Additionally, the induction of defense genes in the RR core response could be ineffective if their expression levels were too weak compared with those seen in resistant MG fruit. We evaluated the levels of gene expression in inoculated RR fruit via a differential expression comparison to inoculated MG fruit. Of all the RR core defense genes identified above, 269/302 were expressed at equal or greater levels in inoculated RR fruit compared with inoculated MG fruit for all three pathogens. Conversely, 33/302 of these defense genes were expressed at higher levels in MG fruit compared with RR fruit for at least one of the three pathogens . These genes are diverse, representing 15 different defense categories. Prominent genes in this category include TAP1 and TAP2 , two peroxidases associated with defensive suberization in tomato ; CHI3 and CHI17 , two chitinases associated with C. fulvum infection ; and the JA biosynthesis gene OPR3 . While it is possible that resistance may be determined by these genes, growing blueberries in containers these results indicated that the differences in immune responses observed between MG and RR fruit are not likely solely responsible for differences in susceptibility, and, therefore, we considered the alternate hypothesis.
We explored the possibility that the increase in susceptibility to fungal pathogens is heavily influenced by a decline of preformed defenses and accumulation of susceptibility factors that occur during fruit ripening prior to pathogen challenge. Due to the complexity of the ripening program, we utilized isogenic non-ripening tomato mutants as tools to identify specific developmental features that are integral to fruit resistance or susceptibility. The Cnr, rin, and nor mutants produce fruit that lack most of the characteristic changes associated with normal ripening, such as color, texture, acidity, sugar accumulation, and ethylene production, but yet are phenotypically different from one another . All three mutant lines likely result from spontaneous gain-of-function mutations in transcription factors with key roles in the regulation of ripening . We inoculated fruit of these mutant genotypes at comparable stages to MG and RR wild-type fruit with B. cinerea, F. acuminatum, and R. stolonifer and measured disease incidence and severity up to 3 dpi . For all three pathogens at both MG-like and RR-like stages, only nor fruit were consistently resistant to infection. MG-like fruit of Cnr were the only unripe fruit susceptible to any pathogen, with both B. cinerea and F. acuminatum able to produce lesions on a significant number of these fruit. Consistent with this, Cnr RR-like were more susceptible than wild-type RR fruit to B. cinerea, with average disease severity nearly twice as great at 3 dpi .
The fruit of rin at both MG-like and RR-like stages showed similar or slightly lower susceptibility to all pathogens when compared with wild-type, with the exception of a significant reduction in disease incidence to F. acuminatum at the RR-like stage. Because some ripening processes may promote susceptibility, others may maintain resistance, and others may have no impact, we hypothesized that the Cnr, rin, and nor mutations differentially affect ripening-associated genes or pathways that are critical to tip the balance towards either susceptibility or resistance. We sequenced mRNA from B. cinerea-inoculated and healthy fruit from the non-ripening mutants at MG-like and RR-like stages at 1 dpi. We chose B. cinerea inoculations because this pathogen showed the clearest differences in susceptibility phenotypes between these genotypes. We first characterized transcriptional responses of mutant fruit to pathogen challenge by using enrichment analysis of defense-related processes to determine if differences in immune responses could explain the distinct susceptibility phenotypes . In most cases, the mutant fruit exhibited similar patterns of defense classification enrichments as wild-type fruit in both stages, with some notable exceptions. Compared with the other genotype–stage combinations, Cnr MG-like responses were deficient in the expression of genes from several prominent defense classifications, including chitin catabolic process , the plant–pathogen interaction and glutathione metabolism pathways, ERF and WRKY transcription factors, and RLK and CAMK genes.
Given that Cnr fruit were the only genotype at the MG-like stage to display susceptibility to B. cinerea infection, it can be suggested that these defense processes may be necessary for resistance in unripe fruit. However, these processes were enriched in the susceptible RR-like fruit of Cnr and rin, as well as wild-type RR fruit, which clearly indicates that they are not sufficient to result in a resistant outcome.For example, the responses of resistant nor fruit in both MG-like and RR-like fruit were noticeably less enriched in ethylene-associated pathways and more enriched in JA-associated pathways. These results suggest that JA-mediated defenses may contribute to tomato fruit resistance in the absence of ethylene, and that the nor mutation may activate JA-associated resistance. In support of this observation, levels of JA in healthy fruit appeared to be linked to resistance: they were highest in RR-like nor fruit, and only nor fruit experienced an increase in JA in the transition from MG-like to RR-like . Ethylene levels increase dramatically during ripening in wild-type fruit, but they remain low in all three non-ripening mutants . However, both Cnr and rin mutants produce ethylene in response to B. cinerea inoculation, with ethylene production in inoculated Cnr MG-like fruit reaching levels nearly three times greater than wild-type MG fruit by 3 dpi. Moreover, ethylene signaling/response genes are highly enriched in Cnr MG-like fruit responses . In contrast, healthy nor fruit did not produce substantial amounts of ethylene at either stage, and inoculation in nor fruit did not appear to induce significant ethylene production as in rin and Cnr fruit. These results indicate that high levels of ethylene are not required for B. cinerea resistance and most likely promote susceptibility. Regardless, the combination of hormone activity and defense gene enrichment suggests that, with the exception of Cnr MG-like fruit, resistance or susceptibility in the non-ripening mutants cannot be merely explained by the presence and/or magnitude of immune responses.
To identify genes that are involved in resistance or susceptibility that change during tomato fruit ripening, we used a differential expression analysis comparing healthy RR/RR-like to healthy MG/MG-like fruit for each wildtype and mutant line. In wild-type fruit, 6574 genes were significantly down-regulated in RR fruit compared with MG, while 5674 genes were significantly up-regulated . We used the susceptibility phenotypes and the transcriptional profiles of the mutant fruit to filter these ripening-associated genes and identify critical preformed defense mechanisms or susceptibility factors. Of the four genotypes, square pots all except nor experience an increase in susceptibility in the transition from MG/MG-like to RR/RR-like fruit. Thus, we selected ripening-associated genes that showed the same expression pattern in wild-type, Cnr, and/or rin, but not nor. This filtering resulted in 2893 down-regulated and 2003 up-regulated genes, respectively. We assumed that effective preformed defenses will decrease during ripening. Thus, the set of filtered down-regulated genes, being those that are highly expressed in healthy MG fruit compared with healthy RR fruit, should contain key genes related to preformed defenses. The filtered down-regulated genes contained 251 defense genes, while up-regulated genes included only 171 defense genes, indicating a net loss of about 80 genes in the transition from MG/MG-like to RR/RR-like susceptible fruit. Furthermore, the 251 defense genes from the filtered down-regulated set were over-represented by functional categories involved in reactive oxygen species response and detoxification, proteolysis, and the biosynthesis of secondary metabolites . These down-regulated ROS-related genes spanned several subfamilies including thioredoxins, glutaredoxins, glutathione S-transferases, and peroxidases. Among the down-regulated proteolytic genes were several subtilisin-like proteases, including SBT3 . Lastly, in addition to several genes involved in the methylerythritol 4-phosphate pathway of terpenoid biosynthesis, two copies of the lignin biosynthesis gene CCoAOMT were also among the filtered down-regulated class, suggesting that cell wall fortification could be inhibited upon infection. These results indicate that ripening involves a loss of multiple defense genes, and that the pre-existing levels of genes involved in ROS regulation, proteolysis, and secondary metabolite biosynthesis may be critical for resistance. Finally, we evaluated filtered up-regulated genes that are highly expressed in healthy RR fruit compared with healthy MG fruit, as they may include potential susceptibility factors. Since there is little scientific literature on classes of genes that may constitute susceptibility factors in plants, we focused on the up-regulated genes that were highly expressed in the RR/ RR-like fruit of the susceptible genotypes. Such genes may have disproportionate impacts on susceptibility due to their high expression. To identify these genes, we calculated average normalized read count values for each gene across WT, Cnr, and rin RR/RR-like fruit. The distribution of these values over the filtered up-regulated genes is a notably long-tailed one with a range of 2.43 to 179 649.29 and an average of 1295.
We identified genes with abnormally high expression values by selecting outliers from a log10-transformed distribution of the data. This resulted in a list of 16 genes . They include several genes previously discovered to be active during tomato fruit ripening, including the flavor volatile biosynthesis gene ADH2 , the carotenoid biosynthesis gene Z-ISO , the pectin-degrading enzymes PG2a and PL , and the ethylene receptor ETR4 , among other genes involved in carbohydrate metabolism. While any of these genes has the potential to impact susceptibility, genes for cell wall-degrading enzymes, such as PL and PG2a, which facilitate fruit softening during ripening, represent especially good candidates given both the importance of cell wall integrity in defense against fungal pathogens and previous research on RNAi-developed mutants in tomato . To validate the impact of PG2a and PL expression in wild-type RR fruit on susceptibility to B. cinerea, we utilized CRISPR-based mutants in each of these genes . RR fruit from these lines are similar in regards to soluble solids content, titratable acidity, and juice pH, but CRISPR-PL fruit are nearly 30% firmer than fruit from the CRISPR-PG2a and azygous WT control lines . In conjunction with these firmness differences, RR fruit of the CRISPR-PL line, but not the CRISPR-PG2a line, demonstrated reduced susceptibility to B. cinerea compared with the azygous line . At 3 dpi, disease incidence in the CRISPR-PL fruit was 56% lower than that in azygous fruit. We conclude that the ripeningassociated pectate lyase enzyme is a major susceptibility factor for B. cinerea infection in tomato fruit.During ripening, fruit may gradually lose either the ability to activate or the effectiveness of components of the plant immune system, defensive hormone production and signaling, and downstream transcriptional responses. Alternatively, ripening processes such as cell wall breakdown, simple sugar accumulation, changes in pH and secondary metabolite composition, and, in climacteric fruit, increased levels of ethylene may impact the fruit’s capability to resist fungal attack . The widespread nature of this phenomenon in diverse fruit pathosystems suggests that ripening-associated susceptibility is likely to be mediated by combinations of the above factors. In tomato, ripening-associated susceptibility has been demonstrated not only for the model necrotrophic pathogen B. cinerea, but for other fungal pathogens including Colletotrichum gloeosporioides , R. stolonifer, and F. acuminatum . Here, for the first time, we identified specific host responses in both resistant unripe and susceptible ripe fruit that are common to multiple pathogens and thus represent core responses to fungal infection. Most prominently, these core responses featured RLKs, WRKY and ERF transcription factors, JA biosynthesis, and chitin catabolism . Some genes that appear in both the MG core and the RR core responses were previously studied components of plant immunity in tomato, including the JA biosynthesis gene LoxD , the subtilisin-like protease SBT3 , the peroxidase CEVI-1 , and the chitinase CHI9 . Though response to inoculation overlaps somewhat with response to wounding in MG fruit , transcriptional profiles , and ethylene measurements indicate that the bulk of inoculation responses are a direct result of fungal attack. This is also evident by the presence of a necrotic ring only in inoculated MG fruit and not in the wounded controls or the inoculated RR fruit, indicating that the unripe fruit is capable of inducing an oxidative burst in response to the pathogen presence . However, most defense genes uncovered were found solely in the RR core response. These included several well-known defense genes that were only expressed in RR fruit, such as WRKY33 , the ERF PTI5 , the RLK TPK1b , and the MAP kinase MPK3 . While the MG core response did contain some defense genes that were not present in the RR core response , expression of most of these genes was also identified in the RR response to one or two pathogens.