TEXT

Top Abstract Text References

Staphylococcus epidermidis is commonly cultured from the conjunctiva and lid margins of normal subjects (4, 13). Because of its ubiquitous nature and relatively low virulence, S. epidermidis has received so far little attention for its role in ocular infections. However, in different studies S. epidermidis has been reported to play a significant role in several ocular external diseases such as chronic blepharitis and suppurative keratitis (9, 10, 12). It is also important to characterize genetically the strains isolated in order to determine the clonality and spread of particularly virulent strains (4, 6, 20). Recently, several reports describing the use of insertion sequences (IS) and pulsed-field gel electrophoresis for genetic typing and comparison of staphylococci and enterococci have appeared (2, 3, 6, 15-17). A 1.8-kb IS-like element from Enterococcus hirae (GenBank accession no. U22540) showed high-level homology (65%) to IS found in S. epidermidis (3, 15). In this study, we analyzed 19 strains of S. epidermidis isolated from patients with eyelid, conjunctival, or corneal infection over a period of 12 months. Antibiotic susceptibility testing and chromosomal DNA hybridization with a 1.1-kb internal IS fragment from E. hirae were performed in an attempt to correlate different kinds of external ocular disorders caused by S. epidermidis.

A retrospective review of all culture-positive cases of chronic blepharitis, acute purulent conjunctivitis, and suppurative keratitis examined at the Institute of Ophthalmology, University of Sassari, Italy, between June 1997 and June 1998 was performed. Cases in which S. epidermidis was the sole isolate were selected. Nineteen strains of S. epidermidis from patients with chronic blepharitis (10), purulent conjunctivitis (8), and suppurative keratitis (3) were analyzed. Specimens were collected from the lower fornix of the palpebral conjunctiva, lid margins, and corneal ulcers by using dacron swabs. Cultures for the detection of bacteria, fungi, and Chlamydia trachomatis were performed according to a previously recommended protocol (21). S. epidermidis isolates were identified by standard biochemical methods (21). Antibiotic susceptibility was determined by agar disk diffusion (Kirby-Bauer method). The following antibiotics were tested: penicillin, methicillin, gentamicin, tetracycline, erythromycin, ciprofloxacin, vancomycin, and teicoplanin. For teicoplanin and vancomycin, results were confirmed by the standard agar dilution method (11). Briefly, overnight cultures of bacteria were plated at a series of dilutions on tryptic soy agar plates containing antibiotic-free medium or a twofold dilution of teicoplanin within the drug concentration range of 0.1 to 128 µl/ml. Plates were incubated at 37°C for 48 h, and the number of bacterial colonies was counted.

The chromosomal DNA of the 19 isolates of S. epidermidis was extracted as previously reported (2). For digestion, DNA was incubated with HindIII restriction enzyme (Promega, Madison, Wis.) according to the recommendations of the manufacturer. Agarose gel electrophoresis was performed with a horizontal gel apparatus (model HE 99; HSI, San Francisco, Calif.). Samples were loaded into wells in a 0.7% agarose gel (Boehringer, Mannheim, Germany) and electrophoresed at 30 V for 14 to 18 h. Electrophoresis was performed at room temperature in 0.04 M Tris-acetate (Boehringer)-0.001 M EDTA (pH 8.0). Gels were stained with a solution of ethidium bromide (8). For hybridization, a 1,122-bp probe containing most of the IS-like sequence was obtained by partially digesting a 7.8-kb EcoRI fragment with the enzymes MboI and MspI (15). The DNA of S. epidermidis was transferred to supported nitrocellulose (Nitroplus 2000; MSI, Westboro, Mass.) by using a vacuum transfer device (ABN, Emeryville, Calif.), and Southern blots were performed by a modification of the method of Southern (19). Hybridization was performed at 68°C, and the blots were washed at 68°C with 0.1× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate [pH 7.0]) and 0.1% sodium dodecyl sulfate. Probes were labeled with the enhanced chemiluminescent gene labeling kit (Amersham International, Little Chalfont, United Kingdom). Autoradiography was carried out at room temperature with Kodak X-RP film. The 1.1-kb band was eluted from the low-melting-point agarose gel after electrophoresis.

The fingerprints produced by the 19 isolates of S. epidermidis were evaluated with Image Master 1D software (Pharmacia Biotech, Uppsala, Sweden) and compared by using Dendron software (Solltech Inc., Oakdale, Iowa). The similarity value was calculated on the basis of band position alone.

All 19 isolates were positive for chromosomal DNA hybridization with the 1.1-kb IS fragment. Twelve different patterns were obtained. The fingerprints and the dendrogram illustrating the relationship between the 19 S. epidermidis strains are shown in Fig. 1. Isolates 3 to 7, 10 to 12, and 17 and 18 showed identical patterns (SA, SC, and SB, respectively). The same hybridization pattern (SA) was found in three isolates from patients with purulent conjunctivitis. Seven of nine isolates from patients with chronic blepharitis showed different fingerprinting profiles. Antibiotic susceptibility and the profiles of the strains did not change after several cultures and appear to have remained stable over a period of 10 months. The antibiotic susceptibility of the 19 isolates of S. epidermidis is summarized in Table 1. Seventeen (89%) strains were penicillin resistant, four (21%) were methicillin resistant, six (32%) were gentamicin resistant, seven (37%) were erythromycin resistant, two (11%) were ciprofloxacin resistant, one (5%) was vancomycin resistant, and one (5%) was teicoplanin resistant. Susceptibility to tetracycline was tested for 16 isolates; nine (56%) were found to be resistant to this antibiotic. In total, only two strains were sensitive to all antibiotics, whereas seven (36%) were resistant to three or more antibiotics. Isolate 9 was found to be resistant to all antibiotics. Isolates 11 and 12 had the same antibiotic susceptibility (resistance to penicillin and gentamicin), a pair of plasmids with the same molecular size (4.5 and 2.3 kb), and the same hybridization pattern (SC). DNA extraction and gel electrophoresis showed the presence of one or more plasmids in 11 cases (data not shown).

View larger version (17K): [in this window] [in a new window]   FIG. 1.   Dendrogram (left) and computer elaboration of fingerprintings (right) of the 19 isolates of S. epidermidis. Letters correspond to patterns, and numbers correspond to patients. Lambda HindIII-digested marker (MII) was used as a molecular weight marker.

The 19 ocular isolates of S. epidermidis analyzed in this study showed highly variable patterns of antibiotic susceptibility. Most isolates were penicillin and tetracycline resistant, whereas resistance to vancomycin, teicoplanin, and ciprofloxacin was found in only one case. Given the high-level broad-spectrum activity against most bacteria and the reduced frequency of ocular toxic effects, ciprofloxacin is currently considered the drug of choice in the therapy for bacterial keratitis (7). However, there is growing evidence for ciprofloxacin-resistant ocular strains of S. epidermidis (14, 18). Vancomycin and teicoplanin are antistaphylococcal antibiotics, to which resistance is rarely seen, and should be considered the drugs of last resort for the therapy of nosocomial gram-positive infections (5). The multiple antibiotic resistance of S. epidermidis is a recognized problem. It might possibly represent a response to prolonged treatment. In our study, seven (36%) isolates were resistant to three or more antibiotics. The detection of one strain of S. epidermidis resistant to all antibiotics tested is cause for concern. The spread of such strains in hospitals may constitute a threat for immunocompromised patients. Multiresistant strains of S. epidermidis often colonize the skin of hospitalized patients and hospital personnel (1).

Widespread skin colonization serves as a potential reservoir for multiresistant isolates that can cause infections, particularly infections of indwelling intravascular devices. Furthermore, these colonizing isolates serve as a reservoir for antibiotic resistance genes that can transfer among coagulase-negative staphylococci and be acquired by Staphylococcus aureus (3). In our study, a high degree of genetic polymorphism was found among the 19 strains of S. epidermidis analyzed. This data is not surprising, as the situation presented did not represent an outbreak and one expects that unrelated patients would have unrelated strains. Isolates 10, 11, and 12 shared the same antibiotic susceptibility (SC), plasmid profile, and hybridization pattern (SC). Strains 5 and 6 generated identical antibiotic susceptibility profiles and IS patterns. All these strains were isolated from patients of different ages living in Sassari, Italy. It seems that there is a correlation between antibiotic susceptibility and IS clusters, although we have found different strains carrying different antibiotic profiles within the same IS family (e.g., strains 17 and 18 and strains 3 and 4). Strains 9, 16, and 17 are resistant to methicillin, but they have different antibiotic profiles and IS fingerprinting patterns, thus suggesting a completely different origin. These strains were isolated from patients from different towns (Osilo, Bonorva, and Sassari, respectively). Strains 3 and 7, which yielded the same hybridization pattern, SA, were isolated from two surgeons working in the same hospital. It is likely that these patients came into contact with each other. Hybridization pattern SA was found on five occasions; in three cases (5 to 7) it was associated with acute purulent conjunctivitis. A high degree of polymorphism was found among the nine isolates from patients with chronic blepharitis. The correlation between IS fingerprinting patterns and antibiotic susceptibility profiles may reveal a common origin for the strains analyzed, but the possible acquisition of plasmids or transposons carrying antibiotic resistance genes may complicate the epidemiological analysis. In conclusion, in our study DNA hybridization with the 1.1-kb insertion sequence from E. hirae appeared to be a rapid (2 days) and sensitive technique for distinguishing between different strains of S. epidermidis; moreover, it does not need expensive equipment like other techniques such as pulsed-field gel electrophoresis. Since the antibiotic susceptibility of S. epidermidis is unpredictable and multiresistant isolates are common, we recommend that antibiotic susceptibility testing be performed in all cases of ocular infections caused by these organisms.