SEARCH

SEARCH BY CITATION

Keywords:

  • Escherichia coli;
  • kdgR;
  • deoR;
  • proteomics;
  • insertion mutation

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgements
  8. References
  9. Supporting Information

Escherichia coli has been used widely in laboratory and the biotech industry. However, the genetic and metabolic characteristics remain inadequately studied, particularly for those strains with extensive genetic manipulations that might have resulted in unknown mutations. Here, we demonstrate a comparative proteomics and genetics approach to identify unknown mutations in E. coli K-12 derivatives. The comparative proteomic and genetic analyses revealed an IS5 disruption of the kdgR gene in two commonly used derivative strains of E. coli K-12, XL1-Blue and DH5α, compared with K-12 wild-type strain W3110. In addition, a controversial deoR mutation was clarified as a wild type in E. coli DH5α using the same approach. This approach should be useful in characterizing the unknown mutations in various mutant strains developed. At the same time, comparative proteomic analysis also revealed the distinct metabolic characteristic of the two derivatives: higher biosynthetic flux to purine nucleotides. This is potentially beneficial for the synthesis of plasmid DNA.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgements
  8. References
  9. Supporting Information

Escherichia coli is widely used in laboratory and industry for producing diverse products such as biochemicals, biopolymers, plasmid DNA, and recombinant proteins (Lee et al., 2005; Park et al., 2008). In particular, plasmid DNA production has attracted considerable attention with the recent increasing demand for plasmid DNA for gene therapies and vaccination (Kutzler & Weiner, 2008). Although numerous E. coli strains are available as potential host strains including XL1-Blue and DH5α, their genetic and metabolic characteristics remain inadequately studied. This might be explained by the fact that the generation of these strains usually involves random mutations, followed by the selection of a particularly wanted phenotype, and often requires many steps of transfer or the deletion of undefined DNA fragments, thus leading to some unintentional and/or undiscovered mutations. These complex genotypes have often been ignored, but they are becoming increasingly important as we are moving into systems-level studies on these strains (Lee et al., 2005; Park et al., 2008).

Comparative proteomics offers a powerful platform technology to study the differentially expressed proteins in response to various genetic and environmental perturbations (Han & Lee, 2003, 2006). This technology has been used for the study of cell physiology and the identification of new biomarkers (Han & Lee, 2003; Meng & Veenstra, 2007). However, to date, there has been no report on the use of comparative proteomics to identify genetic mutations. It was reasoned that mutations in the structural as well as the regulatory genes could be identified by examining the differentially expressed proteins, which can be confirmed by further genetic analysis such as PCR and DNA sequencing. To demonstrate a proof of concept, we performed a comparative proteomic analysis of two E. coli K-12 derivatives XL1-Blue and DH5α with the sequenced wild-type strain. An unknown kdgR mutation was identified in the two derivatives. In addition, a controversial deoR mutation was clarified as a wild type in E. coli DH5α using the same approach.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgements
  8. References
  9. Supporting Information

Escherichia coli strains and growth conditions

The wild-type E. coli K-12 W3110 (Korean Collection for Type Cultures number 2223, Daejeon, Korea) was used as a reference. Two K-12 derivatives, XL1-Blue [endA1 gyrA96 thi-1 recA1 relA1 lac glnV44 [F′∷ Tn10 proAB+lacIqΔ (lacZ)M15] hsdR17(rKmK+)] and DH5α [FendA1 glnV44 thi-1 recA1 gyrA96Φ80dlacZΔM15Δ (lacZYA-argF)U169, hsdR17(rKmK+), λ], were obtained from Stratagene (La Jolla, CA) and Invitrogen Corporation (Carlsbad, CA), respectively. The detailed history and relationships of these strains were described previously (Bachmann, 1987). During strain construction, the two derivatives had undergone a high degree of mutagenesis to obtain several important mutations for routine cloning and plasmid production (Bullock et al., 1987; Grant et al., 1990). All strains were grown in 350-mL Erlenmeyer flasks containing 50 mL of Luria–Bertani (LB) medium at 37 °C and 220 r.p.m. in a shaking incubator. The seed culture was prepared by inoculating a single colony into 10 mL LB medium and cultured overnight at 37 °C and 220 r.p.m. This seed culture (0.5 mL) was used to inoculate the flasks. When OD600 nm reached ∼0.5, cells were harvested by centrifugation at 3500 g for 5 min at 4 °C, and the cell pellets were frozen at −80 °C before proteomic analysis.

Protein preparation, two-dimensional gel electrophoresis, and gel analysis

The frozen cells were washed twice with low-salt washing buffer and subsequently resuspended in a buffer containing 10 mM Tris-HCl (pH 8.0), 1.5 mM MgCl2, 10 mM KCl, 0.5 mM dithiothreitol, and 0.1% w/v sodium dodecyl sulfate (SDS). The cell suspensions were mixed with a lysis buffer consisting of 7 M urea, 2 M thiourea, 40 mM Tris, 65 mM dithiothreitol, and 4% w/v 3-[(3-cholamidopropyl) dimethylammonio]-1-propanesulfonate (CHAPS). Soluble proteins were separated by centrifugation at 13 000 g for 10 min at 4 °C, and the protein concentration was measured using the Bradford method (Bradford, 1976). The proteins (150 μg) were diluted into 340 μL of a rehydration buffer containing 7 M urea, 2 M thiourea, 20 mM dithiothreitol, 2% w/v CHAPS, 0.8% w/v immobilized pH gradient (IPG) buffer (Amersham Biosciences, Uppsala, Sweden), and 1% v/v cocktail protease inhibitor (Roche Diagnostics GmbH, Mannheim, Germany) and then loaded onto Immobiline DryStrip gels (18 cm, pH 3–10 NL; Amersham Biosciences). The loaded IPG strips were rehydrated for 12 h on the Protean IEF Cell (Bio-Rad, Hercules, CA) and focused at 20 °C for 3 h at 250 V, followed by 6000 V until a total of 65 kV h was reached. Following separation in the first dimension, the strips were equilibrated in a solution containing 6 M urea, 0.375 M Tris-HCl (pH 8.8), 20% w/v glycerol, 2% w/v SDS, 130 mM dithiothreitol, and 0.002% w/v bromophenol blue for 15 min at room temperature. The IPG strips were then equilibrated with the buffer described above in which the dithiothreitol was replaced with 135 mM iodoacetamide for 15 min at room temperature. The equilibrated strips were transferred to 12% w/v SDS-polyacrylamide gels. The second dimensional separation was performed using the Protean II xi cell (Bio-Rad) at 35 mA per gel until the bromophenol blue reached the gel tips. Protein spots were visualized using the PlusOne® Silver Staining Kit (Amersham Biosciences), and the stained gels were scanned using the UMAX PowerLook 2100XL Scanner (UMAX Technologies Inc., Dallas, TX). Three gels were prepared from each strain. Spots were detected, quantified, matched, and compared using the pdquest analysis software (version 7.3.1, Bio-Rad). For each comparison (XL1-Blue vs. W3110, DH5α vs. W3110), Student's t-test and a 95% level of confidence were used to detect statistically significant differences. The spots that were differentially expressed by>1.5-fold were identified by gel match or LC–MS/MS (Lee et al., 2006; Xia et al., 2008).

Genetic analysis

Genomic DNA of the three strains was prepared using a DNeasy blood and tissue kit (Qiagen, Valencia, CA). To amplify the kdgR fragment (from 127 bp upstream of the start codon to the stop codon), primers FSkdgRXba (5′-CACTCTAGACTGATATTCACGGTGGATGT-3′, XbaI restriction site underlined) and RSkdgRXho (5′-TATCTCGAGTCAGAACGGATAGTCGTGAT-3′, XhoI restriction site underlined) were designed according to the related sequence of W3110. Similarly, to amplify the deoR fragment, primers FSdeoRXba (5′-CCATCTAGACTGGATATGCTCGGTGGATT-3′, XbaI restriction site underlined) and RSdeoRXho (5′-TATCTCGAGCGTCATCCGGTTATACGTCA-3′, XhoI restriction site underlined) were designed and used in the PCR reactions. The PCR products were first analyzed by agarose gel electrophoresis. Next, each of the PCR products, after digestion with XbaI and XhoI, was cloned into plasmid pBluescript SK (−) (Stratagene). The resulting recombinant plasmids were subjected to DNA sequencing using the M13 Forward and M13 Reverse universal primers. Sequencing was additionally performed using the primers FSkdg (5′-CGAGCGCCCAGTTCAAACAA-3′) and RSkdg (5′-GGGATAACCGAGCTGTCGCA-3′) to uncover the DNA sequence of insertion mutation.

Results and discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgements
  8. References
  9. Supporting Information

Comparative proteomics and genomics reveal the kdgR mutation in E. coli XL1-Blue and DH5α

For each strain, we analyzed three replicates derived from a single culture. The experiments were repeated and the same conclusion was reached using cultures from different single colonies. In total, 19 proteins were differentially expressed and identified through comparative proteomic analysis (Table 1). Of these, four proteins (KdgK, KduI, KduD, and YjgK) showed expression in strains E. coli XL1-Blue and DH5α, but not in strain W3110 (Fig. 1, see Supporting Information, Fig. S1 for full-size gel). Interestingly, gene regulatory analysis indicated that the four proteins are products of genes belonging to the same KdgR regulon (Rodionov et al., 2000, 2004) (Fig. 2). In addition, the expression of Entner–Doudoroff aldolase (Eda), which is partially repressed by KdgR (Murray & Conway, 2005), was upregulated in E. coli XL1-Blue and DH5α compared with W3110 (Figs 1 and 2). Presumably, the constitutive expression of KdgK, KduI, KduD, and YjgK and the partial derepression of Eda resulted from kdgR gene mutation in the chromosomes of E. coli XL1-Blue and DH5α. To test this, the genomic DNA of the three strains was prepared to PCR amplify the kdgR fragment (from 127 bp upstream of the start codon to the stop codon) using the primers FSkdgRXba and RSkdgRXho, respectively. As expected, the kdgR fragment of W3110 was ∼900 bp in size (Fig. 3a). However, the kdgR fragments of XL1-Blue and DH5α were ∼1.2 kb larger, implying insertional mutation in the two K-12 derivatives. To further identify the insertion sequences (ISs), the two kdgR variants were digested with XbaI and XhoI and cloned into plasmid pBluescript SK (−) (Stratagene) for DNA sequencing, respectively. Indeed, DNA sequencing revealed IS5, an insertion element able to transpose within the E. coli genome, in the kdgR coding region in both XL1-Blue and DH5α (Fig. 3b). To rule out that the insertion mutation was due to routine maintenance in our laboratory, the same genetic analysis was applied to the two strains obtained from another laboratory (Prof. Sun Chang Kim, Department of Biological Sciences, KAIST); IS5 disruption of kdgR was also observed (data not shown).

Table 1.   Differentially expressed proteins in Escherichia coli XL1-Blue or DH5α in comparison with the wild-type strain W3110 identified by gel match or LC–MS/MS
Protein nameAccession*Protein description*Mw (kDa)/pI*Sequence coverage (%)Matched peptidesFold change
XL1-BlueDH5α
  • *

    The accession number, description of the identified proteins and calculated pI and Mw are from ExPASy Proteomics Server (http://kr.expasy.org/).

  • The number of peptides matched and sequence coverage are from MASCOT search results of Matrix science (http://www.matrixscience.com/).

  • Fold changes were calculated from triplicate proteome experiments. Details on statistical analysis are described in Materials and methods. The infinity symbol (∞) means the protein spot was undetectable in the 2-D gel of the control strain W3110.

  • §

    Proteins with ‘–’ were identified by gel matching to the reference database SWISS-2DPAGE (http://br.expasy.org/) or EcoProDB (http://eecoli.kaist.ac.kr/main.html). The others were identified by LC–MS/MS.

Carbon compounds degradation
 EdaP0A955KHG/KDPG aldolase22.3/5.572442.82.4
 KdgKP376472-Dehydro-3-deoxygluconokinase34.0/4.923611
 KduDP377692-Dehydro-3-deoxy-d-gluconate-5-dehydrogenase27/5.24226
 KduIQ469384-Deoxy-l-threo-5-hexosulose-uronate ketol-isomerase31.1/5.71197
Purine ribonucleotide biosynthesis
 PurCP0A7D7Phosphoribosylaminoimidazole-succinocarboxamide synthase27.0/5.072772.43.7
 PurDP15640Phosphoribosylamine-glycine ligase45.9/4.9632103.32.7
 PurHP15639Bifunctional purine biosynthesis protein purH57.3/5.5320103.05.2
Salvage of nucleosides and nucleotides
 AddP22333Adenosine deaminase36.4/5.362040.50.7
 CddP0ABF6Cytidine deaminase31.5/5.421640.40.4
 UdpP12758Uridine phosphorylase27.0/5.811120.30.4
Amino acid biosynthesis/degradation
 AspCP00509Aspartate aminotransferase43.6/5.5449191.61.8
 GlyAP0A825Serine hydroxymethyltransferase45.3/6.032292.52.9
 SerC§P23721Phosphoserine aminotransferase39.7/5.373.13.6
Transport/binding proteins
 DppA§B1X8G0Dipeptide transporter; periplasmic-binding component of ABC superfamily60.3/6.210.40.3
 MalE§P0AEX9Maltose-binding periplasmic protein40.7/5.220.60.4
 OppA§B1XAT3Oligopeptide transporter periplasmic subunit61.0/6.050.40.2
 RbsBP02925d-Ribose-binding periplasmic protein28.5/5.992761.71.8
Aminoacyl-tRNA synthetases and modification
 PheSP08312Phenylalanyl-tRNA synthetase α-chain36.8/5.792881721
Hypothetical protein
 YjgKP0AF96Hypothetical protein16.9/5.3181  
image

Figure 1.  Detailed regions of silver-stained two-dimensional gels displaying changes in the expression level of KdgR regulon proteins extracted from Escherichia coli W3110, XL1-Blue, and DH5α.

Download figure to PowerPoint

image

Figure 2.  Operon structures and regulatory sites for the KdgR transcriptional repressor in Escherichia coli. Filled ellipse, KdgR-binding site; filled rectangle, GntR-binding site; open circle, PhoB-binding site.

Download figure to PowerPoint

image

Figure 3.  a) Agarose gel electrophoresis of PCR-amplified kdgR gene fragments from Escherichia coli W3110, XL1-Blue, and DH5α, respectively. (b) DNA sequencing revealed the IS5 insertion mutation in kdgR coding regions of E. coli XL1-Blue and DH5α. The kdgR gene fragments were cloned into plasmid pBluescript SK (−) for sequencing as described in Materials and methods.

Download figure to PowerPoint

Differential insertion mutations have also been observed in other E. coli K-12 strains. For example, in the sequenced MG1655 and DH10B, an insertion of IS3E into the gatR gene leads to the constitutive expression of gatYZABCD operon (Nobelmann & Lengeler, 1996; Durfee et al., 2008). The tdh promoter structure altered by the insertion of IS3 activates a cryptic pathway for threonine metabolism in E. coli PS1236 (Aronson et al., 1989). In a selected E. coli mutant that can grow on propanediol as the sole carbon and energy source, IS5 insertion between fucAO and the fucPIK operon caused the constitutive expression of the fucAO operon (Chen et al., 1989).

The mutation of deoR is a controversial allele in E. coli DH5α (Grant et al., 1990; Durfee et al., 2008). DeoR is involved in the repression of genes related to the transport and catabolism of deoxyribonucleoside nucleotides. None of the proteins encoded by the deoR regulon genes (i.e. deoCABD, nupG, and tsx) was found to be differentially expressed between E. coli DH5α and W3110. It was thus inferred that the deoR gene was wild type in E. coli DH5α. To confirm this, we PCR amplified the deoR gene fragment from the genomic DNA of DH5α and cloned into pBluescipt SK (−) for DNA sequencing. The results showed that the deoR gene is unambiguously wild type in E. coli DH5α. This proved that the previous assumption of a higher transformation rate in E. coli DH5α caused by the mutation of deoR (Hanahan et al., 1991) is improper.

Comparative proteomics implies a higher purine biosynthetic flux of the E. coli XL1-Blue and DH5α

We mapped most of the differentially expressed proteins onto the metabolic pathways of E. coli (Fig. 4). Interestingly, three proteins involved in purine nucleotides biosynthesis (PurD, PurC, and PurH) were upregulated by 2.4–5.2-folds in E. coli XL1-Blue and DH5α. The two proteins leading to glycine formation (SerC and GlyA) were also upregulated, which coincided well with the upregulation of PurD that utilizes glycine as a substrate (Fig. 4). It should be noted that GlyA catalyzes the formation of 5,10-methylene-tetrahydrofolate, which is the major source of C1 units in the cell and a direct precursor of 10-formyl-tetrahydrofolate. Therefore, it is possible that GlyA upregulation allowed a higher metabolic pool to 10-formyl-tetrahydrofolate for purine biosynthesis (via PurH). On the other hand, three enzymes (Cdd, Add, and Udp) involved in the salvage pathway of nucleosides and nucleotides were downregulated in E. coli XL1-Blue and DH5α (Table 1 and Fig. 4). Other differentially expressed proteins include transport or binding proteins (DppA, MalE, OppA, and RbsB) and aminoacyl-tRNA synthetic enzyme (PheS). In particular, ribose transporter protein RbsB showed a significantly higher expression in both XL1-Blue and DH5α, implying an elevated uptake of ribose for the biosynthesis of ribosyl nucleosides ((Baev et al., 2006). Taken together, it appeared that the two derivatives had a higher biosynthetic flux to purine nucleotides, which is potentially beneficial for the production of plasmid DNA.

image

Figure 4.  Mapping onto the Escherichia coli metabolic network differentially expressed proteins in E. coli XL1-Blue and DH5α in comparison with W3110. Upregulated genes are shown in a dark gray box, while downregulated genes are shown in a light gray box. Dashed arrows indicate multistep reactions. Metabolic intermediates are: AIR, 5-aminoimidazole ribotide; GlC, glucose; GLY, glycine; G6P, glucose-6-phosphate; G3P, glyceraldehyde phosphate; IMP, inosine monophosphate; KDG, 2-keto-3-deoxygluconate; KDPG, 2-keto-3-deoxy-6-phospho-d-gluconate; 3PG, 3-phosphoglycerate; 6PG, 6-phosphogluconate; Pi, phosphate; PRPP, 5-phosphoribosyl diphosphate; PYR, pyruvate; R5P, ribose-5-phosphate; SER, serine; XMP, xanthosine-5′-phosphate.

Download figure to PowerPoint

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgements
  8. References
  9. Supporting Information

A previous unknown kdgR mutation by IS5 insertion was identified in E. coli XL1-Blue and DH5α, and a controversial deoR mutation was confirmed as a wild type in E. coli DH5α. We have expanded the application of comparative proteomics for the identification of unknown genetic mutations in genome-unsequenced E. coli K-12 derivatives. Combined comparative proteomic and genetic analyses performed in this study should be useful in linking the genotypes and phenotypes. On the other hand, whole-genome sequencing is becoming increasingly cost-effective. This technology will provide a catalogue of sequence differences, and will allow further analysis such as the classification of the effects of particular mutations on specific phenotypes.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgements
  8. References
  9. Supporting Information

This work was supported by the Converging Research Center Program (2009-0082332) of the Ministry of Education, Science, and Technology (MEST) through the National Research Foundation (NRF). Further support by the World Class University Program (R32-2008-000-10142-0) of the MEST through NRF is appreciated.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgements
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgements
  8. References
  9. Supporting Information

Fig. S1. The typical 2-DE maps of Escherichia coli W3110 (a), XL1-Blue (b) and DH5α (c).

Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

FilenameFormatSizeDescription
FML_2157_sm_supplinformation.doc3048KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.