**4.3 Genome and pathogenicity of** *V. virens*

The availability of *V. virens* genome provides a good basis for charactering its pathogenicity in rice flower. As reported, the genome of UV-8b is approximately 39.4-Mb, encoding 8426 putative proteins [52]. The strain IPU010 possesses a genome of 33.6-Mb, which encodes 6451 predicted proteins [53]. Genome analysis reveals that *V. virens* is evolutionarily closest to the entomopathogen *Metarhizium spp.*, suggesting host jumping from animal kingdom to plant kingdom [52]. Moreover, genome information provides evidence supporting that *V. virens* specifically infects rice flower and has a biotrophic lifestyle, since genes responsible for secreted proteins and secondary metabolism are enriched, while genes associated with polysaccharide degradation and nutrient uptake are diminished [52]. A web-based protein-protein interactive database for *V. virens*-*Oryza sativa* interaction has been released, greatly facilitating investigation of *V. virens* pathogenicity [54]. Putatively, 628 secreted proteins are encoded by *V. virens* genome, 193 of the secreted proteins are predicted to be effectors [52].

Effectors are powerful weapons possessed by pathogens to manipulate host immune system and metabolisms for successful colonization. Characterizing their roles is important for understanding pathogen-host interactions. In *V. virens* genome, a number of genes encoding effector proteins, such as UV\_1261, UV\_2508, and UV\_2286, have been identified to suppress *Burkholderia glumae*-induced cell death [52], whilst UV\_5823 shows ability to suppress plant RNA silencing [55]. On

the contrary, some effectors of *V. virens* could induce cell death or defense response in rice protoplast. For example, UV\_44 induces cell death, and this ability relies on the serine peptidase active sites. UV\_1423 could be *N*-glycosylated, which affects its ability to trigger cell death [56]. However, so far, no effectors have been characterized to function in the flower infection of *V. virens*.

Nevertheless, several virulence factors have been identified in *V. virens*. UvSUN2 is a SUN domain protein; loss of function of this factor results in inability of infecting rice flower, as well as abnormal stress responses and mycelium growth [57]. Mutation in *UvPRO1* increases sensitivity to abiotic stresses and attenuates virulence, in addition to impaired growth rate and sporulation [49]. In contrast, a low-affinity iron transporter encoded by *Uvt3277* negatively regulates virulence in *V. virens* [58].

### **4.4 Host compatible interaction with** *V. virens*

Monitoring host responses to *V. virens* infection could help to uncover the pathogenesis of RFS. In an earlier transcriptome study, a series of differentially expressed genes have been identified in a susceptible rice cultivar 93-11 infected with a field *V. virens* isolate [59]. Among them, genes regulated by Ca2+ or abscisic acid are down-regulated, while genes regulated by Myb or WRKY transcription factors are up-regulated. *OsSWEET11* and *OsSWEET14*, which may be involved in disease susceptibility [60], are also up-regulated by *V. virens* infection. Specifically, many pollen development associated genes are down-regulated by *V. virens* infection, but not responsive to other biotic and abiotic stresses, suggesting that these genes may play unique roles in rice-*V. virens* interaction [59]. Additionally, several transcriptome analyses on compatible rice-*V. virens* interactions have been reported. Genes involved in hydrolase, transporter, and flower development tend to be downregulated in susceptible cultivar Huang-Xiu-Zan upon infection [61]. Expression of many defense-related genes such as *PAL* and *PR* genes could be suppressed in susceptible rice cultivars infected with *V. virens* [42, 62]*.*

As a successful pathogen, *V. virens* should have abilities to set up colonization in rice floral organs and acquire abundant nutrients for propagation, in addition to subvert rice immunity. Transcriptome analysis reveals that genes associated with flower opening, such as *ARF6* and *ARF8* homologs, are down-regulated by *V. virens* infection [42]. This may contribute to inhibition of flower opening during RFS pathogenesis [42]. Furthermore, *V. virens* infection causes failure of ovary fertilization. However, a number of grain-filling-specific genes, such as seed-specific starch synthesis related genes and those encoding seed storage proteins, are activated for high expression in *V. virens*-infected rice spikelets [42]. It is suggested that *V. virens* may be able to mimic fertilization and hijack rice grain-filling system for nutrient supply to pathogen growth and RFS ball formation. This finding is further supported by an independent study [63]. Although the underlying mechanism needs further investigation, the observation of *V. virens* activating rice grain filling could provide a promising explanation why mild *V. virens* infection enhances rice yield traits, including grain weight and filled grain number [9]. Identification and characterization of *V. virens* factors that manipulate rice grain filling should be an interesting research area in the future.
