrook J, Fritsch EF, Maniatis T: Molecular Cloning: a Laboratory Manual. 1989, Cold Spring Harbor, NY, Cold Spring Harbor Laboratory Press, 1076-2ndGoogle Scholar
- Ausubel FA, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA, Struhl K: Current Protocols in Molecular Biology. 1998, New York, Wiley and SonsGoogle Scholar
- Orchard SS, Goodrich-Blair H: Identification and functional characterization of a Xenorhabdus nematophila oligopeptide permease. Appl Env Microbiol. 2004, 70: 5621-5627. 10.1128/AEM.70.9.5621-5627.2004.View ArticleGoogle Scholar
- Wycuff DR, Matthews KS: Generation of an AraC-araBAD promoter-regulated T7 expression system. Anal Biochem. 2000, 277: 67-73. 10.1006/abio.1999.4385.View ArticleGoogle Scholar
- Orchard SS: Oligopeptide and Nucleoside Salvage by the Bacterium Xenorhabdus nematophila. Bacteriology. 2005, Madison, University of WisconsinGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.