ccrABEnt serine recombinase genes are widely distributed in the Enterococcus faecium and Enterococcus casseliflavus species groups and are expressed in E. faecium

The presence, distribution and expression of cassette chromosome recombinase (ccr) genes, which are homologous to the staphylococcal ccrAB genes and are designated ccrABEnt genes, were examined in enterococcal isolates (n=421) representing 13 different species. A total of 118 (28 %) isolates were positive for ccrABEnt genes by PCR, and a number of these were confirmed by Southern hybridization with a ccrAEnt probe (n=76) and partial DNA sequencing of ccrAEnt and ccrBEnt genes (n=38). ccrABEnt genes were present in Enterococcus faecium (58/216, 27 %), Enterococcus durans (31/38, 82 %), Enterococcus hirae (27/52, 50 %), Enterococcus casseliflavus (1/4, 25 %) and Enterococcus gallinarum (1/2, 50 %). In the eight other species tested, including Enterococcus faecalis (n=94), ccrABEnt genes were not found. Thirty-eight sequenced ccrABEnt genes from five different enterococcal species showed 94–100 % nucleotide sequence identity and linkage PCRs showed heterogeneity in the ccrABEnt flanking chromosomal genes. Expression analysis of ccrABEnt genes from the E. faecium DO strain showed constitutive expression as a bicistronic mRNA. The ccrABEnt mRNA levels were lower during log phase than stationary phase in relation to total mRNA. Multilocus sequence typing was performed on 39 isolates. ccrABEnt genes were detected in both hospital-related (10/29, 34 %) and non-hospital (4/10, 40 %) strains of E. faecium. Various sequence types were represented by both ccrABEnt positive and negative isolates, suggesting acquisition or loss of ccrABEnt in E. faecium. In summary, ccrABEnt genes, potentially involved in genome plasticity, are expressed in E. faecium and are widely distributed in the E. faecium and E. casseliflavus species groups.


INTRODUCTION
The emergence of multidrug-resistant hospital-acquired Enterococcus faecium as one of the most important pathogens in the developed world has been a remarkable development in the last two decades (Leavis et al., 2006;Werner et al., 2003). Molecular epidemiological studies and comparative genomic hybridization analyses of E. faecium (Leavis et al., 2007;Werner et al., 2003) have revealed genotypic differences between hospital and community isolates (Leavis et al., 2006). Mixed whole genome arrays demonstrated a distinct genetic make-up of hospital-associated E. faecium with more than 100 extra genes, possibly acquired by horizontal gene transfer (Leavis et al., 2007). The esp virulence gene, located on a putative pathogenicity island, is one of the determinants acquired by hospital-associated E. faecium. These observations, as well as current multilocus sequence typing (MLST) data, strongly indicate that gene flux and recombination contribute significantly to diversification and adaptation of E. faecium (Leavis et al., 2006;van Schaik et al., 2010).
Recombinases facilitate the exchange of DNA fragments within and between bacteria and are thus pivotal in genome plasticity. Staphylococcal cassette chromosome (SCC) elements are vehicles for exchange of genetic information in staphylococci. These elements are characterized by the presence of terminal inverted repeats and unique recombinase genes, and are flanked by direct repeats (Ito et al., 2001(Ito et al., , 2004Katayama et al., 2003). So far, the major group of elements described are SCCmec I-VIII (International Working Group on the Classification of Staphylococcal Cassette Chromosome Elements, 2009) responsible for the spread of methicillin resistance between staphylococci. The movement of SCC elements is dependent on the gene products of the cassette chromosome recombinase genes (ccr), either the ccrA-ccrB complex or the single product of ccrC (Katayama et al., 2000;Noto & Archer, 2006). These proteins are serine recombinases of the resolvase/invertase family which integrate the SCC element in a site-specific manner (Ito et al., 1999). To our knowledge, ccr genes have only been reported in staphylococcal species.
Here, we report for the first time to our knowledge, the presence of ccrAB genes in enterococci, hereby designated ccrAB Ent , and show that they are expressed under standard in vitro growth conditions. Our analyses show that the ccrAB Ent genes are widely distributed in Enterococcus species belonging to the E. faecium and Enterococcus casseliflavus species groups.

Detection of ccrAB
The evolutionary relationships of CcrAB Ent , Ccr of staphylococci (deduced from ccrA1, ccrA2, ccrA3, ccrA4, ccrB1, ccrB3, ccrB4 and ccrC), and three other site-specific recombinases (site-specific integrase of bacteriophage Q-FC1 found in E. faecalis and two site-specific recombinases from Clostridium acetobutylicum ATCC824) were further investigated. These were included because they have been part of previous similar analyses (Ito et al., 2004) and because the ccrA and ccrB, as well as one of the recombinases from C. acetobutylicum (AE001437; locus tag no. CAC 2247), have been annotated as if they were DNA invertase Pin homologue proteins. The full-length ccrB1 of NCTC10442 and ccrB4 of HDE288 were reconstituted as described earlier (Ito et al., 2004). A neighbour-joining tree was constructed using MEGA3 (Kumar et al., 2004) by creating 2000 bootstrap replicates. Sites with gaps/ missing data were excluded during analyses. Recombination within the sequenced regions of ccrA Ent and ccrB Ent was determined by phi test (Bruen et al., 2006 Bacterial DNA extraction for PCR analyses was performed manually by using the InstaGene matrix kit (Bio-Rad) or the GenoM-48 robotic workstation using GenoPrep DNA from blood, standard kit (Genovision). DNA for hybridization purposes was isolated using guanidium isothiocyanate (Dahl & Sundsfjord, 2003).
For long range PCR, 2 U DNA polymerase enzyme rTth XL (Perkin Elmer) was used per reaction and 1.4 mM Mg(OAc) 2 in a standard XL PCR mix, or a 0.76 Pfu Ultra mix (Stratagene) with 2.5 U Pfu Ultra polymerase per reaction. DNA sequencing was performed using BigDye 3.1 technology (Applied Biosystems). Real-time PCR was performed using ABI Prism 7300 real-time PCR system (PE Biosystems) and TaqMan universal mastermix (Applied Biosystems).
Detection of ccrAB Ent genes and PCR linkage to surrounding genes. ccrAB Ent genes were detected by PCR, using the primer pairs FA-RA and FB-RB, respectively (Table 2), and genes in selected isolates were detected by Southern hybridization and DNA sequencing. PCRs were also performed on 13 of 14 ccrAB Ent -positive E. faecium isolates selected for MLST as well as two ccrAB Ent -positive E. faecium animal isolates from Norway, to search for the presence and conservation of gene synteny in the surrounding genes (Table 2 and Fig. 1a). Primers and probes were designed using E. faecium DO sequences as template.
Expression analysis of ccrAB Ent genes by real-time quantitative PCR. To analyse if ccrAB Ent genes are expressed, E. faecium DO was grown aerobically in BHI broth at 37 uC for 18-24 h. Subsequently the culture was diluted 1 : 50 in BHI broth and grown with agitation to OD 600 0.7 or to stationary phase (grown overnight). The cell suspension was centrifuged and the cells were immediately frozen on dry ice or liquid nitrogen before adding an RNA stabilizing solution, RNA later (Ambion). Alternatively, RNA later or RNA protect (Qiagen) was added directly to the inoculum, according to the manufacturer's instructions. RNA extraction was performed by using the RNeasy mini kit (Qiagen) using a prolonged lysis step of 45 min with 10 mg lysozyme and 10 U mutanolysin in a total volume of 100 ml. On-column DNase treatment was performed according to the manufacturer's instructions. A successive removal of DNA was performed using Turbo DNase (Ambion) according to the manufacturer's instructions. RNA integrity was determined by agarose gel electrophoresis. Reverse transcription of the total RNA was performed using the ABRTR1 primer and the High Capacity cDNA Reverse Transcription kit (Applied Biosystems) or Superscript III RNase Hreverse transcriptase (Invitrogen). Real-time PCR was performed on the cDNA using primers ccrAFre, ccrARre, ccrBFre, ccrBRre, recAFre, recARre, pbp5Fre, pbp5Rre, adkFre and adkRre, and probes ccrA Ent , ccrB Ent , recA, pbp5 and adk ( Table 2). Expression of ccrAB Ent genes was compared with the expression of recA, pbp5 and adk. Ten-fold serial dilutions of E. faecium DO genomic DNA were used to make standard curves to determine PCR efficiency, using the equation: E510 (21/slope) 21. The PCR efficiencies ranged from 88 to 104 % in one assay and 99 to 100 % in a second assay and were considered similar enough to be able to compare only C t (threshold cycle) values for a semiquantitative relative measurement of expression. The expression experiments were performed in three triplicates; a no template control (NTC) and a minus reverse transcriptase control (2RT) was included in each assay. The 2RT controls were in the range of an acceptable difference from the cDNA expression analysis (.5C t difference).
Analysis of ccrAB Ent mRNA linkage by RT-PCR. RNA isolation was performed as described above. RNA was treated with the DNAfree kit (Ambion). Reverse transcription of total RNA was performed with SuperScript III reverse transcriptase (Invitrogen) using primers CcrBRTR1 or CcrBxR. RT-PCR without reverse transcriptase was performed on total RNA to check for DNA contamination. Linkage of ccrA Ent and ccrB Ent mRNAs as a bicistronic mRNA was analysed by PCRs on cDNAs using primers located in ccrA Ent (CcrARTR1 and CcrAxF) and ccrB Ent (CcrBRTR1 and CcrBxR) ( Fig. 1b and Table 2).

RESULTS AND DISCUSSION
The evolutionary relationships of CcrAB Ent , Ccr of staphylococci and three other site-specific recombinases were further investigated. The phylogenetic analyses revealed an evolutionary relationship between CcrA Ent and CcrB Ent from enterococci and the staphylococcal CcrAB cluster (Fig. 2). However, the low identity score between the enterococcal and staphylococcal proteins does not support a recent horizontal transfer of the ccr genes between these species.
ccrAB Ent genes are expressed in E. faecium Analyses of ccrAB Ent gene expression were performed during both the exponential and stationary phase of E. faecium DO grown in rich medium. Both genes were expressed in approximately the same amounts in exponential phase. ccrAB Ent genes were expressed .70-fold lower than the pbp5, recA and adk genes (Supplementary   S1). The mRNA abundance of ccrAB Ent was lower in stationary phase than in exponential phase. ccrAB Ent gene sequences (GenBank accession nos FJ572967-FJ573039) from E. faecium (n514), E. hirae (n510 for ccrA Ent and 11 for ccrB Ent ), E. durans (n510), E. gallinarum (n51) and E. casseliflavus (n51) isolates were aligned and a neighbour-joining phylogenetic tree was made with 2000 bootstrap replicates using the P-distance model (Fig. 3). The ccrAB Ent genes both clustered into two major clades represented by the majority of E. faecium (clade I) and E. hirae (clade II) isolates, respectively. With 7 of 10 isolates clustering in clade II, E. hirae appears to be slightly more dispersed between the two ccrA Ent clades. ccrAB Ent from the E. gallinarum and E. casseliflavus isolates clustered in clade II with the majority of ccrAB Ent from the E. hirae isolates. In E. durans, 6 of 10 ccrA Ent genes clustered in clade I, while 7 of 10 ccrB Ent clusters were in clade II. Except for ccrA Ent from E. faecium E1304, the ccrAB Ent genes of the human isolates clustered in clade I whereas the animal isolates were found in both clades. Incongruence between ccrA Ent and ccrB Ent phylogenies within an isolate was noted for 11 isolates, all of animal origin (Fig. 3, isolates marked with asterisks). Phi tests revealed no statistically significant evidence for recombination within the sequenced regions of the ccrA Ent and ccrB Ent genes. However, the incongruence suggests recombination of the ccr Ent genes outside the sequenced regions of the two genes. Incongruence between these genes has also been seen for S. aureus (Ito et al., 2004).
ccrAB Ent genes were only found in isolates belonging to the E. faecium and E. casseliflavus species groups that belong to the same tree branch in phylogenetic trees based on enterococcal 16S and sodA gene diversity (Devriese et al., 1993;Poyart et al., 2000). The absence of ccrAB Ent in the other species could be explained by the low number of isolates tested, except for E. faecalis, or by a lack of integration sites recognized by ccrAB Ent in the strains not belonging to the E. faecium or E. casseliflavus groups. Alternatively, their ccrAB Ent genes may exhibit such a low sequence identity to the ccrAB Ent genes identified in this study that they are missed using the PCR and hybridization conditions used in the present study.
Variations of the ccrAB Ent genes and the surrounding region between selected E. faecium isolates PFGE analysis and Southern hybridization of 76 E. faecium isolates with the ccrA Ent probe confirmed the PCR results.
One ccrA Ent PCR-negative strain (399/F98/A1) was ccrA Entpositive by Southern blot hybridization (data not shown) indicating that sequence diversity affects PCR amplification. Also, XbaI analyses of ccrA Ent and ccrB Ent genomic regions revealed heterogeneity and only one copy of ccrA Ent . The ccrA Ent probe hybridized to an approximately 10 kb fragment in DO, TUH7-55, E1304 and E1293 isolates; however, the ccrA Ent -positive fragment of E0470 and E0745 was approximately 24 kb (data not shown). To investigate this in more detail, the presence of ccrAB Ent flanking genomic genes identified in the DO genome was determined by multiple PCRs in 15 ccrAB Ent -positive and 16 ccrAB Ent -negative isolates (Fig. 1a). Examinations of the ccrAB Ent surrounding region in several isolates showed a variable pattern of the ccrAB Ent flanking sequences with hospital-associated isolates showing most sequence similarity with the DO sequence (Table 3). All 31 isolates were positive for the tnp gene-specific PCR (tnp belongs to the IS30 family) as well as for orf1 PCR and three ccrAB Entpositive isolates of different sequence types (STs) were also positive for the REP factor gene PCR. This REP factor gene harbours a REP_trans domain belonging to superfamily pfam02486. This family represents probable topoisomerases that makes a sequence-specific single stranded nick in the origin of replication. Plasmid REPs, phage REPs (RstAs) and transposon REPs (Cro/CI transcriptional regulators) belong to this family. Long-range PCRs confirmed linkage of these genes with ccrAB Ent and conservation of gene synteny surrounding ccrAB Ent with the exception of isolates 64/F99/H6, 399/F99/A10, 399/F99/ H8, and S399/F99/A14, for which linkage of tnp-orf1 and orf1-ccrB Ent was not confirmed. Furthermore, a ccrB Ent -ccrA Ent linkage was not shown in 64/F99/H6 (Table 3 and Fig. 1a). The inability to link genes that were positive on gene-specific PCRs may indicate that the region between these genes is larger than expected or that the specific genes are located at other regions in the genome. The transposase of the IS30 family is, for instance, located at more than one site in E. faecium DO. Annotation of contig 655 (http:// maple.lsd.ornl.gov/cgi-bin/JGI_microbial/contig_viewer.cgi? org=efae&chr=08jun04&contig=Contig655&sort=left_bp, on 21 June 2010) also indicates that the ccrAB Ent genes are located in a region containing several transposases. The   (Wang & Archer, 2010) and we have showed that the ccrAB Ent genes are expressed in E. faecium DO. It has been postulated that SCC may carry the genes conferring methicillin resistance but may also enable genetic exchange of other genes among staphylococcal species (Katayama et al., 2003). However, to our knowledge, no studies have provided direct experimental evidence for intercellular transfer of SCC between staphylococci.
DNA sequencing of the ccrAB Ent , tnp and orf1 of the 15 ccrAB Ent -positive isolates showed 94-100 % and 96-100 % sequence identity in ccrA Ent and ccrB Ent genes (GenBank accession nos FJ572978-FJ572981, FJ572997-FJ573001, FJ573014-FJ573018, FJ573032-FJ573036), respectively, while sequences of orf1 and tnp were 100 % identical in all isolates (data not shown). According to Hanssen et al. (2004), up to 4 % variation within the ccrAB genes has been observed for a given staphylococcal species. The ccrAB genes found in SCCmec types II and IV can vary up to 5 % at the nucleotide level (Noto & Archer, 2006). Since both ccrAB Ent genes and the staphylococcal ccrAB genes show sequence variations within the recombinase genes, which have the same gene synteny and variable surrounding regions, we hypothesize that they may have similar functions in contributing to excision and integration of surrounding genes within the genome and possibly also mobilization of surrounding genes between cells. The asterisks indicate isolates in which ccrA Ent and ccrB Ent belong to different clades. All sequences were aligned using CLUSTAL W. The neighbour-joining phylogenetic tree was made with MEGA4.0 using 2000 bootstrap replicates and the P-distance model. Bootstrap values higher than 80 % are shown at the branches. The scale bar indicates genetic distance in substitutions per site. The two main clades of ccrA Ent and ccrB Ent are indicated. The dataset consisted of 547 nt with 494 parsimony-informative sites for ccrA Ent and 513 nt with 227 parsimony-informative sites for ccrB Ent .
Investigation of possible association between ccrAB Ent and ST within E. faecium of human origin MLST analyses of E. faecium isolates (n539) revealed that the ccrAB Ent genes are dispersed among different STs (Table 1). Ten of 29 (34 %) hospital-related E. faecium isolates were ccrAB Ent -positive, while 4 of 10 (40 %) nonhospital-related isolates were positive. Furthermore, specific STs within hospital-related strains were represented by both ccrAB Ent -positive and -negative isolates ( Table 1), suggesting that ccrAB Ent genes are acquired and not a part of the core genome.

Concluding remarks
Cassette chromosome recombinases may be important in recombination and genome plasticity in enterococci.
Expression analyses indicate that the recombinase genes are active in E. faecium DO and thus, may play a role in the recombination or movement of genetic elements. Further investigation of the ccrA Ent and ccrB Ent will be essential to reveal the contribution of these genes for recombination and mobilization events in enterococci.