• No results found

Distribution of Genes Encoding MSCRAMMs and Pili in Clinical and Natural Populations of Enterococcus faecium

N/A
N/A
Protected

Academic year: 2020

Share "Distribution of Genes Encoding MSCRAMMs and Pili in Clinical and Natural Populations of Enterococcus faecium"

Copied!
6
0
0

Loading.... (view fulltext now)

Full text

(1)

0095-1137/09/$08.00⫹0 doi:10.1128/JCM.02283-08

Copyright © 2009, American Society for Microbiology. All Rights Reserved.

Distribution of Genes Encoding MSCRAMMs and Pili in Clinical and

Natural Populations of

Enterococcus faecium

Jouko Sillanpa

¨a

¨,

1,2

† Vittal P. Prakash,

1,2

†‡ Sreedhar R. Nallapareddy,

1,2

† and Barbara E. Murray

1,2,3

*

Division of Infectious Diseases, Department of Internal Medicine,1Center for the Study of Emerging and Re-Emerging Pathogens,2

and Department of Microbiology and Molecular Genetics,3University of Texas Medical School, Houston, Texas 77030

Received 28 November 2008/Returned for modification 16 January 2009/Accepted 27 January 2009

Enterococcus faeciumhas recently emerged as an important cause of nosocomial infections. We previously

identified 15 predicted surface proteins with characteristics of MSCRAMMs and/or pili and demonstrated that their genes were frequently present in 30 clinicalE. faeciumisolates studied; one of these,acm, has been studied in further detail. To determine the prevalence of the other 14 genes among variousE. faeciumpopulations, we have now assessed 433E. faeciumisolates, including 264 isolates from human clinical infections, 69 isolates from stools of hospitalized patients, 70 isolates from stools of community volunteers, and 30 isolates from animal-related sources. A variable distribution of the 14 genes was detected, with their presence ranging from 51% to 98% of isolates. While 81% of clinical isolates carried 13 or 14 of the 14 genes tested, none of the community group isolates and only 13% of animal isolates carried 13 or 14 genes. The presence of these genes was most frequent in endocarditis isolates, with 11 genes present in all isolates, followed by isolates from other clinical sources. The number of genes significantly associated with clinical versus fecal or animal origin (P0.04 to <0.0001) varied from 10 to 13, depending on whether comparisons were made against individual clinical subgroups (endocarditis, blood, and other clinical isolates) or against all clinical isolates combined as one group. The strong association of these genes with clinical isolates raises the possibility that their preservation/ acquisition has favored the adaptation ofE. faeciumto nosocomial environments and/or patients.

Enterococci, commensal members of the intestinal flora in humans and animals, have emerged over the last 3 decades as important hospital-associated opportunistic pathogens causing a variety of infections, including urinary tract infections, sur-gical site infections, bacteremia, and infective endocarditis (20, 21). Several species of enterococci have been recognized, but more than 90% of infections are caused by only two species,

Enterococcus faecalis and Enterococcus faecium. Until the

1990s, infections due toE. faecaliswere reported to outnumber those due to E. faecium by 9 to 1, but more recently, the incidence ofE. faeciumhas increased dramatically, and it has been found as the causative organism in 20 to 36% of entero-coccal infections in some U.S. hospitals (8, 14, 25, 33). Simi-larly, hospital-associatedE. faeciuminfections also appear to be an emerging phenomenon in some European countries (32). Coinciding with this trend, treatment ofE. faecium infec-tions has increasingly been hampered by the spread of strains resistant to aminoglycosides,␤-lactams, and glycopeptides, the three major classes of drugs of therapeutic choice. Lately,E.

faeciumstrains resistant to other classes of newly developed

antibiotics have also emerged (17), narrowing available thera-peutic options.

We and others recently analyzed the genome sequence of

endocarditis-derivedE. faeciumstrain TX0016 and identified 22 genes, named fms (E. faecium surface protein-encoding) genes, predicted to encode LPXTG family cell wall-anchored surface proteins (13, 28). We subsequently identified 15 of the 22 genes, includingacm (fms8), a previously described gene encoding a collagen adhesin, as harboring characteristic fea-tures of MSCRAMMs and/or pili, such as predicted structural organization into modules with immunoglobulin-like folding, and demonstrated that all 15 genes occurred frequently among 30 clinically derivedE. faeciumisolates studied (24, 28). Using a larger collection of diverseE. faeciumisolates, we detected production of Acm predominantly in isolates derived from clinical sources (23), despite the gene being present in almost all isolates. Furthermore, deletion ofacmresulted in signifi-cant attenuation in an experimental endocarditis model (22), indicating involvement in E. faecium pathogenesis. Another MSCRAMM-encoding gene that we have characterized,fms10

(subsequently renamedscm), encodes a second collagen-bind-ing MSCRAMM inE. faecium. Recombinant Scm differs from Acm in collagen type specificity by showing affinity for collagen type V instead of types I and IV, which are specifically bound by Acm. Eleven of the remaining 13 genes were found to be clustered into four separate genomic loci, each with an adja-cent class C sortase gene characteristic of pilus-encoding genes of gram-positive bacteria, and were predicted to form four distinct pili (28). One of these clusters, E. faecium ebpABC

(ebpABCfm), is transcribed as a single operon, and we

demon-strated polymerization of its predicted major pilus subunit protein, EbpCfm, into a high-molecular-weight complex (28). A

recent study by Hendrickx et al. showed that antibodies against EbpCfm(also referred to as PilB) and Fms21 (also referred to

as PilA) stained pili on the surface ofE. faecium(12).

* Corresponding author. Mailing address: Division of Infectious Diseases, Department of Internal Medicine, University of Texas Med-ical School at Houston, 6431 Fannin Street, MSB 2.112, Houston, TX 77030. Phone: (713) 500-6745. Fax: (713) 500-6766. E-mail: bem.asst @uth.tmc.edu.

† J.S., V.P.P., and S.R.N. contributed equally to this study. ‡ Present address: Department of Pathology, Baylor College of Medicine, Houston, TX 77030.

Published ahead of print on 4 February 2009.

896

on May 16, 2020 by guest

http://jcm.asm.org/

(2)

Gene association studies have suggested that acquisition of resistance determinants to antibiotics, in particular to ampicil-lin and subsequently vancomycin, as well as acquisition of other traits, such as a functionalacmgene,espEfm(encoding an

enterococcal surface protein), and hylEfm (hyaluronidase-like

gene), may have facilitated colonization, infection, and/or transmission, although the order of these events cannot cur-rently be determined (16, 23, 26). Hendrickx and colleagues recently demonstrated enrichment of 5 of 22 genes encoding cell wall-anchored proteins in a major hospital-associated genogroup, designated clonal complex 17 (CC17) (16, 34), in a study of 131E. faeciumisolates (12, 13). Hence, it is reasonable to hypothesize that some of the 14 MSCRAMM- and/or pilus-encoding genes (in addition to acm) may also contribute to pathogenesis and/or provideE. faeciumwith a selective advan-tage in hospital settings, for example, by conferring adherence properties on host tissues. Enrichment of these traits may also have contributed to the recent nosocomial success of CC17.

In the present study, we tested a large collection ofE. fae-ciumisolates to assess the global distribution of the recently identified 14 MSCRAMM- and/or pilus-encoding genes in var-ious clinical and natural populations ofE. faecium.The gene distribution and association of the isolates with a clinical or nonclinical origin were also analyzed.

MATERIALS AND METHODS

Bacterial isolates, species identification, and growth conditions.A total of 433 clinical and nonclinicalE. faeciumisolates were included in this study. These isolates were collected between 1973 and 2005 from diverse geographic locations (the United States, Argentina, Brazil, France, Belgium, Poland, The Nether-lands, Spain, Norway, and China), including epidemic and outbreak strains from around the world (1, 2, 4, 6, 7, 9–11, 15, 18, 19, 23, 26, 27, 29, 31). Isolates known to be identical by pulsed-field gel electrophoresis to other isolates in this collec-tion were excluded. TheE. faeciumisolates were divided into different groups based on their source of isolation. Among the 264 human clinical isolates (clin-ical group), 15 were derived from blood cultures of endocarditis patients (endo-carditis subgroup) and 57 were from blood cultures of nonendo(endo-carditis patients (blood subgroup). The remaining 192E. faeciumclinical isolates, which were collected from specimens of bile, bone, catheters, cervix, cerebrospinal fluid, placenta, peritoneal fluid, sputum, urine, and wounds, were grouped as other clinical isolates (OCI subgroup). Sources of nonclinical isolates included stools of hospitalized patients (n⫽69; hospital stool group) and community volunteers (n⫽70; community stool group) and animals or animal products (n⫽30; animal group). Forty-three of these isolates were previously studied by multilocus se-quence typing and/or pulsed-field gel electrophoresis and assessed for the pres-ence of thepurK1allele (a marker for CC17), resulting in 26 being identified as belonging to CC17 and 17 being identified as non-CC17 isolates (23).

All of theE. faeciumisolates, initially identified to the species level by bio-chemical tests, were confirmed by high-stringency colony hybridization using intragenicefafmandaac(6)-Iiprobes, as previously described (9, 29). TheE.

faeciumisolates were grown on brain heart infusion (Difco Laboratories, Detroit, MI) broth/agar and incubated overnight at 37°C.

MSCRAMM- and/or pilus-encoding DNA probes.Genomic DNA of the se-quencedE. faeciumstrain TX0016 was extracted using a DNeasy blood and tissue kit (Qiagen Inc., Valencia, CA); this DNA was used for PCR amplification of intragenic regions of the 14 MSCRAMM- and/or pilus-encoding genes, using primers listed elsewhere (28). Radiolabeled DNA probes were prepared from the purified PCR products with the RadPrime DNA labeling system (Invitrogen, Carlsbad, CA) according to the protocol supplied by the manufacturer. Unin-corporated labeled nucleotides were removed using Micro Bio-Spin P-30 col-umns (Bio-Rad, Hercules, CA).

Colony hybridization.Colony lysate membrane preparation and hybridization conditions for enterococci were described in our previous studies (5, 29). In brief, E. faeciumisolates were inoculated onto sterile nylon membranes (GE Health-care, Piscataway, NJ) placed on brain heart infusion agar plates. The sequenced strainE. faeciumTX0016 was included as a positive control. After overnight growth at 37°C, colonies were lysed on the membrane after treatment with

lysozyme (10 mg/ml) and mutanolysin (4 units/ml). Genomic DNA on the mem-brane was then denatured, washed, and fixed. Hybridizations were carried out under high-stringency conditions (hybridization buffer containing 50% formamide, 5⫻Denhardt’s solution, 5⫻SSC [1⫻SSC is 0.15 M NaCl plus 0.015 M sodium citrate, pH 7.2], 0.1% sodium dodecyl sulfate [SDS], and 100␮g/ml calf thymus DNA) at 42°C overnight, followed by three washes with 2⫻SSC and 0.1% SDS at room temperature (15 min each) and two washes with 0.1⫻SSC and 0.1% SDS at 50°C (15 min each).

Statistical analysis.The distributions of the 14 MSCRAMM- and/or pilus-encoding genes inE. faeciumisolates originating from different clinical or non-clinical groups/subgroups were compared using two-tailed Fisher’s exact test.P values of⬍0.05 were considered statistically significant.

RESULTS

Overall distribution of predicted MSCRAMM- and/or pilus-encoding genes in E. faecium isolates. We and others have shown that acm is present in ⬎99% of isolates of diverse origins (3, 23). The occurrence of the remaining 14 MSCRAMM- and/or pilus-encoding genes among 264 human clinical isolates, 69 stool isolates from hospitalized patients, 70 stool isolates from community volunteers, and 30 animal-de-rived isolates, collected over a period of 33 years from four continents, is summarized in Table 1. Among all isolates, the percentages of isolates showing hybridization to the individual gene probes ranged from 51% to 98%, but there was substan-tial variation between different groups, ranging from 67%

(fms18) to 99% (fms13) in the clinical group, from 38%

(fms18) to 99% (fms13) in the hospitalized patient-derived

stool isolate group, from 7% (fms18) to 94% (fms17) in the community volunteer-derived stool isolate group, and from 33% (fms15) to 100% (fms13) in the animal group. The highest frequency of these genes was detected in the endocarditis sub-group, with 11 of the genes being carried by 100% of the isolates. The remaining three genes (fms18,fms20, andfms21) were found in 53%, 80%, and 93% of endocarditis isolates, respectively (Table 1).

When total numbers of genes were compared between isolates, there were 3,372 genes present of 3,696 possible (91%) among the 264 clinical isolates, 709 genes present of 966 possible (73%) among the 69 stool isolates from hospitalized patients, 597 genes present of 980 possible (61%) among the 70 stool isolates from community volunteers, and 274 genes present of 420 possible (65%) among the 30 isolates from animal or animal products. Pairwise comparisons of these results showed a highly significant difference between the clinical group and each of the three non-clinical groups (Pⱕ0.0001; Fisher’s exact test).

Distributions of different combinations of the 14 genes.As shown in Fig. 1, larger numbers of the MSCRAMM- and/or pilus-encoding genes were found in clinical than in nonclinical isolates. As anticipated, stool isolates from hospitalized pa-tients were found to carry a smaller number of the 14 genes than isolates from the clinical group but a larger number than isolates from the two nonclinical groups; this was likely due to fecal samples from hospitalized patients consisting of a mixture of both nosocomially derived and community-derived organ-isms.

Further examination of the hybridization profiles showed that the presence/absence of the 14fmsgenes resulted in 82 observed gene combinations among the 433 isolates. The three most common combinations accounted for 57% of all isolates. More specifically, over one-half of the isolates either had all 14

on May 16, 2020 by guest

http://jcm.asm.org/

(3)

genes present (28%) or were missing just fms18 (15%) or

fms20(14%). The remaining combinations were more evenly

distributed among different isolates. The clinical subgroups contained most of the isolates carrying all or 13 of the genes

(81% of clinical isolates), whereas such isolates were not found in the community stool group and were relatively rarely found (13%) in the animal group (Table 2). Furthermore, over two-thirds (70%) of the community stool isolates carried only 8 to 11 of the 14 genes studied.

Comparison of the distribution of each MSCRAMM- and/or pilus-encoding gene in infection-derived E. faecium isolates versus that in community isolates or animal isolates. (i) Dis-tribution of the 11 MSCRAMM- and/or pilus-encoding genes clustered at four different genomic loci of TX0016. (a)ebpfm

[image:3.585.49.541.81.324.2]

operon genes.The ebpAfm, ebpBfm, and ebpCfm (pilB) genes TABLE 1. Distribution of 14 predicted MSCRAMM- and/or pilus-encoding genes among 433 temporally and geographically diverseE. faeciumisolates

Gene

% of isolates with gene

Endocarditis (n⫽15)

Blood (n⫽57)

OCIa (n⫽192)

Total clinical (n⫽264)

Stool (hospital) (n⫽69)

Stool (community) (n⫽70)

Animal (n⫽30)

Total (n⫽433)

scm 100 96 97 97 80 46b 93 86

fms15 100 100 97 98 78 29b 33b 79

fms18 53 77 65 67 38 7b 37c 51

ebpoperon genes

ebpAfm 100 96 96 96 72 73

b 77c 87

ebpBfm 100 96 96 96 72 73b 77c 87

ebpCfm(pilB) 100 96 96 96 72 73

b 77c 87

fms11-fms19-fms16cluster

fms11 100 96 90 92 68 43b 47b 77

fms19 100 96 90 92 68 43b 47b 77

fms16 100 96 90 92 68 43b 47b 77

fms14-fms17-fms13clustere

fms14 100 96 96 97 77 87c 73b 90

fms17 100 96 96 97 86 94 77c 93

fms13 100 100 98 99 99 93c 100 98

fms21-fms20clusterd

fms21(pilA) 93 95 89 91 88 79c 80 88

fms20 80 79 64 68 61 71 50 66

a

Other clinical isolates from various tissues/infections (see Materials and Methods). b

P⬍0.05 against each of the three clinical subgroups (endocarditis, blood, and OCI groups). c

P⬍0.05 against the combined clinical group (total). d

Genes in these clusters did not coexist in some of the isolates.

[image:3.585.300.540.546.723.2]

FIG. 1. Percentages of isolates in different clinical and nonclinical groups that were positive for one or more of the 14fmsgenes.

TABLE 2. Distribution ofE. faeciumisolates based on the number of genes present

No. of genes present

% of isolates

Clinical (n⫽264)

Stool (hospital)

(n⫽69)

Stool (community)

(n⫽70)

Animal (n⫽30)

Total (n⫽433)

14 40 25 0 0 28

13 41 22 0 13 30

12 6 0 6 20 6

11 3 7 17 7 6

10 3 9 16 7 6

9 1 3 10 7 3

8 2 13 27 13 9

7 2 3 4 10 3

6 2 1 9 13 4

5 0.4 6 7 3 3

4 0 10 1 7 2

3 0 1 3 0 1

2 0 0 0 0 0

1 0 0 0 0 0

on May 16, 2020 by guest

http://jcm.asm.org/

(4)

were found to coexist, consistent with our previous demonstra-tion of transcripdemonstra-tional expression of this gene cluster as a single operon (Table 1) (28). TheebpABCfmoperon was found in all

but a few clinical isolates (96 to 100%) but was less frequently (73 to 77%) found in nonclinical isolates (community and animal isolates). There was a significant difference for each of the three clinical subgroups versus the community fecal group (P⫽0.019 to⬍0.0001) and for the blood and OCI subgroups versus animal group isolates (P⫽0.0070 and 0.0012, respec-tively).

(b) fms11-fms19-fms16 cluster. Similar to the ebpABCfm

operon and in agreement with our earlier genome analysis predicting the organization of thefms11-fms19-fms16cluster as an operon, all three genes were carried by each of the isolates that was positive for any one of the genes (Table 1). These three genes were present in most isolates of the three clinical subgroups (ranging from 90 to 100%) but were less common among community (43%) and animal (47%) isolates, present-ing a highly significant statistical difference when each of the clinical subgroups was compared against either the community or animal isolates (P⫽0.0002 to 0.0001).

(c)fms14-fms17-fms13cluster.Although the majority of

iso-lates in all groups carried thefms14-fms17-fms13cluster (Table 1) (96% of clinical isolates, 83% of community stool isolates, and 73% of animal isolates), there was still a significant differ-ence between two of the clinical subgroups (blood isolates and OCI) and the community stool or animal group (P⫽0.0002 to 0.038). However, although analysis of the TX0016 genome predicted organization of this cluster as a pilus-encoding operon, 10% (45/433 isolates) of isolates did not hybridize to one or two of the gene probes of this cluster while hybridizing to the other probe(s). Most (35 [78%]) of these discordant isolates were derived from nonclinical sources.

(d)fms21-fms20cluster.Two other genes were also found in

very close proximity to each other in the genome of TX0016, namely,fms20and a major pilus subunit-encoding gene,fms21

(pilA) (12, 13, 28). However, they are separated by a putative class C sortase gene and another open reading frame that lacks a cell wall anchoring motif. The majority of all clinical isolates (91%) carriedfms21, whilefms20was less common thanfms21

and the second least common gene overall (afterfms18);fms20

andfms21genes coexisted in only 66% of the 433 isolates.

(ii) Distribution of the three fmsgenes not clustered with other MSCRAMM- and/or pilus-encoding genes of TX0016.

As shown in Table 1, the nonclustered MSCRAMM-encoding genes of the sequenced TX0016 genome (scm, fms15, and

fms18) showed various distributions in different groups, albeit

with a considerably higher frequency of each gene in each of the three clinical subgroups than in the community-derived stool isolate group. No significant variation between the dif-ferent clinical subgroups (endocarditis isolates, blood isolates, and OCI) was observed (Table 1). It is notable that carriage of

fms18was less frequent in the different groups than that of all

other genes. Although there was a clear-cut association be-tween the presence ofscmand a clinical origin (97% of clinical isolates versus 46% of community-derived isolates; P

0.0001), thescmgene was also overrepresented in isolates of animal origin (93%). In contrast,fms15and fms18were de-tected in a minority of animal isolates, and the difference between animal isolates and each of the three clinical

sub-groups was highly significant forfms15(P⬍0.0001), as was the difference between animal isolates and both the blood and OCI subgroups forfms18(P⫽0.0004 to 0.004).

Presence of the 14 genes in CC17.We previously analyzed 43 isolates studied here for their ancestral relationships and iden-tified 26 of them as CC17 isolates and 17 as non-CC17 isolates (23). Among these, 11 of the 14 genes were present in 100% of the CC17 isolates, while the remaining three genes, fms18,

fms21, andfms20, were present in 58%, 92%, and 81% of the

CC17 isolates, respectively (Table 3). A widely variable but generally less frequent occurrence was detected in non-CC17 isolates, ranging from 29% to 100%. Overall, 9 of the 14 genes showed a significantly higher carriage in CC17 isolates than in non-CC17 isolates, while no statistically significant enrichment among CC17 isolates was observed for five of the genes.

DISCUSSION

The ability of bacteria to infect and cause disease is often dependent on specific genes that distinguish more pathogenic strains from their less pathogenic relatives. In gram-positive organisms, such as streptococci and staphylococci, these virulence-associated determinants include surface-anchored MSCRAMMs (e.g., collagen binding protein Cna of

Staphylococ-cus aureus), surface pili (e.g., fibronectin-collagen-T

antigen-car-rying pili ofStreptococcus pyogenes), secreted proteins or enzymes (e.g., gelatinase ofE. faecalis), genes involved in the synthesis of toxins (e.g., Panton-Valentine leukocidin genes ofS. aureus), and capsular exopolysaccharides (e.g., theStreptococcus pneumoniae

capsule). Recent reports, including one from our laboratory, have assessed the distribution of newly identified or predicted surface protein genes, including MSCRAMM- and/or pilus-encoding genes, in various groups ofE. faeciumisolates, with each

sub-TABLE 3. Comparison of gene distributions of CC17 and non-CC17 isolates

Gene

No. (%) of isolates with gene

CC17 (n⫽26) Non-CC17 (n⫽17)

scm 26 (100)a 11 (65)

fms15 26 (100)a 5 (29)

fms18 15 (58)a 3 (18)

ebpoperon genes

ebpAfm 26 (100)a 13 (76)

ebpBfm 26 (100)

a 13 (76)

ebpCfm(pilB) 26 (100)a 13 (76)

fms11-fms19-fms16cluster

fms11 26 (100)a 5 (29)

fms19 26 (100)a 5 (29)

fms16 26 (100)a 5 (29)

fms14-fms17-fms13cluster

fms14 26 (100) 15 (88)

fms17 26 (100) 16 (94)

fms13 26 (100) 17 (100)

fms21-fms20cluster

fms21(pilA) 24 (92) 15 (88)

fms20 21 (81) 13 (76)

aP0.05 against non-CC17 isolates in our study.

on May 16, 2020 by guest

http://jcm.asm.org/

[image:4.585.300.542.87.324.2]
(5)

group consisting of relatively few isolates that were collected from various human and animal sources (12, 13, 28). Results from these studies showed that eight MSCRAMM- and/or pilus-encod-ing genes (ebpAfm,ebpBfm,ebpCfm,scm,fms13,fms15,fms17, and

fms21) were widely distributed and present in 82 to 94% of

iso-lates; the remaining six genes were found in 26 to 56% isolates, whileacmwas previously shown to be present in⬎99% of isolates (23).

To determine the presence of these 14 MSCRAMM- and/or pilus-encoding genes in variousE. faeciumisolates from clin-ical and nonclinclin-ical environments, we employed an expanded collection of 433 geographically and temporally diverseE. fae-ciumisolates. Analysis of the total number of these 14 genes in each group showed a considerably higher carriage rate (91%) for the clinical group than for the community stool and animal groups (61% and 65%, respectively), indicating an overall ac-cumulation and/or higher level of preservation of these genes in isolates associated with infections in humans. A similar conclusion can be derived from the hybridization data for 131 isolates recently presented by Hendrickx et al. (12, 13) in which an estimated 88% of the total number of possible genes were carried by clinical and hospital outbreak-associated isolates and 56 to 76% of possible genes were carried by isolates from other sources. This is in contrast toacm, which was carried by ⬎99% of all isolates, although the Acm protein was expressed almost exclusively by clinical isolates and was often present as a pseudogene in nonclinical isolates (23).

Independent analysis of the distribution of each of the genes studied here also pointed to a strong association with a clinical source of isolation for three nonclustered genes (scm,fms15,

andfms18) and two of the four pilus-encoding clusters (ebpABCfm

and fms11-19-16). Taken together, these nine genes were

significantly enriched in each of the three clinical subgroups compared to the community stool group. With the exception ofscm, these genes also occurred significantly more often in two or more clinical subgroups than in the animal group. When the clinical subgroups were combined and compared with the community stool and/or animal group, a significant association with clinical origin was also observed for four additional genes

(fms14,fms17,fms13, andfms21). Although statistical analyses

confirmed a significant enrichment of these genes in each of the infection-derived groups (P ⬍ 0.0001 to 0.019), we ac-knowledge that it would be an oversimplification to directly infer function from the presence or absence of a given gene, particularly since some of the genes may be nonfunctional, as evidenced by previous findings ofacm,fms15,fms16, andfms19

as pseudogenes in some isolates (23, 28). Although direct parisons with the results of Hendrickx et al. (12, 13) are com-plicated by the lack of information on the presence of each gene in the different clinical and nonclinical groups, our results generally agree with these reports, which showed similar en-richment of these genes. Subsequent assessment of gene dis-tribution in a subset of our isolates that were previously cate-gorized as either CC17 or non-CC17 isolates showed more enrichment in CC17 isolates than that found by Hendrickx et al. (13), with 11 of the 14 genes present in 100% (26) of CC17 isolates, while the remaining 3 genes were present in 58 to 92% of CC17 isolates. Furthermore, we detected some genes in non-CC17 isolates more frequently, e.g., fms14 and fms20, which were present in 88% and 76% of non-CC17 isolates,

respectively, in our collection, compared to 41% and 42% of non-CC17 isolates, respectively, in the study of Hendrickx et al. (13). On the other hand, two of the genes,scmandfms15, were present at lower rates in non-CC17 isolates (65% and 29%, respectively) in our study than the previously published rates of 92% and 77% for non-CC17 isolates (13), but this may be a result of our small non-CC17 isolate sample size.

As shown in Table 1, some of the isolates (mostly from nonclinical groups) did not hybridize to one or two genes of the

fms14-17-13cluster, suggesting the possibility of a high

recom-bination rate in this genomic region, which could lead to either divergent gene sequences or deletion of the genes altogether. Interestingly, the variation in thefms14-17-13cluster resulted mainly from the absence of fms14and/or fms17 (39/45 iso-lates), while the putative pilus backbone subunit gene,fms13, was missing in only 6 of the 45 divergent isolates. Similarly,

fms20was absent in 25% of isolates positive for the predicted

major pilin gene, fms21, and rarely (0.5% [2/433 isolates]) existed independently. While we were not able to utilize a low-stringency hybridization approach to detect divergent al-leles due to apparent cross-reactivity with other pilus-encoding genes, PCR analyses using multiple intra- and intergenic primer pairs also failed to amplify these gene fragments from the selected isolates tested (data not shown). Similar variant pilus subunit-encoding gene clusters have been reported for other gram-positive organisms (30).

As expected from the individual gene results, larger numbers of genes were found to be carried by clinical isolates than by nonclinical isolates (Fig. 1). Concurrent with this, clinical iso-lates showed less variability in their gene content, with the majority of them (80% [212/264 isolates]) exhibiting only three different gene combinations of the 82 combinations found among all isolates, while the three most common combinations for the community stool and animal groups accounted for only 34% (24/70 isolates) and 27% (8/30 isolates) of the isolates, respectively. Accumulation of these genes in clinical isolates may be a marker of their clonal origin, a potential factor that influences the distribution of these genes as well as enrichment of specific gene combinations. On the other hand, the consid-erable variability seen in nonclinical isolates, e.g., the presence of pilus clusters in isolates with otherwise heterogeneous con-tents of the 14 genes, may have resulted from horizontal gene transfer between isolates that are not closely related evolution-arily.

In summary, our observation of the more frequent occur-rence of many of the MSCRAMM- and pilus-encoding genes in clinical isolates than in fecal isolates from community vol-unteers and, generally, animal isolates suggests that their pres-ervation and/or enrichment may have contributed to the sur-vival and dissemination ofE. faeciumin hospital environments. Our ongoing characterization of these genes aims at revealing their potential role(s) inE. faeciumpathogenesis, such as ad-herence to host cells and tissues.

ACKNOWLEDGMENTS

This work was supported in part by NIH grant R01 AI067861 from the Division of Microbiology and Infectious Diseases, NIAID, to Bar-bara E. Murray.

on May 16, 2020 by guest

http://jcm.asm.org/

(6)

REFERENCES

1.Boyce, J. M., S. M. Opal, J. W. Chow, M. J. Zervos, G. Potter-Bynoe, C. B. Sherman, R. L. Romulo, S. Fortna, and A. A. Medeiros.1994. Outbreak of multidrug-resistantEnterococcus faeciumwith transferablevanBclass van-comycin resistance. J. Clin. Microbiol.32:1148–1153.

2.Bush, L. M., J. Calmon, C. L. Cherney, M. Wendeler, P. Pitsakis, J. Poup-ard, M. E. Levison, and C. C. Johnson.1989. High-level penicillin resistance among isolates of enterococci. Implications for treatment of enterococcal infections. Ann. Intern. Med.110:515–520.

3.Camargo, I. L., M. S. Gilmore, and A. L. Darini.2006. Multilocus sequence typing and analysis of putative virulence factors in vancomycin-resistant and vancomycin-sensitiveEnterococcus faeciumisolates from Brazil. Clin. Micro-biol. Infect.12:1123–1130.

4.Cereda, R. F., A. C. Gales, S. Silbert, R. N. Jones, and H. S. Sader.2002. Molecular typing and antimicrobial susceptibility of vancomycin-resistant Enterococcus faeciumin Brazil. Infect. Control Hosp. Epidemiol.23:19–22. 5.Coque, T. M., J. E. Patterson, J. M. Steckelberg, and B. E. Murray.1995. Incidence of hemolysin, gelatinase, and aggregation substance among enterococci isolated from patients with endocarditis and other infections and from feces of hospitalized and community-based persons. J. Infect. Dis.171:1223–1229.

6.Coque, T. M., J. F. Tomayko, S. C. Ricke, P. C. Okhyusen, and B. E. Murray. 1996. Vancomycin-resistant enterococci from nosocomial, community, and animal sources in the United States. Antimicrob. Agents Chemother.40: 2605–2609.

7.Coudron, P. E., C. G. Mayhall, R. R. Facklam, A. C. Spadora, V. A. Lamb, M. R. Lybrand, and H. P. Dalton.1984.Streptococcus faeciumoutbreak in a neonatal intensive care unit. J. Clin. Microbiol.20:1044–1048.

8.Deshpande, L. M., T. R. Fritsche, G. J. Moet, D. J. Biedenbach, and R. N. Jones.2007. Antimicrobial resistance and molecular epidemiology of van-comycin-resistant enterococci from North America and Europe: a report from the SENTRY antimicrobial surveillance program. Diagn. Microbiol. Infect. Dis.58:163–170.

9.Duh, R. W., K. V. Singh, K. Malathum, and B. E. Murray.2001. In vitro activity of 19 antimicrobial agents against enterococci from healthy subjects and hospitalized patients and use of anacegene probe fromEnterococcus faecalisfor species identification. Microb. Drug Resist.7:39–46.

10.Handwerger, S., B. Raucher, D. Altarac, J. Monka, S. Marchione, K. V. Singh, B. E. Murray, J. Wolff, and B. Walters.1993. Nosocomial outbreak due toEnterococcus faeciumhighly resistant to vancomycin, penicillin, and gentamicin. Clin. Infect. Dis.16:750–755.

11.Harthug, S., A. Digranes, O. Hope, B. E. Kristiansen, A. G. Allum, and N. Langeland.2000. Vancomycin resistance emerging in a clonal outbreak caused by ampicillin-resistantEnterococcus faecium. Clin. Microbiol. Infect. 6:19–28.

12.Hendrickx, A. P., M. J. Bonten, M. van Luit-Asbroek, C. M. Schapendonk, A. H. Kragten, and R. J. Willems.2008. Expression of two distinct types of pili by a hospital-acquiredEnterococcus faeciumisolate. Microbiology154: 3212–3223.

13.Hendrickx, A. P., W. J. van Wamel, G. Posthuma, M. J. Bonten, and R. J. Willems.2007. Five genes encoding surface-exposed LPXTG proteins are enriched in hospital-adaptedEnterococcus faeciumclonal complex 17 iso-lates. J. Bacteriol.189:8321–8332.

14.Iwen, P. C., D. M. Kelly, J. Linder, S. H. Hinrichs, E. A. Dominguez, M. E. Rupp, and K. D. Patil.1997. Change in prevalence and antibiotic resistance ofEnterococcusspecies isolated from blood cultures over an 8-year period. Antimicrob. Agents Chemother.41:494–495.

15.Kim, W. J., R. A. Weinstein, and M. K. Hayden.1999. The changing molec-ular epidemiology and establishment of endemicity of vancomycin resistance in enterococci at one hospital over a 6-year period. J. Infect. Dis.179:163– 171.

16.Leavis, H. L., M. J. Bonten, and R. J. Willems. 2006. Identification of

high-risk enterococcal clonal complexes: global dispersion and antibiotic resistance. Curr. Opin. Microbiol.9:454–460.

17.Lentino, J. R., M. Narita, and V. L. Yu.2008. New antimicrobial agents as therapy for resistant gram-positive cocci. Eur. J. Clin. Microbiol. Infect. Dis. 27:3–15.

18.Miranda, A. G., K. V. Singh, and B. E. Murray.1991. DNA fingerprinting of Enterococcus faeciumby pulsed-field gel electrophoresis may be a useful epidemiologic tool. J. Clin. Microbiol.29:2752–2757.

19.Montecalvo, M. A., H. Horowitz, C. Gedris, C. Carbonaro, F. C. Tenover, A. Issah, P. Cook, and G. P. Wormser.1994. Outbreak of vancomycin-, ampi-cillin-, and aminoglycoside-resistantEnterococcus faeciumbacteremia in an adult oncology unit. Antimicrob. Agents Chemother.38:1363–1367. 20.Murray, B. E.2000. Vancomycin-resistant enterococcal infections. N. Engl.

J. Med.342:710–721.

21.Murray, B. E., and G. M. Weinstock.1999. Enterococci: new aspects of an old organism. Proc. Assoc. Am. Physicians111:328–334.

22.Nallapareddy, S. R., K. V. Singh, and B. E. Murray.2008. Contribution of the collagen adhesin Acm to pathogenesis ofEnterococcus faeciumin exper-imental endocarditis. Infect. Immun.76:4120–4128.

23.Nallapareddy, S. R., K. V. Singh, P. C. Okhuysen, and B. E. Murray.2008. A functional collagen adhesin gene,acm, in clinical isolates ofEnterococcus faeciumcorrelates with the recent success of this emerging nosocomial pathogen. Infect. Immun.76:4110–4119.

24.Nallapareddy, S. R., G. M. Weinstock, and B. E. Murray.2003. Clinical isolates ofEnterococcus faeciumexhibit strain-specific collagen binding me-diated by Acm, a new member of the MSCRAMM family. Mol. Microbiol. 47:1733–1747.

25.National Nosocomial Infections Surveillance System.2004. National Noso-comial Infections Surveillance (NNIS) System report, data summary from January 1992 through June 2004. Am. J. Infect. Control32:470–485. 26.Rice, L. B., L. Carias, S. Rudin, C. Vael, H. Goossens, C. Konstabel, I. Klare,

S. R. Nallapareddy, W. Huang, and B. E. Murray.2003. A potential viru-lence gene,hylEfm, predominates inEnterococcus faeciumof clinical origin.

J. Infect. Dis.187:508–512.

27.Sapico, F. L., H. N. Canawati, V. J. Ginunas, D. S. Gilmore, J. Z. Mont-gomerie, W. J. Tuddenham, and R. R. Facklam.1989. Enterococci highly resistant to penicillin and ampicillin: an emerging clinical problem? J. Clin. Microbiol.27:2091–2095.

28.Sillanpaa, J., S. R. Nallapareddy, V. P. Prakash, X. Qin, M. Hook, G. M. Weinstock, and B. E. Murray.2008. Identification and phenotypic charac-terization of a second collagen adhesin, Scm, and genome-based identifica-tion and analysis of 13 other predicted MSCRAMMs, including four distinct pilus loci, inEnterococcus faecium. Microbiology154:3199–3211. 29.Singh, K. V., T. M. Coque, G. M. Weinstock, and B. E. Murray.1998. In vivo

testing of anEnterococcus faecalis efaAmutant and use ofefaAhomologs for species identification. FEMS Immunol. Med. Microbiol.21:323–331. 30.Telford, J. L., M. A. Barocchi, I. Margarit, R. Rappuoli, and G. Grandi.

2006. Pili in gram-positive pathogens. Nat. Rev. Microbiol.4:509–519. 31.Thal, L., S. Donabedian, B. Robinson-Dunn, J. W. Chow, L. Dembry, D. B.

Clewell, D. Alshab, and M. J. Zervos.1998. Molecular analysis of glycopep-tide-resistantEnterococcus faeciumisolates collected from Michigan hospi-tals over a 6-year period. J. Clin. Microbiol.36:3303–3308.

32.Top, J., R. Willems, S. van der Velden, M. Asbroek, and M. Bonten.2008. Emergence of clonal complex 17Enterococcus faeciumin The Netherlands. J. Clin. Microbiol.46:214–219.

33.Treitman, A. N., P. R. Yarnold, J. Warren, and G. A. Noskin.2005. Emerging incidence ofEnterococcus faeciumamong hospital isolates (1993 to 2002). J. Clin. Microbiol.43:462–463.

34.Willems, R. J., J. Top, M. van Santen, D. A. Robinson, T. M. Coque, F. Baquero, H. Grundmann, and M. J. Bonten.2005. Global spread of vanco-mycin-resistantEnterococcus faeciumfrom distinct nosocomial genetic com-plex. Emerg. Infect. Dis.11:821–828.

on May 16, 2020 by guest

http://jcm.asm.org/

Figure

TABLE 1. Distribution of 14 predicted MSCRAMM- and/or pilus-encoding genes among 433 temporally and geographically diverse E
TABLE 3. Comparison of gene distributions of CC17 andnon-CC17 isolates

References

Related documents

The results depicted that some of socio economic variables: Gender and Religious Orientation affect emotional intelligence but Age, Educational Qualification,

Longest Approximate Time to end (LATE) scheduler [4] is mainly focusing on speculative execution of tasks. In speculative execution, when a task execution is

Grammatical aspect, event components, and their relationship with the temporal and causal structure of events.. An example of trial images for the sentence “The man was

Root mean square pricing errors (RMPSE) are reported. momentum factors) capture some orthogonal source of priced risk. The first-stage FM regression shows that the factor loadings

We illustrate the importance of the proposed distribution over the McDonald Weibull, beta Weibull, modified Weibull, transmuted Weibull and exponentiated Weibull distributions by

Vehicle ad hoc network (VANET) nodes are self organized and have self maintained information without any centralized infrastructure authority or any centralized data of servers[3]

If these drugs must be used together in childhood, careful monitoring of plasma levels of carbamazepine is essential, to avoid the syndrome of toxicity which is now recognized and

This study was carried out to describe the thyroid volume in Healthy Sudanese’s subjects aged 6-18years by ultrasound and to correlate this volume with