**4.2 Genetic transformation of** *V. virens*

*Protecting Rice Grains in the Post-Genomic Era*

**4. Pathogenesis of RFS**

**Figure 2.**

**4.1 Infection process of** *V. virens*

*(H) are courtesy of Prof. Dongwei Hu from Zhejiang University, China.*

The RFS pathogen *V. virens* specifically attacks rice flowers to form RFS balls, causing economically important disease. The infection process of *V. virens* in rice flower has been identified cytologically as follows: at late booting stage of rice, spores of *V. virens* come into contact with developing spikelets and germinate on their surface, or epiphytically grown hyphae reach the surface of developing spikelets. The hyphae could not penetrate the spikelet but extend into the inner space of a spikelet via the gap between the lemma and the palea [38]. After entering, the pathogen primarily infects the stamen filaments intercellularly [39], probably due to loose alignment of cells and flexible cell walls [40]. Lodicules and stigma could also be attacked, although to a lesser extent [39, 41]. However, no infection structures, such as appressorium and haustorium, can be detected during infection. Along with time, mycelia grow to enclose all the floral organs, then protrude out of the spikelet, and ultimately form a ball-shape colony covered with chlamydospores. At late stage of infection or in a RFS ball, stamen filaments are replaced by mycelia, but the ovary and lodicules remain intact, suggesting that they may contribute to the formation of RFS ball [42]. In addition, the hyphae of *V. virens* could not extend into pedicles and stems connect-

*Morphology of the rice false smut pathogen. Colony of Ustilaginoidea virens (anamorph) in PSA medium from top view (A) and back view (B). (C) Colonies of U. virens in PSB. (D) A single colony of U. virens in PSB. (E) Conidia of U. virens in PSB. (F) Chlamydospores of U. virens from false smut balls. (G) Sclerotia of Villosiclava virens (telemorph). (H) Sclerotia of V. virens germinate to produce fruit bodies. Images (G) and* 

ing the spikelets, and no anatomic changes are detected in pedicles [39].

Although RFS disease symptoms are observed at rice grains due to pathogen infections of spikelets, *V. virens* also grows on other rice organs without obvious symptoms. At the germination stage of rice seed, chlamydospores of *V. virens* could germinate on coleoptiles and the hyphae are able to extend intercellularly between epidermal cells [43]. At seedling stage, chlamydospores could also germinate on the surface of roots and grow in the intercellular space of root epidermal cells [44, 45]. More recently, a detailed observation on *V. virens*-infected rice roots indicates that the cellulose microfibrils of epidermal cell wall are very loose, similar to those of stamen filaments, and thus are prone to be infected [46]. However, the hyphae of *V. virens* are stopped by sclerenchyma layer from entering into endodermis and phloem tissues [46]. Again,

**92**

Genetic manipulation is essential to clarify the pathogenicity of *V. virens* and its interaction with rice. Several transformation techniques have been successfully applied to *V. virens*. Electroporation, the process in which a strong electric pulse is applied to an organism in order to transiently increase membrane permeability, has the advantages of being rapid and inexpensive. Through electroporation on conidia, an enhanced green fluorescence protein (eGFP)-expressing *V. virens* strain was obtained, which was able to infect rice flowers and form RFS balls [47]. Polyethylene glycol (PEG)-mediated approach is typically more efficient than electroporation and generally yields a higher percentage of stable transformants. Ashizawa and colleagues [38] engineered a GFP-tagged *V. virens* strain via PEGmediated transformation on protoplasts, and identified the infection route of *V. virens* in rice spikelets with this strain. *Agrobacterium tumefaciens*-mediated transformation (ATMT) is a fast and easy way to transfer foreign DNA into fungal cells. A few reports have recorded the establishment and optimization of ATMT procedure on *V. virens* conidia, and construction of T-DNA mutant libraries [48, 49], which facilitate characterizing virulence factors in this pathogen. For example, Yu et al. [48] obtained a T-DNA insertion library with 5600 hygromycin-resistant transformants, and identified 37 mutants with impaired pathogenicity. Targeted gene knockout with PEG- or ATMT-mediated transformation has been tried on *V. virens*; however, the homologous gene replacement frequency was very low [50]. Very recently, Liang and colleagues [51] established a CRISPR-Cas9 system, which remarkably increased the frequency of homologous gene replacement. The knockout efficiency could be as high as 50–90% for some *V. virens* genes.
