Construction of a Novel Shuttle Vector for Tetragenococcus species based on a Cryptic Plasmid from Tetragenococcus halophilus

A cryptic plasmid (pTH32) was characterized from Tetragenococcus halophilus 32, an isolate from jeotgal, Korean traditional fermented seafood. pTH32 is 3,198 bp in size with G+C content of 35.84%, and contains 4 open reading frames (ORFs). orf1 and orf2 are 456 bp and 273 bp in size, respectively, and their translation products showed 65.16% and 69.35% similarities with RepB family plasmid replication initiators, respectively, suggesting the rolling-circle replication (RCR) mode of pTH32. orf3 and orf4 encodes putative hypothetical protein of 186 and 76 amino acids, respectively. A novel Tetragenococcus-Escherichia coli shuttle vector, pMJ32E (7.3 kb, Emr), was constructed by ligation of pTH32 with pBluescript II KS(+) and an erythromycin resistance gene (ErmC). pMJ32E successfully replicated in Enterococcus faecalis 29212 and T. halophilus 31 but not in other LAB species. A pepA gene, encoding aminopeptidase A (PepA) from T. halophilus CY54, was successfully expressed in T. halophilus 31 using pMJ32E. The transformant (TF) showed higher PepA activity (49.8 U/mg protein) than T. halophilus 31 cell (control). When T. halophilus 31 TF was subculturd in MRS broth without antibiotic at 48 h intervals, 53.8% of cells retained pMJ32E after 96 h, and only 2.4% of cells retained pMJ32E after 14 days, supporting the RCR mode of pTH32. pMJ32E could be useful for the genetic engineering of Tetragenococcus and Enterococcus species.


Bacterial Strains, Plasmids and Culture Conditions
Bacterial strains and plasmids used in this work are listed in Table 1. Tetragenococcus species were cultivated in lactobacilli MRS broth (Acumedia, USA) or agar (1.5%, w/v) containing NaCl (5%, w/v) under anaerobic conditions at 30 o C. Enterococcus, Lactobacillus, Lactococcus and Weissella species were cultivated in the same media without NaCl. Escherichia coli DH5α was cultivated in Luria-Bertani (LB, Acumedia, USA) medium at 37 o C with aeration. The following antibiotics were added to the media as selective agents when required: ampicillin (Sigma-Aldrich, USA) (100 μg/ml for E. coli) and erythromycin (Sigma-Aldrich, USA) (200 μg/ml for E. coli, 50 μg/ml for Enterococcus and Tetragenococcus, 5 μg/ml for Lactobacillus, Lactococcus, and Weissella strains).

DNA Manipulations
Plasmid DNA was prepared from LAB by the method of O'Sullivan and Klaenhammer [19] and from E. coli DH5α using a commercial kit (FavorPrep plasmid extraction mini kit, Favorgen, Vienna, Austria). Restriction enzymes, T4 DNA ligase, and alkaline phosphatase (Promega, USA) were used according to the protocols provided by the suppliers. Polymerase chain reaction (PCR) amplifications were carried out with 10 to 100 ng DNA template and 10 pmol primers using Ex Taq DNA polymerase (Takara, Japan). Plasmid DNA and PCR amplicons were verified by agarose gels electrophoresis, stained with EcoDye DNA staining solution (Biofact, Korea) and visualized under UV light.

Construction of pMJ32E and Transformation of LAB and E. coli DH5α with pMJ32E
pTH32 was digested with HindIII and then ligated with pBluescript II KS(+) (Stratagene, USA). The This study recombinant (pMJ32) was obtained. Erythromycin resistance gene (ermC) was PCR amplified from pJY33E by using primer pair: ErmC-F (5′-CCCTCTAGACTTATCGGATAATAA-3′, XbaI site underlined) and ErmC-R (5′-GGGGGATCCACAAAAAATAGGCAC-3′, BamHI site underlined) [11]. The amplified ermC was inserted into pMJ32 at XbaI and BamHI sites. E. coli DH5α competent cells were prepared and transformed by standard protocols [20]. The plasmid DNA and E. coli competent cells were mixed, kept on ice for 1 min and transferred to a 0.1 cm electroporation cuvette. Electroporation was performed with a Gene Pulser II (BioRad, USA) under the following conditions: peak voltage, 2.1 kV; capacitance, 25 μF; resistance, 200 Ω. After pulsing, cells were resuspended in 1 ml of LB broth, incubated for 1 h at 37 o C, and spreaded onto LB plates containing antibiotics. Competent cells of Lactobacillus plantarum KCTC3104, Lactococcus lactis subsp. cremoris MG1363, and Weissella confusa CB1 were prepared and transformed according to a previous report [21]. Competent cells of Enterococcus and Tetragenococcus genera were prepared and transformed according to a previous report [22].

Expression of pepA from T. halophilus CY54
A pepA gene encoding aminopeptidase A (PepA) was PCR amplified from T. halophilus CY54 chromosomal DNA by using pepA-F (5′-CCCGTCGACTGCTGCTTCATAATG-3′, SalI site underlined) and pepA-R (5′-TTTCTCGAGCTGGCAACCTCAATC-3′, XhoI site underlined) primers [23]. The PCR amplified pepA was 2.6 kb in size, including its own promoter, and inserted into pMJ32E at SalI and XhoI sites. The resulting pMJ32EpepA was introduced into T. halophilus 31 and E. coli DH5α by electroporation, respectively. T. halophilus 31 and E. coli DH5α transformants (TFs) carrying pMJ32EpepA were grown in MRS broth (with 5% NaCl) and LB broth, respectively. Cells were harvested at the end of logarithmic growth phase by centrifugation (5,000 ×g for 15 min at 4 o C), washed 3 times with phosphate-buffered saline (PBS) and resuspended in PBS. T. halophilus cells resuspended in PBS with lysozyme (10 mg/ml) were incubated for 1 h at 37 o C before subjected to ultrasonication (Bandelin Electronic, Berlin, Germany). E. coli cells were disrupted by ultrasonication without lysozyme pretreatment. Disrupted cells were centrifuged at 12,000 ×g for 15 min at 4°C, and the supernatants were used as the cell extracts for PepA activity measurements. The protein concentration of the cell extract was determined by the Bradford method [24]. PepA activities in the cell extracts were measured according to previous report using glutamyl p-nitroanilide (glu-pNA) as the substrate [25]. One unit (U) of PepA was defined as the amout of enzyme that released 1 nmol of p-nitroaniline per minute under the assay conditions.

Stability of pMJ32E in T. halophilus 31
Stability of pMJ32E was examined by a method as described previously [10]. T. halophilus 31 TFs harbouring pMJ32E were grown in MRS broth with NaCl (5%) and erythromycin (Em, 50 μg/ml) for 48 h at 30 o C. Then the culture was used to inoculate (1%, v/v) fresh MRS broth containing NaCl (5%) but no Em, and the inoculated culture was incubated for 48 h at 30 o C. Subculturing was repeated up to 7 times without Em selection. Aliquots of each culture were spreaded onto MRS plates containing NaCl (5%) only and MRS plates containing NaCl (5%) and Em (50 μg/ml), respectively. The percentage of cells harboring pMJ32E was calculated by dividing the number of cells on MRS plates with Em by the number of cells on MRS plates without Em, and multiplied by 100.

Nucleotide Sequence
The complete nucleotide sequence of pTH32 was deposited to GenBank with the accession number MZ687077.

Results and Discussion
Characterization of pTH32 T. halophillus 32 was one of the isolates from jeotgal products, Korean traditional fermented seafoods by professor Lee group at Kyonggi university, Korea and kindly provided for this work. T. halophillus 32 possesses a small-sized plasmid, pTH32. Due to the small size, pTH32 was chosen as the frame for a Tetragenococcus -E. coli shuttle vector. pTH32 was digested with several restriction enzymes and it was found that pTH32 had a single recognition site for EcoRI, HindIII, NdeI, and NcoI. When pTH32 was double digested with EcoRI and HindIII, 2.2 and 1.0 kb fragments were generated. Each fragment was cloned into pUC19, resulting in pUCTH32L (pUC19 with 2.2 kb fragment) and pUCTH32S (pUC19 with 1.0 kb insert). The nucleotide sequences of the inserts were determined and combined into one sequence. pTH32 was 3,198 bp in length, and the G+C content was 35.84%, and it was within the known range of G+C contents of Tetragenococcus species [26].
Sequence analysis revealed the presence of 4 ORFs, 6 palindromic sequences (inverted repeat, IR I-VI), and 5 direct repeats (DR I-V) (Fig. 1). DRs and IRs observed in many plasmids are involved in the regulation of the replication processes of plasmids [27]. Many DRs and IRs are present upstream of orf1, the region where regulation of replication processes is expected to occur. orf1 can potentially encode a protein of 151 amino acids showing 65.16% sequence similarity with a RepB family plasmid replication initiator protein from Enterococcus faecium (WP_159373655.1). orf2 can potentially encode a protein of 90 amino acids showing 69.35% sequence similarity with a RepB family plasmid replication initiator protein from Staphylococcus aureus (WP_224670973.1). RepB proteins are known to have nicking-closing (topoisomerase I-like) activities for supercoiled DNA and are found among rolling circle replicating (RCR) plasmids [28]. orf1 and orf2 might form an operon since the distance between the end of ORF1 and the start of ORF2 is close, 62 nucleotides. Possible promoter sequences (-35 and -10) are located upstream of orf1 (underlined in Fig. 1). IRII located immediate downstream of ORF2 might serve as a transcription terminator. It can form an hairpin-loop structure with the free energy of -15.86 kcal/mol as determined by mfold program (http://www.unafold.org/mfold/applications/dna-folding-form.php). orf3 can potentially encode a protein of 185 amino acids showing 98.38% sequence similarity with a hypothetical protein from T. halophilus (WP_012478268.1). orf4 can potentially encode a protein of 74 amino acids showing 69.44% sequence similarity with a hypothetical protein from T. halophilus (WP_176446071.1). The functions of hypothetical proteins are unknown because they do not show any significant homologies to known proteins in the database. Considering the small size and significant homologies with RepB proteins, pTH32 was expected to replicate via RCR mode.

Construction of pMJ32E and Transformation of LAB
A Tetragenococcus -E. coli shuttle vector, pMJ32E, was constructed as described above (Fig. 2). Transformation of several LAB genera was tried with pMJ32E. pMJ32E successfully replicated in Enterococcus faecalis 29212 and Tetragenoccus halophilus 31. Transformation efficiency was 1.2 × 10 2 and 1.1 × 10 1 CFU/μg DNA for E. faecalis 29212 and T. halophilus 31, respectively. These efficiencies were lower than those reported for vectors based on pUCL287 replicon, which were in the range of 1.14 × 10 3 -2.1 × 10 4 CFU/μg plasmid [14,15]. T. halophilus 31 was isolated from jeotgal together with T. halophilus 32, and T. halophilus 31 harbors a 5 kb plasmid (results not shown). Among the several Tetragenococcus species tested, T. halophiluds 31 was the only strain transformed by pMJ32E. Transformation efficiency for T. halophilus 31 was quite low, but it could be improved if efforts to optimize the transformation conditions are tried. pMJ32E seems to have narrow host range since other LAB genera were not transformed with pMJ32E. Another possibility is that transformation of other LAB genera with pMJ32E might be more difficult compared with Tetragenococcus and other closely related LAB such as Enterococcus. More studies are necessary to clarify this issue.

Expression of pepA from T. halophilus CY54 in T. halophilus 31
To test the potential of pMJ32E as a cloning vector for Tetragenococcus species, expression of pepA gene in a heterologous host was tried. pMJ32EpepA was constructed as described above (Fig. 2). T. halophilus 31 [pMJ32EpepA] showed PepA activity of 49.8 U/mg protein, which was much higher than 9.1 U/mg protein of T. halophilus 31 harboring intact pMJ32E (control). E. faecalis 29212 [pMJ32EpepA] showed PepA activity of 39.2 U/mg protein, which was much higher than 20.2 U/mg protein of E. faecalis 29212 harboring intact pMJ32E (control). E. coli DH5α [pMJ32EpepA] showed PepA activity of 380 U/mg protein, which was also higher than 94 U/mg protein of E. coli DH5α harboring pMJ32E (control) ( Table 2). The results proved the usefulness of pMJ32E as a cloning vector for T. halophilus and E. faecalis.

Stability of pMJ32E in T. halophilus 31
The segregational stability of pMJ32E in T. halophilus 31 was monitored under the absence of Em selection (Fig. 3). After the first 28 generations, 53.8% cells retained pMJ32E and the portion of cells which lost pMJ32E increased rapidly thereafter. After 100 generations, only 2.4% of the cells still retained pMJ32E. The results are comparable to those observed among many RCR-type plasmids. RCR plasmids usually have low segregational stabilities due to the accumulation of ssDNA intermediates [27]. Theta type plasmids, which do not produce ssDNA intermediates, are more stable than RCR plasmids [27]. For example, a shuttle vector pFBYC018E based on a RCR plasmid (pFR18) showed just 3% stability after 100 generations [29]. pL11 based on another RCR plasmid (pSMA23) showed 2% and 6% stability in Lb. casei and Lb. gasseri, respectively, after 210 generations [30]. Whereas pSJ33E, a theta-replicating plasmid from Leuconostoc mesenteroides SY2, was maintained 100% in L. mesenteroides SY1 after 100 generations [10]. A shuttle vector, pRCEID-LC7.6, constructed by ligation of a theta replicating plasmid from Lactobacillus casei with pUC19 showed 16% stability in L. casei RCEID02 after 200 generations [31]. The segregational stability of pMJ32E in T. halophilus 31 was lower than those of thetareplicating vectors but similar to those of RCR plasmids, supporting the rolling circle replication of pMJ32E.
The copy number of pMJ32E in T. halophilus 31 was qualitatively compared with those of pSJE (a theta-type plasmid) in Levilavtobacillus brevis 2.14 [10] and pJY33E (a RCR-plasmid) in L. brevis 2.14 [11]. Three cultures were grown in MRS broth (10 ml) under the same conditions. When the OD 600 reached 0.6, plasmid DNA was prepared and dissolved in 40 μl distilled water. The same amount of DNA samples (5 μl) were analyzed by agarose gel electrophoresis and stained with ethidium bromide (Fig. 4). The results indicated that the copy number of pMJ32E was similar with those of pSJE and pJY33E.
A small cryptic plasmid, pTH32, was isolated from T. halophilus 32 and characterized. A novel E. coli -Tetragenococcus shuttle vector, pMJ33E, was constructed based on pTH32. pMJ32E successfully replicated in T. halophilus 31 and E. faecalis 29212. But no TFs were obtained for other LAB genera, which might be due to the narrow host range of pTH32 or failure to find optimum transformation conditions. A pepA gene from