2024 Volume 21 Issue 4 Article ID: e210024
JCVI-syn3B (syn3B), a minimal synthetic bacterium that only possesses essential genes, facilitates the examination of heterogeneous gene functions in minimal life. Conventionally, Escherichia coli is used to construct DNA fragments for gene transfer into the syn3B genome through Cre/loxP system. However, the construction process is challenging and time-consuming due to various issues, including the inhibition of E. coli growth and unexpected recombination, especially with AT-rich DNA sequences such as those found in Mycoplasma genes. Therefore, in this study, we aimed to develop a new transformation method to overcome these issues. We assembled the vector and target DNA fragments using an in vitro homologous recombination system and subsequently transferred the products into the syn3B genome. We obtained approximately 103~104 recombinant colonies per milliliter of the original culture in eight days, which is four days shorter than the conventional period, without any recombination issues, even for AT-rich DNA. This method may be applicable to other gene manipulation systems based on Cre/loxP system.
A rapid and trouble free method was developed to transfer genes to the genome of minimal synthetic bacterium JCVI-syn3B through Cre/loxP system. This method can be applied to Cre/loxP-based gene manipulation system in various research fields.
JCVI-syn3B (syn3B) is a “synthetic” bacterium based on a small bacterium, Mycoplasma mycoides, which was established to answer the question what is the minimal life? The necessity of individual M. mycoides genes were judged by bioinformatics and experiments. The syn3B genome was designed by a computer, synthesized chemically, assembled in E. coli and yeast, and then transplanted to a related species, Mycoplasma capricolum [1]. The syn3B genome contains a locus of loxP site to which external genes can be efficiently transferred via the Cre recombinase activity (Fig. 1) [2–4]. Therefore, syn3B is a useful platform for both basic and applied researches [3,5–8]. Our group is discussing the origin of cell motility, based on syn3B experiments [5,9]. In the transformation of syn3B by external genes requires a plasmid containing the target DNA, Cre recombinase, loxP sites, and puromycin resistance gene [2]. We construct these plasmids using Escherichia coli (Fig. 1) as described below. (a, day 1) Polymerase chain reaction (PCR)-amplified vector and target DNA fragment were assembled via homologous recombination-based methods using the NEBuilder® and In-Fusion® systems. The constructs were then used to transform E. coli cells. (b, day 2) Transformed colonies were inoculated into a liquid medium. (c, days 3–5) Plasmids were isolated and sequenced. (d, days 6–9) Syn3B was transformed with the plasmid. (e, days 10–12) Colonies were cultured in a liquid medium and target gene transfer was confirmed via direct PCR. This method exhibited some drawbacks, including poor E. coli growth and unexpected recombination. Moreover, we suffered from unexpected gene recombination in transfer of some Mycoplasma and related species (class Mollicutes), possibly due to their high AT content (67–76%) [10,11]. Therefore, in this study, we aimed to explore four transfer methods without E. coli to develop a new E. coli-free approach that is four days faster than the conventional method for gene transfer.
JCVI-syn3B (GenBank, CP069345.1) and E. coli (DH5α) were cultured in the SP4 [1,2] and Luria–Bertani media at 37°C, respectively.
Preparation of DNA FragmentsMycoplasma mobile genome was isolated as previously described [12]. Vector and focused M. mobile DNA fragments were amplified from pSD128 DNA [3] and M. mobile genomic DNA using primers P1–P2 and P3–P4 (Table 1), respectively. The DNA fragments were assembled using the NEBuilder® HiFi DNA Assembly Master Mix (New England Biolabs, Inc., MA, USA) and In-Fusion HD® Cloning Kit (Takara Bio Inc., Kusatsu, Japan). Then, the assembled product was amplified via PCR using primers P5 and P6 and circularized via self-assembly with NEBuilder®.
| Name | Sequence |
|---|---|
| P1 | 5'-GTCAAAGATCATACCGTTCG-3' |
| P2 | 5'-AGCTAGCTCAGTCCTTATTT-3' |
| P3 | 5'-AAATAAGGACTGAGCTAGCTTTGAATCGAACTGAATATGC-3' |
| P4 | 5'-CGAACGGTATGATCTTTGACTACTCATAATATGCTACCTC-3' |
| P5 | 5'-GGTTGGACTCAAGACGATAG-3' |
| P6 | 5'-GTCTTGAGTCCAACCGGCCGGCCCGGAGGCGTCCC-3' |
| P7 | 5'-GTAAATTCCCAAATTATTCCATCA-3' |
| P8 | 5'-TATGTGATAATGCCAATCGCTAAG-3' |
Transformation of JCVI-syn3B was performed as previously described [5]. In short, competent cells were incubated with respective DNA for transfer for 15 min on ice, treated with 70% PEG6000, and then incubated in SP4 medium at 37°C for 3 h. Cells were then plated on a SP4 (containing horse serum instead of FBS) plate supplemented by 3 μg/mL puromycin. After incubation at 37°C for four days, the number of transformed colonies was counted. The transformation efficiency is shown as values relative to those of 100 ng pSD128 plasmid. This is because the efficiency largely depended on serum. For further studies, the transformed colonies were picked and inoculated into 100 μL of SP4 medium containing puromycin (3 μg/mL) for 18–24 h to make stocks, and confirmed via PCR using primers P7 and P8.
To skip DNA manipulation in E. coli, we assembled the DNA fragments using an in vitro homologous recombination system and amplified them using PCR to obtain large amount of uniform DNA fragments, beneficial for transformation. Four types of DNAs were tried as shown in Fig. 2. As our target DNA, we used a 1,916-bp fragment containing the 1,536-bp gene encoding phosphoglycerate kinase from M. mobile (MMOB4530) and its flanking region (at 565,247–567,161 nt on the genome) [13–17].
First, we tried an assembled and amplified DNA fragments (Fig. 2(i)). Initially, we performed assembly using the In-Fusion® system, but the size of the resulting target fragments was shorter than the full length (Fig. 2B). In-Fusion® system has no DNA polymerase and ligase for the gap-filling activity that prevents mutagenesis during filling [18,19]. Therefore, the region ended by the target DNA and P5 or P6 should be amplified by the PCR (Fig. 2B right). This assumption was supported by observations that the amplified fragment changed with the length of target fragment. Subsequently, we used NEBuilder® with an in-built gap-filling activity [20]. Using NEBuilder®, we successfully amplified the DNA fragments (Fig. 2B). Upon transforming syn3B with 100 ng of the PCR product, 392±124 colonies (n=6) were obtained. However, the efficiency was approximately 1/50 of that achieved with a plasmid (pSD128 including target DNA) isolated from E. coli (Fig. 3A(i)). We investigated the cause for this low efficiency by performing transformation using circularized DNA.
PCR products were circularized by NEBuilder® (Fig. 2(ii)). The amount of DNA converted to a form other than the original linear form was estimated from the agarose gel electrophoresis band intensity (Fig. 2B), which indicated a 94.6% conversion. The DNA mixture was used for transformation (Fig. 3A(ii)). Notably, 6,458±1,428 colonies from 425 μL cultures (n=6) were obtained. The transformation efficiency was approximately 3/10 that achieved with the plasmid isolated from E. coli. These results suggest that circularized DNA achieves a transformation efficiency similar to that achieved with the supercoiled plasmids isolated from E. coli. Next, we examined the products assembled without PCR amplification.
Transformation Using Assembled Circular DNAThe product assembled using NEBuilder®, which was expected to be circular, was directly transferred into the syn3B genome (Fig. 2(iii)). The DNA product assembled using 30 ng vector and 28 ng target DNA was transferred into syn3B, resulting in 2,067±262 colonies from 425 μL cultures (n=6; Fig. 3A(iii)). This shows the advantage of circular DNA in the uptake process. Next, we assembled the product using In-Fusion® (Fig. 2(iv)) and obtained 1,282±418 colonies from 425 μL cultures (n=6; Fig. 3A(iv)), indicating plasmid instability caused by the gap region (P=0.005 via Student’s t-test). Here, the transformation efficiency could not be directly compared with that of the conventional method as the precise amount of assembled DNA was unknown. However, sufficient number of colonies were obtained for gene transfer using the NEBuilder® and In-Fusion® products.
Unexpected Recombination in the TransformantsWe examined the transfer of the target DNA into loxP via direct PCR using primers P7 and P8 (Fig. 1) for 18 transformants, and repeated the whole process of this examination for six times (Fig. 3B). For the colonies obtained using the conventional method, we observed PCR bands at the correct positions in 97% of colonies. However, compared to that of the conventional method, success rates (74, 75, 93, and 92%, n=108) were low for the linear (Fig. 2(i)), circular (Fig. 2(ii)), NEBuilder®-assembled (Fig. 2(iii)), and In-Fusion®-assembled (Fig. 2(iv)) DNAs, respectively. This decrease in success rate may be due to unexpected amplification during PCR, because the decrease depended on number of PCR application and the success rate was 97% when the plasmid cloned in E. coli was used. Additionally, transformation efficiency of our direct transformation method was lower than that of conventional transformation using plasmids isolated from E. coli.
Originally, we started this method to overcome problems in transferring AT-rich Mollicutes (Mycoplasma and related species) DNA into the syn3B genome. To date, we have constructed 58 syn3B transformants with AT content of 65–76%. Moreover, using the established method, we successfully transferred a 14.5-kb fragment consisting of eight genes of M. mobile into the syn3B genome. We also transferred the genes of Haloplasma contractile into syn3B using the new method [21,22], which is impossible with E. coli-based gene manipulation systems. Approximately 58 kb of the transferred DNA region was sequenced from 18 transformants. Two clones each with single-nucleotide deletions and substitutions were observed. These mutations were possibly caused by PCR, as similar mutations have been reported in E. coli-based recombination of PCR products.
Cre/loxP system is used for genome and gene manipulation widely for microorganisms, animals, plants, and human [23–25]. Although our study focused on basic research of syn3B, the direct assembly method may be applicable for other targets.
Overall, we developed an in vitro direct rapid method for gene transfer to the JCVI-syn3B strain appreciable for Cre/loxP recombination. Our method is particularly useful for DNA sequences unstable or harmful to E. coli, and also appreciable for various systems based on Cre/loxP recombination.
The authors declare no conflicts of interest.
AU performed most of the experiments. Mimura contributed to mutation ratio of transformants. AU and Miyata wrote an early version of the manuscript. All authors contributed to the project and manuscript.
All data generated/analyzed in this study are included in this article.
A preliminary version of this work, DOI: https://www.biorxiv.org/content/10.1101/2024.08.01.606163v1, was deposited in the bioRxiv on August 1, 2024.
This study was supported by JST CREST grant (grant number JPMJCR19S5 to Miyata), and Grant-in-Aid for JSPS Fellows (grant number JP24KJ0189 to HK).
