Involvement of Adherence and Adhesion Staphylococcus epidermidis Genes in Pacemaker Lead-Associated Infections

Abstract

We explored three genes of attachment (fbe and atlE) and adhesion (ica) in 27 and 10 Staphylococcus epidermidis strains involved in pacemaker-related infections (PMI) and intravascular-catheter-related infections (IVCI), respectively, and in 25 saprophytic strains. The detection rates of fbe and atlE were identical in PMI and IVCI strains, but ica detection rates were identical in PMI and saprophytic strains.

Staphylococcus epidermidis is the most frequent causal microorganism isolated in infections of implanted medical devices, particularly in intravascular-catheter infections (IVCI) and in pacemaker-related infections (PMI). Adherence to polymer surfaces and their capacity for biofilm formation contribute to the pathogenesis of infections of implanted medical devices. Biofilm formation may be divided into two phases. First, the primary attachment involves multiple physicochemical, protein, and polysaccharide factors. Genetic analysis has revealed that the atlE gene (6) and the fbe gene (9) are involved in attachment. In the second phase, the attached bacteria proliferate and accumulate in the biofilm. Production of a polysaccharide intercellular adhesin is essential for biofilm accumulation (10). Polysaccharide intercellular adhesin is encoded by the ica operon (7). The importance of these genes in PMI is yet unknown. The aim of the present study was to compare S. epidermidis strains obtained from patients with PMI and IVCI with saprophytic strains. For most strains, genes mediating biofilm formation, such as fbe, atlE, and ica, were detected.

Twenty-seven S. epidermidis strains were isolated from the leads of 24 patients referred for chronic PMI according to the modified Duke criteria (8). The time interval between the last procedure at the pacemaker site and referral was 27.4 ± 29.1 months (mean ± standard deviation). Three portions of each pacemaker lead were placed in sterile phosphate-buffered saline and vortexed for 1 min, and a 0.1-ml aliquot of the suspension was plated onto sheep blood agar and chocolate agar under O₂ and CO₂ atmospheres, respectively. To avoid contaminant strains, we excluded patients with infections with species other than the S. epidermidis isolates and limited the analysis to S. epidermidis isolates found on all the lead segments. Ten strains from 10 patients with IVCI were obtained from cultures of catheters performed by rinsing the distal segment of the catheter with 1 ml of broth and inoculating 100 μl of the broth onto blood agar. The cultures were considered to be significant when the bacterial count was ≥10³ CFU/ml. As a negative control, 25 saprophytic strains were isolated from the hands of healthy volunteers who did not attend the hospital. Species identification was performed with an automated system (API Expression System; bioMérieux, Marcy l'Etoile, France) according to the instructions of the manufacturer.

A standard protocol for DNA extraction was used as described previously (11). Four primer pairs were designed to amplify fragments from the fbe, atlE, and icaADB genes (Table 1). PCR detection of the icaADB cluster was performed by amplification of a DNA region partially covering the icaA, icaD, and icaB loci. Amplification was performed with a Perkin-Elmer model 2400 thermal cycler (Applied Biosystems, Les Ulis, France). One pair of primers and one probe were designed to amplify fragments from the icaB gene for real-time PCR (TaqMan). The ica sequence used was taken from the GenBank database of the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov). The primers and probe specific for icaB were derived from the icaADBC gene and designed by using Primer Express software version 2.0 (Applied Biosystems, Foster City, Calif.). An artificial mimic was engineered by replacing the TaqMan probe site with a larger irrelevant DNA fragment, which allowed discrimination of the icaB gene by using two-color TaqMan probe reporters. When the mimic was added to the samples, successful amplification of this mimic demonstrated the absence of substances inhibitory to PCR, thereby validating negative results. Amplification and detection were performed with an ABI Prism 7700 sequence detection system (Applied Biosystems). Post-PCR data analysis was performed with Sequence Detector software (Applied Biosystems).

TABLE 1.

DNA sequences of amplification primers

Gene	Primera	DNA sequence (5′-3′)	Product size (bp)	Reference
fbe	fbe-F	306-TAAACACCGACGATAATAACCAAA-329	495	9
	fbe-R	778-GGTCTAGCCTTATTTTCATATTCA-801
atlE	atlE-F	3497-CAACTGCTCAACCGAGAACA-3516	682	4, 6
	atlE-R	4158-TTTGTAGATGTTGTGCCCCA-4178
icaADB	icaADB-F	1893-TTATCAATGCCGCAGTTGTC-1913	546	4, 7
	icaADB-R	2388-GTTTAACGCGAGTGCGCTAT-2408
icaB	icaB-F	3040-CGAATCCGTCCCATTCCTTT-3059	81
	icaB-R	3094-ATGCCGATAACTATAGAATTCCACGTA-3121
	icaB-P	FAM-TAGCGTTTCAAATGCATCAT

^aF, forward primer; R, reverse primer; P, probe.

Data were analyzed by using Student's unpaired t tests, chi-square and Fisher's exact tests, or analysis of variance. A P value of ≤0.05 was considered significant.

Detection rates for the fbe, atlE, and icaADB genes by PCR are reported in Table 2. For the healthy volunteers, the attachment genes were found in almost 50% of the strains studied. In contrast to the results for healthy volunteers, the attachment genes were detected in all or almost all of the S. epidermidis strains from patients with PMI or IVCI (P < 0.0032). The detection rates of the icaADB gene among the three categories of strains (PMI related, IVCI related, and saprophytic) showed great differences and were significantly higher in strains from patients with IVCI (P < 0.0002). TaqMan PCR analysis using a probe for the icaB locus confirmed the results of PCR detection of icaADB. Only two positive strains (one PMI-related and one saprophytic strain) for icaADB were found negative by TaqMan for icaB. False negatives due to substances that were inhibitory to PCR were excluded. Strains isolated from patients with different clinical presentations, such as fever (15 patients) and vegetations (17 patients), had the same PCR profiles.

TABLE 2.

PCR results for the three groups of S. epidermidis strains

Gene	PMI strains (n = 27)a	P for PMI vs IVCI P	IVCI strains (n = 10)a	P for IVCI vs Sb	Saprophytic strains (n = 25)a	P for PMI vs S	Overall P
fbe	24 (88.9)	NSc	10 (100)	0.014	14 (56)	0.011	0.0032
atlE	24 (88.9)	NS	10 (100)	0.007	13 (52)	0.0053	0.0012
icaADB	7 (25.9)	0.0007	9 (90)	0.0002	5 (20)	NS	0.0002

^aResults are given in numbers (percentages) of strains with the indicated gene. n, total number of strains.

^bS, saprophytic.

^cNS, not significant.

In the present study, the attachment genes were almost always present in strains recovered from patients with PMI and IVCI. The percentage of atlE-positive S. epidermidis strains was comparable to that reported by Frebourg et al. (4). There was no agreement between authors of earlier reports for a potent role of the ica locus in the pathogenic mechanisms. Ziebuhr et al. (12) reported the presence of the ica locus in 85% of S. epidermidis blood culture isolates compared to 6% of saprophytic isolates. Galdbart et al. (5) found ica in 44 of 45 strains from patients with prosthetic-material-related joint infections and 2 of 23 strains from eight healthy patients. The ica locus prevalence was only 49% for S. epidermidis strains studied by Arciola et al. (1). Thus, all these authors concluded that targeting the ica locus was useful as a diagnostic marker to distinguish between invasive and contaminating isolates of S. epidermidis (1, 2, 4, 5). However, these results disagree with those published by de Silva et al. (3),who found no significant differences in ica operons between pathogenic and nonpathogenic isolates. Such discrepancies between studies might be explained by the choice of strains tested that had been recovered from patients with IVCI (1, 4) or obtained from stock collections or from blood culture isolates (3-5, 12).

Because ica is found in a relatively low proportion of PMI isolates, we must wonder about false-negative results with PMI strains. We have worked directly with extracted DNA, and false-negatives due to substances that are inhibitory to PCR have been excluded. Two different PCR techniques have been used, with different primers giving gene products of different sizes from different parts of the icaADBC gene, and the results with these two PCRs were comparable. The frequency of false negatives due to slight differences in gene sequence was probably low.

Recent studies (4, 12) have proposed that ica locus detection to discriminate between contaminating and infecting S. epidermidis strains might be useful and should be added to clinical criteria used for the diagnosis of IVCI. The adhesion ability created by the ica locus is not an advantage for diagnosis of PMI. The two genes of attachment were more interesting, but their presence in 50% of the saprophytic strains probably limited their clinical value. One limitation of our study might be the potential contamination of the few strains analyzed and, to a lesser extent, by the small number of isolates recovered.

The data reported here indicate an important role of the genes of attachment in the pathogenic mechanisms of PMI. We have no evidence for a role of the ica locus. Detection of the ica locus by PCR is not useful in discriminating between invasive and contaminating S. epidermidis strains when PMI is suspected.