Dissecting and engineering of the TetR family regulator SACE_7301 for enhanced erythromycin production in Saccharopolyspora erythraea

Background Saccharopolyspora erythraea was extensively utilized for the industrial-scale production of erythromycin A (Er-A), a macrolide antibiotic commonly used in human medicine. Yet, S. erythraea lacks regulatory genes in the erythromycin biosynthetic gene (ery) cluster, hampering efforts to enhance Er-A production via the engineering of regulatory genes. Results By the chromosome gene inactivation technique based on homologous recombination with linearized DNA fragments, we have inactivated a number of candidate TetR family transcriptional regulators (TFRs) and identified one TFR (SACE_7301) positively controlling erythromycin biosynthesis in S. erythraea A226. qRT-PCR and EMSA analyses demonstrated that SACE_7301 activated the transcription of erythromycin biosynthetic gene eryAI and the resistance gene ermE by interacting with their promoter regions with low affinities, similar to BldD (SACE_2077) previously identified to regulate erythromycin biosynthesis and morphological differentiation. Therefore, we designed a strategy for overexpressing SACE_7301 with 1 to 3 extra copies under the control of PermE* in A226. Following up-regulated transcriptional expression of SACE_7301, eryAI and ermE, the SACE_7301-overexpressed strains all increased Er-A production over A226 proportional to the number of copies. Likewise, when SACE_7301 was overexpressed in an industrial S. erythraea WB strain, Er-A yields of the mutants WB/7301, WB/2×7301 and WB/3×7301 were respectively increased by 17%, 29% and 42% relative to that of WB. In a 5 L fermentor, Er-A accumulation increased to 4,230 mg/L with the highest-yield strain WB/3×7301, an approximately 27% production improvement over WB (3,322 mg/L). Conclusions We have identified and characterized a TFR, SACE_7301, in S. erythraea that positively regulated erythromycin biosynthesis, and overexpression of SACE_7301 in wild-type and industrial S. erythraea strains enhanced Er-A yields. This study markedly improves our understanding of the unusual regulatory mechanism of erythromycin biosynthesis, and provides a novel strategy towards Er-A overproduction by engineering transcriptional regulators of S. erythraea. Electronic supplementary material The online version of this article (doi:10.1186/s12934-014-0158-4) contains supplementary material, which is available to authorized users.


Background
Saccharopolyspora erythraea, a Gram-positive industrial actinomycetes, produces a valuable macrolide antibiotic erythromycin A (Er-A). Er-A and its semi-synthetic derivatives, such as clarithromycin, azithromycin, dirithromycin, roxithromycin, and telithromycin, are widely used in medicine to treat infections caused by pathogenic Grampositive bacteria [1]. Considering its high industrial and medical importance, enhancement of Er-A production has been performed by metabolic engineering or classical mutagenesis methods over the past 60 years [2][3][4]. Also, as a model actinomycete system, S. erythraea has been used for studying the biosynthesis and combinatorial biosynthesis of polyketide antibiotics, which are synthesized by modular type I polyketide synthase (PKS) [5,6]. The erythromycin biosynthetic pathway has been investigated by genetic and biochemical approaches [7,8]. The erythromycin biosynthetic gene (ery) cluster contains 20 genes arranged in four major polycistronic units, spanning over 60 kb of DNA [9]. Unusually, the ery cluster lacks any regulatory genes in S. erythraea, hampering efforts to improve Er-A production by modulating gene regulation pathways.
Based on a DNA microarray strategy, the industrial overproducer strain was found to express the entire ery cluster several days longer than the wild-type strain, suggesting that there exist some regulators for controlling erythromycin biosynthesis [10]. In vitro and in vivo approaches led to the identification of BldD (SACE_2077), a key developmental regulator in actinomycetes [11], controlling erythromycin biosynthesis and morphological differentiation in S. erythraea [12]. BldD was found to bind to the promoter regions of ery cluster. However, the action of BldD on the promoters of ery genes was much weaker than on its own promoter [12]. Recently, a putative regulatory protein, SACE_5599, was proved to positively control erythromycin production and morphological differentiation in S. erythraea [13]. Nevertheless, the knowledge of molecular regulatory mechanisms controlling erythromycin biosynthesis remains limited.
In this study, we identified and characterized a novel TFR (SACE_7301) that positively regulated erythromycin biosynthesis. SACE_7301 was found to increase the transcription of erythromycin biosynthetic gene eryAI and the resistance gene ermE by binding to their promoter regions in S. erythraea. Further overexpression of SACE_7301 led to enhanced Er-A titers in the wild-type S. erythraea A226 and the high-yield industrial strain WB.

SACE_7301 positively regulates the erythromycin biosynthesis
Since extensive investigations have proved that TFRs are involved in the antibiotic biosynthesis in actinomycetes through gene inactivation and bioassay experiments [33], we have identified several TFRs pertinent to erythromycin production in S. erythraea, including the repressor SACE_3986 [32], and the activator SACE_7301 currently studied. By homologous recombination of linearized fragments ( Figure 1A), SACE_7301 was replaced by the thiostrepton resistance gene (tsr) in S. erythraea A226, and the desired mutant, named as ΔSACE_7301, was isolated and confirmed by PCR analysis ( Figure 1B).
Erythromycin titers of A226 and ΔSACE_7301 were estimated by fermentation and bioassay. ΔSACE_7301 displayed lower inhibitory activity against Bacillus subtilis relative to A226, but had slightly higher levels than the bldD deletion mutant ( Figure 1C), a previously identified regulator of erythromycin biosynthesis [12]. Furthermore, A226 and ΔSACE_7301 were cultivated at 30°C for up to 6 days in the R5 liquid fermentation medium, and the extracts of those cultures were analyzed by HPLC. Compared with the parental strain A226, ΔSACE_7301 had a 38% reduction in the Er-A yield in 6 day fermentation period (from 58 mg/L to 36 mg/L). ( Figure 1D). However, the two strains had comparable growth rates and cell densities ( Figure 1E). Thus, SACE_7301 is postulated to positively regulate erythromycin production in S. erythraea.
In order to exclude the possibility that the decreased erythromycin production was due to a random mutation in other chromosome loci of S. erythraea, ΔSACE_7301 was complemented with a copy of SACE_7301 (pZMW7301). The Er-A yield of the complemented strain ΔSACE_7301/ pZMW7301 was nearly recovered to the parental level ( Figure 1F). To further substantiate the role of SACE_7301 as an activator for erythromycin production, we introduced pZMW and pZMW7301 into A226, obtaining the control strain A226/pZMW and the SACE_7301-overexpressed strain A226/pZMW7301, respectively. As detected by HPLC analysis, the Er-A yield of A226/pZMW-7301 was 64 mg/L, that was 10% and 14% increased relative to those of A226 (58 mg/L) and A226/pZMW (56 mg/L) respectively ( Figure 1F). Taken together, our results verified that SACE_7301 acted as a positive regulator controlling the erythromycin production in S. erythraea.

SACE_7301 is not involved in the morphological differentiation of S. erythraea
Previously, BldD and SACE_5599 were identified to concurrently regulate erythromycin biosynthesis and morphological differentiation in S. erythraea [12,13]. In order to discern whether SACE_7301 was also associated with the morphological differentiation, we inoculated A226 and its derivative mutants on R3M agar medium for observing sporulation. ΔSACE_7301, ΔSACE_7301/pZMW7301 and A226/pZMW7301 on R3M plates were identical to A226 in the rate of aerial mycelia formation, whereas ΔbldD showed the defect of aerial hyphae as expected (data not shown). These results further substantiate that SACE_7301 acts exclusively as a positive regulator for the synthesis of erythromycin in S. erythraea.

Overexpression of SACE_7301 increases the transcription of eryAI and ermE
To investigate the effect of SACE_7301 overexpression on the transcription of ery genes, we chose erythromycin biosynthetic gene eryAI (encoding polyketide synthase I) and the resistance gene ermE (encoding rRNA methyltransferase) for transcriptional comparison between control strain A226/pZMW and overexpressed strain A226/ pZMW7301 on the second or fourth day of growth. When grown for 4 days, A226/pZMW7301 exhibited 2.2-and 1.7-folds transcription increases detected by qRT-PCR for eryAI and ermE in comparison to A226/ pZMW ( Figure 2B). Similar transcriptional levels of eryAI and ermE were found for the two strains in 2-day growth ( Figure 2A). The data supports the interpretation that SACE_7301 plays a positive role in the erythromycin biosynthesis by increasing the transcription of several structural and resistance genes.

SACE_7301 binds to the promoter regions of eryAI and ermE
TetR family members often regulate the transcription of target genes by specifically binding to their promoters [33]. To examine whether SACE_7301 might directly regulate the transcription of eryAI and ermE, a full-length SACE_7301 gene was expressed in Escherichia coli BL21 (DE3) using pET22b, and the purified SACE_7301 protein fused with His tag (Additional file 1: Figure S1) was used for in vitro binding experiments by EMSA.
6-FAM labeled eryAI and ermE promoter DNA fragments were respectively mixed with purified His 6 -tagged SACE_7301. At concentrations equal or greater than 1 μM of SACE_7301, mobility shifts were evident. Meantime, it was found that a 50-fold excess of unlabeled probes markedly competed with labeled probes for binding to SACE_7301 ( Figure 3A-B). As a negative control, a non-specific DNA, poly dIdC, was used to compete with the 6-FAM labeled probes, and the shifted bands did not disappear ( Figure 3A-B), confirming that SACE _7301 specifically bound to the promoter regions of eryAI and ermE.
To evaluate the relative affinity of SACE_7301 to abovementioned probes, we expressed recombinant BldD protein in E. coli BL21 (DE3), and purified it with His 6 -tag (Additional file 1: Figure S1). EMSAs showed that the concentrations of BldD shifting the two promoter probes of eryAI and ermE were similar to that of SACE_7301 ( Figure 3C-D), indicating that SACE_7301 and BldD had the same degree of affinities to the two promoters. Therefore, the combined data indicates that SACE_7301 up-regulates the transcription of eryAI and ermE by directly interacting with their promoter regions.
Effects of SACE_7301 overexpression on erythromycin biosynthesis in S. erythraea A226 As overexpression of SACE_7301 with an extra copy displayed an increase of Er-A production in S. erythraea A226 as described above, we proposed that increased copy numbers of SACE_7301 would lead to additional increases in Er-A levels. SACE_7301 under the control of PermE* was cloned into pSET152, creating 1, 2, or 3 copies of SACE_7301, introduced into A226 ( Figure 4A), to generate A226/7301, A226/2×7301 and A226/3×7301 strains respectively. By fermentation and HPLC analysis, the three strains all presented Er-A product improvement ( Figure 4B). Furthermore, the Er-A production exhibited a positive correlation with the copy number of SACE_7301, reaching the highest 1.44 folds with 3 copies ( Figure 4B). When cultured in the R5 liquid medium, these strains had comparable growth rates and cell densities (Additional file 1: Figure S2), suggesting that the enhanced erythromycin yield in SACE_7301-overexpressed strains was not caused by the increase of biomass.
In addition, qRT-PCR was used to compare the transcriptional levels of SACE_7301, eryAI and ermE between A226 and its SACE_7301-overexpressed strains. With the increase of SACE_7301 copy number, the transcripts of SACE_7301 in A226/7301, A226/2×7301 and A226/3×7301 were found to be up-regulated by 4.6-, 8.4-and 16.5-folds relative to that of A226 ( Figure 4C). Likewise, the transcripts of eryAI and ermE in those mutants were gradually increased compared to the original level ( Figure 4C). These results indicated that overexpression of SACE_7301 with 1 to 3 extra copies in S. erythraea associated with a stepwise increase of Er-A production by promoting the transcription of eryAI and ermE. In conjunction with RNA changes, A226/3×7301 also demonstrated increased resistance to erythromycin over A226 when cultured for 2 days with addition of greater than 160 μg/ml erythromycin (Additional file 1: Figure S3).
Also, we cultured A226/3×7301 in the R5 liquid medium for 6 days without addition of apramycin, and found that the size of the gene expression cassette containing the 3 copies of SACE_7301 from plasmid pSET152-3×7301 did not change by PCR analysis (Additional file 1: Figure S4), indicating that the engineered strain had a stably inserted SACE_7301.

Improved Er-A production by overexpression of SACE_7301 in an industrial strain WB
To examine the applicability of overexpression of SACE_7301 for enhancing erythromycin production in industrially-relevant strains, SACE_7301 was introduced into an industrial S. erythraea WB strain with 1, 2 or 3 extra copies under the control of PermE*, and the desired mutants WB/7301, WB/2×7301 and WB/3×7301 were obtained. In flask experiments with an industrial medium for 6 days, the Er-A yields of WB/7301 (685 mg/L), WB/2×7301 (755 mg/L) and WB/3×7301 (832 mg/L) were respectively increased by 17%, 29%, and 42% in comparison to that of WB (586 mg/L) ( Figure 5A). Likewise, no significant change of cell growth was detected for these strains in comparison (Additional file 1: Figure S2).
We also tested the reproducibility of one of the engineered strains in a larger scale by selecting the highestyield strain WB/3×7301. Following 8-day fermentation at a 5 L fermentor scale, WB/3×7301 exhibited a strikingly higher Er-A yield (4,230 mg/L), that was 27% greater than WB (3,322 mg/L). Since this strain maintains the characteristics of its parental strain, we conclude that the additional Er-A production from SACE_7301-engineered industrial strains will be of significant commercial value.

Discussion
TFRs form homodimers that generally act as transcriptional repressors to regulate the transcription of genes involved in antibiotic biosynthesis or resistance [34], but little is known about the TFR as a positive regulator for controlling the antibiotic biosynthesis in actinomycetes. Recently, the TetR family regulatory gene gouR, situated in the gene cluster of gougerotin biosynthesis, was characterized to be a positive regulator for modulating the gougerotin production by coordinating its biosynthesis and export in Streptomyces graminearus [27]. In this work, we identified a novel TFR (SACE_7301) from S. erythraea, and proved that it played a positive role in regulating erythromycin biosynthesis.
SACE_7301 was shown to specifically bind to the promoter regions of eryAI and ermE in vitro, and enhance their transcription. SACE_7301, located approximately 8.1 Mb in the chromosome of S. erythraea, is not closely positioned with the ery cluster (GenBank Accession No. NC-009142. 778,214-832,825 nt, 0.8 Mb) and therefore it became evident that the gene product of SACE_7301 might function as a trans-acting factor remotely regulating erythromycin biosynthesis. This regulation pattern   may be similar to that of BldD, a previously identified regulator for erythromycin biosynthesis [12]. As SACE_7301 exhibited low affinities to eryA and ermE promoters, it was difficult to determine the tight binding sites through DNase I footprinting assay. On the other hand, moderate increase in Er-A production was achieved by overexpression of SACE_7301 with an extra copy, similar to BldD in S. erythraea A226 [35]. These data indicate that SACE_7301 is likely a positive regulator that directly acts on the promoters of ery genes.
To further improve the production of erythromycin, an overexpression tactic was designed and utilized for overexpressing SACE_7301 under PermE* in wild-type strain A226 and industrial overproducer WB. Along with the increase of SACE_7301 copy number, SACE_7301, eryAI and ermE were significantly up-regulated, and Er-A yields of those mutants rose step by step. However, these changes in transcription level do not linearly correspond to the changes in Er-A production. A similar result was also reported for the TetR family regulator GouR in S. graminearus [27].
Overexpression of bldD or SACE_5599 could improve erythromycin production in the wild-type S. erythraea strain, but not in the industrial overproducer [13]. Considering that WB is already an industrial strain, it is likely that overexpression of SACE_7301 will have commercial value when applied in other erythromycin high-yield S. erythraea strains. In other examples, the overexpression approach was utilized to engineer the halogenase gene ctcP, resulting in improvement of chlortetracycline production in industrial S. aureofaciens [37]. Overexpression of tailor genes eryK and eryG was also performed to enhance the production and purity of Er-A in an industrial S. erythraea [4]. To our knowledge, this is the first report for tandemly expressing transcriptional regulators to enhance antibiotic production in actinomycetes. We anticipate that the strategy of overexpression of additional regulators will be generally applicable for improvement of other antibiotics in industry.

Conclusion
The novel TFR SACE_7301 positively regulates erythromycin biosynthesis in S. erythraea, and it activated the transcription of erythromycin biosynthetic gene eryAI and the resistance gene ermE by interacting with their promoters. Furthermore, when SACE_7301 was overexpressed in wildtype or industrial S. erythraea strains with 1 to 3 extra copies, Er-A yields had stepwise enhancement in comparison to the parental strains. Our data here form a new understanding of the unusual regulatory mechanism of erythromycin biosynthesis, and provide a valuable means to improve erythromycin production.

Strains, plasmids, and culture conditions
All strains and plasmids used in this study are listed in Table 1. S. erythraea and its derivatives were grown either in TSB liquid medium for seed culture, DNA extraction and protoplast preparation, or on R3M agar medium for protoplast regeneration or phenotypic observation with thiostrepton or apramycin when appropriate [38,39]. E. coli and B. subtilis strains were cultured in Luria-Bertani (LB) liquid medium or on LB plates at 37°C [40]. The general techniques in E. coli and S. erythraea were performed as described [38,40].
Using the primers 7301-P5 and 7301-P6 (Table 2), a 0.66kb DNA fragment containing a full-length SACE_7301 was amplified by PCR with the genomic DNA of A226 as a template. The PCR product was cleaved with NdeI/EcoRV, and inserted into the corresponding sites of integrative plasmid pZMW [42], yielding pZMW-7301. By PEG-mediated protoplast transformation, pZMW7301 was introduced into the ΔSACE_7301 mutant and the parental strain A226, respectively. The complemented strain ΔSACE_7301/ pZMW7301 and overexpression strain A226/pZMW7301 were obtained by apramycin resistance screening and confirmed by PCR analysis with the primers apr-F and apr-R ( Table 2).

Overexpression of SACE_7301 in S. erythraea
In order to introduce 1 to 3 extra copies of SACE_7301 into the chromosome of S. erythraea, the combined DNA fragment containing PermE* and SACE_7301 from the plasmid pZMW7301 was amplified using the primer pair ermE-F and 7301-R ( Table 2). The amplified product was cloned into pBluescript II SK (+) via XbaI and SpeI sites ( Figure 4A), and the correct sequence was confirmed by DNA sequencing. The fragment digested with XbaI and SpeI was repeatedly subcloned into XbaI site of pSET152 to generate pSET152-7301, pSET152-2×7301 and pSET152-3×7301, respectively ( Figure 4A). Those plasmids were successively introduced into S. erythraea A226 or WB, and corresponding SACE_7301-overexpressed strains were obtained by apramycin screening and PCR confirmation.

Fermentation and erythromycin assay
For flask fermentation of A226 and its derivatives, spores from R3M plates cultured for 4 days were incubated in 250 ml baffled flasks containing 30 ml TSB medium for 2 days at 30°C. Then, 3 ml cultures were inoculated into 250 ml baffled flasks containing 30 ml R5 liquid fermentation medium [38], and shaken for 6 days at 30°C (220 rpm). For flask fermentation of the erythromycin high-yielding strain WB and its derivatives, a liquid industrial medium was used as previously described [32]. After 24 h fermentation, 0.3 ml of n-propanol was added into the cultures and further incubated for 5 days at 30°C (220 rpm).
For bioreactor cultures of WB and its SACE_7301-overexpressed strain WB/3×7301, the fermentation experiments were carried out in a 5 L fermentor (Baoxing, Shanghai, China) containing 3 L liquid industrial medium. Seed cultures for the cultivation in bioreactors were prepared as described above, using 450 ml industrial seed medium in 2 L baffled flasks. Bioreactors were inoculated with 30 vol.% seed culture. Dissolved oxygen concentration and pH were monitored using autoclavable electrodes (Mettler Toledo, Switzerland). Dissolved oxygen was maintained about 50% with increasing agitation and aeration rate during the bioprocess, and the pH was kept at 7.0 to 7.2. Foaming was controlled by automatic addition of antifoam. Samples (50 ml) were taken every 24 h for Er-A yield analysis with HPLC as described previously [44]. HPLC analysis was performed on Agilent 1260 HPLC system equipped with an Agilent Extend-C18 column (5 μm, 250 × 4.6 mm), which was equilibrated with 40% solution A (acetonitrile, chromatographic grade) and 60% solution B (potassium dihydrogen phosphate, 0.032 mol/L, pH 6.8). An isocratic program was carried out at a flow rate of 1.0 ml/min at 29°C using UV detector at 210 nm.
For bioassay-based analysis, A226, ΔSACE_7301 and ΔbldD were grown in 30 ml R5 liquid medium in 250 ml flasks for 6 days at 30°C. 5 μl fermentation supernatants from those liquid cultures were added to LB agar plates, which were sprayed with an overnight culture of B. subtilis PUB110. The plates were incubated at 37°C for 12 h, and the erythromycin production was estimated by detecting the growth-inhibition zones as previously demonstrated [31]. Moreover, Er-A produced by A226, WB and their derivatives were quantitatively analyzed by HPLC.
Transcriptional analysis by quantitative real-time PCR (qRT-PCR) The relative transcriptional levels of eryAI and ermE, from erythromycin biosynthetic gene cluster [7], were determined by qRT-PCR analysis. Specific primers were designed as listed in Table 2. According to the manufacturer's instructions of a RNA extraction/purification kit (SBS), total RNA was extracted from S. erythraea A226 and its derivatives after 2 or 4 days of growth on R5 agar medium [38]. Isolated RNA was treated with DNase I (MBI Fermentas), and reverse transcription was achieved using a cDNA synthesis kit (MBI Fermentas). Quantitative real time PCR reactions were performed on the Applied Biosystems Step-One Plus system with Maxima™ SYBR Green/ROX qPCR Master Mix (MBI Fermentas). The hrdB gene encoding the major sigma factor in S. erythraea was used as an internal control, and relative quantification was evaluated using a comparative cycle threshold (C T ) method [45].

Overexpression and purification of SACE_7301 and BldD
As the original SACE_7301 gene could not be expressed in E. coli BL21 (DE3) initially, we optimized the N-terminal codons during the amplification of SACE_7301 using the primers 7301-P9 and 7301-P10 ( Table 2). The PCR product was cloned into NdeI and HindIII restriction sites of pET22b to generate pET22b-7301. Then, pET22b-7301 was introduced into E. coli BL21 (DE3) for protein production. Gene expression was induced with 1 mM isopropylthiogalactoside (IPTG) for 5 h at 30°C. Cells were harvested and followed by sonication treatment. Purification of His 6 -tagged protein was carried out by use of Ni-NTA Sepharose chromatography [46]. The concentration of purified protein was quantified by BCA assays, and its purity was judged by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) analysis. Similarly, bldD was amplified with the primers bldD-F and bldD-R (Table 2), and expressed in E. coli BL21 (DE3).

Electrophoretic mobility shift assays (EMSAs)
EMSAs were carried out in the light of previously described methods [47]. Using the genomic DNA from S. erythraea A226 as a template, the promoter regions of eryAI and ermE were successively amplified by PCR with their respective primers ( Table 2) labeled at 5′ end by the 6-isomer of carboxyfluorescein (6-FAM). Those labeled DNA fragments were independently mixed with purified His 6 -tagged SACE_7301 or BldD proteins. The binding reaction system consisted of 10 mM Tris (pH 7.5), 5 mM MgCl 2 , 50 mM EDTA, 60 mM KCl, 10 mM DTT, 10% glycerol, 150 ng labeled probes and 0.5-4 μM purified SACE_7301 or BldD. For competitive inhibition of the binding reaction, 7.5 μg unlabeled above-mentioned DNA fragments or 7.5 μg poly dIdC were added into that reaction system, respectively. After incubation on ice for 10 min, the reactants were run on an 8% TBE polyacrylamide gel (Bio-Rad) with 0.5 × TBE as a running buffer at 30 mA for 1 h.

Additional file
Additional file 1: Figure S1. Purification of His 6 -tagged SACE _7301 and BldD. Left lane 116 KDa protein ladder. Figure S2. Time courses of dry weight of mycelia (DWM) of low-yielding S. erythraea strain A226, industrial overproducer WB, and their derivatives. Mean values of at least three replicates were shown, with the standard deviation indicated by error bars. Figure S3. Growth of A226 and A226/3×7301 treated with erythromycin. The two strains were cultured in the R5 liquid medium for 2 days by addition of four concentrations of erythromycin (40 μg/ml, 80 μg/ml, 160 μg/ml and 320 μg/ml), respectively, and their dry weights of mycelia (DWM) were successively measured. Mean values of at least three replicates were shown, with the standard deviation indicated by error bars. Figure S4. PCR analysis of the cassette containing 3 copies of PermE*-SACE_7301 from A226/3×7301 using the primers Test-F and Test-R. The specific primers flanking above-mentioned cassette of the plasmid pSET152-3×7301 were designed as shown in Table 2. A226/3×7301 was cultured in the R5 liquid medium for 6 days without apramycin. Lanes: M: 5, 000-bp DNA ladder, +: positive control (pSET152-3×7301 as a template), -: negative control (Genomic DNA of A226 as a template), 1 to 6: Genomic DNA of A226/ 3×7301 cultured for 1 to 6 days, respectively, 7: Genomic DNA of the mixture of A226/7301, A226/2×7301 and A226/3×7301 with equal amount.