Shuttle Vector System for Methanococcus maripaludis with Improved Transformation Efficiency

Abstract

We have identified an open reading frame and DNA element that are sufficient to maintain shuttle vectors in Methanococcus maripaludis. Strain S0001, containing ORF1 from pURB500 integrated into the M. maripaludis genome, supports a significantly smaller shuttle vector, pAW42, and a 7,000-fold increase in transformation efficiency for pURB500-based vectors.

Methanococcus maripaludis S2 is a mesophilic methanogen that has one of the best-developed genetic systems of any organism in the domain Archaea. Because the genome of M. maripaludis S2 has been completely sequenced, this strain is the strain of choice for genetic manipulation (5). In addition to integrative vectors (7, 8) and a markerless mutagenesis protocol that allows the deletion of nonessential genomic regions and the stable introduction of mutations or foreign genes (7), M. maripaludis can be transformed with a shuttle vector using a polyethylene glycol (PEG)-based transformation protocol (11). The entire cryptic plasmid pURB500 isolated from M. maripaludis C5 (10) was used to construct a 12,691-bp shuttle vector, pDLT44, which was reported to produce 3.3 × 10⁷ transformants/μg by PEG transformation in M. maripaludis JJ (10). pDLT44 has been used as the parent for a series of additional shuttle vectors, namely, pWLG30 (4), pWLG40 (6), and the pLW40 series (3). However, the smallest existing shuttle vectors (pLW40) are still over 10 kb in size, making cloning difficult and potentially limiting transformation efficiency. M. maripaludis S2 does not appear to support the same high levels of transformation as originally reported for M. maripaludis JJ either in our hands or in other laboratories (K. Jarrell, personal communication), so we sought to improve transformation rates by manipulating the shuttle vector.

It was hypothesized that ORF1 is required for plasmid maintenance, probably functioning as a Rep-type protein (10; reviewed in reference 2). We carried out a reanalysis of the open reading frames (ORFs) on pURB500 (see Table S1 in the supplemental material) that supported this notion. To develop a smaller shuttle vector, we simultaneously tested whether ORF1 was required for plasmid maintenance and whether it could fulfill its role in trans. A 2.8-kb fragment containing ORF1 plus 150-bp upstream flanking region (assumed to contain the promoter for this ORF) was PCR amplified from pLW40 (3) using oligonucleotides ORF1_5prime (5′-AAAGGCGCGCCTCCCTGAAGAAGAAGAGAG-3′) and ORF1_3prime (5′-AAAGGCGCGCCAGTTCCATTTTACCACC-3′). The product was ligated into pGEM-T (Promega) to produce pSS1 (see Table S2 in the supplemental material). The inserted sequence displayed a single T-to-C mutation changing isoleucine 9 to methionine in the protein. This mutation exists in pLW40 but is not present in the parental plasmid, pWLG30. pLW40 plasmid containing either mutant or wild-type (WT) ORF1 sequences were found to be transformed with comparable efficiency (data not shown).

The ORF1-containing fragment was excised from pSS1 using AscI and subcloned into the same site in pBLPrt (7) to form the ORF1 knock-in plasmid, pSS2. This construct targets integration of the insert to the chromosomal upt gene. M. maripaludis S2 strain Mm900 was transformed by markerless mutagenesis (7) using pSS2 to produce a new strain (S0001) thought to contain the natural promoter for ORF1 and the ORF1 gene. Colonies were selected, put into liquid McCas (1, 7), and grown overnight. Genomic DNA from M. maripaludis S0001 was recovered and digested with BsmI, and the successful generation of a knock-in strain was confirmed by Southern blotting (9).

To test the effect of ORF1 on plasmid maintenance, pLW40 was digested with XhoI and PacI. The resulting 7.4-kb backbone fragment was treated with Klenow fragment, gel purified, and ligated to produce pSS3, a plasmid in which ORF1 had been deleted. To determine whether pSS3 could transform wild-type (Mm900) or ORF1-containing (S0001) strains, 5 ml of M. maripaludis cells at an optical density at 600 nm (OD₆₀₀) of 0.7 to 1.0 were transformed (11) with 5 μg of plasmid in anaerobic TE buffer (10 mM Tris, 1 mM EDTA). In parallel, the same strains were transformed with TE buffer or with the parental plasmid, pLW40. Transformed cells were allowed to recover overnight in the absence of selection and subjected to 10-fold serial dilutions, and then 4-μl aliquots were spotted onto McCas or McCas supplemented with 2.5 μg/ml puromycin before being incubated in a pressure vessel containing H₂/CO₂ (4:1) at 20 lb/in² for 3 or 4 days at 37°C (Fig. 1). As expected, all cells grew well in the absence of selection (Fig. 1A and C). In the Mm900 strain, only cells transformed with the pLW40 plasmid were able to grow in the presence of puromycin, and only a few colonies were observed in the undiluted sample (Fig. 1B, pLW40, 10⁰). In contrast, when transformed with either pLW40 or pSS3, the S0001 strain harboring a genomic copy of ORF1 was able to grow in the presence of puromycin (Fig. 1D). This result suggests not only that ORF1 is required to support the proliferation of pURB500-based plasmids but also that ORF1 is able to function in trans.

FIG. 1.

ORF1 is essential for plasmid maintenance but can function in trans. (A to D) Wild-type (Mm900) (A and B) and ORF1-containing (S0001) (C and D) M. maripaludis strains were transformed with plasmid pLW40 (containing ORF1) or pSS3 (with ORF1 deleted) and plated as 10-fold dilutions on medium in the absence (A and C) or presence (B and D) of puromycin. A no-vector control (TE buffer [TE]) was transformed at the same time. Cells could grow on McCas plus puromycin only if the vector was maintained. Only pLW40 could be maintained in strain Mm900 (B), but strain S0001 could also maintain plasmid pSS3 (D).

Tumbula et al. (10) had previously noted that pURB500 possessed two “ORFLESS” regions and speculated that one of these regions was likely to encode the origin of replication for this plasmid. As the plasmid in which ORF1 had been deleted (pSS3) possesses both of these ORFLESS regions, we produced two additional plasmids that contained either the ORFLESS1 region (pAW42 [Fig. 2 A]) or the ORFLESS2 region and part of ORF3 (pAW40 [Fig. 2B]). pSS3 was digested with BbsI and SpeI to remove ORFLESS1 and part of ORF3. The 6.1-kb fragment was treated with Klenow fragment, gel purified, and ligated to generate pAW40. pSS3 was digested with PciI and Acc65I to remove ORFLESS2 and ORF3. The 4.9-kb fragment was treated with Klenow fragment, gel purified, and ligated to produce pAW42. We tested whether either of these plasmids could be maintained by strain S0001 by plating transformed cells on medium in the absence and presence of puromycin (Fig. 2C and D). Only pAW42 was able to support growth on puromycin, suggesting that the ORFLESS1 region encodes the origin of replication for pURB500. Further analysis of the ORFLESS1 region reveals the presence of repeat sequences that could form part of a putative replication origin (see Fig. S1 in the supplemental material). Since both ORFLESS2 and ORF3 are deleted in pAW42, we can conclude that they are not required for plasmid maintenance in the presence of selection. Our Escherichia coli/M. maripaludis S2 shuttle vector, pAW42 (4,952 bp), is ∼5.3 kb smaller than other expression vectors available for this organism (pLW40, 10,231 bp). The reduction in size makes this vector considerably easier to manipulate than those previously used.

FIG. 2.

The ORFLESS1 region is essential for plasmid maintenance. (A to D) Plasmids containing either the ORFLESS1 region (pAW42) (A) or the ORFLESS2 region (pAW40) (B) were transformed into M. maripaludis strain S0001 and plated as 10-fold dilutions on medium in the absence (C) or presence (D) of puromycin. Positive (pSS3) and no-vector controls (TE) were transformed at the same time. Only the vector containing ORFLESS1 (pAW42) could be maintained.

To test the transformation efficiency of pAW42 and pLW40, we compared the number of puromycin-resistant colonies produced when M. maripaludis S0001 was transformed with pAW42 to those observed when M. maripaludis Mm900 was transformed with pLW40. We calculated transformation efficiencies and found that the M. maripaludis S0001/pAW42 combination (5.3 × 10⁶ transformants/μg) is transformed approximately 7,000-fold more efficiently than the M. maripaludis Mm900/pLW40 system (750 transformants/μg). Interestingly, pLW40 is also transformed much more efficiently in strain S0001 than in strain Mm900 (Fig. 1), supporting the notion that any pURB500-based vector will be more effectively transformed in the S0001 strain. Expression of ORF1 from a genomic context likely enhances plasmid replication in M. maripaludis S2, regardless of whether the plasmid carries its own copy of ORF1. The most likely explanation for this phenomenon is that the plasmid can be replicated immediately on entry to the cell because ORF1 protein (ORF1p) is already present.