ARTICLE Enhancing E. coli Isobutanol Tolerance Through Engineering Its Global Transcription Factor cAMP Receptor Protein (CRP) Huiqing Chong,1 Hefang Geng,1 Hongfang Zhang,1 Hao Song,1 Lei Huang,2 Rongrong Jiang1 1
School of Chemical & Biomedical Engineering, Nanyang Technological University, 62 Nanyang Drive, Singapore 637459, Singapore; telephone: 65-65141055; fax: 65-67947553; e-mail: [email protected] 2 Institute of Biological Engineering, Department of Chemical and Biological Engineering, Zhejiang University, Hangzhou, Zhejiang, P. R. China
ABSTRACT: The limited isobutanol tolerance of Escherichia coli is a major drawback during fermentative isobutanol production. Different from classical strain engineering approaches, this work was initiated to improve E. coli isobutanol tolerance from its transcriptional level by engineering its global transcription factor cAMP receptor protein (CRP). Random mutagenesis libraries were generated by error-prone PCR of crp, and the libraries were subjected to isobutanol stress for selection. Variant IB2 (S179P, H199R) was isolated and exhibited much better growth (0.18 h1) than the control (0.05 h1) in 1.2% (v/v) isobutanol (9.6 g/L). Genome-wide DNA microarray analysis revealed that 58 and 308 genes in IB2 had differential expression (>2-fold, p < 0.05) in the absence and presence of 1% (v/v) isobutanol, respectively. When challenged with isobutanol, genes related to acid resistance (gadABCE, hdeABD), nitrate reduction (narUZYWV), flagella and fimbrial activity (lfhA, yehB, ycgR, fimCDF), and sulfate reduction and transportation (cysIJH, cysC, cysN) were the major functional groups that were upregulated, whereas most of the down-regulated genes were enzyme (tnaA) and transporters (proVWX, manXYZ). As demonstrated by single-gene knockout experiments, gadX, nirB, rhaS, hdeB, and ybaS were found associated with strain isobutanol resistance. The intracellular reactive oxygen species (ROS) level in IB2 was only half of that of the control when facing stress, indicating that IB2 can withstand toxic isobutanol much better than the control. Biotechnol. Bioeng. 2014;111: 700–708. ß 2013 Wiley Periodicals, Inc.
Huiqing Chong and Hefang Geng have contributed equally to this paper. Correspondence to: R. Jiang Contract grant sponsor: Ministry of Education, Singapore Contract grant number: MOE2012-T2-2-117 Received 15 May 2013; Revision received 16 September 2013; Accepted 10 October 2013 Accepted manuscript online 12 October 2013; Article first published online 6 November 2013 in Wiley Online Library (http://onlinelibrary.wiley.com/doi/10.1002/bit.25134/abstract). DOI 10.1002/bit.25134
700
Biotechnology and Bioengineering, Vol. 111, No. 4, April, 2014
KEYWORDS: transcriptional engineering; global regulator; isobutanol tolerance; cAMP receptor protein; strain engineering; biofuel
Introduction Among various biofuels, bioethanol, and biobutanol are the most researched bioalcohols in recent years (Cascone, 2008; Girio et al., 2010). Biobutanol is a family of four-carbon alcohols with advantageous characteristics over bioethanol, that is, high energy density, low corrosivity, and low hydroscopicity (Bruno et al., 2009). Owing to its successful production in titers of more than 20 g/L in E. coli, isobutanol has emerged as an alternative biofuel in addition to 1-butanol and ethanol lately (Atsumi et al., 2008). It can be directly used as gasoline additive or for the synthesis of fuel additive, methyl tertiary-butyl ether (MTBE) (Hoflund et al., 1999). However, isobutanol toxicity towards microbial hosts becomes one of the major concerns during fermentation– 8 g/L isobutanol is already sufficient to impair E. coli growth (Brynildsen and Liao, 2009). Efforts have been made to alleviate this stress such as in situ removal of isobutanol and the final isobutanol titers was reported to reach 50 g/L (Baez et al., 2011). Besides isobutanol removal, obtaining an isobutanol-tolerant strain may also help maximize isobutanol production. Improving E. coli isobutanol tolerance was achieved previously through classical strain engineering approaches such as adaptive evolution (Minty et al., 2011; Smith and Liao, 2011). Metabolic engineering tools, including deletion of tnaA, gatY, acrAB, marCRAB, and yhbJ (Atsumi et al., 2010), as well as individual knock-out of acrR (Luhe et al., 2012), have also been employed to enhance strain isobutanol tolerance. In recent years, transcriptional engineering approaches start to emerge in strain engineering (Alper et al., 2006; Park et al., 2003), which address the limitations of ß 2013 Wiley Periodicals, Inc.
classical strain engineering and metabolic engineering approaches. Several transcription factors have been engineered to alter various cell phenotypes so far, including zinc-finger containing artificial transcription factors (Park et al., 2005), sigma factor (Alper and Stephanopoulos, 2007), H-NS (Hong et al., 2010b), Hha (Hong et al., 2010a), and exogenous transcription factor IrrE (Wang et al., 2012a). A recent publication revealed that global transcription factor cAMP receptor protein (CRP) is one of the major regulators involved in E. coli isobutanol stress response (Brynildsen and Liao, 2009). CRP can regulate more than 400 genes in E. coli (Keseler et al., 2011), and many metabolic processes (Ma et al., 2004). Our group has successfully engineered CRP in the past to improve strain performance under various stresses (Basak and Jiang, 2012; Basak et al., 2012; Chong et al., 2013; Zhang et al., 2012a,b). In this work, we intend to improve E. coli isobutanol tolerance through error-prone PCR of CRP. Here, global regulator CRP was first engineered by errorprone PCR and the random mutagenesis libraries were exposed to isobutanol stress for mutant isolation. DNA microarray was carried out on the best mutant, and 13 genes were selected for single-gene knockout and tested for their isobutanol sensitivity. Intracellular reactive oxygen species (ROS) level and mutant CRP DNA activation properties to various CRP-dependent promoters were also investigated.
Materials and Methods Materials The host strain E. coli DH5a Dcrp was constructed by knocking out crp from E. coli DH5a (Invitrogen, Carlsbad, CA) according to a previously established protocol (Baba et al., 2006). Keio knockout strains (Baba et al., 2006) were purchased from Coli Genetic Stock Center (Yale University), including JW0474-1 (BW25113 DybaS), JW1488-7 (DgadB), JW1487-3 (DgadC), JW5669-1 (DhdeB), JW2733-1 (DcysI), JW3480-2 (DgadE), JW3483-1 (DgadW), JW3484-1 (DgadX), JW1668-1 (DynhG), JW2466-1 (DhyfA), JW3328-1 (DnirB), JW3547-6 (DyiaK), and JW3876-1 (DrhaS). Cells were cultured in LBGMg medium consisting of Luria-Bertani (LB) medium (Merck Millipore, Damstadt, Germany) supplemented with 10 mM MgSO4.7H2O (Merck Millipore) and 1 g/L glucose (Sigma Aldrich, St. Louis, MO). Restriction enzymes and pfu DNA polymerase were obtained from Fermentas (Burlington, Canada). T4 DNA ligase was from New England Biolabs (Ipswich, MA). QIAquick gel extraction kit and QIAprep spin miniprep kit were from QIAGEN (Hilden, Germany), while DNA Clean & Concentrator kit was from Zymo Research (Irvine, CA). Plasmids pKSCP (containing native crp operon) and pKSC (blank plasmid without native crp operon) were from our previous work (Zhang et al., 2012a). Construction of Error-Prone PCR Libraries Error-prone PCR was performed using GeneMorph1 II Random Mutagenesis Kit (Agilent Technologies, Santa
Clara, CA) according to manufacturer’s protocols. Primers 50 -GAGAGGATCCATAACAGAGGATAACCGCGCATG-30 and 50 -AGATGGTACCAAACAAAATGGCGCGCTACCAGGTAACGCGCCA-30 were used to amplify crp with Bam HI and Kpn I restriction sites (underlined), respectively. 30-ng DNA template was used and the error-prone PCR program was set as follows: 2 min at 95 C, 30 cycles of 95 C for 1 min, 62 C for 1 min, followed by 72 C for 1 min, and 10 min at 72 C. The PCR products were purified and digested with Bam HI and Kpn I. Random mutagenesis libraries were generated by inserting the mutated crp fragments into the corresponding restriction sites of plasmid pKSC using T4 DNA ligase. Transformation was carried out by electroporation with E. coli DH5a Dcrp competent cells using an Eppendorf1 multiporator (Hamburg, Germany). Isolation of Isobutanol-Tolerant Mutants In order to isolate isobutanol-tolerant CRP mutants, the transformation mixture was inoculated into LBGMg medium containing 0.8% (v/v) isobutanol. LBGMg medium was chosen for cultivation as it was reported before that MgSO4 has stabilization effect over cell membrane and can improve cell growth in organic solvents (Aono, 1998). Cells were cultured in a shaker incubator (New Brunswick Scientific, Shanghai, China) at 37 C, 200 rpm for 20 h before being diluted (1.0%) into fresh LBGMg medium with elevated isobutanol concentration (up to 1.1%). After three to five successive subcultures, diluted cell culture was spread onto LB-agar plates. Individual colonies were randomly picked for miniprep and DNA sequencing. The mutated crp of the plasmid from the best mutant was digested and re-ligated with freshly prepared pKSC backbone, resulting in plasmid pIB2. pIB2 was then re-transformed into fresh E. coli DH5a Dcrp host strain, generating the mutant strain used in this study that was designated as IB2. Mutant Growth Under Isobutanol Stress Overnight cell culture (1%, v/v) was inoculated into fresh LBGMg medium containing 0%, 1%, and 1.2% (v/v) isobutanol. Cell growth at 37 C was recorded by monitoring its absorbance at 600 nm. Triple experiments were carried out on the mutant and the control. DNA Microarray Total cellular RNA were isolated from both the mutant and the control using PureLink1 RNA Mini Kit and PureLink1 DNase (Life Technologies, Carlsbad, CA) according to manufacturer’s protocols, with cells grown for 3 h in the absence of isobutanol and for 6 h in the presence of 1% (v/v) isobutanol. The quality and integrity of the extracted RNA was verified by a NanoDrop 1000 spectrophotometer (Thermo Scientific, Waltham, MA) at 230 nm and gel electrophoresis. Two biological replicates of mutant and the control were sent to Genomax Technologies (Singapore)
Chong et al.: Enhancing E. coli Isobutanol Tolerance Biotechnology and Bioengineering
701
for DNA microarray assay using Agilent SurePrint E. coli 8 15K slides. The obtained data was analyzed by Agilent Genespring GX software and the p-values were calculated with unpaired Student t-test. Assay of Isobutanol Sensitivity with Single-Gene Knockout Strains The kanamycin resistance gene of BW25113 DybaS, DgadB, DgadC, DhdeB, DcysI, DgadE, DgadW, DgadX, DynhG, DhyfA, DnirB, DyiaK, and DrhaS strains was removed with plasmid pCP20 according to a previously established protocol (Cherepanov and Wackernagel, 1995). The native crp operon of these strains were knocked out using l-Red recombination technique (Datsenko and Wanner, 2000). Plasmid HKpETDuet obtained from our previous work (Chong et al., 2013) was double digested with Eco RI and PacI to obtain the kanR DNA cassette H1-Kan-H2, which has 500 bp upstream and downstream homology regions of the crp operon. 350-ng DNA cassette H1-Kan-H2 was electroporated into these strains for crp knockout. Strain isobutanol sensitivity was examined in LBGMg medium at 37 C, 200 rpm with or without 1.2% (v/v) isobutanol. Survival The survival of the mutant and the control was studied by exposing both exponential phase cells (OD600 0.6) to 3% (v/v) isobutanol. Aliquots of cell culture were drawn at different time points after isobutanol exposure (up to 60 min), subjected to serial dilutions in LB medium, and spread onto LB-agar plates. “Survival” is defined as the percentage of colony forming unit (CFU) in the presence of isobutanol to the CFU in the absence of isobutanol.
Results Isolation of Isobutanol-Tolerant Mutant In order to isolate isobutanol-tolerant mutants, error-prone libraries comprising more than 106 clones were challenged with isobutanol stress. About 30 individual colonies were randomly picked, sequenced, and their growth performance was tested under isobutanol stress. The best mutant had two amino acid mutations in CRP, S179R and H199R. To eliminate the possibility of tolerance enhancement due to chromosomal mutations or plasmid-borne mutations, the newly transformed strain IB2 (DH5a Dcrp þ pIB2) was used in this study. DH5a Dcrp strain harboring plasmid pKSCP (containing native crp operon) was denoted as the control. The isobutanol tolerance of IB2 was verified by testing its growth against the control in the absence and presence of isobutanol with DH5a Dcrp þ pKSC (blank plasmid without native crp operon) & native DH5a (containing native crp) þ pKSC as reference (Fig. 1). In the absence of isobutanol, both IB2 and the control exhibited similar growth rate of 0.4 h1 (Fig. 1A). Similar phenomenon was
702
Biotechnology and Bioengineering, Vol. 111, No. 4, April, 2014
Figure 1. Growth profile of IB2, the control, native DH5a þ pKSC, and DH5aDcrp þ pKSC in (A) 0% isobutanol; (B) 1% (v/v) isobutanol; (C) 1.2% (v/v) isobutabol. All cells were grown in LBGMg medium at 37 C, 200 rpm. All experiments were performed in triplicates.
also observed in other stress-tolerant CRP mutants (Basak and Jiang, 2012; Zhang et al., 2012a). The presence of isobutanol stress resulted in the growth difference between IB2 and the control (Fig. 1B and C). When the isobutanol concentration increased to 1.2% (9.6 g/L), the growth rate of the control declined drastically to 0.05 h1, whereas that of IB2 only dropped slightly to 0.18 h1, still better than the native DH5a strain containing blank plasmid pKSC. The DH5a Dcrp strain harboring blank plasmid pKSC had the worst growth at all isobutanol concentration tested. To further demonstrate the isobutanol resistance of IB2, we exposed the mutant to 3% (v/v) isobutanol (24 g/L) and compared its survival to the control over a period of 60 min (Fig. 2). At the end of the 60-min period, IB2 displayed approximately 100-fold higher survival than the control. The cross-tolerance of the CRP mutant towards other short-chain alcohols, namely 1-propanol (C3), 1-butanol (C4), and 1-pentanol (C5), was also explored. As shown in Figure 3, there was noticeable growth difference between IB2 and the control with the alcohols tested. DNA Microarray and Quantitative Real-Time Reverse Transcription PCR (RT-PCR) Genome-wide DNA microarray assay was carried out on both IB2 and the control under non-stress and stress conditions. In the absence of isobutanol stress, 58 genes exhibited differential expression (>2-fold, p value
School of Chemical & Biomedical Engineering, Nanyang Technological University, 62 Nanyang Drive, Singapore 637459, Singapore; telephone: 65-65141055; fax: 65-67947553; e-mail: [email protected] 2 Institute of Biological Engineering, Department of Chemical and Biological Engineering, Zhejiang University, Hangzhou, Zhejiang, P. R. China
ABSTRACT: The limited isobutanol tolerance of Escherichia coli is a major drawback during fermentative isobutanol production. Different from classical strain engineering approaches, this work was initiated to improve E. coli isobutanol tolerance from its transcriptional level by engineering its global transcription factor cAMP receptor protein (CRP). Random mutagenesis libraries were generated by error-prone PCR of crp, and the libraries were subjected to isobutanol stress for selection. Variant IB2 (S179P, H199R) was isolated and exhibited much better growth (0.18 h1) than the control (0.05 h1) in 1.2% (v/v) isobutanol (9.6 g/L). Genome-wide DNA microarray analysis revealed that 58 and 308 genes in IB2 had differential expression (>2-fold, p < 0.05) in the absence and presence of 1% (v/v) isobutanol, respectively. When challenged with isobutanol, genes related to acid resistance (gadABCE, hdeABD), nitrate reduction (narUZYWV), flagella and fimbrial activity (lfhA, yehB, ycgR, fimCDF), and sulfate reduction and transportation (cysIJH, cysC, cysN) were the major functional groups that were upregulated, whereas most of the down-regulated genes were enzyme (tnaA) and transporters (proVWX, manXYZ). As demonstrated by single-gene knockout experiments, gadX, nirB, rhaS, hdeB, and ybaS were found associated with strain isobutanol resistance. The intracellular reactive oxygen species (ROS) level in IB2 was only half of that of the control when facing stress, indicating that IB2 can withstand toxic isobutanol much better than the control. Biotechnol. Bioeng. 2014;111: 700–708. ß 2013 Wiley Periodicals, Inc.
Huiqing Chong and Hefang Geng have contributed equally to this paper. Correspondence to: R. Jiang Contract grant sponsor: Ministry of Education, Singapore Contract grant number: MOE2012-T2-2-117 Received 15 May 2013; Revision received 16 September 2013; Accepted 10 October 2013 Accepted manuscript online 12 October 2013; Article first published online 6 November 2013 in Wiley Online Library (http://onlinelibrary.wiley.com/doi/10.1002/bit.25134/abstract). DOI 10.1002/bit.25134
700
Biotechnology and Bioengineering, Vol. 111, No. 4, April, 2014
KEYWORDS: transcriptional engineering; global regulator; isobutanol tolerance; cAMP receptor protein; strain engineering; biofuel
Introduction Among various biofuels, bioethanol, and biobutanol are the most researched bioalcohols in recent years (Cascone, 2008; Girio et al., 2010). Biobutanol is a family of four-carbon alcohols with advantageous characteristics over bioethanol, that is, high energy density, low corrosivity, and low hydroscopicity (Bruno et al., 2009). Owing to its successful production in titers of more than 20 g/L in E. coli, isobutanol has emerged as an alternative biofuel in addition to 1-butanol and ethanol lately (Atsumi et al., 2008). It can be directly used as gasoline additive or for the synthesis of fuel additive, methyl tertiary-butyl ether (MTBE) (Hoflund et al., 1999). However, isobutanol toxicity towards microbial hosts becomes one of the major concerns during fermentation– 8 g/L isobutanol is already sufficient to impair E. coli growth (Brynildsen and Liao, 2009). Efforts have been made to alleviate this stress such as in situ removal of isobutanol and the final isobutanol titers was reported to reach 50 g/L (Baez et al., 2011). Besides isobutanol removal, obtaining an isobutanol-tolerant strain may also help maximize isobutanol production. Improving E. coli isobutanol tolerance was achieved previously through classical strain engineering approaches such as adaptive evolution (Minty et al., 2011; Smith and Liao, 2011). Metabolic engineering tools, including deletion of tnaA, gatY, acrAB, marCRAB, and yhbJ (Atsumi et al., 2010), as well as individual knock-out of acrR (Luhe et al., 2012), have also been employed to enhance strain isobutanol tolerance. In recent years, transcriptional engineering approaches start to emerge in strain engineering (Alper et al., 2006; Park et al., 2003), which address the limitations of ß 2013 Wiley Periodicals, Inc.
classical strain engineering and metabolic engineering approaches. Several transcription factors have been engineered to alter various cell phenotypes so far, including zinc-finger containing artificial transcription factors (Park et al., 2005), sigma factor (Alper and Stephanopoulos, 2007), H-NS (Hong et al., 2010b), Hha (Hong et al., 2010a), and exogenous transcription factor IrrE (Wang et al., 2012a). A recent publication revealed that global transcription factor cAMP receptor protein (CRP) is one of the major regulators involved in E. coli isobutanol stress response (Brynildsen and Liao, 2009). CRP can regulate more than 400 genes in E. coli (Keseler et al., 2011), and many metabolic processes (Ma et al., 2004). Our group has successfully engineered CRP in the past to improve strain performance under various stresses (Basak and Jiang, 2012; Basak et al., 2012; Chong et al., 2013; Zhang et al., 2012a,b). In this work, we intend to improve E. coli isobutanol tolerance through error-prone PCR of CRP. Here, global regulator CRP was first engineered by errorprone PCR and the random mutagenesis libraries were exposed to isobutanol stress for mutant isolation. DNA microarray was carried out on the best mutant, and 13 genes were selected for single-gene knockout and tested for their isobutanol sensitivity. Intracellular reactive oxygen species (ROS) level and mutant CRP DNA activation properties to various CRP-dependent promoters were also investigated.
Materials and Methods Materials The host strain E. coli DH5a Dcrp was constructed by knocking out crp from E. coli DH5a (Invitrogen, Carlsbad, CA) according to a previously established protocol (Baba et al., 2006). Keio knockout strains (Baba et al., 2006) were purchased from Coli Genetic Stock Center (Yale University), including JW0474-1 (BW25113 DybaS), JW1488-7 (DgadB), JW1487-3 (DgadC), JW5669-1 (DhdeB), JW2733-1 (DcysI), JW3480-2 (DgadE), JW3483-1 (DgadW), JW3484-1 (DgadX), JW1668-1 (DynhG), JW2466-1 (DhyfA), JW3328-1 (DnirB), JW3547-6 (DyiaK), and JW3876-1 (DrhaS). Cells were cultured in LBGMg medium consisting of Luria-Bertani (LB) medium (Merck Millipore, Damstadt, Germany) supplemented with 10 mM MgSO4.7H2O (Merck Millipore) and 1 g/L glucose (Sigma Aldrich, St. Louis, MO). Restriction enzymes and pfu DNA polymerase were obtained from Fermentas (Burlington, Canada). T4 DNA ligase was from New England Biolabs (Ipswich, MA). QIAquick gel extraction kit and QIAprep spin miniprep kit were from QIAGEN (Hilden, Germany), while DNA Clean & Concentrator kit was from Zymo Research (Irvine, CA). Plasmids pKSCP (containing native crp operon) and pKSC (blank plasmid without native crp operon) were from our previous work (Zhang et al., 2012a). Construction of Error-Prone PCR Libraries Error-prone PCR was performed using GeneMorph1 II Random Mutagenesis Kit (Agilent Technologies, Santa
Clara, CA) according to manufacturer’s protocols. Primers 50 -GAGAGGATCCATAACAGAGGATAACCGCGCATG-30 and 50 -AGATGGTACCAAACAAAATGGCGCGCTACCAGGTAACGCGCCA-30 were used to amplify crp with Bam HI and Kpn I restriction sites (underlined), respectively. 30-ng DNA template was used and the error-prone PCR program was set as follows: 2 min at 95 C, 30 cycles of 95 C for 1 min, 62 C for 1 min, followed by 72 C for 1 min, and 10 min at 72 C. The PCR products were purified and digested with Bam HI and Kpn I. Random mutagenesis libraries were generated by inserting the mutated crp fragments into the corresponding restriction sites of plasmid pKSC using T4 DNA ligase. Transformation was carried out by electroporation with E. coli DH5a Dcrp competent cells using an Eppendorf1 multiporator (Hamburg, Germany). Isolation of Isobutanol-Tolerant Mutants In order to isolate isobutanol-tolerant CRP mutants, the transformation mixture was inoculated into LBGMg medium containing 0.8% (v/v) isobutanol. LBGMg medium was chosen for cultivation as it was reported before that MgSO4 has stabilization effect over cell membrane and can improve cell growth in organic solvents (Aono, 1998). Cells were cultured in a shaker incubator (New Brunswick Scientific, Shanghai, China) at 37 C, 200 rpm for 20 h before being diluted (1.0%) into fresh LBGMg medium with elevated isobutanol concentration (up to 1.1%). After three to five successive subcultures, diluted cell culture was spread onto LB-agar plates. Individual colonies were randomly picked for miniprep and DNA sequencing. The mutated crp of the plasmid from the best mutant was digested and re-ligated with freshly prepared pKSC backbone, resulting in plasmid pIB2. pIB2 was then re-transformed into fresh E. coli DH5a Dcrp host strain, generating the mutant strain used in this study that was designated as IB2. Mutant Growth Under Isobutanol Stress Overnight cell culture (1%, v/v) was inoculated into fresh LBGMg medium containing 0%, 1%, and 1.2% (v/v) isobutanol. Cell growth at 37 C was recorded by monitoring its absorbance at 600 nm. Triple experiments were carried out on the mutant and the control. DNA Microarray Total cellular RNA were isolated from both the mutant and the control using PureLink1 RNA Mini Kit and PureLink1 DNase (Life Technologies, Carlsbad, CA) according to manufacturer’s protocols, with cells grown for 3 h in the absence of isobutanol and for 6 h in the presence of 1% (v/v) isobutanol. The quality and integrity of the extracted RNA was verified by a NanoDrop 1000 spectrophotometer (Thermo Scientific, Waltham, MA) at 230 nm and gel electrophoresis. Two biological replicates of mutant and the control were sent to Genomax Technologies (Singapore)
Chong et al.: Enhancing E. coli Isobutanol Tolerance Biotechnology and Bioengineering
701
for DNA microarray assay using Agilent SurePrint E. coli 8 15K slides. The obtained data was analyzed by Agilent Genespring GX software and the p-values were calculated with unpaired Student t-test. Assay of Isobutanol Sensitivity with Single-Gene Knockout Strains The kanamycin resistance gene of BW25113 DybaS, DgadB, DgadC, DhdeB, DcysI, DgadE, DgadW, DgadX, DynhG, DhyfA, DnirB, DyiaK, and DrhaS strains was removed with plasmid pCP20 according to a previously established protocol (Cherepanov and Wackernagel, 1995). The native crp operon of these strains were knocked out using l-Red recombination technique (Datsenko and Wanner, 2000). Plasmid HKpETDuet obtained from our previous work (Chong et al., 2013) was double digested with Eco RI and PacI to obtain the kanR DNA cassette H1-Kan-H2, which has 500 bp upstream and downstream homology regions of the crp operon. 350-ng DNA cassette H1-Kan-H2 was electroporated into these strains for crp knockout. Strain isobutanol sensitivity was examined in LBGMg medium at 37 C, 200 rpm with or without 1.2% (v/v) isobutanol. Survival The survival of the mutant and the control was studied by exposing both exponential phase cells (OD600 0.6) to 3% (v/v) isobutanol. Aliquots of cell culture were drawn at different time points after isobutanol exposure (up to 60 min), subjected to serial dilutions in LB medium, and spread onto LB-agar plates. “Survival” is defined as the percentage of colony forming unit (CFU) in the presence of isobutanol to the CFU in the absence of isobutanol.
Results Isolation of Isobutanol-Tolerant Mutant In order to isolate isobutanol-tolerant mutants, error-prone libraries comprising more than 106 clones were challenged with isobutanol stress. About 30 individual colonies were randomly picked, sequenced, and their growth performance was tested under isobutanol stress. The best mutant had two amino acid mutations in CRP, S179R and H199R. To eliminate the possibility of tolerance enhancement due to chromosomal mutations or plasmid-borne mutations, the newly transformed strain IB2 (DH5a Dcrp þ pIB2) was used in this study. DH5a Dcrp strain harboring plasmid pKSCP (containing native crp operon) was denoted as the control. The isobutanol tolerance of IB2 was verified by testing its growth against the control in the absence and presence of isobutanol with DH5a Dcrp þ pKSC (blank plasmid without native crp operon) & native DH5a (containing native crp) þ pKSC as reference (Fig. 1). In the absence of isobutanol, both IB2 and the control exhibited similar growth rate of 0.4 h1 (Fig. 1A). Similar phenomenon was
702
Biotechnology and Bioengineering, Vol. 111, No. 4, April, 2014
Figure 1. Growth profile of IB2, the control, native DH5a þ pKSC, and DH5aDcrp þ pKSC in (A) 0% isobutanol; (B) 1% (v/v) isobutanol; (C) 1.2% (v/v) isobutabol. All cells were grown in LBGMg medium at 37 C, 200 rpm. All experiments were performed in triplicates.
also observed in other stress-tolerant CRP mutants (Basak and Jiang, 2012; Zhang et al., 2012a). The presence of isobutanol stress resulted in the growth difference between IB2 and the control (Fig. 1B and C). When the isobutanol concentration increased to 1.2% (9.6 g/L), the growth rate of the control declined drastically to 0.05 h1, whereas that of IB2 only dropped slightly to 0.18 h1, still better than the native DH5a strain containing blank plasmid pKSC. The DH5a Dcrp strain harboring blank plasmid pKSC had the worst growth at all isobutanol concentration tested. To further demonstrate the isobutanol resistance of IB2, we exposed the mutant to 3% (v/v) isobutanol (24 g/L) and compared its survival to the control over a period of 60 min (Fig. 2). At the end of the 60-min period, IB2 displayed approximately 100-fold higher survival than the control. The cross-tolerance of the CRP mutant towards other short-chain alcohols, namely 1-propanol (C3), 1-butanol (C4), and 1-pentanol (C5), was also explored. As shown in Figure 3, there was noticeable growth difference between IB2 and the control with the alcohols tested. DNA Microarray and Quantitative Real-Time Reverse Transcription PCR (RT-PCR) Genome-wide DNA microarray assay was carried out on both IB2 and the control under non-stress and stress conditions. In the absence of isobutanol stress, 58 genes exhibited differential expression (>2-fold, p value
