|Home | About | Journals | Submit | Contact Us | Français|
Genomic DNA from methicillin-resistant Staphylococcus aureus isolates recovered from pigs and their caretakers proved resistant to SmaI digestion, leading to uninterpretable results in standard pulsed-field gel electrophoresis. This is the result of a yet unknown restriction/methylation system in the genus Staphylococcus with the recognition sequence CCNGG.
Many microorganisms protect themselves against negative influences of foreign DNA by using restriction endonucleases that cut the DNA at a specific recognition sequence. To protect their own DNA from being digested, this is usually modified by methylation of specific adenine or cytosine residues at the corresponding recognition sites. A wide variety of restriction enzymes and DNA methyltransferases has been recognized throughout the microbial world. For an up-to-date online overview, see http://rebase.neb.com/. Pulsed-field gel electrophoresis (PFGE) is the gold standard in uncovering epidemiological relationships between microbial isolates. For Staphylococcus aureus, this is routinely performed using the restriction endonuclease SmaI. Occasionally, some isolates will not yield an interpretable result with PFGE analysis because of technical problems associated with degradation of the genomic DNA (4). This may lead to only very faint bands, degraded bands, or no banding patterns at all.
Upon investigating pig farming as a possible source of methicillin-resistant S. aureus in The Netherlands, we encountered a different technical problem with methicillin-resistant S. aureus isolates from pigs and their caretakers that proved to be nontypeable by PFGE using SmaI (5). This problem was specifically associated with restriction enzyme digestion, since genomic DNA of these isolates appeared as a single high-molecular-weight band close to the position where it was loaded on the gel. We speculated that this could be the result of DNA methylation, since it has been established that the activity of the restriction enzyme SmaI is blocked by the presence of 5-methylcytosine at specific sites in its recognition sequence CCCGGG (http://rebase.neb.com/). This was investigated by performing PFGE analysis by following established procedures (1) using the restriction enzyme SmaI or XmaI. These two restriction enzymes are neoschizomers, i.e., they cut the same recognition sequence but at a different position. Unlike SmaI, the activity of its neoschizomer XmaI is only partially reduced on methylated DNA. An unrelated control strain (ATCC 29213) was used in all experiments. Genomic DNA ofthe pig farming isolates is clearly protected from digestion bySmaI but not by XmaI, whereas DNA of the control strain wascut by both enzymes, yielding identical banding patterns (Fig. (Fig.1).1). These results indicate that the DNA from these isolates is methylated and that the type of modification may involve 5-methylcytosine. To substantiate this finding, bisulfite sequencing was used to identify the presence of 5-methylcytosines in genomic DNA from these isolates. Bisulfite treatment of nonmethylated cytosines leads to conversion to uracil, whereas 5-methylcytosine is resistant to bisulfite treatment (2). PCR amplification and subsequent DNA sequence analysis of bisulfite-treated DNA can thus reveal the presence of 5-methylcytosine residues: a 5-methylcytosine will appear as a regular C in the resulting sequence trace, whereas a nonmethylated cytosine will appear as a T. Genomic DNA was subjected to sodium bisulfite treatment using the EZ DNA methylation kit (BaseClear, Leiden, The Netherlands). PCR primers were designed to amplify a relatively GC-rich genomic fragment containing at least one SmaI site. Primer sequences for bisulfite-treated DNA were 5′-TGGTGGGATATTATTTAGTTGTGTT-3′ and 5′-ACTTAAATACTTTCAACACTTATCCC-3′. Primer sequences for untreated DNA were 5′-TGGTGGGATACTACCCTAGC-3′ and 5′-GCTTAGATGCTTTCAGCACTT-3′. Amplification of these fragments was performed under standard PCR conditions (10 min at 94°C, followed by 35 cycles of 30 s at 94°C, 30 s at 50°C, and 1 min at 72°C) using FastStart Taq DNA polymerase (Roche Diagnostics, Almere, The Netherlands). The obtained fragments were purified and sequenced on a MegaBACE 500 platform as described elsewhere (3). In Fig. Fig.2,2, changes in the genomic DNA sequence upon sodium bisulfite treatment are shown. Almost all of the original C residues are converted to T residues, as can be expected for unmethylated cytosines. However, a number of C residues still remain present. Alignment of the context of nonconverted C residues showed that all nonconverted C residues are part of a consensus sequence CC*NGG, where C* is the position affected by methylation (Fig. (Fig.2).2). None of the examined C residues that were converted to thymine residues matched this consensus sequence. Furthermore, in the control strain, all original cytosines were converted to thymine residues (not shown). For the genus Staphylococcus as a whole, no enzyme with such a recognition site has yet been reported (http://rebase.neb.com/).
Apart from DNA degradation, methylation of genomic DNA adds another condition leading to uninterpretable PFGE analysis using the standard restriction endonuclease SmaI. However, as shown, this can easily be overcome by adopting existing protocols to include the use of restriction enzymes not influenced by the presence of methylated residues. In addition, since methylation does not affect the ability of DNA to be amplified by PCR, restriction enzyme-independent PCR-based fingerprinting protocols may also overcome the problems associated with DNA methylation.