Logo of jbacterPermissionsJournals.ASM.orgJournalJB ArticleJournal InfoAuthorsReviewers
J Bacteriol. Oct 2006; 188(19): 6858–6868.
PMCID: PMC1595521

Comparative Genomic Hybridization Analysis of Enterococcus faecalis: Identification of Genes Absent from Food Strains

Abstract

Enterococcus faecalis, a member of the natural microbiota of animal and human intestinal tracts, is also present as a natural contaminant in a variety of fermented foods. Over the last decade, E. faecalis has emerged as a major cause of nosocomial infections. We investigated the genetic diversity in 30 clinical and food isolates, including strains V583 and MMH594, in order to determine whether clinical and food isolates could be distinguished. Data were obtained using comparative genomic hybridization and specific PCR with a total of 202 probes of E. faecalis, selected using the available V583 genome sequence and part of the MMH594 pathogenicity island. The cognate genes encoded mainly exported proteins. Hybridization data were analyzed by a two-component mixture model that estimates the probability of any given gene to be either present or absent in the strains. A total of 78 genes were found to be variable, as they were absent in at least one isolate. Most of the variable genes were clustered in regions that, in the published V583 sequence, related to prophages or mobile genetic elements. The variable genes were distributed in three main groups: (i) genes equally distributed between clinical and dairy food isolates, (ii) genes absent from dairy food-related isolates, and (iii) genes present in MMH594 and V583 strains only. Further analysis of the distribution of the last gene group in 70 other isolates confirmed that six of the probed genes were always absent in dairy food-related isolates, whereas they were detected in clinical and/or commensal isolates. Two of them corresponded to prophages that were not detected in the cognate isolates, thus possibly extending the number of genes absent from dairy food isolates. Genes specifically detected in clinical isolates may prove valuable for the development of new risk assessment markers for food safety studies and for identification of new factors that may contribute to host colonization or infection.

Enterococci are ubiquitous low-GC percentage gram-positive bacteria encountered in various environments, including animal and human intestinal tracts, soil, plants and water. They are found as members of the natural microbiota of a variety of fermented food products such as artisanal cheeses and fermented sausages and reportedly play an important role in food processing (23, 28). Enterococci have been consumed for centuries, and both Enterococcus faecalis and Enterococcus faecium species are subdominant members of the digestive microbiota in human. However, isolates of both species are emerging as major causes of nosocomial infections, including urinary tract and abdominal infections, bacteremia, and endocarditis in patients with severe underlying diseases or an impaired immune system (40). E. faecalis causes 60 to 80% of enterococcal infections. Less than 2% of infections are due to strains resistant to clinically relevant antibiotics, ampicillin and vancomycin, implying that other genetic factors are necessary for E. faecalis infection and virulence potential (30).

Phenotypic studies suggest that E. faecalis strains vary in their colonization and invasion abilities and, thus, likely vary in their virulence potential (14, 26, 39). Overall genetic diversity of E. faecalis has been reported using various molecular typing methods (for review, see reference 18; also J. C. Ogier and P. Serror, submitted for publication). While diversity might explain the presence of E. faecalis in various environmental niches and contribute to virulence, little is known about the identity and the distribution of the variable genes. Among the dozen of E. faecalis putative virulence factors reported (for a review, see references 27 and 35), sets of known and potential virulence factors (e.g., aggregation substance, enterococcal surface protein [Esp], cytolysin toxin [Cyl], and gelatinase [GelE]) are widespread among various collections of isolates, including food-associated isolates (3, 10, 14, 16, 17, 23, 33, 53). The findings that E. faecalis virulence genes are detected in food-associated isolates calls for safety assessment measures (10, 16, 23). Up to now, studies of E. faecalis genetic diversity have focused on just three chromosomal regions, known as the pathogenicity island (PAI) (43, 54), the fsr locus, and the capsular polysaccharide gene clusters (32, 41, 48). Recent sequence availability of the clinical E. faecalis V583 genome (46) and of the pathogenicity island of strain MMH594 (54) has opened the way to explore E. faecalis genome diversity using DNA array technology for transcriptome analysis (1) and comparative genomic hybridization (CGH). CGH between pathogenic and nonpathogenic or avirulent isolates within a single species has proven useful for delineating putative bacterial pathogen determinants (7, 8, 12, 13, 15, 37, 51, 56). Despite the widespread use of CGH, only few statistical approaches for data analysis have been developed. The most common data analysis methods use a constant ratio value as a threshold for assigning genes into either the absent or present categories. This threshold is usually empirically determined from a comparison of reference strains for which the gene distribution is known. This method supposes a high reproducibility between arrays. However, inherent differences between membranes and the fluctuating efficiency of DNA labeling are responsible for variation between arrays. The mixture model has proven to be a powerful statistical method to classify genes in a finite number of groups (11, 24, 38, 45).

To investigate genetic diversity of clinical and food isolates of E. faecalis on a larger scale, we explored 30 E. faecalis isolates from clinical and food origins for the presence of 202 genes. For this purpose, we performed comparative genomic hybridization on a focused macroarray containing 186 genes, mainly encoding exported proteins, and specific PCR on 16 genes. We applied a statistical approach based on the mixture model to ensure reliable DNA array data analysis that would distinguish the presence or absence of genes in a given isolate.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

The E. faecalis strains used in this study are listed in Table Table1.1. These include the type strain and isolates of clinical (15 isolates), food (13), and commensal (1 fecal isolate from a healthy volunteer) origins. Clinical strains were obtained from the Centre Hospitalier de Versailles (CHV), the Centre Hospitalier Universitaire of Lyon (25), or the Health Sciences Center at University of Oklahoma. Food isolates were provided by the CNRZ (National Centre for Zootechnical Research) collection at Jouy-en-Josas and the INRA collection at Aurillac. The commensal isolate studied was obtained from the Health Sciences Center at the University of Oklahoma. To assess the distribution of genes scored as absent in food isolates, 70 additional E. faecalis clinical, food, and commensal isolates from diverse locations (Argentina, Egypt, England, France, and the United States) were analyzed by specific PCR. A description of these strains is available upon request.

TABLE 1.
Enterococcus faecalis isolates by source of isolation and MLVA type

Diversity of the 30 isolates was first determined by multilocus variable-number tandem-repeat analysis (MLVA) using a modified procedure developed by Titze-de-Almeida and collaborators (57). For each PCR, 40 ng of total DNA was suspended in 20 μl containing 12 pmol of region-specific pairs of primers (MWG Biotech, Courtabeuf, France), a 0.2 mM concentration of each deoxynucleotide triphosphate, 1× T.Pol incubation buffer (10 mM Tris-HCl [pH 9.0], 50 mM KCl, 0.1% Triton X-100, 0.2 mg/ml bovine serum albumin) containing 1.5 mM MgCl2 (Qbiogene, Illkirch, France) and 0.5 U of Taq DNA polymerase (Qbiogene). Reactions were performed in a Mastercycler gradient (Eppendorf). PCR was carried out as previously described (57) with 15 cycles of touchdown amplification-denaturation steps, except that 10 cycles of standard PCR with annealing at 55°C were added to increase the PCR product yields. PCR product sizes were analyzed on 1.5% agarose gels to determine the number of repeats. MLVA data were analyzed with the classification multivariate method of StatGraphics Plus, version 5.1. Equal weights were given to large and small numbers of differences in the number of repeats within a particular locus. MLVA analysis of the 30 isolates using aceB, espC, efa3, efa5, and efa6 repeats (56) resulted in 19 distinct MLVA types (Table (Table11).

To investigate the distribution of genes enriched in clinical and food isolates, a total of 70 isolates were analyzed by PCR in addition to the 30 isolates analyzed by DNA array hybridization. All strains were grown at 37°C without shaking in brain heart infusion broth. Detection of gelatinase activity was determined on Todd-Hewitt (Difco Laboratories) agar plates containing 3% gelatin and revealed as described previously (51).

Probe and primer design and macroarray construction.

The macroarray was designed using 211 E. faecalis genes, of which 205 genes were from strain V583 and 6 were from strain MMH594. This selection includes three housekeeping genes used as positive controls (gyrA, dnaN, and rpoB corresponding to EF0002, EF0006, and EF3238 in V583, respectively). Three genes used as negative controls were from E. coli (cheZ) and Saccharomyces cerevisiae (YAL058C-A and YAL047C).

Oligonucleotide primers were designed to amplify PCR fragments ranging from 200 bp to 500 bp. Most of the primers were identical to those used in the V583 microarray designed by Aakra and collaborators (1). In order to perform two-stage PCR amplification (47), each primer was designed with a nonvariable adaptamer sequence at the 5′ end. The adaptamer sequences were 5′-TACCTTCTCGAGGGGAC and 5′-ACCCTCTCGTGGGCAG for forward and reverse primers, respectively. The first PCR step was performed in 50 μl of reaction mixture containing 20 ng of genomic DNA, a 200 μM concentration each deoxyribonucleotide, a 300 nM concentration of each primer (MWG Biotech), and 1.5 U of Taq DNA polymerase (Qbiogene). Amplification consisted of a denaturation step at 94°C of 5 min, followed by 2 cycles of 30 s at 94°C, 30 s at 56°C, and 1 min at 72°C and 25 cycles of 30 s at 94°C and 1 min 30 s at 72°C, with a final step of 1 min at 72°C. PCR products were purified after electrophoresis in 1.2% low-melting-temperature agarose using a DNA gel extraction kit (Millipore, Bedford, Mass.). Second-round PCR products were generated by reamplification of the purified first-round PCR products using the adaptamer sequences as primers. Reactions were carried out with 1 μl of 100-fold diluted PCR product, a 200 μM concentration of each deoxynucleotide triphosphate, a 300 nM concentration of each primer, and 2 U of Qbiogene Taq DNA polymerase. Amplification was achieved by denaturation at 94°C for 5 min, followed by 30 cycles of 94°C for 30 s, 54°C for 30 s, and 72°C for 30 s, with an additional extension period at 72°C for 1 min. Before spotting, PCR products were precipitated with 2.5 volumes of absolute ethanol and resuspended in 3× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate). PCR product concentration was adjusted to 200 ng/μl after DNA quantification by gel electrophoresis and Fluoroskan Ascent (Thermo Electron Corporation, Boston, Mass.). PCR samples were printed in duplicate on Nylon membranes (Hybond-N+; Amersham Biosciences, Buckinghamshire, England), using a Qbot macroarrayer (Genetix, Hampshire, United Kingdom) by the Centre de Ressources Biologiques GADIE (INRA, Jouy-en-Josas, France). After spotting, DNA was submitted to a denaturation step by treating membranes with 0.5 M NaOH-1.5 M NaCl, followed by a neutralization step with 1 M Tris-HCl-M NaCl, and membranes were rinsed four times with milliQ water. Membranes were dried for 1 h at room temperature and for 2 h at 80°C.

The specificity of DNA macroarray probes was analyzed by sequencing the 214 amplified DNA fragments with an ABI PRISM 3700 DNA analyzer from Applied Biosystems. In total, 186 E. faecalis genes, among which 180 were from strain V583 and 6 from strain MMH594, were considered for data macroarray analysis. Sixteen probes, which had to be excluded from the macroarray because of mixed PCR products (EF0511, EF0540, EF0605, EF0776, EF1420, EF2253, EF2525, PAIEF0047, and PAIEF0053) or putative cross-hybridization (EF0355, EF1896, EF1992, EF2250, EF2347, EF3256, and EFA0047), were analyzed using PCR. Their presence in the 30 isolates was examined by specific PCR amplification with primer pairs used for amplification of the cognate PCR probes. A full list of primers, PCR product sizes, and their nucleotide sequences are available as supplemental material (see Table S1 in the supplemental material).

DNA methods.

Total DNA was extracted from 4 ml of late-exponential phase cultures as previously described (19). About 50 ng of genomic DNA was labeled with 50 μCi of α-32P-labeled dATP (Amersham Biosciences, Orsay, France) using the RadPrime DNA labeling system (Invitrogen, Carlsbad, Calif.). Labeled genomic DNA was then purified from unincorporated nucleotides on Microspin S-200 HR or ProbeQuant G-50 micro columns (Amersham Biosciences) according to manufacturer's instructions.

Membranes were soaked in 2× SSC and prehybridized for 2 h at 42°C in hybridization buffer (6× SSC, 50% formamide, 1% sodium dodecyl sulfate [SDS], 10× Denhart's solution [0.2% Ficoll, 0.2% polyvinylpyrrolidone, 0.2% bovine serum albumin]) and 15 μg/ml of denatured and sonicated salmon sperm DNA. Labeled genomic DNA was denatured at 100°C for 5 min and hybridized for 16 h at 42°C. Membranes were washed using conditions of increasing stringency: 10 min at 42°C in 6× SSC-50% formamide-0.1% SDS, 10 min at 60°C in 2× SSC-0.1% SDS, 10 min at 60°C in 0.2× SSC-0.1% SDS, and 10 min at 65°C in 0.1× SSC-0.1% SDS. They were then sealed in Saran Wrap and exposed to Storage Phosphor GP screens (Amersham Biosciences) for 3 to 5 days. Screens were scanned at 100-μm resolution using a Storm Imager (Amersham Biosciences). Signal intensities were quantified using Imagene software, version 5.5 (BioDiscovery, El Segundo, CA).

Specific PCR amplifications were performed using conditions similar to those for probe preparation, except that the initial denaturation step was followed by five cycles of 30 s at 94°C, 30 s at 56°C, and 1 min at 72°C. This modification was introduced so as to allow amplification in the case of mismatches due to potential sequence variations between isolates.

To test whether the absence of the EF1420 and EF2144 genes was linked to the absence of complete prophages, we analyzed the junction regions of prophages 03 and 05 of V583 in strains that lacked the prophage genes. The primer pair OEF168 (5′-TGGGGTTAATCCATTTGACC-3′) and OEF169 (5′-TGACAGCTAAACAGTATGCG-3′) and the pair OEF170 (5′-AAATCTGTCATTCCAGCGAC-3′) and OEF171 (5′-TTTGACGATTACTCGTCGC-3′) were used to amplify the junction region between EF1416 and EF1490 and between EF2083 and EF2146, respectively.

Data analysis.

The bimodal distribution of hybridization signals highlights the existence of two different gene populations corresponding in fact to their presence or absence. Each gene population fits a Gaussian distribution model. The hybridization signal distribution, f(x), is modeled by a mixture defined as the weighted sum of the two Gaussian distributions (one for each population): f(x) = pN(x121) + (1 − p)N(x222), where x, the hybridization signal, is the log-transformed signal mean intensity, p is the proportion of genes in the first class, and N(xi2i) is the Gaussian density of probability with mean μi and variance σ2i of population i (i = 1 or 2).

The parameters of the mixture model are estimated by maximum likelihood using an expectation maximization (EM) algorithm implemented in the library MCLUST (20, 21) available in the R statistical environment (http://www.r-project.org). The maximization step of the EM algorithm generates an estimation of the parameters (mean and variance) of each Gaussian component and of the mixing proportion p. The expectation step computes the probabilities that the observed signals belong to either class. The EM algorithm was applied using several initial values, and parameters maximizing the likelihood of the model were retained. Gene classification was based on the probability that the relevant probe belonged to the population of high or low hybridization signals (i.e., respectively, present or absent genes). The macroarray data were analyzed using Class2G (http://migale.jouy.inra.fr/class2g), an R plug-in based on the R package MCLUST and integrated in the BioArray Software Environment (49).

RESULTS AND DISCUSSION

Macroarray data analysis.

In setting up comparative genomic hybridization for E. faecalis, we first developed a method to distinguish positive from negative hybridization signals using a normal mixture model with two Gaussian components. Comparative genomic hybridization of E. faecalis strains was performed on a set of 186 E. faecalis genes, which mainly encode putative membrane proteins, lipoproteins, cell wall surface proteins, or secreted proteins (see Table S1 in the supplemental material). We specifically focused on bacterial cell surface proteins that are known to participate in adaptation and/or survival in various environments. Among them were included E. faecalis known virulence genes (gelE, cylB, cspI, and asa1). We also took care to select genes at different genome locations. Total DNA isolated from 30 strains including V583 and MMH594 strains as positive controls was used for macroarray hybridization. Raw and analyzed hybridization data are available in the supplemental material (see Table S2). No hybridization was detected when E. faecium DNA was used as target (data not shown), showing that the conditions used for gene detection with the array were specific. In our macroarray, EF3238, encoding the beta subunit of RNA polymerase, shares the highest DNA sequence identity with the E. faecium counterpart (85%). Consequently, we estimated that the sequence identity of genes classified as present was above 85%. According to our data and the data of other investigators (42, 44), nucleotide sequence identity between gene homologs within the E. faecalis species is at least 98%. The classification of genes as absent from a given isolate was therefore considered as unambiguous. Classification of a gene as present or absent often relies on empirical cutoff ratios calculated from independent experiments with a reference strain. However, due to inherent experimental variability, this analysis method leads to the misclassification of genes (36). To ensure reliable data analysis of our hybridization results—and since our primary aim was to classify genes in two groups, i.e., present and absent—we analyzed distribution of the hybridization signal of each array by a normal mixture model with two Gaussian components using Class2G software (see Materials and Methods). Hybridization signals clearly segregated into two groups, strong and weak, indicating the presence or absence of the specific target genes. As expected, hybridization signals of strain V583 (Fig. (Fig.1A)1A) exhibited an almost normal distribution due to the fact that most of the probed genes were selected from that genome. In contrast, a bimodal distribution was observed for the tested strains, indicating that some genes were absent, as shown in Fig. Fig.1B1B.

FIG. 1.
Distribution and density of probability of hybridization signals modeled by a mixture model with two Gaussian components obtained with the V583 (A) and VE14531 (B) strains.

We observed some overlap between the two populations, such that an accurate threshold to classify genes as present (class 1) or absent (class 0) could not be ascertained. For final classification, genes were classified as present (probability of belonging to class 1 was ≥0.9), absent (probability of belonging to class 1 was ≤0.1) or ambiguous (probability of belonging to class 1 was between 0.1 and 0.9). This allows detection of potentially unreliable classification (ambiguous genes) that can be further investigated to ensure results. The status of the genes classified as ambiguous was established by 249 PCR amplifications. Overall, 121 genes were confirmed present in some isolates. In addition, to estimate the true-positive rate (ratio of positives correctly classified over the total positives) and the false-positive rate (ratio of negatives incorrectly classified over total negatives), we performed 1,130 specific PCR amplifications. The true-positive and the false-positive rates had an average of 84% (median, 90%) and 16% (median, 12%), respectively. False-positive misclassification may result from (i) biased choice of PCR amplifications, as half of them were performed on genes classified as ambiguous in some isolates, (ii) the attraction of the larger group (class 1), and (iii) cross-hybridizations between fragment probes and homologous genomic sequences as paralogs. Indeed, as the E. faecalis V583 genome carries a number of paralogs (46), it is likely that some of the genes classified as present correspond to paralogs with different cellular roles. Closer analysis of our probes revealed that probes EF0146, EF0487, EF0492, EF3076, and EF3253 may cross-hybridize as they have V583 paralogs sharing more than 89% identity at the nucleotide level. Their detection in the isolates thus indicates the presence of at least one member of the corresponding gene family.

The data analysis method used in this work is particularly adapted for comparative genomic hybridization, especially for DNA-DNA macroarray data in the context of gene classification. Its main advantage over commonly used methods is that it consists of an array-to-array analysis that limits the impact of experimental variations between strains. This is because it requires neither arbitrarily or experimentally fixed threshold values related to reference strain(s) nor extensive data normalization. Furthermore, it provides a measure of confidence in the gene classification.

Identification and distribution of the variable genes.

Results of the macroarray data on 186 E. faecalis probes plus the 16 genes analyzed by PCR (see Material and Methods and Table S3 in the supplemental material) as tested on 30 isolates are shown in Fig. Fig.2.2. Of the 202 E. faecalis genes, a total of 124 were detected in all 30 isolates, indicating that they are conserved. The number of absent genes in comparison to strains V583 and MMH594 varied between 68 (in food isolate VE14585) and 30 (in clinical isolate VE14514). In total, 78 genes (Table (Table2)2) proved to be variable, and their distribution in isolates was classified in three groups.

FIG. 2.
Detection of a selection of E. faecalis genes in 30 isolates from clinical and food origins. The individual chromosomes are displayed vertically, and genes are ordered according to their organization in the reference strain V583 and in MMH494 PAI. The ...
TABLE 2.
Distribution of 78 genes identified as variable between E. faecalis isolates

Group I.

The largest group comprised 51 genes that were equally detected in both food and clinical isolates, indicating that these genes cannot be used to determine strain origins (Table (Table2).2). For example, gelE and asa1, corresponding to EF1818 and EFA0047 in V583, are characterized as virulence factors yet are detected in several food isolates. These results agree with previous studies, although a lower incidence of these genes in food isolates was reported (16, 53). Nevertheless, gelatinase activity was not detected in 38 and 50% of the food and clinical isolates of our collection, respectively (Table (Table1).1). Indeed, various isolates carrying gelE reportedly fail to produce gelatinase, likely due to a deletion of the fsr cluster region (16, 41, 48). The low frequency of gene EF1825, which is included in the 23.9-kb deletion comprising the fsr locus, suggests that such deletions may have occurred (Table (Table1).1). For GelE-positive strains that do not carry EF1825, the deletion may have occurred such that the fsr genes are still present. Similarly, genes cpsI (EF2487), cpsH (EF2488), and EF2175 belong or are close to the capsular polysaccharide biosynthesis gene clusters cps (EF2485 to EF2495) and epa (EF2180 to EF2200) that are implicated in E. faecalis virulence (29, 58, 59). These genes are also found in food isolates.

The low incidence of asa1 and cpsI in clinical isolates observed here contrasts with previous studies (3, 10, 29), possibly reflecting differences in the origins of the isolates. However, the low asa1 frequency in this study may be explained by a higher primer specificity since primers used in previous reports also matched with the other aggregation substance-encoding genes prgB (2, 9) and asp1 (9) of plasmids pCF10 and pPD1, respectively. This observation may explain why isolates VE14568 and VE14571 were found negative for asa1 compared to the study of Archimbaud et al. (3). We therefore suggest that the frequency of asa1 has been overestimated in previous studies.

Conservation of the genes encoding factors important for virulence may suggest that selection for the maintenance of such traits may exist in their natural environment, which might be as diverse as soil or plants, and in insect, reptile, bird, or mammal digestive tracts (for a review, see reference 2). Although all the above factors are known to enhance virulence in animal or cellular models, they may not be sufficient for E. faecalis pathogenicity.

Group II.

The variable genes of the second group were more often found in clinical isolates than in food isolates. Thirteen were absent in the dairy food isolates, and two genes were more frequent in clinical isolates (Table (Table2).2). Genes EF0552 and EF0553 belong to a putative operon encoding a xylose-containing oligosaccharide phosphotransferase system (PTS) transporter (60) and could confer the ability to colonize particular biotopes, such as plants, that provide xylose-rich polymers.

We examined 70 additional enterococcal isolates to confirm the classification of genes scored as absent from a total of 50 dairy food strains. Specific PCR analysis was performed using gene-specific primer pairs. Interestingly, cylB that codes the ABC transporter of cytolysin, turned out to be present in a particular class of food isolates corresponding to ewe and goat cheese isolates (Table (Table3).3). The higher rates of cylB in food isolates in previous studies may be due to sampling enriched in these food groups (16, 52, 53). Gene EF2170, which is part of the cps cluster, was significantly underrepresented in dairy food isolates. Differential distribution of cps genes could impact the cell wall polysaccharide composition and thereby contribute to serological differences among isolates (31, 32).

TABLE 3.
Distribution of 11 genes among a collection of 100 isolatesa

Six of the PCR-probed genes were not detected in any of the dairy food isolates tested, whereas they were found in 13 to 21% of the clinical or commensal isolates (Table (Table3).3). Interestingly, three of them (EF0573, EF0592, and EF0605) were codetected in the same clinical isolates, suggesting that the isolates may derive from a common ancestor. Among these, EF0592, an adhesin-like encoding gene, was reported to be exclusively associated with clinical isolates (10). Genes EF1420 and EF2144 encode two nonhomologous putative lipoproteins that are specific to E. faecalis and are embedded within prophages. When tested in several food isolates, the entire prophages were absent (see below), raising the possibility that up to 139 genes are coordinately absent. To our knowledge, E. faecalis phages have been scarcely studied; however, it is tempting to speculate that, as previously proposed for several bacteria, prophages may encode fitness factors that confer a benefit under peculiar ecological conditions (4).

This is the first report of identification of E. faecalis genes absent from food isolates. Their low incidence in clinical and commensal isolates suggests that these genes may act as putative fitness factors. However, their detection in food isolates may be a valuable indicator of potential risk.

Group III.

The third group comprises genes absent in all isolates except reference strain V583 and related isolate MMH594 (Table (Table3).3). Eight genes (excepting V583-specific EF2513) may be good markers for the common lineage of MMH594 and V583. Some of them may be particularly mobile as they are close to or part of putative mobile elements (compiled in Table Table4).4). Since bacterial pathogenesis is a multifactorial process, strain-specific genes may contribute to the epidemic spread of this lineage. These genes are likely to constitute clonal markers that should be traced to follow their spread among isolates.

TABLE 4.
Putative mobile genetic elements predicted in V583 genomea

Strain variability and gene conservation.

Among the genes analyzed in this work, 61% (124/202) were shared by all isolates tested and may be part of the “core” set of E. faecalis genes. Conservation of putative adhesion proteins suggests that they may be related to a global requirement for adhesion to survive in different ecological niches. Among these, the cell wall-anchored proteins with tandem repeats of the immunoglobulin fold, encoded by the EF0089, EF1093, EF1099, EF1269, and EF2224 genes, were recently reported to be ubiquitous among E. faecalis (43, 55). The high-affinity dicarboxylate carbohydrate transport system encoded by the EF0429-EF0431 operon is rarely found in gram-positive bacteria but is totally conserved in E. faecalis. In the V583 genome, it is located in a region rich in putative carbohydrate metabolism genes that may encode important functions in particular environments. Regarding the diversity of the natural environments of E. faecalis, conserved genes are probably inherent to certain of its lifestyles. Some, if not all, of these genes may be required for gastrointestinal lifestyles in humans and/or animals.

Cluster analysis on the set of the 78 variable genes revealed three strain clusters (Fig. (Fig.3).3). Two clusters comprised isolates of different geographical and environmental origins. Based on our data, the urine isolate VE14514 from France is related to MMH594 and V583, mainly due to the conservation of the PAI genes probed, indicating that the PAI might be spread worldwide (43). Interestingly, besides sharing PAI genes, strain clusters I and III share EF2168, EF2170, and EF2175 and lack the putative operon EF2682-EF2686 and EF3023 and EF3075, respectively (Fig. (Fig.2).2). We also identified loci that distinguish the two isolates MMH594 and V583. Strain MMH594 was found to lack genes EF0153, EF0163, EF0164, EF2237, and EF2513. This suggests that V583 may have diverged from MMH594, not only by the acquisition of vanB and the deletion of the 17-kb region which encompass esp (43, 54) but also by acquisition of at least part of the mobile element efaC2 (Table (Table4).4). In addition, detection of EF2282 and EF2347 in MMH594 indicates that vanB genes have been transferred in an existing mobile element, leading to a complex structure identified as a putative integrative and conjugative element (5). This suggests that, independently of the evolution of the PAI region, other genes clearly contribute to define E. faecalis lineages.

FIG. 3.
Dendrogram showing the relationship between 30 E. faecalis isolates. Clustering analysis was performed using the StatGraphics program. The gene dissimilarity used is the squared Euclidean distance obtained from a binary matrix (genes × isolates) ...

Not surprisingly, comparative genome hybridization reveals greater strain diversity than MLVA. Although we did distinguish food isolates by the absence of six genes, these genes were insufficient to cluster the isolates in a separate category (Fig. (Fig.3).3). These results suggest that E. faecalis isolates may be intrinsically related due to the fact that food isolates are likely to result from fecal animal contamination. Food isolates may have factors for establishment in their natural host, making it difficult to distinguish them from those of clinical origin.

Global genetic plasticity relative to the reference genome.

When examined in the V583 genome, half of the variable genes are scattered over the chromosome, while 38 were adjacent or located in discrete regions predicted to be mobile genetic elements (Table (Table44).

Two genes (EF1420 and EF2144) tested in this work were located in putative prophages. Interestingly, they were absent in all food isolates tested (group II). Their absence may correspond to the absence of the cognate prophage. To further investigate the corresponding regions, we designed oligonucleotides to amplify a fragment spanning the predicted insertion site (46). Sequence analysis of the amplicons from strains that lacked EF1420 and EF2144 allowed us to identify two deletions of 48,373 bp and 43,518 bp, respectively. Consequently, we could delineate the integration site of prophages 03 and 05 of V583 strain. Prophage 03 was integrated between EF1416 and EF1490 in the 3′ end of EF1416, which encodes a glucose-6-phosphate isomerase. Prophage 05 was integrated in the 3′ end of a gene encoding tRNAThr2. Prophages 03 and 05 are flanked by a 76-bp repeat and a 15-bp repeat, respectively, corresponding to their attachment sites (Table (Table4).4). These results demonstrate that prophages 03 and 05 may have been acquired recently since they are absent in several natural strains. These two prophages, which were not detected in the analyzed dairy food isolates, account for a total of 135 genes. Most of them are of unknown function, and, as stated above, they may contribute to E. faecalis adaptation to various environments including the host. Further investigation of their distribution is required to ascertain their absence in food isolates.

The E. faecalis PAI region is of particular interest since it was clearly associated with two isolates responsible for multiple infections (54). Our data confirmed that the putative pathogenicity island exhibits a high variability and is largely disseminated among isolates, in keeping with a recently published study (43). In this study, 8 of the 30 strains lacked the 20 genes present in the PAI of MMH594. Several PAI genes (PAIEF0053, esp, EF0540, EF0577, and EF0582) were detected in about half of the isolates; others (EF0501, EF0518, cylB, cylA, PAIEF0050, EF0573, EF0592, and EF0605) had a lower incidence. This variability confirms the modular structure of this mobile genetic element. In contrast to a scenario where PAI would have evolved through deletion events only, our data suggest that PAI may evolve by gain and loss of partial PAI segments, as suggested by Shankar and coworkers (54).

All these results demonstrate that the genetic variability of E. faecalis isolates relates to gene loss and acquisition to not only the PAI region but also other genomic regions including capsular polysaccharide clusters, prophages, and putative mobile genetic elements. Analysis of the percent GC content of variable genes (Table (Table2)2) suggests that several of them have been recently acquired by horizontal transfer from distant organisms like gram-negative bacteria. However, the majority may result from intraspecies transfer. The efficient E. faecalis conjugation systems are likely to facilitate formation of variants leading to genome flexibility. As an opportunistic bacteria, it is likely that E. faecalis pathogenesis results from the coordinated expression of diverse fitness and virulence factors favoring its adaptation to the hostile environment of the host and its antibacterial defenses (6). We speculate that E. faecalis genes identified as overrepresented in or specific to clinical isolates may constitute putative fitness factors by conferring an advantage to E. faecalis to establish in humans. Further genome sequencing of multiple E. faecalis isolates from various origins would enhance our progress in the identification of specific sequences that may encode adaptation factors to adverse environmental conditions and contribute to enhanced pathogenicity.

In summary, this is the first large-scale study of E. faecalis genome diversity that gives a first picture of the stable versus variable regions in the E. faecalis chromosome. We applied a more robust statistical method for gene distribution analysis by macroarray, which could be of general use. Several discrete regions of variability were identified, including two V583 putative prophages whose the integration sites were mapped. Our data confirm the modular structure of the PAI region with subregions conserved in more that half of the isolates. A significant finding of this work is that six of the probed genes appear to be absent from the dairy food isolates. Even if these genes are not ubiquitous in clinical isolates, they may constitute good markers for risk assessment regarding E. faecalis isolates that are found in fermented products. Genes specifically detected or overrepresented in clinical isolates are of particular interest as they may be fitness factors contributing to the development of human infection. Since E. faecalis pathogenesis is a multifactor process, we believe that extended genomic studies will allow identification of the genes needed for E. faecalis colonization ability and pathogenicity.

Supplementary Material

[Supplemental material]

Acknowledgments

We are very grateful to A. Gruss, M. A. Petit from UBLO, and F. Rodolphe from MIG for careful reading of the manuscript. We thank P. Y. Allouch, M. Gilmore, M. C. Montel, J. Anba, D. B. Clewell, J. J. Gratadoux, and E. Vandenesch for providing the isolates; we thank V. Loux and R. Bossy for genome analysis using the AGMIAL package. P.S. is indebted to A. Gross for her support.

This work was supported by the Institut National de la Recherche Agronomique and the Agence Française de Sécurité Sanitaire et Environnementale. S.B. was supported by a doctoral fellowship from the Institut National de la Recherche Agronomique and the Région Ile-de-France.

Footnotes

Supplemental material for this article may be found at http://jb.asm.org/.

REFERENCES

1. Aakra, A., H. Vebo, L. Snipen, H. Hirt, A. Aastveit, V. Kapur, G. Dunny, B. Murray, and I. F. Nes. 2005. Transcriptional response of Enterococcus faecalis V583 to erythromycin. Antimicrob. Agents Chemother. 49:2246-2259. [PMC free article] [PubMed]
2. Aarestrup, F. M., P. Butaye, and W. Witte. 2002. Nonhuman reservoirs of enterococci, p. 55-99. In M. S. Gilmore, D. B. Clewell, P. Courvalin, G. M. Dunny, B. E. Murray, and L. B. Rice (ed.), The enterococci: pathogenesis, molecular biology, and antibiotic resistance. ASM, Washington, D.C.
3. Archimbaud, C., N. Shankar, C. Forestier, A. Baghdayan, M. S. Gilmore, F. Charbonne, and B. Joly. 2002. In vitro adhesive properties and virulence factors of Enterococcus faecalis strains. Res. Microbiol. 153:75-80. [PubMed]
4. Brussow, H., C. Canchaya, and W.-D. Hardt. 2004. Phages and the evolution of bacterial pathogens: from genomic rearrangements to lysogenic conversion. Microbiol. Mol. Biol. Rev. 68:560-602. [PMC free article] [PubMed]
5. Burrus, V., G. Pavlovic, B. Decaris, and G. Guedon. 2002. The ICESt1 element of Streptococcus thermophilus belongs to a large family of integrative and conjugative elements that exchange modules and change their specificity of integration. Plasmid 48:77-97. [PubMed]
6. Casadevall, A., and L. A. Pirofski. 1999. Host-pathogen interactions: redefining the basic concepts of virulence and pathogenicity. Infect. Immun. 67:3703-3713. [PMC free article] [PubMed]
7. Chan, K., S. Baker, C. C. Kim, C. S. Detweiler, G. Dougan, and S. Falkow. 2003. Genomic comparison of Salmonella enterica serovars and Salmonella bongori by use of an S. enterica serovar Typhimurium DNA microarray. J. Bacteriol. 185:553-563. [PMC free article] [PubMed]
8. Chen, T., Y. Hosogi, K. Nishikawa, K. Abbey, R. D. Fleischmann, J. Walling, and M. J. Duncan. 2004. Comparative whole-genome analysis of virulent and avirulent strains of Porphyromonas gingivalis. J. Bacteriol. 186:5473-5479. [PMC free article] [PubMed]
9. Clewell, D. B., P. K. Tomich, M. C. Gawron-Burke, A. E. Franke, Y. Yagi, and F. Y. An. 1982. Mapping of Streptococcus faecalis plasmids pAD1 and pAD2 and studies relating to transposition of Tn917. J. Bacteriol. 152:1220-1230. [PMC free article] [PubMed]
10. Creti, R., M. Imperi, L. Bertuccini, F. Fabretti, G. Orefici, R. Di Rosa, and L. Baldassarri. 2004. Survey for virulence determinants among Enterococcus faecalis isolated from different sources. J. Med. Microbiol. 53:13-20. [PubMed]
11. Dean, N., and A. E. Raftery. 2005. Normal uniform mixture differential gene expression detection for cDNA microarrays. BMC Bioinformatics 6:173. [PMC free article] [PubMed]
12. Dobrindt, U., F. Agerer, K. Michaelis, A. Janka, C. Buchrieser, M. Samuelson, C. Svanborg, G. Gottschalk, H. Karch, and J. Hacker. 2003. Analysis of genome plasticity in pathogenic and commensal Escherichia coli isolates by use of DNA arrays. J. Bacteriol. 185:1831-1840. [PMC free article] [PubMed]
13. Doumith, M., C. Cazalet, N. Simoes, L. Frangeul, C. Jacquet, F. Kunst, P. Martin, P. Cossart, P. Glaser, and C. Buchrieser. 2004. New aspects regarding evolution and virulence of Listeria monocytogenes revealed by comparative genomics and DNA arrays. Infect. Immun. 72:1072-1083. [PMC free article] [PubMed]
14. Dupre, I., S. Zanetti, A. M. Schito, G. Fadda, and L. A. Sechi. 2003. Incidence of virulence determinants in clinical Enterococcus faecium and Enterococcus faecalis isolates collected in Sardinia (Italy). J. Med. Microbiol. 52:491-498. [PubMed]
15. Dziejman, M., E. Balon, D. Boyd, C. M. Fraser, J. F. Heidelberg, and J. J. Mekalanos. 2002. Comparative genomic analysis of Vibrio cholerae: genes that correlate with cholera endemic and pandemic disease. Proc. Natl. Acad. Sci. USA 99:1556-1561. [PMC free article] [PubMed]
16. Eaton, T. J., and M. J. Gasson. 2001. Molecular screening of Enterococcus virulence determinants and potential for genetic exchange between food and medical isolates. Appl. Environ. Microbiol. 67:1628-1635. [PMC free article] [PubMed]
17. Elsner, H. A., I. Sobottka, D. Mack, M. Claussen, R. Laufs, and R. Wirth. 2000. Virulence factors of Enterococcus faecalis and Enterococcus faecium blood culture isolates. Eur. J. Clin. Microbiol. Infect. Dis. 19:39-42. [PubMed]
18. Facklam, R. R., M. G. S. Carvalho, and L. M. Teixeira. 2002. History, taxonomy, biochemical characteristics, and antibiotic susceptibility testing of enterococci, p. 1-54. In M. S. Gilmore, D. B. Clewell, P. Courvalin, G. M. Dunny, B. E. Murray, and L. B. Rice (ed.), The enterococci: pathogenesis, molecular biology, and antibiotic resistance. ASM Press, Washington, D.C.
19. Fouet, A., and A. L. Sonenshein. 1990. A target for carbon source-dependent negative regulation of the citB promoter of Bacillus subtilis. J. Bacteriol. 172:835-844. [PMC free article] [PubMed]
20. Fraley, C., and A. E. Raftery. 1999. MCLUST: software for model-based cluster analysis. J. Classification 16:297-306.
21. Fraley, C., and A. E. Raftery. 2002. Model-based clustering, discriminant analysis, and density estimation. J. Am. Stat. Assoc. 97:611-631.
22. Franke, A. E., and D. B. Clewell. 1981. Evidence for a chromosome-borne resistance transposon (Tn916) in Streptococcus faecalis that is capable of “conjugal” transfer in the absence of a conjugative plasmid. J. Bacteriol. 145:494-502. [PMC free article] [PubMed]
23. Franz, C. M., A. B. Muscholl-Silberhorn, N. M. Yousif, M. Vancanneyt, J. Swings, and W. H. Holzapfel. 2001. Incidence of virulence factors and antibiotic resistance among enterococci isolated from food. Appl. Environ. Microbiol. 67:4385-4389. [PMC free article] [PubMed]
24. Garrett, E. S., and G. Parmigiani. 2003. Statistical methods for qualitative analysis of gene expression, p. 362-387. In G. Parmigiani, E. S. Garrett, R. A. Irizarry, and S. L. Zeger (ed.), The analysis of gene expression data. Spinger-Verlag, New York, N.Y.
25. Gauduchon, V., L. Chalabreysse, J. Etienne, M. Celard, Y. Benito, H. Lepidi, F. Thivolet-Bejui, and F. Vandenesch. 2003. Molecular diagnosis of infective endocarditis by PCR amplification and direct sequencing of DNA from valve tissue. J. Clin. Microbiol. 41:763-766. [PMC free article] [PubMed]
26. Gentry-Weeks, C., M. Estay, C. Loui, and D. Baker. 2003. Intravenous mouse infection model for studying the pathology of Enterococcus faecalis infections. Infect. Immun. 71:1434-1441. [PMC free article] [PubMed]
27. Gilmore, M. S., P. S. Coburn, S. R. Nallapareddy, and B. E. Murray. 2002. Enterococcal virulence, p. 301-354. In M. S. Gilmore, D. B. Clewell, P. Courvalin, G. M. Dunny, B. E. Murray, and L. B. Rice (ed.), The enterococci: pathogenesis, molecular biology, and antibiotic resistance. ASM Press, Washington, D.C.
28. Giraffa, G. 2003. Functionality of enterococci in dairy products. Int. J. Food Microbiol. 88:215-222. [PubMed]
29. Hancock, L. E., and M. S. Gilmore. 2002. The capsular polysaccharide of Enterococcus faecalis and its relationship to other polysaccharides in the cell wall. Proc. Natl. Acad. Sci. USA 99:1574-1579. [PMC free article] [PubMed]
30. Hancock, L. E., and M. S. Gilmore. 2000. Pathogenicity of enterococci, p. 251-258. In V. A. Fischetti (ed.), Gram-positive pathogens. American Society for Microbiology, Washington, D.C.
31. Huebner, J., Y. Wang, W. A. Krueger, L. C. Madoff, G. Martirosian, S. Boisot, D. A. Goldmann, D. L. Kasper, A. O. Tzianabos, and G. B. Pier. 1999. Isolation and chemical characterization of a capsular polysaccharide antigen shared by clinical isolates of Enterococcus faecalis and vancomycin-resistant Enterococcus faecium. Infect. Immun. 67:1213-1219. [PMC free article] [PubMed]
32. Hufnagel, M., L. E. Hancock, S. Koch, C. Theilacker, M. S. Gilmore, and J. Huebner. 2004. Serological and genetic diversity of capsular polysaccharides in Enterococcus faecalis. J. Clin. Microbiol. 42:2548-2557. [PMC free article] [PubMed]
33. Huycke, M. M., and M. S. Gilmore. 1995. Frequency of aggregation substance and cytolysin genes among enterococcal endocarditis isolates. Plasmid 34:152-156. [PubMed]
34. Huycke, M. M., C. A. Spiegel, and M. S. Gilmore. 1991. Bacteremia caused by hemolytic, high-level gentamicin-resistant Enterococcus faecalis. Antimicrob. Agents Chemother. 35:1626-1634. [PMC free article] [PubMed]
35. Kayaoglu, G., and D. Orstavik. 2004. Virulence factors of Enterococcus faecalis: relationship to endodontic disease. Crit. Rev. Oral Biol. Med. 15:308-320. [PubMed]
36. Kim, C. C., E. A. Joyce, K. Chan, and S. Falkow. 2002. Improved analytical methods for microarray-based genome-composition analysis. Genome Biol. 3:RESEARCH0065. [Online.] http://genomebiology.com/2002/3/11/research/0065. [PMC free article] [PubMed]
37. Koide, T., P. A. Zaini, L. M. Moreira, R. Z. Vencio, A. Y. Matsukuma, A. M. Durham, D. C. Teixeira, H. El-Dorry, P. B. Monteiro, A. C. da Silva, S. Verjovski-Almeida, A. M. da Silva, and S. L. Gomes. 2004. DNA microarray-based genome comparison of a pathogenic and a nonpathogenic strain of Xylella fastidiosa delineates genes important for bacterial virulence. J. Bacteriol. 186:5442-5449. [PMC free article] [PubMed]
38. McLachlan, G. J., R. W. Bean, and D. Peel. 2002. A mixture model-based approach to the clustering of microarray expression data. Bioinformatics 18:413-422. [PubMed]
39. Mohamed, J. A., W. Huang, S. R. Nallapareddy, F. Teng, and B. E. Murray. 2004. Influence of origin of isolates, especially endocarditis isolates, and various genes on biofilm formation by Enterococcus faecalis. Infect. Immun. 72:3658-3663. [PMC free article] [PubMed]
40. Mundy, L. M., D. F. Sahm, and M. Gilmore. 2000. Relationships between enterococcal virulence and antimicrobial resistance. Clin. Microbiol. Rev. 13:513-522. [PMC free article] [PubMed]
41. Nakayama, J., R. Kariyama, and H. Kumon. 2002. Description of a 23.9-kilobase chromosomal deletion containing a region encoding fsr genes which mainly determines the gelatinase-negative phenotype of clinical isolates of Enterococcus faecalis in urine. Appl. Environ. Microbiol. 68:3152-3155. [PMC free article] [PubMed]
42. Nallapareddy, S. R., R. W. Duh, K. V. Singh, and B. E. Murray. 2002. Molecular typing of selected Enterococcus faecalis isolates: pilot study using multilocus sequence typing and pulsed-field gel electrophoresis. J. Clin. Microbiol. 40:868-876. [PMC free article] [PubMed]
43. Nallapareddy, S. R., H. Wenxiang, G. M. Weinstock, and B. E. Murray. 2005. Molecular characterization of a widespread, pathogenic, and antibiotic resistance-receptive Enterococcus faecalis lineage and dissemination of its putative pathogenicity island. J. Bacteriol. 187:5709-5718. [PMC free article] [PubMed]
44. Naser, S. M., F. L. Thompson, B. Hoste, D. Gevers, P. Dawyndt, M. Vancanneyt, and J. Swings. 2005. Application of multilocus sequence analysis (MLSA) for rapid identification of Enterococcus species based on rpoA and pheS genes. Microbiology 151:2141-2150. [PubMed]
45. Newton, M., and C. Kendziorski. 2003. Parametric empirical Bayes methods for microarrays, p. 254-259. In G. Parmigiani, E. S. Garrett, R. A. Irizarry, and S. L. Zeger (ed.), The analysis of gene expression data. Spinger-Verlag, New York, N.Y.
46. Paulsen, I. T., L. Banerjei, G. S. Myers, K. E. Nelson, R. Seshadri, T. D. Read, D. E. Fouts, J. A. Eisen, S. R. Gill, J. F. Heidelberg, H. Tettelin, R. J. Dodson, L. Umayam, L. Brinkac, M. Beanan, S. Daugherty, R. T. DeBoy, S. Durkin, J. Kolonay, R. Madupu, W. Nelson, J. Vamathevan, B. Tran, J. Upton, T. Hansen, J. Shetty, H. Khouri, T. Utterback, D. Radune, K. A. Ketchum, B. A. Dougherty, and C. M. Fraser. 2003. Role of mobile DNA in the evolution of vancomycin-resistant Enterococcus faecalis. Science 299:2071-2074. [PubMed]
47. Richmond, C. S., J. D. Glasner, R. Mau, H. Jin, and F. R. Blattner. 1999. Genome-wide expression profiling in Escherichia coli K-12. Nucleic Acids Res. 27:3821-3835. [PMC free article] [PubMed]
48. Roberts, J. C., K. V. Singh, P. C. Okhuysen, and B. E. Murray. 2004. Molecular epidemiology of the fsr locus and of gelatinase production among different subsets of Enterococcus faecalis isolates. J. Clin. Microbiol. 42:2317-2320. [PMC free article] [PubMed]
49. Saal, L. H., C. Troein, J. Vallon-Christersson, S. Gruvberger, A. Borg, and C. Peterson. 2002. BioArray software environment (BASE): a platform for comprehensive management and analysis of microarray data. Genome Biol. 3:SOFTWARE0003.1-0003.6. [Online.] http://genomebiology.com/2002/3/8/software/0003. [PMC free article] [PubMed]
50. Sahm, D. F., J. Kissinger, M. S. Gilmore, P. R. Murray, R. Mulder, J. Solliday, and B. Clarke. 1989. In vitro susceptibility studies of vancomycin-resistant Enterococcus faecalis. Antimicrob. Agents Chemother. 33:1588-1591. [PMC free article] [PubMed]
51. Salama, N., K. Guillemin, T. K. McDaniel, G. Sherlock, L. Tompkins, and S. Falkow. 2000. A whole-genome microarray reveals genetic diversity among Helicobacter pylori strains. Proc. Natl. Acad. Sci. USA 97:14668-14673. [PMC free article] [PubMed]
52. Semedo, T., M. Almeida Santos, P. Martins, M. F. Silva Lopes, J. J. Figueiredo Marques, R. Tenreiro, and M. T. Barreto Crespo. 2003. Comparative study using type strains and clinical and food isolates to examine hemolytic activity and occurrence of the cyl operon in enterococci. J. Clin. Microbiol. 41:2569-2576. [PMC free article] [PubMed]
53. Semedo, T., M. A. Santos, M. F. Lopes, J. J. Figueiredo Marques, M. T. Barreto Crespo, and R. Tenreiro. 2003. Virulence factors in food, clinical and reference enterococci: a common trait in the genus? Syst. Appl. Microbiol. 26:13-22. [PubMed]
54. Shankar, N., A. S. Baghdayan, and M. S. Gilmore. 2002. Modulation of virulence within a pathogenicity island in vancomycin-resistant Enterococcus faecalis. Nature 417:746-750. [PubMed]
55. Sillanpaa, J., Y. Xu, S. R. Nallapareddy, B. E. Murray, and M. Hook. 2004. A family of putative MSCRAMMs from Enterococcus faecalis. Microbiology 150:2069-2078. [PubMed]
56. Smoot, J. C., K. D. Barbian, J. J. Van Gompel, L. M. Smoot, M. S. Chaussee, G. L. Sylva, D. E. Sturdevant, S. M. Ricklefs, S. F. Porcella, L. D. Parkins, S. B. Beres, D. S. Campbell, T. M. Smith, Q. Zhang, V. Kapur, J. A. Daly, L. G. Veasy, and J. M. Musser. 2002. Genome sequence and comparative microarray analysis of serotype M18 group A Streptococcus strains associated with acute rheumatic fever outbreaks. Proc. Natl. Acad. Sci. USA 99:4668-4673. [PMC free article] [PubMed]
57. Titze-de-Almeida, R., R. J. L. Willems, J. Top, I. Pereira Rodrigues, R. Fonseca Ferreira, I. I., H. Boelens, M. C. C. Brandileone, R. C. Zanella, M. S. Soares Felipe, and A. van Belkum. 2004. Multilocus variable-number tandem-repeat polymorphism among Brazilian Enterococcus faecalis strains. J. Clin. Microbiol. 42:4879-4881. [PMC free article] [PubMed]
58. Xu, Y., B. E. Murray, and G. M. Weinstock. 1998. A cluster of genes involved in polysaccharide biosynthesis from Enterococcus faecalis OG1RF. Infect. Immun. 66:4313-4323. [PMC free article] [PubMed]
59. Xu, Y., K. V. Singh, X. Qin, B. E. Murray, and G. M. Weinstock. 2000. Analysis of a gene cluster of Enterococcus faecalis involved in polysaccharide biosynthesis. Infect. Immun. 68:815-823. [PMC free article] [PubMed]
60. Zuniga, M., I. Comas, R. Linaje, V. Monedero, M. J. Yebra, C. D. Esteban, J. Deutscher, G. Perez-Martinez, and F. Gonzalez-Candelas. 2005. Horizontal gene transfer in the molecular evolution of mannose PTS transporters. Mol. Biol. Evol. 22:1673-1685. [PubMed]

Articles from Journal of Bacteriology are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...