A single copy integration vector that integrates at an engineered site on the Staphylococcus aureus chromosome

Identical sequence alteration in L54a attP and attB affects integration specificity

In the λ site-specific recombination system, a single bp alteration within the 7 bp overlap sequence, in either attP or attB, drastically reduces the efficiency of recombination. However, the identical change in the partner site restores recombination efficiency [4, 20]. Based on this, the λ site-specific recombination system has been developed for recombinant DNA applications (for example, the GATEWAY^® cloning system developed by Invitrogen Life Science Technologies).

Since the staphylococcal phage L54a Int-mediated recombination system closely parallels that of λ, we hypothesized that a similar alteration in integration specificity could be achieved in the L54a system [4, 20]. This would allow us to develop a single-copy cloning vector that would integrate at a region of our choice with minimal interference on normal biological function of the cells. To this end, we first evaluated 12 intergenic regions for transcriptional activity by RT-PCR. The intergenic region between SAOUHSC00937 and SAOUHSC00938 (ORF designations are according to S. aureus strain NCTC8325 annotation) was initially selected for insertion of a mutated attB site. We detected very little transcriptional activity in this region in strain 8325-4 cultures grown for 2 h, 4 h or 16 h (data not shown), although two sRNAs are known to be encoded within this region [17].

We then tested whether identical sequence alterations of L54a attP and attB affect L54a integration specificity. As shown in Figure 1, a pair of 7-bp imperfect inverted repeats, flanking a 3 bp spacer sequence, is located in the L54a attP and attB sites [8]. The inverted repeats could serve as the binding sites for L54a Int reminiscent of λ Int core binding half sites. The spacer region is similar to the overlap region of λ. Accordingly, we altered 2 bps within the core of the attP and attB pair by overlapping PCR. The altered attP was named attP 1 (Figure 1). The 1617-bp PCR fragment containing attP 1 and the adjacent int gene was cloned into pCL52.2 resulting in plasmid pLL3961.

PCR was used in order to determine whether pLL3961 had integrated at attB 1 or the wild type attB site. Out of 4 integrants that we tested, 3 were integrated at the correct location, attB 1 (results not shown). However, in one of the integrants the plasmid had integrated at the wild type L54a attB site in the geh gene. These results indicate that the altered att sites can be recognized by L54a Int but that the specificity of integration is less than perfect. These results also indicate that the size of the L54a attB site is less than 21 bp in length, which is shorter than the 27 bp that we previously determined [13].

Developing pLL102 with high fidelity integration at an engineered attB site

The above results suggest that phage integration specificity could be altered by introducing the identical bp changes into the core sequence of attP and attB, but 2-bp alteration is not enough to ensure high fidelity. Therefore, to increase the specificity, we replaced 5 bp of the att core sequence as shown in Figure 1. The new att site was named attP 2. The attP 2 fragment was then cloned into pCL25 [2], replacing the wild type L54a attP fragment, to form pLL102 (Figure 2). Since there is still a low level of transcription at the attB 1 site described above, we tested additional intergenic regions for insertion of attB 2 into the chromosome. We found that transcriptional activity in the region between SAOUHSC00009 and SAOUHSC00010 of the 8325-4 chromosome was undetectable at 3 different time points (after 2, 4 and 16 h of culture) by RT-PCR (data not shown). The 21 bp attB 2 site was then inserted into the intergenic region between G and C at coordinates 14208 and 14209 (NCTC8325 genome) of strain RN4220 (Figure 1) by pKOR1 allele replacement. The resultant strain, CYL12348, was confirmed by PCR. Plasmid pYL112Δ19, which encodes L54a Int [3], was introduced into CYL12348 to generate strain CYL12349. Plasmid pLL102 was then electroporated into CYL12349. Examination of 22 transformants by PCR (Figure 3) showed that all the transformants carried pLL102 integrated at the attB 2 site. None of the transformants had an insertion at the original L54a attB site. These results indicate that pLL102 has a high degree of specificity for the attB 2 site. The fact that in the attP 2/attB 2 pair, the alteration was made in all 3 nucleotides in the spacer region and 2 nucleotides in the mismatched inverted repeat (Figure 1) suggests that these nucleotides are not needed for L54a Int function. These results therefore reaffirm that L54a Int behaves very similarly to that of λ Int suggesting conservation of Int functionality across different species as diverse as E. coli and S. aureus.

Integration of pLL102 at the attB2 site does not affect transcription of the flanking genes

SAOUHSC00009 (NCTC8325 genome) is annotated as serS which encodes seryl-tRNA synthetase, whereas SAOUHSC00010 is a hypothetical branched chain amino acid transporter. The two genes are transcribed in the same orientation with SAOUHSC00009 being upstream of SAOUHSC00010 (Figure 1). The location of the L54a attB 2 site in the chromosome was selected because we detected no transcription in this region using RT-PCR. The intergenic region is 649 bp in length and there is a rho-independent terminator close to the 3'end of SAOUHSC00009. In addition, there is no known sRNA in this region [16, 17]. To avoid interfering with the transcription of the flanking genes, we positioned the attB 2 insertion site 31 bp downstream of the terminator and 618 upstream of the SAOUHSC00010 start site ATG. Although our RT-PCR experiment did not detect transcripts flanking the insertion site, it is still possible that an artificial insertion at this site could cause alteration in gene expression in the flanking genes. To determine whether the insertion of the attB 2 site and the subsequent integration of pLL102 affect transcription of the flanking genes, we performed qRT-PCR of the adjacent genes. We found no significant differences in SAOUHSC00009 or SAOUHSC00010 mRNA levels between RN4220, CYL12349 (i.e., RN4220-attB 2) and CYL12376 (i.e., RN4220 attB 2(pLL102)) (data not shown) indicating that the insertion of attB 2 or the integration of pLL102 at the attB 2 site does not affect expression of the surrounding genes.

Bacterial strains, plasmids and culture conditions

RT-PCR and qRT-PCR section

RNA for RT-PCR and qRT-PCR was isolated as described [22]. RT-PCR was performed using the Quantitech Reverse Transcription Kit from Qiagen Inc., Valencia, CA. RT-PCR using primers OU9R7, OU9R8, OU9R9 and OU9R10 (Table 1) was performed to probe for transcripts in the intergenic region between SAOUHSC00009 and SAOUHSC00010 using RNA isolated from 4, 8 and 18 h cultures of 8325-4.

To determine whether the insertion of the attB 2 site and the subsequent integration of pLL102 affected transcription of the flanking genes, we performed qRT-PCR as described [28] of the adjacent genes using primers SAO9F3 and SAO9R3 for SAOUHSC00009 and SAO10F3 and SAO10F3 for SAOUHSC00010. RNA for qRT-PCR was isolated from exponential phase and overnight cultures of strains RN4220, CYL12349 and CYL12376.

Generation of modified attP and attB

The attP 1 and attB 1 sites were created by altering two bp within the core of attP and attB, respectively, by overlapping PCR using primers L54aInt1, L54aAttP1, L54aAttP2 and L54aAttP3 (Table 1), and L54a phage DNA as template. The altered attP and attB were named attP 1 and attB 1, respectively (Figure 1). The 1617-bp fragment containing attP 1 and the adjacent int gene was cloned into pCL52.2 resulting in plasmid pLL3961. All PCR-amplified fragments were verified by DNA sequencing.

The altered attB sequence, named attB 1, was introduced into the bacterial chromosome by the pKOR1 system as previously described [21, 22] using RN4220 DNA as template and primers OU937R9, OU937R10, OU937R11 and OU937R12 (Table 1). The 21-bp attB 1 fragment was inserted between bp coordinates 911590 and 911591 (S. aureus NCTC8325 genome) of the bacterial chromosome. This is within the intergenic region between SAOUHSC00937 and SAOUHSC00938 of strain RN4220, 29 bp downstream of sRNA RsaF [17, 23]. The insertion was verified by PCR using primer pairs OU937attBf/OU937R9 and OU937attBr/OU937R12. The resultant strain, CYL12337, was used as the recipient for integration of pLL3961. The integration was confirmed by PCR using primer pairs OU937R3/L54aAttP3 and OU937R8/L54aAttP3. Primer pairs scv2.1/L54aAttP3 and scv1/L54aAttP4 were used to determine whether a plasmid had integrated at the wild type attB site.

The attP and attB sites with 5 bp substitutions were constructed by overlapping PCR using primers L54aAttP4, L54aAttP5, L54aAttP6 and L54aAttP7 (Table 1). The amplified fragments were verified by DNA sequencing. The new att sites were named attP 2 and attB 2 (Figure 1). The 289-bp attP 2 fragment was then cloned into Cla I-Bgl II-digested pCL25 replacing the 350-bp wild type L54a attP fragment to form pLL102 (Figure 2). The 21-bp attB 2 site was then inserted into the intergenic region between G and C at coordinates 14208 and 14209 (NCTC8325 genome) of strain RN4220 (Figure 1) using the pKOR1 allele replacement vector and primers OU9R1, OU9R2, OU9R3 and OU9R4. The resultant strain, CYL12348, was confirmed by PCR using primer pairs OU9R1/OU9R6 and OU9R4/OU9R5. Plasmid pYL112Δ19, which encodes L54a Int [3], was introduced into CYL12348 by phage 52A transduction and selection for chloramphenicol resistance to generate strain CYL12349. Plasmid pLL102 was then electroporated into CYL12349 and transformants were selected with 3 μg/ml tetracycline.