Since neutrophils are prevalent among infiltrates and are effective IL-17 producers, as reported in this report and others [36, 37], and are strongly recruited by

IL-17, the positive feedback loop is likely initiated by chemokine-producing resident corneal cells. This attribute explains the rapid fungal growth in immunocompetent BALB/c mice. In the corneas of nude mice, however, the lack of chemokine production leads to decreased leukocyte infiltration, which in turn hampers fungal expansion in the cornea. Our survey of chemokine expression in inoculated corneas confirmed that nude mice are deficient BGJ398 mouse in overall chemokine production (Fig. 6D and E). Furthermore, the CXCL2 supplementation experiments in both nude and BALB/c mice (Fig. 7) provided further support for this hypothesis. Since both APCs in the stroma [9, 10] and corneal epithelial cells as well as

mTOR inhibitor keratinocytes [38-40] are the potential resources of such cytokine/chemokines, the exact mechanisms accounting for the decreased ability of nude mice corneas to produce chemokines and IL-6 (e.g. one of the Th17-inducing factors) upon fungal challenge deserve further investigation. Another apparent issue is that immunodeficient nude mice or CD4+ T-cell-depleted mice did not develop CaK while previous reports have shown that HIV/AIDS patients are more likely to develop FK [14-16]. This might occur because HIV infections deplete CD4+ T cells gradually and partially. Nevertheless, the FK model employs a large pathogen load directly injected into stroma of CD4-null mice. The differences in antimicrobial mechanisms between humans and mice might reconcile pheromone the above inconsistency. Notably, the immunocompetent mice in this study were able to recover from CaK in 3 weeks without treatment, but untreated human patients with FK usually lose corneal function soon after symptoms emerge. Thus, more studies are

required to determine whether IL-17 activity in murine CaK is conserved in FK in humans, including HIV carriers. Given the well-established fact that Th17 cells are a major source of IL-17, and our results showing that CD4-deficient mice did not develop CaK, it is tempting to speculate that IL-17 and Th17 cells functionally converge in the CaK formation pathway. However, based on the difference in the number of CD4+ T cells and neutrophils in BALB/c corneas with CaK (Fig. 5), together with the fact that exogenous CXCL2 reconstituted sensitivity of nude mice to CaK (Fig. 7), we hypothesize that CaK development is neutrophil dependent, especially in the early phase of infection. This neutrophil-dominated response might occur with Th17 cells, as in BALB/c mice, or independent of Th17 cells, as in CXCL2-sensitized nude mice. Similar to our study, Karthikeyan et al.