The pathogenetic role of soluble products of Trichomonas vaginalis growth in culture is controversial. To evaluate this role, T. vaginalis was grown in broth and cell culture and the cell-free filtrate was applied to fresh cell culture monolayers. When adjusted to pH 6.5, filtrates obtained from 22-h culture growth totally disrupted McCoy, HEp-2, human foreskin fibroblast, and Chinese hamster ovary cell monolayers within 6 h. These detached ceils remained >90% viable. This cell-detaching factor (CDF) was heat and acid labile, with a pH optimum of 6.5. CDF has trypsinlike activity which disrupts monolayer cells, but cells do not die if the pH is controlled. CDF was purified by ethanol precipitation, ammonium sulfate fractionation, and ion-exchange and gel filtration column chromatography. A 200,000-molecular-weight glycoprotein which was also immunogenic by immunoblot with human sera reactive to T. vaginalis was isolated in this manner. This confirms the presence of a specific soluble CDF derived from T. vaginalis whose application may be important as a diagnostic tool and in further studies of pathogenesis.

Extracellular protease activity was detected in serum-free culture filtrates of Trichomonas vaginalis. The activity was demonstrated by hydrolysis of hide powder azure and possessed the characteristics of cysteine type proteases: inhibition by N-ethyl maleimide, Cu2+, antipain, N-tosyl-L-phenylalanine chloromethyl ketone, N-tosyl-L-lysine chloromethyl ketone, leupeptin, chymostatin, and iodoacetamide, and enhancement by cysteine, EDTA, and dithiothreitol. The activity was optimal at acid pH and the protease was also active on peptide nitroanilides with arginine derivatives. Purification of this activity by ethanol precipitation, ammonium sulfate fractionation, ion exchange chromatography, and gel filtration resulted in the isolation of two proteases estimated by sodium dodecyl sulfate - polyacrylamide gel electrophoresis to have molecular masses of 60 and 30 kilodaltons (kDa), respectively. The larger molecular mass protease broke down during purifications to two subunits of approximately 23 and 43 kDa, as determined by gel electrophoresis. Rabbit sera derived by immunization with the 23-kDa subunit cross-reacted by immunoblot with the 60- and 43-kDa subunits, but not with the 30-kDa protease. These soluble products of T. vaginalis growth could be important pathogenically in establishing T. vaginalis infection in the normally acid (pH less than or equal to 4.5) environment of the vagina.

Recent work with a mouse model of Trichomonas vaginalis infection indicated that estrogenized BALB/c mice that were preinfected with Lactobacillus acidophilus showed a greater duration of T. vaginalis infection as compared to a control group of mice that were not treated with L. acidophilus. To examine the interaction between T. vaginalis and L. acidophilus further we performed in vitro competitive growth assays between the 2 species. Although the addition of L. acidophilus to the T. vaginalis cultures slowed the growth of the protozoa, the added bacteria did not increase trichomonad death. However, T. vaginalis had a deleterious effect on L. acidophilus growth in combined cultures when compared to matched controls. Using an initial inoculum of 10(5)/ml, at 40 hr the control L. acidophilus concentrations had grown to 1.3 x 10(7)/ml. However, in combined culture with T. vaginalis, L. acidophilus concentrations at 40 hr had fallen to 7.8 x 10(5)/ml and 6.1 x 10(4)/ml for the 10:1 (T. vaginalis at 10(4)) and 1:1 (T. vaginalis at 10(5) test ratios, respectively (P < 0.01). This demonstrates that T. vaginalis can cause the concentration of L. acidophilus to fall in vitro and may explain why the concentration of L. acidophilus in the vagina falls in trichomoniasis.

We have previously isolated two extracellular cysteine proteases (60 and 30 kDa, respectively) from the cell filtrate of an isolate of Trichomonas vaginalis. In this study the clinical presentation of 12 clinical isolates of T. vaginalis was correlated with their protease activity. All 12 isolates produced a 60-kDa protease as demonstrated by immunoblotting. However, only 5 of 12 isolates produced a 30-kDa protease in the extracellular filtrate. Protease activity did not differentiate between isolates obtained from symptomatic versus asymptomatic women. The 60-kDa protease, which breaks down into a 43- and 23-kDa subunit, was detected by immunoblotting with anti-23-kDa cross-reacting rabbit serum in the vaginal washes of 3/3 women with active T. vaginalis infection, 0/1 woman cured of T. vaginalis and 0/2 control women with vaginal candidiasis. These results suggest that the 60-kDa protease is found in all isolates testes in vitro, is present in active disease and may be important in the pathogenesis of T. vaginalis infection.

Recent work has shown that Trichomonas vaginalis produces a cell-detaching factor (CDF) that causes detachment of monolayer cells in vitro. To study the role of CDF as a pathogenic marker of disease, we studied the production of CDF in 12 clinical isolates of T. vaginalis. These isolates were also utilized in the mouse subcutaneous assay of Honigberg, which is the standard for pathogenicity of T. vaginalis. The isolates were divided into three groups based on clinical presentation (asymptomatic [n = 4], moderate [n = 4], and severe symptoms [n = 4]). CDF was assessed by harvesting the supernatant from the growth of T. vaginalis in cell culture and filtering the supernatant through a 0.45-,Im-pore-size filter. The filtrate was applied in a microtiter cytotoxicity assay. The mouse subcutaneous assay did not significantly differentiate among the isolates. However, CDF was strongly associated with clinical presentation by two-way, repeated-measure analysis of variance (P = 0.025). Thus, CDF appears to correlate with clinical presentation and may be an important virulence marker in T. vaginalis pathogenesis.