Host barriers and responses to Uncinula necator in developing grape berries

Grape berries are highly susceptible to powdery mildew 1 week after bloom but acquire ontogenic resistance 2 to 3 weeks later. We recently demonstrated that germinating conidia of the grape powdery mildew pathogen (Uncinula necator) cease development before penetration of the cuticle on older resistant berries. The mechanism that halts U. necator at that particular stage was not known. Several previous studies investigated potential host barriers or cell responses to powdery mildew in berries and leaves, but none included observation of the direct effect of these factors on pathogen development. We found that cuticle thickness increased with berry age, but that ingress by the pathogen halted before formation of a visible penetration pore. Cell wall thickness remained unchanged over the first 4 weeks after bloom, the time during which berries progressed from highly susceptible to nearly immune. Autofluorescent polyphenolic compounds accumulated at a higher frequency beneath appressoria on highly susceptible berries than on highly resistant berries; and oxidation of the above phenolics, indicated by cell discoloration, developed at a significantly higher frequency on susceptible berries. Beneath the first-formed appressoria of all germinated conidia, papillae occurred at a significantly higher frequency on 2- to 5-day-old berries than on 30- to 31-day-old fruit. The relatively few papillae observed on older berries were, in most cases (82.8 to 97.3%), found beneath appressoria of conidia that had failed to produce secondary hyphae. This contrasted with the more abundantly produced papillae on younger berries, where only 35.4 to 41.0% were located beneath appressoria of conidia that had failed to produce secondary hyphae. A pathogenesis-related gene (VvPR-1) was much more highly induced in susceptible berries than in resistant berries after inoculation with U. necator. In contrast, a germin-like protein (VvGLP3) was expressed within 16 h of inoculation in resistant, but not in susceptible berries. Our results suggest that several putative barriers to infection, i.e., cuticle and cell wall thickness, antimicrobial phenolics, and two previously described pathogenesis-related proteins, are not principal causes in halting pathogen ingress on ontogenically resistant berries, but rather that infection is halted by one or more of the following: (i) a preformed physical or biochemical barrier near the cuticle surface, or (ii) the rapid synthesis of an antifungal compound in older berries during the first few hours of the infection process.