Studies to date show that vaginal and oral epithelial cell anti-Candida
activity requires cell contact and a putative carbohydrate moiety (10
). Although a specific effector carbohydrate had yet to be identified, the role for a carbohydrate comes from the sensitivity of the antifungal activity to periodic acid that cleaves carbohydrates to aldehydes and thus releases the majority of the sugar moieties from the cell surface. The present study focused on a better understanding of the properties and mechanisms of this potentially important innate host defense activity. The vast majority of data were obtained from oral epithelial cells, although vaginal epithelial cells were employed as well in many of the designs to confirm the similar action by both types of cells.
Previous studies showed that adherence of epithelial cells to Candida
was not affected by periodic acid treatment (10
), indicating that the putative carbohydrate moiety was directly involved in the anti-Candida
activity rather than a secondary effect of adherence (10
). Despite the effective cleavage of sulfated polysaccharides, sialic acid residues, or glucose- and mannose-containing carbohydrates, a specific carbohydrate moiety was not identified (10
Recognizing the difficulty in identifying a specific effector carbohydrate, we reasoned that supernatants containing the putative carbohydrate(s) could be used to better understand the properties of the moiety and the mechanism of action. Results showed that supernatants containing the putative carbohydrates incubated with Candida
could not compete for the inhibitory activity by fresh epithelial cells, using both a pretreatment or direct treatment design. The moiety equivalence to that on epithelial cells (5 : 1 ratio equivalence) was used in principle in these assays and as such reduces the possibility of a role for released carbohydrate. However, we recognize that the released moiety may not resemble the intact moiety such that it was unable to bind to receptors on Candida
in the same manner as that present on the epithelial cells. Alternatively, the concentration of the released moiety may not have been saturating. In any event, similar results were seen for both oral and vaginal cells confirming, as with all previous studies (10
), that oral and vaginal epithelial cells function similarly.
Taken together, results suggest that the effector moiety on oral and vaginal epithelial cells may not be a carbohydrate and that periodic acid is simply inactivating the cells instead. Support for this comes from a retrospective analysis of a large number of experiments conducted over several years where it was determined that no matter what level of carbohydrate was stripped from the cells by periodic acid, the same abrogation of growth inhibition was observed. Thus, an acid-labile property of the epithelial cell antifungal activity became a distinct alternative possibility. This was confirmed by a final series of studies that evaluated the effects of two additional acids. In one experiment, treatment of oral or vaginal epithelial cells with either TFMS or HCl abrogated the anti-fungal activity without liberating appreciable carbohydrates or affecting epithelial cell viability. In a second design, performing the standard [3
H]-glucose uptake assay under acidic conditions resulted in reduced oral epithelial cell antifungal activity. Therefore, future mechanistic studies will focus on the interaction of Candida
and epithelial cells in the presence or absence of acid treatment. These studies also provided a possible explanation for the weaker activity of vaginal epithelial cells compared to oral epithelial cells (14
). The reduction of the oral cell activity in an acidic environment suggests that the lower activity by human vaginal epithelial cells may be due to the general acidic microenvironment of the vaginal cavity. An alternative explanation is that the restriction of Candida
to blastoconidia at the lower pH influenced the epithelial cell activity. However, this is unlikely since a similar study performed at room temperature that restricted Candida
to the blastoconidia had no effect on the activity (13
), and the acidic pH in these studies had no effect on the growth of Candida
alone despite being confined to blastoconidia.
Additional information regarding the mechanism of action was garnered from studies using fixed epithelial cells. Previously we reported that the anti-Candida
activity was partially resistant to fixation (14
). In those studies, Trypan blue was used as the indicator of viability. As such, the small numbers of positively stained cells suggested that the fixed cells were still viable when tested, or at least impermeable to Trypan blue. However, the present study clearly shows that fixed cells that had antifungal activity were indeed non–viable, as evidenced by nuclear staining by propidium iodide. Together, these results suggest that the antifungal activity requires intact (impermeable to Trypan blue) but not necessarily live epithelial cells. The requirement for Trypan blue impermeability had been suggested for some time as growth inhibition was never observed if the cells were not deemed viable by Trypan blue dye exclusion (15
). This was confirmed in the present study by the positive correlation of antifungal activity with Trypan blue dye exclusion from fixed and unfixed cells tested immediately after fixation compared to after 4 days in culture. But obviously the requirement for Trypan blue impermeability does not infer a requirement for live cells. Of note, the lack of a requirement for live cells is consistent with the lack of any evidence for intracellular signals (tyrosine kinases or phospholipase C) in the epithelial cell antifungal activity, as well as a lack of involvement by microtubules and microfilaments (Fidel, unpublished data).
The results of this study provide a considerable amount of new information regarding the requirements and mechanism for the epithelial cell antifungal activity. The static antifungal activity, while not having a strict requirement for live cells, does require that the putative effector moiety be adequately present on intact cells. Furthermore, the activity appears to be acid-labile rather than mediated by an effector carbohydrate, although we understand that the two may not be mutually exclusive. In this regard, it is possible that the experiments designed and conducted involved a carbohydrate that was not able to be evaluated in the absence of the epithelial cells. However, our most recent data suggest that surface proteins (or glycoproteins) extracted from the epithelia cells inhibit Candida growth (unpublished data). Thus, while carbohydrates may be involved at some level, it does not appear that they are involved exclusively. We propose that the static activity by epithelial cells may represent a sophisticated symbiotic relationship between the host and Candida, where the host benefits by Candida staying ‘in check’, and Candida benefits by not being killed. As for the growth inhibition mechanism, inasmuch as he moiety on the epithelial cells can exert some action on Candida, it is equally possible that, as a result of attachment to the epithelial cells, Candida elicits a selfcontrolling response that halts its growth. Studies are in progress to further investigate the effector moiety and to evaluate the interaction between epithelial cells and Candida at the molecular level.