One step engineering of T7-expression strains for protein production

Abstract

The T7-expression system has been very useful for protein expression in Escherichia coli. However, it is often desirable to over-express proteins in species other than E. coli. Here, we constructed an inducible broad-host-range T7-expression transposon, which allows simple one-step construction of T7-expression strains in various species, providing the option to over-express proteins of interest in a broader host-range. This transposon contains the T7 RNA polymerase driven by the lacUV5 promoter, which is repressed by the lac-repressor. Leaky expression is prevented by the presence of T7-lysozyme on this construct. The complete T7-expression system is flanked by mariner transposon repeats of the suicidal R6Kγori plasmid, pBT20-Δbla. Stable integration of the whole system is possible by a one-step selection for a Flp-excisable Gm^R-marker. We showed the engineering of E. coli, Pseudomonas aeruginosa, Erwinia carotovora, Salmonella choleraesuis, Agrobacterium tumefaciens, and Chromobacterium violaceum strains with this construct and demonstrated the expression of the Burkholderia pseudomallei Asd protein in these hosts, by induction with isopropyl-β-D-thiogalactopyranoside (IPTG).

Introduction

For many laboratories, the T7-expression system in E. coli is still the first choice for protein expression. A primary reason for the success of this system is that target genes are cloned under the control of the very strong T7-promoter, which is not recognized by E. coli RNA polymerase [1-3]. Therefore, virtually no expression occurs until a source of T7 RNA polymerase is provided. In addition, leaky expression is further prevented by T7-lysozyme [4]. The gene encoding for the target protein is essentially “off”, and this reduces plasmid instability due to the production of proteins potentially toxic to the host cell. Many genes, difficult to express in E. coli promoter-based systems (e.g., tac, lac and trc), have been stably cloned and expressed in the T7-expression system in E. coli [5]. Hence, the T7-expression system in E. coli has been very useful to many laboratories.

However, expression in E. coli is not desirable in many instances due to numerous potential problems, including codon usage, toxicity, post-translational modification, compartmentalization, folding, solubility, and others. Thus, it may be desirable to over-express proteins in the native or alternative hosts to resolve these problems and produce enough proteins for subsequent purification. There has been very limited use of the T7-expression system in species other than E. coli [6-8], due to the lack of constructs available for easy engineering of expression strains in other species.

Over-expression of target proteins in their native hosts, other than E. coli, may resolve protein expression problems. Here, we have constructed an inducible broad-host-range T7-expression system, which could be used to create T7-expression strains of various species in one simple step. Protein expression problems are numerous and having an array of expression hosts to test the expression of target proteins is the first step towards a resolution.

Materials and Methods

Bacterial strains, plasmids, and media

The bacterial strains and plasmids used in this study are listed in Table 1. E. coli, E. carotovora, S. choleraesuis, A. tumefaciens, and C. violaceum strains were grown in LB (Difco). P. aeruginosa strains were grown in Pseudomonas Isolation Agar (PIA, Difco). Antibiotics were added to the media for selection and plasmid maintenance as follows: for E. coli, E. carotovora, S. choleraesuis, A. tumefaciens, and C. violaceum, gentamycin (Gm) 15 μg/ml was used; for E. coli, E. carotovora and S. choleraesuis, ampicillin (Ap) 110 μg/ml was used; 10 μg/ml tetracycline (Tet) was used for A. tumefaciens; and 50 μg/ml chromophenicol (Cm) was used for C. violaceum; for P. aeruginosa, Gm 150 μg/ml and carbenicillin (Cb) 500 μg/ml were used. For the growth of E. coli HPS1-mob-Δasd-pir116 strain, 100 μg/ml of diaminopimelic acid (DAP) was supplied.

Table 1.

Bacterial strains and plasmids^*

Strains/plasmids		Relevant properties	Reference
Strains
E. coli	DH5α	Φ80dlacZΔM15, Δ(lacZYA-argF)U169 deoR recA1 endA1 hsdR17 (rK- mK+) supE44 λ- thi-1 gyrA96 relA1	Invitrogen
	DH5α-T7pol-Gm-FRT	GmR, E. coli T7-expression strain	This study
	HPS1-pir116	e14- (mcrA) recA1 endA1 gyrA96 thi-1 hsdR17 supE44 relA1Δ(lac-proAB) rif zxx::miniTn5Lac4 (lacIq lacZΔM15) uidA::pir116 zdg-232::Tn10	Laboratory collection
	HPS1-mob-Δasd-pir116	e14- (mcrA) recA1 endA1 gyrA96 thi-1 hsdR17 supE44 relA1Δ(lac-proAB) rif zxx::miniTn5Lac4 (lacIq lacZΔM15) Δasd::FRT uidA::pir116 zdg-232::Tn10	Laboratory collection
P. aeruginosa	PAO1	Prototroph	[16]
	PAO1-T7pol-Gm-FRT	GmR, P. aeruginosa T7-expression strain	This study
	PAO1-T7pol-FRT	P. aeruginosa T7-expression strain with the GmR-cassette excised	This study
E. carotovora	EC14	Prototroph	[17]
	EC-T7pol-Gm-FRT	GmR, E. carotovora T7-expression strain	This study
S. choleraesuis	ATCC #14028	Prototroph	[18]
	SC-T7pol-Gm-FRT	GmR, S. choleraesuis T7-expression strain	This study
A. tumefaciens	NTL4	Prototroph	[19]
	AT-T7pol-Gm-FRT	GmR, A. tumefaciens T7-expression strain	This study
C. violaceum	CV026	cvil::mini-Tn5	[20]
	CV-T7pol-Gm-FRT	GmR, C. violaceum T7-expression strain	This study
Plasmids	pPS1072	ApR; pUC19 with T7pol and lacI	[7]
	pLysS	ApR; T7-lys containing plasmid	Novagen
	pPS1072-lys	ApR; T7-lys cloned into pPS1072	This study
	pPS856	GmR, ApR ; GmR cassette flanked by FRTs	[10]
	pPS1072-lys-Gm	GmR, ApR ; GmR cassette cloned into pPS1072-lys	This study
	pCD11PKS	CmR; R6Kγori plasmid	[21]
	pR6K-lys-Gm	GmR; R6Kγori and attP cloned into pPS1072-lys-Gm	This study
	pUC19	ApR; cloning vector	[22]
	pUC19-T7pol-lacI	GmR, ApR; GmR, lacI and T7pol cloned into pUC19	This study
	pUC19-T7pol-lacI-lys	GmR, ApR; T7-lys cloned into pUC19-T7pol-lacI	This study
	pBT20	GmR, ApR; mariner transposon	[14]
	pBT20-Δbla	GmR; pBT20 with bla gene deleted	This study
	pBT20-Δbla-T7pol	GmR; T7-expression system transposon plasmid	This study
	pFLP2	ApR; flp containing plasmid	[10]
	pPS910	ApR; B. pseudomallei strain 1026b asd containing plasmid	Schweizer, H. P.
	pET19b	ApR; T7-expression plasmid	Novagen
	pET19b-asd	ApR; asd cloned into pET19	This study
	pET19b-asd-ΔlacI	ApR; pET19b-asd with deletion of lacI	This study
	pUCP30NT	GmR; broad-host-range T7-expression plasmid	Laboratory collection
	pUCP30NT-asd	GmR; asd cloned into pUCP30NT	This study
	pBBR1MCS-Cm	CmR; broad-host-range cloning vector	[23]
	pBBR1MCS-Cm-asd	CmR; asd cloned into pBBR1MCS-Cm	This study
	pBBR1MCS-Tet	TetR; broad-host-range cloning vector	[15]
	pBBR1MCS-Tet-asd	TetR; asd cloned into pBBR1MCS-Tet	This study

^*For plasmids and strains constructed in this study, please see text for further details.

Recombinant DNA techniques

Standard molecular biology procedures were used as described [9]. Restriction enzymes, T4 DNA ligase, T4 DNA polymerase were purchased from New England Biolabs and used as recommended by the supplier. Taq and Pfu DNA polymerases (Stratagene) were used for PCR. Restriction digests and PCR products were run on 1% agarose gel for separation and eluted by utilizing the Zymoclean Gel DNA Recovery Kit (Zymo Research Corp.). Zyppy Plasmid Miniprep kit (Zymo Research Corp.) was used for plasmid purification.

Construction of pBT20-Δbla-T7pol

T7 gene 3.5, T7-lysozyme encoding gene (T7-lys), was amplified from pLysS (Novagen) by PCR using oligos 5′-GGGAAGCATGCGTGGTCTCCCTTTAGTGAGTTC-3′ and 5′-GGAATGGTGCATGCAAGGAGATG-3′, and the PCR product was digested with SphI (underlined) and cloned into pPS1072 digested with the same enzyme. The resulting plasmid pPS1072-lys was cut with SalI and blunt-ended, and the gentamycin-resistance-cassette (Gm^R-cassette) of pPS856 (SacI digested and blunt-ended) was cloned into it, resulting in pPS1072-lys-Gm. Next, R6Kγori and λ-attP were amplified from pCD11PKS using oligos 5′-AGATGTGTATAAGAGACAGCCCGGGATCAGCAGTTCAACCTGTTG-3′and 5′-AGATGTGTATAAGAGACAGCCCGGGAAATCAAATAATGATTTTATTTTG-3′, which was then ligated into the SmaI + HindIII digested and blunt-ended pPS1072-lys-Gm to create pR6K-lys-Gm (Fig. 1A). pR6K-lys-Gm was digested with BamHI and the 4.4 kb fragment, containing the T7-pol and the lacI genes (Fig. 1A), was cloned into the BamHI site of pUC19. The resulting plasmid pUC19-T7pol-lacI was cut with SmaI, blunt-ended, purified and then cut with MluI. In parallel, pR6K-lys-Gm was cut with BsaI, blunt-ended, purified, and then cut with MluI. The 2.2 kb fragment, containing the Gm^R-cassette and the T7-lys genes, was recovered and then ligated with the 6.6 kb fragment of pUC19-T7pol-lacI to construct pUC19-T7pol-lacI-lys (Fig. 1B).

Figure 1.

Genetic maps of T7 vectors. (A) and (B) are intermediates of the T7 transposon construct. (C) Indicates the T7 over-expression transposon flanked by mariner repeats. The mariner transposase recognizes two 28bp inverted mariner repeats and integrates the whole region between these two repeats into the chromosome of the host cell. Abbreviations: aacC1, gentamycin acetyl transferase encoding gene; bla, β-lactamase encoding gene; FRT, Flp-recombination-target; lacI, lac repressor; MR, mariner repeat; ori, ColE1 origin; oriT, conjugal origin of transfer; P_lac, lacUV5 promoter; R6Kγori, R6K origin of replication; T7-lys, T7-lysozyme; T7-pol, T7 RNA polymerase; Tnp, mariner transposase. These plasmid sequences have been deposited in GenBank, and the accession numbers are EF153730 (A), EF153731 (B), and EF153732 (C).

The final T7-system on a transposon plasmid was constructed in several steps. pBT20 was digested with SpeI and NotI, blunt-ended, self-ligated, and transformed into HPS1-pir116 to delete the β-lactamase encoding gene, yielding pBT20-Δbla. The pUC19-T7pol-lacI-lys (Fig. 1B), was digested with SacI + HindIII, and blunt-ended; this 6.2 kb fragment, carrying the T7-lysozyme, Gm^R, LacI and T7 RNA polymerase encoding gene, was introduced between the mariner repeats of pBT20-Δbla (XbaI digested and blunt-ended) to yield pBT20-Δbla-T7pol (Fig. 1C).

Construction of expression vectors

The asd gene was amplified from pPS910 (B. pseudomallei asd containing plasmid) using oligos 5′-GGTTACATATGAACGTAGGTCTCGTAGGTTG-3′ and 5′-TTTCGGGATCCGTACGCGTCGTAATCGCGTAGT-3′. The 1.1 kb asd-fragment was digested with NdeI + BamHI (underlined) and ligated into the pET19b and pUCP30NT vectors digested with the same enzymes to construct pET19b-asd (Fig. 3A) and pUCP30NT-asd (Fig. 3B), respectively.

Figure 3.

Genetic maps of the T7-expression vectors. (A) pET19b-based T7-expression vector, used for the strains in which the ColE1 origin can replicate including E. coli, E. carotovora and S. choleraesuis. pUC-based high-copy origins should not be used in these strains, because of the leaky expression of target proteins (see text). (B) Broad-host-range T7-expression vector utilized in this study for P. aeruginosa. (C) Broad-host-range T7-expression vector used for A. tumefaciens. (D) Broad-host-range T7-expression vector used for C. violaceum. (E) pET19b-asd with the lacI deletion used for E. coli, E. carotovora, S. choleraesuis. The target asd gene, encoding the Asd protein of B. pseudomallei, was introduced downstream of the T7-promoter (P_T7), and upstream of T7-terminator. Abbreviations: Cm, chromophenicol resistant cassette; T7T, T7-terminator; Tet, tetracyclin resistant cassette.

We constructed pBBR1MCS-Tet-asd, pBBR1MCS-Cm-asd, and pET19b-asd-ΔlacI in several steps. For the construction of pBBR1MCS-Tet-asd (Fig. 3C), the pET19b-asd was digested with XbaI, blunt-ended, purified, and cut with HindIII; and then, the 1.3 kb fragment containing ribosomal-binding-site (RBS), asd, and T7-terminator was cloned into pBBR1MCS-Tet digested with SmaI + HindIII. To construct the T7-expression vector pBBR1MCS-Cm-asd (Fig. 3D), the pET19b-asd was digested with XbaI + HindIII, and the 1.3kb fragment containing RBS, asd and T7-terminator was cloned into pBBR1MCS-Cm digested with the same enzymes. pET19b-asd was digested with BglII + BspMI, blunt-ended, and the fragment without lacI was self-ligated to yield pET19b-asd-ΔlacI (Fig. 3E).

Construction of T7-expression strains

E. coli HPS1-mob-Δasd-pir116 strain was used as the conjugal donor to introduce the T7-transposon of pBT20-Δbla-T7pol into E. coli, P. aeruginosa, E. carotovora, S. choleraesuis, A. tumefaciens, and C. violaceum. All strains were grown to log-phase prior to conjugation. Conjugation was done by mixing 0.7 ml of the donor strain and 0.7 ml of each recipient strain, individually. The tube was centrifuged at 9,000 × g for 1 minute, and all but 30 μl of the supernatant and cell pellet was discarded. The 30 μl mixture of cell suspension was spotted onto cellulose acetate filters (Sartorius) on LB agar plates and incubated at 37°C for 6 hours. Filters were then vortexed in 1 ml LB broth, and dilutions were plated on LB + Gm15 or PIA + Gm150 plates without DAP to prevent growth of the E. coli donor. Gm^R transconjugants were purified and screened through PCR using oligos 5′-AGCTCCTCGAGCTTGAACGAATTGTTAGGTGGC-3′ and 5′-GCGCGTCTAGACATAAGCCTGTTCGGTTCG-3′, which anneal upstream and downstream of the Gm^R-cassette, respectively.

To express the target asd gene on a Gm^R plasmid (pUCP30NT-asd), the P. aeruginosa strain, PAO1-T7pol-Gm-FRT, was transformed with pFLP2 to excise the Gm^R-cassette from this strain as previously described [10]. pFLP2 was cured by streaking PAO1-T7pol-Gm-FRT harboring pFLP2 on PIA supplemented with 5% sucrose. Colonies from the PIA sucrose plates were patched onto PIA ± Gm150 to confirm the excision of the Gm^R-cassette, in addition to patching on PIA + Cb500 plates to make sure pFLP2 was cured. pUCP30NT-asd was conjugated into PAO1-T7pol using the HPS1-mob-Δasd-pir116 donor strain and selected on PIA + Gm150.

To transform plasmids into E. coli, E. carotovora, S. choleraesuis, A. tumefaciens, and C. violaceum, electro-competent cells were prepared essentially as previously described [11]. All five different T7-expression strains were grown at 37 °C in LB broth; in mid-log phase (OD₆₀₀∼ 0.6) flasks were placed in ice water and chilled for 15 minutes. And then cells were harvested, washed twice in ice-cold 1 mM HEPES, pH 7.0, and resuspended at 10¹¹ cells/ml in 10% glycerol. 40 μl aliquots were flash-frozen in dry ice/ethanol bath and stored at -80 °C. The cells were thawed on ice and electroporated with 1 μl of plasmid DNA, recovered in 1 ml SOC at 37°C for 1∼2 hours, and then plated on LB plates containing appropriate antibiotics. pET19b-asd or pET19b-asd-ΔlacI were electroporated into competent cells of DH5α-T7pol-Gm-FRT, EC-T7pol-Gm-FRT and SC-T7pol-Gm-FRT, which were selected on LB + Ap110. pBBR1MCS-Tet-asd was electroporated into AT-T7pol-Gm-FRT and selected on LB + Tet10, and pBBR1MCS-Cm-asd was electroporated into CV-T7pol-Gm-FRT and selected on LB + Cm50.

Protein expression in newly engineered strains

Expression of the B. pseudomallei Asd protein from T7-expression strains by IPTG induction was performed from whole-cell extracts as described below. All six T7-expression strains harboring the different T7-expression vectors were grown overnight at 37°C in 3 ml of LB broth with the appropriate antibiotics. The overnight cultures were diluted 1:100 into fresh LB broth with antibiotic and grown at 37°C to OD₆₀₀ of 0.8∼1.0, and then induced with 1 mM IPTG for 4 hours. One ml of each cell culture was harvested, resuspended in 200 μl of 2 × SDS-PAGE loading buffer and boiled for 5 minutes. Total cellular proteins were run along with Broad-Range Protein Markers (New England Biolabs) or Kaleidoscope Prestained Standards (BioRad) on a 10% SDS-PAGE gel and visualized by quick staining with Coomassie Brilliant Blue R-250 [12].

Stability Study

Studies were carried out in E. carotovora and S. choleraesuis to test the maintenance of the T7-transposon insertion on the chromosome, in non-selective condition. The overnight cultures (∼10⁹ CFU/ml) of two T7-expression strains, EC-T7pol-Gm-FRT and SC-T7pol-Gm-FRT, were diluted 10⁶ times (∼10³ CFU/ml) with LB media and grown for 24 hours, to ∼10⁹ CFU/ml, individually, and dilutions plated on LB ± Gm15. This process was repeated 5 times and the colonies were counted for all the plates. Since 10⁶ dilution is approximately ∼2²⁰, each time 10⁶ dilution was made, this was taken as 20 generations. The percent of Gm^R colonies were calculated based on the total number of colonies on LB media.

RNA isolation and Real-Time RT-PCR

Cells of DH5 α-T7pol-Gm-FRT, EC-T7pol-Gm-FRT, AT-T7pol-Gm-FRT, and CV-T7pol-Gm-FRT were harvested at mid-log phase (OD₆₀₀ 0.8∼1.0) for RNA isolation. Total mRNA was isolated using Qiagen RNeasy mini kit (Qiagen), with an additional on-column DNaseI treatment. Final RNA concentrations and purities were determined on a Beckman DU7500 spectrophotometer (A_260/280 between 1.8 and 2.0). 4 μg RNA aliquots were treated with 2 units of RQ1 DNaseI (Promega) for the second time in 2 μl of iScript cDNA synthesis buffer (BioRad) for 10 minutes at 37°C. DnaseI was inactivated at 70°C for 10 minutes and then the samples were chilled on ice. Then 14 μl of iScript cDNA synthesis buffer, 4 μl iScript Reverse Transcriptase (BioRad) and 52 μl of diethyl pyrocarbonate (DEPC) double destilled water (DDW) was added to the same tube. cDNA was synthesized according to recommendations by BioRad. The final product were taken up to 500 μl with DDW (no DEPC), and 10 μl of 100 × dilutions were used as templates for each quantitative real-time RT-PCR run as described below.

Primers for each gene in the real-time RT-PCR reactions were designed using Integrated DNA Technologies Primer Quest software (http://www.idtdna.com/). Forward and reverse primer pairs 5′-AACTCCCGATGAAACCGGAAGACA-3′/ 5′-ATGAACTCAAGGCTGATACGGCGA-3′ and 5′-GCAATCTTTGTTCACTGCTCGGCT-3′/5′-AAAGTGGTATCCCACATCGAGCCA-3′ were used for T7pol and T7-lys, respectively.

Supermixes for all reactions were made and aliquoted into sub-supermixes for each gene assayed. Each real-time PCR reaction contained 10 μl of cDNA, 12.5 μl of iQ™ SYBR Green Supermix (BioRad), 120 nM of each forward and reverse primers, in a final volume of 25 μl. Real-time PCR was performed on the iCycler iQ™ (BioRad) with the following protocol: denaturation (95°C for 2 minutes), 55 cycles of amplification and quantification (95°C for 20 seconds, 60°C for 30 seconds and 72°C for 20 seconds). Melt curve analysis was performed on all reactions and revealed single peak.

We followed the method of Peirson et al. [13], as previously reported, to give more accurate quantitative real-time PCR data. Real-time RT-PCR values were averaged over 8 replicates for each gene and fold-changes were calculated using the amplification plot method and the available macro for Data Analysis of Real-Time PCR (DART-PCR) [13]. Accordingly, the average efficiencies of each gene are within 3.7% differences, allowing accurate analysis. The expression level of T7-pol gene was taken as 1 for each T7-expression strains being tested, and the expression of T7-lys was normalized relative to this value.

Results and Discussion

The various components of pBT20-Δbla-T7pol make it easy and convenient to construct and screen for the T7 over-expression strains. After introduction of pBT20-Δbla-T7pol into the recipient cell, the mariner transposase (Fig. 1C) will be expressed and integrates the region between the two mariner repeats (Fig. 1C) randomly [14] onto the host chromosome, resulting in the T7 RNA polymerase driven by the lacUV5 promoter. The lacI and T7-lysozyme were also included (Fig. 2) to prevent the leaky expression of T7 RNA polymerase. The Gm^R-cassette (Fig. 1C) on the transposon of pBT20-Δbla-T7pol aids in the selection of this vector and the transposition, in addition to allowing the screening of T7 over-expression strains by PCR. The identical FRT sequences (Fig. 1C) flanking the Gm^R-cassette allow the excision of the Gm^R-cassette by Flp recombinase to recycle this resistant marker [10].

Figure 2.

Overview of the T7 protein over-expression system. (A) When IPTG is absent, Lac repressor encoded by lacI on the transposon of T7-expression strain binds to the lac-operator, preventing the expression of T7 RNA polymerase by host RNA polymerase. However, the interaction between LacI and lac-operator is not completely tight, and the inherent leakiness of the lac-promoter allows some degree of transcription of T7-pol in the uninduced state. This leaky expression of T7 RNA polymerase is further controlled by its natural inhibitor, T7-lysozyme. In addition, the expression of the target protein is also suppressed by the LacI binding to the plasmid-located lac-operator on the pET19b. (B) Upon induction with IPTG, lac-repressor is removed from the lac-operator, the amount of T7-lysozyme present does not fully titrate out the highly expressed T7 RNA polymerase driven by the lac-promoter. The addition of IPTG also removes the LacI from the expression vector. Consequently, the T7 RNA polymerase binds to the T7-promoter (free of LacI) upstream of the target gene to initiate the expression of the target Asd protein. Abbreviations: asd, B. pseudomonallei strain 1026b aspartate-semiaidehyde dehydrogenase; lacO, lac-operator; P_T7, T7-promoter.

With the T7-expression transposon on pBT20-Δbla-T7pol, we engineered inducible T7-expression strains of various species in one simple step. We chose six different species, E. coli (γ-Proteobacteria), E. carotovora (γ-Proteobacteria), S. choleraesuis (γ-Proteobacteria), P. aeruginosa (γ-Proteobacteria), A. tumefaciens (α-Proteobacteria), and C. violaceum (β-Proteobacteria) to test this T7-transposon. Here, the utility of this T7-expression transposon was successfully tested in all six different species, with the appropriate expression vectors. These T7-expression strains were constructed by easily introducing the T7-expression transposon, pBT20-Δbla-T7pol, and selecting for the transposition events on media containing Gm.

Four expression vectors, pET19b-asd, pUCP30NT-asd, pBBR1MCS-Tet-asd, and pBBR1MCS-Cm-asd (Fig. 3) were constructed for the over-expression of the target B. pseudomallei Asd protein in various species. pET19b-asd, which has the pBR322-based ColE1 origin of replication (Fig. 3A), can be used in E. coli and related species. The T7-expression vector pUCP30NT-asd (Fig. 3B) has the pUC-based ColE1 origin allowing cloning in E. coli, in addition to the broad-host-range ori1600-rep origin. And then, the Gm^R-cassette of PAO1-T7pol::Gm-FRT was removed using Flp recombinase in order to recycle the same antibiotic marker for use on the T7-expression vector, pUCP30NT-asd (Fig. 3B). Broad-host-range T7-expression vectors pBBR1MCS-Tet-asd (Fig. 3C) and pBBR1MCS-Cm-asd (Fig. 3D) have ori-IncC and Rep protein, which have been shown to replicate in all Gram-negative species tested [15]. The cultures of DH5α-T7pol-Gm-FRT/pET19b-asd, EC-T7pol-Gm-FRT/pET19b-asd, SC-T7pol-Gm-FRT/pET19b-asd, PAO1-T7pol-FRT/pUCP30NT-asd, AT-T7pol-Gm-FRT/pBBR1MCS-Tet-asd, and CV-T7pol-Gm-FRT/pBBR1MCS-Cm-asd were grown to log-phase and induced with IPTG. All six strains showed significant expression of the Asd protein, as compared to the cultures without IPTG induction (Fig. 4), indicating that the T7-expression system is functional in all six species. The transposon located lacI is sufficient and no lacI on the expression vector is required, because the single copy of the lacI on the chromosome inhibit detectable expression of Asd when low (∼2-5) to medium (∼15-20) copy T7-expression vectors were used (Fig. 4B, 4E, and 4F). We also tested the stability of the integrated transposon. The E. carotovora and S. choleraesuis T7-expression strains were used for stability studies, as described in Materials and Methods, and 100% maintenance of the integrated T7-transposon was shown in both species, in the non-selectable condition after 100 generations (Fig. 5).

Figure 4.

Over-expression of Asd (∼39 KDa) in novel engineered T7-expression strains. Arrows indicate the over-expressed B. pseudomallei Asd protein upon induction with IPTG in six different T7-expression strains. The T7-expression strains with the respective expression vectors are DH5α-T7pol-Gm-FRT/pET19b-asd (A), EC-T7pol-Gm-FRT/pET19b-asd (B), SC-T7pol-Gm-FRT/pET19b-asd (C), PAO1-T7pol-FRT/pUCP30NT-asd (D), AT-T7pol-Gm-FRT/pBBR1MCS-Tet-asd (E), and CV-T7pol-Gm-FRT/pBBR1MCS-Cm-asd (F).

Figure 5.

Stability study of the integrated T7-transposon in E. carotovora and S. choleraesuis. EC-T7pol-Gm-FRT and SC-T7pol-Gm-FRT were grown in the absence of Gm for 100 generations, which indicate the transposon generates a stable integration.

Although the single copy of lacI on transposon is sufficient to inhibit low and medium copy number plasmids, it is not enough to completely inhibit very high (∼500-700) copy plasmids (e.g. pUC-derived plasmids). For example, pUCP30NT-asd only works properly in the species in which the pUC-based ColE1 origin is not functional, when the ori1600-Rep acts as the origin of replication. We initially used pUCP30NT-asd for T7-expression strains of E. carotovora, S. choleraesuis, and P. aeruginosa and were able to detect inducible expression of the Asd protein only in P. aeruginosa strain PAO1-T7pol (Fig. 4B). However, we did not see a difference in the over-expression of the Asd protein of pUCP30NT-asd in the E. carotovora and S. choleraesuis T7-expression strains when comparing induced and uninduced conditions, although the Asd band was clearly present on 10% SDS-PAGE gel in both conditions (data not shown). Since E. carotovora and S. choleraesuis are closely related to E. coli, and the pUC-based ColE1 origin is functional in these species, we predicted the reason for no difference in induced versus uninduced conditions of target protein is that pUCP30NT-asd is a high copy plasmid (∼500-700 copies per cell) in E. carotovora and S. choleraesuis. There was constitutive expression of the target Asd protein, because the high copy number of the lac-operator titrated out the lac-repressor (LacI) and causes constitutive protein expression without IPTG induction. This is not desirable because the constitutive expression of target protein may be toxic to the host cells. However, this problem was solved by using pET19b-asd (pBR322-based ColE1 origin, ∼15-20 copies per cell) instead of pUCP30NT-asd for E. carotovora and S. choleraesuis T7-expression strains, and the regulated expression of Asd protein with IPTG induction is observable in Fig. 4A, 4C, and 4D. Hence, low (∼2-5) to medium (∼15-20) copy plasmids should be utilized in novel hosts to tightly control expression prior to induction, and pUC-derived plasmids should not be used. To further demonstrate that the expression of a single copy of lacI gene on the chromosome is sufficient to repress protein expression, we deleted the lacI from the pET19b-asd and tested the new vector pET19b-asd-ΔlacI in E. coli, E. carotovora and S. choleraesuis. The over-expression of Asd protein with pET19b-asd-ΔlacI in all three T7-expression strains, DH5α-T7pol-Gm-FRT, EC-T7pol-Gm-FRT and SC-T7pol-Gm-FRT, were detected only in induced conditions, as shown in Fig. 6.

Figure 6.

Deletion of plasmid-borne lacI shows sufficient repression of the T7 system by the transposon-linked lacI. Over-expression of Asd protein with pET19b-asd-ΔlacI were detected in E. coli (A), E. carotovora (B) and S. choleraesuis (C) T7-expression strains. Arrows indicate the over-expressed Asd protein upon induction with IPTG in DH5α-T7pol-Gm-FRT (A), EC-T7pol-Gm-FRT (B), and SC-T7pol-Gm-FRT (C), using the constructed T7-expression vector pET19b-asd-ΔlacI.

Finally, to determine if the T7-lysozyme expressed is sufficient to inhibit basal expression of T7-polymerase, Real-time RT-PCR experiments were performed on four T7-expression strains (E. coli, E. carotovora, A. tumefaciens and C. violaceum). In all four species, the T7-lys was induced several folds (4.5∼7.5) more than T7-pol (Fig. 7), suggesting that the expression of T7-lysozyme is high enough to prevent the leaky expression of T7-polymerase. Also, the successful repression and induction of Asd protein (Fig. 4 and 6) indicates that the level of T7-lysozyme supplied was not too high, resulting in interfering with induction of target genes.

Figure 7.

Over expression of T7 lysozyme relative to the leaky expression of T7 RNA polymerase in uninduced condition. Real-time RT-PCR data for T7-pol and T7-lys genes in the T7-expression strains, E. coli (A), E. carotovora (B), A. tumefaciens (C), and C. violaceum (D). (A) The relative transcriptional levels of T7 lysozyme and T7 RNA polymerase encoding genes were determined in E. coli T7-expression strain, DH5α-T7pol-Gm-FRT, using real-time RT-PCR. The numbers were generated from 8 replicates of each gene, and the averages were shown with standard errors, after adjusting the transcription level of T7-pol to 1. Similar experiments were performed on EC-T7pol-Gm-FRT (B), AT-T7pol-Gm-FRT (C) and CV-T7pol-Gm-FRT (D).

In summary, there are several advantages of this T7 protein over-expression system. First, there is no antibiotic needed to grow the T7-expression strains, since the construct is stably integrated into the chromosome. The lacI and T7-lys expressed on the same transposon prevents the detectable leaky expression of target protein. In addition, the Gm^R-cassette on the construct allows the screening of transposon mutants, and the FRTs flanking the Gm^R-cassette allows the excision of Gm^R-cassette with Flp-recombinase after transposition. The recycling of the Gm^R-cassette is important for reusing the same antibiotic resistant-cassette in the expression vector, especially when there is a limited number of useful antibiotic markers for certain species. With this T7-expression construct, we have easily create novel over-expression strains of various species in one simple step, which allowed protein expression in different expression hosts other than E. coli. Because protein expression problems are numerous and unique to each protein, this novel transposon construct will allow engineering of an array of different T7 over-expression hosts to test and potentially find a resolution.