Repeated Recovery of Staphylococcus saprophyticus From the Urogenital Tracts of Women: Persistence Vs. Recurrence

Objective: The purpose of this study was to determine whether colonization was persistent or recurrent in a small group of women who had repeated recovery of Staphylococcus saprophyticus from their urogenital tracts. Methods: Paired isolates of S. saprophyticus from each of the study subjects were genotypically typed by plasmid fingerprinting and comparison of chromosomal-DNA restriction fragment-length polymorphism patterns by field-inversion gel electrophoresis (FIGE) and contour-clamped homogenous electric-field (CHEF) electrophoresis. Results: All isolates of S. saprophyticus from the study subjects were classified as genetically unique by each of the typing methods. Conclusions: The subjects experienced recurrent colonization with different isolates of S. saprophyticus. These findings may have broader implications regarding the pathogenesis and recurrence of S. saprophyticus urinary-tract infection.

(UTI) in young women. [1][2][3] Investigators in the United States and Europe have demonstrated that S. saprophyticus causes up to 42% of UTIs in this population. [1][2][3] Infection with S. saprophyticus frequently involves the upper urinary tract and recurrence is not unusual. 1-A UTI with gram-negative enteric bacilli is often preceded by periurethral colonization from a fecal reservoir. In addition, it has been shown that women who experience repeated UTIs due to gram-negative organisms gen-erally have recurrent infections with different strains of bacteria, rather than recurrent infections with the same strain. 4 In contrast, the pathogenesis of S. saprophyticus UTI is less clear. Some investigators have been unable to recover S. saprophyticus from mucosal sites, whereas others have stated that colonization of the periurethral membranes correlates well with infection. The persistence or recurrence of colonization has not been studied. In a previous study, we demonstrated that approximately 7% of asymptomatic women were colonized by S. saprophyticus, s Repeated recovery of S. saprophyti-cus was observed in a smaller number of subjects. This study was undertaken to ascertain whether these women experienced persistent colonization of the urogenital and gastrointestinal tracts with the same strain of bacteria vs. recurrent colonization with different strains of S. saprophyticus.

Patients
In a previously published study designed to identify women colonized with S. saprophyticus, 19 of the 257 women studied were found to harbor S. saprophyticus in their urogenital tracts, s The subjects were identified prospectively from a population of women presenting for routine gynecologic care in an office-based gynecology practice. Of the 19 colonized women, 4 were recultured up to 4 times during the 12-month study. Three of these subjects had S. saprophyticus recovered on at least 2 follow-up visits and make up the study population for this report.

Bacteria
Bacterial isolates were Gram-stained and tested for the production of catalase and coagulase. Coagulase-negative staphylococci were identified to species level, as described by Kloos and Schleifer. 6 Briefly, coagulase-negative staphylococci resistant to 5 Ig of novobiocin/ml were classified as S. saprophyticus if they were urease positive; produced acid aerobically from maltose, sucrose, and trehalose; and did not produce acid from xylose.
Preparation of Plasmid DNA Plasmid DNA was purified by the modified use of a commercial kit (Magic Miniprep, Promega Corp., Madison, WI). Briefly, the isolates of S. saprophyticus were incubated overnight on trypticase soy agar (Bectin Dickinson, Baltimore, MD) plates at 37C. The bacteria were harvested and suspended in 300 I1 of TES (50 I-tM Tris, 5 IM EDTA, 50 mM NaC1) buffer. Fifty microliters of lysostaphin (Sigma Chemical Co., St. Louis, MO) solution (1 mg/ml) was added and the bacteria/ lysostaphin solution was incubated overnight at 37C without agitation to achieve lysis of the cells.
Then, 300 I1 of lysis solution (0.2 M NaOH and 1% sodium dodecyl sulfate) was added and 300 I1 of neutralization solution (2.55 M potassium acetate, pH 4.8) was added. The tubes were inverted several times and the cellular debris was removed by centrifugation. The supernatant was treated with DNA purification resin (Promega Corp.) and the resin/DNA mixture was injected into the minicolumn according to the manufacturer's instructions. The column was washed with buffer containing 200 mM of NaC1, 20 mM of Tris-HCL, and 5 mM of EDTA. Next, the minicolumn was centrifuged in a microcentrifuge tube to dry the resin, and the DNA was eluted from the minicolumn with 50 I*1 of TE (10 IM Tris-HC1 (pH 7.5), mM EDTA) buffer. The plasmid preparations were analyzed by electrophoresis on 0.7% agarose gels.
Preparation of Chromosomal DNA The chromosomal DNA was prepared by a modification of the procedure reported by Goering and Winters. 7 Briefly, S. saprophyticus isolates were grown overnight at 37C with shaking in 5-ml volumes of trypticase soy broth (Bectin Dickinson). A 1.0-ml aliquot of cells was transferred to a microcentrifuge tube, harvested by centrifugation, washed with 1 subjects reported a previous UTI, but none had experienced a symptomatic UTI in the preceding year. The paired S. saprophyticus isolates were recovered an average of 84 days apart. The rectum was colonized in all of the 3 repeatedly culturepostive women. In addition, subject had S. saprophyticus recovered from both the urine and rectum on one of the visits. None of the women had urinary-tract symptoms, experienced a UTI during the 12-month period of observation, or received antibiotics.

Plasmid DNA Analysis
The plasmid pattern observed for each of the pairs of isolates is shown in Figure 1. Each of the isolates exhibited a distinct plasmid pattern. The number of plasmid bands ranged from to 7.

Chromosomal DNA Analysis
The chromosomal DNA RFLP analysis is demonstrated by CHEF in Figure 2. Each isolate appeared to have a distinct pattern with a number of dissimilar bands between paired isolates. FIGE (not shown) also revealed distinct differences between the paired isolates. repeated recovery of S. saprophyticus in 3 of the 4 colonized subjects in whom it was sought suggests that it is a relatively common phenomenon in the small subset of women (approximately 7%) colonized by S. saprophyticus. It also suggests that this subset of women may have some predisposing condition for colonization that, to date, remains undefined.

DISCUSSION
Of additional interest is our use of molecular typing methods to examine questions of pathogenesis and epidemiology. This report is the first to use molecular typing methods to answer questions regarding the pathogenesis of S. saprophyticus colonization. Through use of these techniques, we were able to conclusively demonstrate the diversity of the strains and prove recurrent colonization with different strains of bacteria, which is particularly important in cases of recurrent infection for which treatment failure is considered and recommendations for longer courses of therapy are entertained. From these data, we made the following conclusions: 1) molecular typing techniques are valuable tools in the analysis of the epidemiology and pathogenesis of S. saprophyticus colonization and infection; 2) S. saprophyticus can be repeatedly recovered from the urogenital/gastrointestinal tracts of a small proportion of asymptomatic women; and 3) although a sampling error cannot be excluded, these data indicate that these subjects experienced recurrent colonization with different strains of S. saprophyticus rather than persistent colonization with the same strain. In this regard, it is interesting to note that investigators in Sweden have recovered S.
saprophyticus from a variety of foods, which may serve as the route by which the gastrointestinal tracts of these women are recurrently colonized. There are some limitations of this study. First, a sampling error cannot be excluded, as only single colonies were chosen for storage and further evaluation in the first study. It is possible, therefore, that the subjects were colonized with several strains of S. saprophyticus and, due to our method of selecting only single colonies for evaluation, we have introduced a bias. Second, none of the women developed a symptomatic UTI. Therefore, we may be inaccurately extrapolating from colonization to infection.
R UPP ET AL.
Third, it is difficult to draw any final conclusions from a small group of patients.