Combinatorial modulation of galP and glk gene ... - Springer Link

Appl Microbiol Biotechnol (2012) 93:2455–2462 DOI 10.1007/s00253-011-3752-y

APPLIED GENETICS AND MOLECULAR BIOTECHNOLOGY

Combinatorial modulation of galP and glk gene expression for improved alternative glucose utilization Jiao Lu & Jinlei Tang & Yi Liu & Xinna Zhu & Tongcun Zhang & Xueli Zhang

Received: 18 September 2011 / Revised: 6 November 2011 / Accepted: 21 November 2011 / Published online: 13 December 2011 # Springer-Verlag 2011

Abstract Phosphoenolpyruvate (PEP) is an important precursor for anaerobic production of succinate and malate. Although inactivating PEP/carbohydrate phosphotransferase systems (PTS) could increase PEP supply, the resulting strain had a low glucose utilization rate. In order to improve anaerobic glucose utilization rate for efficient production of succinate and malate, combinatorial modulation of galactose permease (galP) and glucokinase (glk) gene expression was carried out in chromosome of an Escherichia coli strain with inactivated PTS. Libraries of artificial regulatory parts, including promoter and messenger RNA stabilizing region (mRS), were firstly constructed in front of β-galactosidase gene (lacZ) in E. coli chromosome through λ-Red recombination. Most regulatory parts selected from mRS library had constitutive strengths under different cultivation conditions. A convenient one-step recombination method was then used to modulate galP and glk gene expression with different regulatory parts. Glucose utilization rates of strains modulated with either galP or glk all increased, and the rates had a positive relation with expression strength of both genes. Combinatorial modulation had a synergistic effect on glucose utilization rate. The highest rate (1.64 g/L h) was tenfold higher than PTS− strain and 39% higher than the

Electronic supplementary material The online version of this article (doi:10.1007/s00253-011-3752-y) contains supplementary material, which is available to authorized users. J. Lu : J. Tang : Y. Liu : T. Zhang Tianjin University of Science & Technology, Tianjin, China J. Lu : J. Tang : Y. Liu : X. Zhu : X. Zhang (*) Key Laboratory of Systems Microbial Biotechnology, Institute of Tianjin Industrial Biotechnology, Chinese Academy of Sciences, Tianjin, China e-mail: [email protected]

wild-type E. coli. These modulated strains could be used for efficient anaerobic production of succinate and malate. Keywords Regulatory parts . Modulation of gene expression . Glucose utilization . PTS . Escherichia coli

Introduction Glucose is the most widely used substrate in the fermentation industry, and high glucose utilization rate is very important for cost-effective production. In Escherichia coli, glucose is firstly transported into cytoplasm through the PEP/carbohydrate phosphotransferase systems (PTS) (Flores et al. 1996; Postma et al. 1996; Zhang et al. 2009a). The system is composed of the soluble and non sugar-specific protein components enzyme I and HPr. A phosphate group is transferred from PEP to EI and HPr, then to sugar-specific enzymes IIA, IIB, and IIC (HernandezMontalvo et al. 2003; Postma et al. 1996). IIC is an integral membrane permease that recognizes and transports the sugar molecules. The glucose-specific PTS proteins are soluble IIAGlc (encoded by crr gene) and membrane-bound IICBGlc permease (encoded by ptsG gene). PEP is an important precursor for synthesis of many industrially important chemicals, such as succinate, malate, and aromatic compounds. For the native glucose utilization pathway in E. coli, half of the PEP produced is used for glucose uptake and phosphorylation (Postma et al. 1996; Zhang et al. 2009a). PTS was usually inactivated to increase PEP supply to improve metabolic fluxes towards desired products (Flores et al. 1996; Wang et al. 2006; Yi et al. 2003; Zhang et al. 2009b, 2010, 2011). However, the resulting PTS− strain grew very slowly and had a low glucose utilization rate (Flores et al. 1996), which was not suitable

2456

for industrial production. Two strategies had been used to increase cell growth and glucose utilization rate in a PTS− strain. Adaptive evolution was used to obtain spontaneous mutants from PB11 that lacked the ptsHIcrr operon but grew faster on glucose than PTS− parental strain (Flores et al. 1996). The improved glucose utilization rate was identified to be caused by functional galactose permease and glucokinase (Flores et al. 1996; Flores et al. 2002). One of these mutants, PB12, had been used to successfully improve synthesis of aromatic compounds, such as 3-deoxy-D-arabino-heptulosonate-7-P (Baez et al. 2001), shikimate (Escalante et al. 2010) and L-phenylalanine (Baez-Viveros et al. 2004). The other strategy was to increase transcriptional levels of galP and glk gene in PTS− strains through plasmid overexpression (Hernandez-Montalvo et al. 2003; Wang et al. 2006). The engineered strain with ptsG deletion and galP overexpression successfully increased succinate production (Wang et al. 2006). However, plasmid overexpression has several disadvantages for engineering of genetically stable strains (Jarboe et al. 2010; Keasling 2008). Plasmid maintenance was a metabolic burden on the host cell, especially for high-copy number plasmids (de la Cueva-Mendez and Pimentel 2007), and only few natural unit-copy plasmids had the desirable genetic stability (Keasling 2008). In addition, only low-copy number plasmids had replication that was timed with the cell cycle; thus, it was difficult to maintain consistent copy number in all cells (Keasling 2008). On the other hand, gene expression was mainly modulated by inducible promoters, such as lac, trc, tac, and ara. Expensive inducers [isopropyl-β-D-thio-galactoside (IPTG) and arabinose] were needed for these promoters, which would be not feasible for large-scale production of bulk chemicals. It is thus desired to modulate gene expression directly in chromosome with constitutive regulatory parts. Promoter libraries had been developed as a solution, which could provide constitutive promoters with a wide range of strength (Alper et al. 2005; Meynial-Salles et al. 2005; Solem and Jensen 2002). In this study, libraries of artificial regulatory parts, including promoter and messenger RNA (mRNA) stabilizing region (mRS), were constructed in front of β-galactosidase gene (lacZ) in E. coli chromosome through λ-Red recombination (Datsenko and Wanner 2000). A convenient one-step recombination method was then used to modulate galP and glk gene expression directly in chromosome to improve alternative glucose utilization in a PTS− strain.

Materials and methods Strains, medium and growth conditions Strains used in this study are listed in Table 1. During strain construction, cultures were grown aerobically at 30°C, 37°C,

Appl Microbiol Biotechnol (2012) 93:2455–2462

or 39°C in Luria broth (per liter: 10 g Difco tryptone, 5 g Difco yeast extract, and 5 g NaCl) containing 20 g glucose/l. For βgalactosidase assay, colonies were inoculated into 15× 100 mm tubes containing 3 ml LB and grown at 37°C and 220 rpm overnight. Seed culture was then inoculated into 15× 100 mm tubes containing 3 ml LB (with the initial OD 0.05) and grown at 37°C and 220 rpm for 4 h. To investigate whether expression strength was affected by glucose repression and medium components, several representative colonies were cultivated aerobically in LB medium with 50 g glucose/ l and in AM1 medium (Martinez et al. 2007) with 50 g glucose/l for β-galactosidase assay. For the native lacZ promoter, 0.1 mM IPTG was added after 1 h growth, followed by additional 3 h growth. These representative colonies were also cultivated anaerobically in AM1 medium with 50 g glucose/l to obtain expression strength under anaerobic condition. Colonies were inoculated into 5-ml tubes fully filled with culture medium and grown at 37°C without shaking overnight. Seed culture was then inoculated into 5-ml tubes fully filled with culture medium (with the initial OD 0.05) and grown at 37°C without shaking for 6 h.

Construction of the first promoter library (P-Lib1) The native E. coli lacZ promoter in chromosome was designed to be replaced by a promoter library through homologous recombination, and the promoter strength can be measured through β-galactosidase assay. Linear DNA fragments were used to facilitate homologous recombination. The upstream region of lacI gene (up50, from −50 to 0 relative to translational start site of lacI) was used as the left homologous arm and the ribosomal binding site (RBS) and partial coding region of lacZ gene (RBS::lacZ’, from −18 to +32 relative to the translational start site of lacZ) was used as the right arm (Fig. 1a). P-Lib1 was designed based on bacteriophage lambda PL promoter (Love et al. 1996), with a sequence of TTATCTCTGGCGGTGTTGACA(N17)GATACT GAG CAC. P-Lib1 had constant −35 and −10 consensus sequences (bold) while randomizing sequences of the separating spacers. The other part of P-Lib1 had the same sequence as PL promoter. Two steps of PCR amplification were used to obtain DNA fragments for recombination (Fig. 1a). DNA fragment I was amplified from plasmid DNA of pKD4 using primer set lacI-FRT/PL1-FRT, which was then used as template to amplify DNA fragment II using primer set lacI-FRT/lacZ-R. DNA fragment II was used for recombination, replacing lacI::P-lacZ::mRS fragment in E. coli chromosome (Fig. 1a). After overnight growth, colonies were picked for β-galactosidase assay.

Appl Microbiol Biotechnol (2012) 93:2455–2462 Table 1 Escherichia coli strains and plasmids used in this work

2457 Relative characteristics

Sources

ATCC 8739 P1-1

Wild type ATCC 8739, FRT-Km-FRT::P1-1::lacZ

Zhang et al. 2009b This work

P2-15

ATCC 8739, FRT-Km-FRT::P2-15::lacZ

This work

M1-12

ATCC 8739, FRT-Km-FRT::M1-12::lacZ

This work

M1-64


This work

M1-30


This work

M1-46 M1-37

ATCC 8739, FRT-Km-FRT::M1-46::lacZ ATCC 8739, FRT-Km-FRT::M1-37::lacZ

This work This work

M1-93


This work

M1-162 PTSGalP12 GalP37 GalP93 Glk12 Glk37 GalP12-Glk12 GalP12-Glk37 GalP37-Glk12 GalP37-Glk37 GalP93-Glk12 GalP93-Glk37

ATCC 8739, FRT-Km-FRT::M1-162::lacZ ATCC 8739, ΔptsI PTS−, FRT-Km-FRT::M1-12::galP PTS−, FRT-Km-FRT::M1-37::galP PTS−, FRT-Km-FRT::M1-93::galP PTS−, FRT-Km-FRT::M1-12::glk PTS−, FRT-Km-FRT::M1-37::glk PTS−, FRT::M1-12::galP, FRT-Km-FRT::M1-12::glk PTS−, FRT::M1-12::galP, FRT-Km-FRT::M1-37::glk PTS−, FRT::M1-37::galP, FRT-Km-FRT::M1-12::glk PTS−, FRT::M1-37::galP, FRT-Km-FRT::M1-37::glk PTS−, FRT::M1-93::galP, FRT-Km-FRT::M1-12::glk PTS−, FRT::M1-93::galP, FRT-Km-FRT::M1-37::glk

This This This This This This This This This This This This This

bla γ β exo (Red recombinase), temperature-conditional replicon bla; FRT-Km-FRT bla flp temperature-conditional replicon and FLP recombinase

Datsenko and Wanner (2000)

Strains

work work work work work work work work work work work work work

Plasmids pKD46 pKD4 pFT-A

Construction of the second promoter library (P-Lib2) In order to increase the recombination efficiency, the left homologous arm was prolonged from 50 to 500 bp. The native mRS, RBS, and partial coding region of lacZ gene (from −38 to +12 relative to the translational start site of lacZ) of E. coli were used as the right homologous arm (Fig. 1b). The sequence of P-Lib2 is TTATCTCTGGCGGTGTTGACA(N14)TGRTATAAT(N6). The −35 consensus region and the upstream 15 nucleotides (italicized) had the same sequence as P-Lib1, while sequence of the −10 consensus region (bold) was changed to TATAAT, and the six nucleotides downstream were randomized. DNA fragment III was amplified from genomic DNA of one recombinant in P-Lib1 (P1-1) using primer set lacI-up500/pL-down-35, which was then used as template to amplify DNA fragment IV using primer set lacI-up500/lacZ-PL-R2. DNA fragment IV was used for recombination, replacing lacI::P-lacZ

Datsenko and Wanner (2000) Posfai et al. (1997)

fragment in E. coli chromosome (Fig. 1b). After overnight growth, colonies were picked for β-galactosidase assay. Construction of mRS library In order to increase mRNA stability after transcription, mRS library (M-Lib1) was constructed between promoter and RBS. P2-15, which had the highest strength in P-Lib2, was used for M-Lib1 construction. The left homologous arm was the same as in P-Lib2, while RBS and partial coding region of lacZ gene (from −12 to +38 relative to the translational start site of lacZ) was used as the right arm (Fig. 1c). The sequence of M-Lib1 is TATAATTGAGCC(N18)GTTT AAACCAGGAAACAGCT. Eighteen random nucleotides and a PmeI site were put between P2-15 (italicized) and the native RBS of lacZ gene (bold). DNA fragment V was amplified from the genomic DNA of P2-15 using primer set lacIup500/pL-down3, which was then used as template to amplify DNA fragment VI using primer set lacI-up500/lacZ-PL-R3C.

2458


Fig. 1 Constructing libraries of artificial regulatory parts. a Constructing the first promoter library (P-Lib1) by replacing the native promoter and mRS of E. coli lacZ with P-Lib1 and mRS of pL (m-L); b constructing the second promoter library (P-Lib2) by replacing the

native promoter of E. coli lacZ with P-Lib2; c constructing mRS library (M-Lib1) by replacing the native promoter and mRS of E. coli lacZ with an artificial promoter (P2-15) and M-Lib1

DNA fragment VI was used for recombination, replacing lacI::P-lacZ::mRS fragment in E. coli chromosome (Fig. 1c). After overnight growth, colonies were picked for βgalactosidase assay.

nucleotides upstream of the native regulatory region followed by 20 nucleotides homologous with FRT, while sequence of antisense primer included 50 nucleotides downstream of translational start codon followed by 15 nucleotides homologous with RBS of lacZ gene (Fig. 3). DNA fragments were amplified from genomic DNA of representative recombinants in M-Lib1 using primer set galP-up-FRT/galP-RBS-down and electroporated into competent cell of PTS− strain with pKD46. After overnight growth on LB plate with 50 mg kanamycin/l, several colonies were picked for PCR verification using primer set galP-up-FRT/galP-381R, followed by sequencing. Right colonies were designated as GalP12, GalP37, and GalP93. In order to further modulate glk gene expression, the FRT-Km-FRT cassette was removed using flippase treatment as described previously (Datsenko and Wanner 2000). Modulation of glk was the same as galP, and primers used are listed in online resource 1 (ESM_1).

Enzyme assays Exponentially grown cells were harvested by centrifugation (7,000×g for 5 min, 4°C). Cells were washed twice in 50 mM sodium phosphate buffer (pH 7.0), and the activity of βgalactosidase was measured as described by Miller (1992). One-step recombination method for modulation of galP and glk gene expression The ptsI gene of E. coli ATCC 8739 was deleted seamlessly as described previously (Jantama et al. 2008; Zhang et al. 2009b) to obtain a PTS− strain. GalP gene of PTS− strain was modulated by three different regulatory parts (M1-12, M1-37, and M1-93). A same primer set was used for modulation (Fig. 3). Sequence of sense primer included 50

Fermentation Strains were cultivated anaerobically for measuring growth and glucose utilization rate. Seed culture for fermentation

A

1 0.8 0.6 0.4 0.2 0

Colonies in P-Lib1

B

Analysis Cell growth was monitored by measuring the optical density at 550 nm. Glucose concentration was measured by highperformance liquid chromatography (Zhang et al. 2007).

Beta-galactosidase activity

was prepared by inoculating fresh colonies into a 250-ml flask containing 100 ml LB and 20 g glucose/l. After 16 h growth (37°C and 120 rpm), this culture was harvested by centrifugation (7,000×g for 5 min). Cells were washed twice in NBS medium (Martinez et al. 2007) and inoculated into a small fermentation vessel (400 ml) containing 250 ml NBS medium with 50 g glucose/l to provide an inoculum of 0.017 g dry cell weight per liter. Anaerobic fermentations were maintained at pH 7.0 by the automatic addition of 6 M potassium hydroxide.

2459

Beta-galactosidase activity


4 3 2 1 0

Results Construction of artificial regulatory parts libraries Near 250 colonies grew up after electroporation of 200 ng DNA fragment II into E. coli ATCC 8739 with pKD46. The recombination efficiency was 1,250 recombinants per microgram DNA. Forty colonies were randomly picked and grown in LB medium for β-galactosidase assay. The native E. coli lacZ promoter was also tested, which was 1.9 U/mg when induced with 0.1 mM IPTG and 0.02 U/mg without IPTG induction. The promoter strengths of P-Lib1 varied from 0.05 to 0.7 times induced lacZ promoter. Most strengths were