Vulvovaginal candidiasis (VVC) affects up to 70 to 75% of women once in their life and presents with pruritus and vaginal discharge (13). Bacterial vaginosis (BV) is a common aberrant condition associated with depletion of lactobacilli, elevated vaginal pH, and overgrowth of Atopobium, Mobiluncus, Prevotella, Gardnerella, and Megasphaera species (1, 8, 17).

Hydrogen peroxide (H₂O₂) has been considered a key factor in Lactobacillus antagonism against pathogens. This compound generates cytotoxic reactive oxygen species, superoxide anions, and hydroxyl radicals in the vaginal fluid (6, 9).

In the present study, 64 healthy women (control group), 68 women diagnosed with VVC, and 64 women diagnosed with BV signed an informed consent statement (Ethics Review Board of the Centro de Saúde Escola da Faculdade de Medicina de Ribeirão Preto—Universidade de São Paulo; CSE-FMRP-USP protocol 0146). This study was registered online at “Comissão Nacional de Ética em Pesquisa” (CONEP document 070202), Brazil. Exclusion criteria included immunosuppression, pregnancy, current use of antibiotics or antifungals, menses during sample collection, and diagnosis of trichomoniasis.

The examining physician detected the presence of vaginal discharge, determined vaginal pH (Acilit indicator strip, pH 0 to 6; Merck, Germany), and collected three vaginal samples. Healthy subjects had no vaginal discharge or signs or symptoms of infections and had vaginal Gram-stained smears dominated by lactobacilli. Subjects were diagnosed with VVC by the presence of vaginal discharge and/or vaginal itching and burning plus being positive for Candida by wet mount preparations with 10% potassium hydroxide, Gram staining, or culture. Subjects diagnosed with BV fulfilled the criteria proposed by Amsel et al. (3) and Nugent et al. (11).

Samples were diluted with saline, plated on de Man Rogosa Sharpe (MRS) agar (Oxoid, Basingstoke, United Kingdom), and incubated for 48 h at 37°C aerobically and anaerobically. Colonies with different morphologies that were catalase- and oxidase-negative, gram-positive rods were processed by PCR-ARDRA.

Chromosomal bacterial DNA was obtained from overnight cultures of Lactobacillus sp. grown in 10 ml of MRS broth (Difco Laboratories, Detroit, MI) (10). For DNA extraction, the Wizard SV genomic DNA purification system kit (Promega Corporation, Madison, WI) was used. Amplification of the 16S-to-23S intergenic spacer region was performed as reported previously (10, 16). Restriction digestion of 16S-to-23S short intergenic spacer regions of the lactobacilli was performed with SphI, NcoI, NheI, SspI, SfuI, EcoRV, DraI, VspI, HincII, EcoRI, HindIII, and AvrII (Promega, New England Biolabs, Roche, and Invitrogen). The presence of three intergenic spacer regions (corresponding to long, medium, or short) was confirmatory for identification of Lactobacillus (see Table S1 in the supplemental material). For Lactobacillus reuteri and Lactobacillus vaginalis, two different restriction sites were observed in the 16S-to-23S spacer region.

DGGE was performed as previously described by Burton and Reid (5). Lactobacillus primers and the amplification conditions utilized were described previously (19). Sequences of the reamplified fragments were determined by the dideoxy chain termination method (Robarts Institute, London, Canada). Analysis of the partial 16S rRNA sequences was conducted by using the GenBank database and the BLAST algorithm. Identities of isolates were determined on the basis of the highest score.

H₂O₂ production by lactobacilli (expressed as negative, 1 to 3, 3 to 10, 10 to 30, or 30 to 100 mg/liter) was measured was with Merckoquant peroxide test strips (Merck, Darmstadt, Germany) (20).

One-way analysis of variance (P < 0.05) and the chi-square test (P < 0.25) were used for comparisons. Pairwise between-group comparisons were performed with differences being considered statistically significant at the 0.017 level to allow for a Bonferroni adjustment for multiple comparisons. Comparisons of Lactobacillus species per patient and production of H₂O₂ were assessed by Kruskal-Wallis test (P < 0.05). Whenever differences were observed, the Wilcoxon two-sample test was performed with a critical level of P < 0.025. Agreement between PCR-ARDRA and PCR-DGGE was assessed using the κ agreement coefficient. SAS software, version 9.1 (SAS Institute Inc., Cary, NC), was used for all tests.

The three groups of patients had similar mean ages and behavioral characteristics (P > 0.05) (data not shown).

Control and VVC groups did not differ in Lactobacillus counts (P = 0.543) (Table ​(Table1),1), in agreement with Sobel and Chaim (14), but contrary to Zdolsek et al. (21), who found more lactobacilli in VVC patients. Healthy and VVC groups exhibited higher counts than the BV group (P < 0.001).

Of the 426 isolates Gram stained and evaluated biochemically, 262 bacilli were gram positive and catalase and oxidase negative: 173 Lactobacillus isolates were identified by PCR-ARDRA, with at least one species in 87.5%, 88.2%, and 32.8% of the control, VVC, and BV groups, respectively (maximum of three Lactobacillus species per subject). This compares with 98.4%, 94.1%, and 57.8% for DGGE (maximum of two species per subject). No statistical difference was observed in the rate of vaginal Lactobacillus colonization in normal and VVC groups (P = 0.897); but these differed from the BV group (P < 0.001). The prevalence of each species is shown in Fig. 1A, B, and C.

In this first-ever study of vaginal lactobacilli from Brazilian women, Lactobacillus species detected were similar to those from subjects with distinctly different environments, diets, and geographic locations, namely Nigeria, Canada, and Sweden (4, 5, 18). This is quite remarkable and warrants studies on the origins of these organisms. PCR-ARDRA did not detect L. brevis, L. iners, L. plantarum, and L. sakei, whereas PCR-DGGE did not identify L. agilis, L. coleohominis, L. ruminus, and L. salivarius in any sample. L. iners does not grow on MRS (7), which explains failure to detect it by PCR-ARDRA.

Few lactobacilli did not produce H₂O₂ (1.4% in healthy subjects and 2.6% in VVC subjects) (P < 0.05) (Table ​(Table2),2), in agreement with others (2, 6, 12), suggesting that H₂O₂ does not per se protect against yeast infection. Possibly antifungal cyclic dipeptides, pyroglutamic acid, and lactones produced by Lactobacillus spp. (15) help protect the host. In contrast, 31.8% of isolates from the BV group did not produce H₂O₂ (P > 0.05).

In summary, there are relatively few Lactobacillus species which are commonly found in the vagina. The importance of H₂O₂ as a key anti-infective remains to be fully determined.