Engineering false smut resistance rice via host-induced gene silencing of two chitin synthase genes of Ustilaginoidea virens.

Rice false smut (RFS), caused by the pathogenic ascomycete fungus Ustilaginoidea virens (U. virens), is one of the severely devastating diseases worldwide (Sun et al., 2020; Zhang et al., 2014). U. virens infects rice flowers causing over 25% yield loss, also producing poisonous mycotoxins that threaten the health of humans and animals (Fan et al., 2016; Zhang et al., 2014). So far, rice varieties showing resistance to RFS are rarely mined and none resistant genes are genetically confirmed (Zhang et al., 2019). Host‐induced gene silencing (HIGS) is a powerful strategy to control diseases, where small interference RNAs (siRNAs) are produced by plants to silence essential genes of pathogens. Application of HIGS to control fungal diseases is successfully reported in several crops (Hou and Ma, 2020). Here, we evaluated the potential of HIGS for engineering rice against RFS by silencing chitin synthase genes of U. virens.

Chitin synthases are key enzymes controlling the formation of chitin, which is a major component of fungal cell wall and plays vital roles in hyphal growth, fungal morphogenesis and pathogenesis (Lenardon et al., 2010). The U. virens genome contains nine chitin synthase genes Uv8b_7958, Uv8b_3908, Uv8b_7948, Uv8b_4757, Uv8b_3222, Uv8b_3223, Uv8b_3210, Uv8b_7677 and Uv8b_1761 (Zhang et al., 2014), which are named as UvChs1‐UvChs9 (Figure 1a). These UvChs genes had differential expression patterns during the colonization of U. virens on rice spikelets. UvChs2 and UvChs5 were significantly activated, whereas others were slightly activated upon U. virens infection (Figure 1b), indicating that UvChs2 and UvChs5 had possibly major roles in the process of U. virens infection.

Figure 1

Generation of strong resistance to RFS via HIGS‐mediated silencing two chitin synthase genes. (a) Phylogenetic tree of fungal chitin synthase. (b) Transcription patterns of nine UvChs genes after U. virens infection. dpi, days post infection. (c) Transcripts of UvChs genes in UvChs2RNAi and UvChs5RNAi U. virens strains. (d–h) Comparison of number of conidiation (d), conidiation germination (e), chitin content (f), chitin synthase activity (g) and number of RFS balls per panicle (h) between non‐transgenic WH17 and UvChs2RNAi or UvChs5RNAi U. virens strains. (i) Microscopic assay of siRNAs migrating into U. virens cells. (j) Comparison of number of RFS balls per panicle between wild‐type WDR64 and HIGSChs2, HIGSChs5 or HIGSChs2/5 transgenic lines. (k) Transcripts of UvChs2 and UvChs5 in spikelets of HIGSChs2, HIGSChs5 or HIGSChs2/5 transgenic lines after U. virens infection. (l) Agronomic traits of wild‐type WDR64 and HIGSChs2, HIGSChs5, HIGSChs2/5 or OsSWEET11:HIGSChs2/5 transgenic lines. (m) Comparison of number of RFS balls per panicle between wild‐type WDR64 and OsSWEET11:HIGSChs2/5 transgenic lines. (n) Transcripts of UvChs2 and UvChs5 in spikelets of OsSWEET11:HIGSChs2/5 transgenic lines after U. virens infection. (o) Length distribution and abundance of siRNAs targeting UvChs2 and UvChs5 in OsSWEET11:HIGSChs2/5 transgenic lines. (p) Accumulation of siRNAs in OsSWEET11:HIGSChs2/5 lines by polyacrylamide gel electrophoresis‐northern assay. Data represent mean ± SD. Asterisks indicate significant difference determined by two‐tailed Student’s t‐test (**P < 0.01 or *P < 0.05).

To verify the potential roles of UvChs2 and UvChs5 on chitin content and pathogenesis of U. virens, we selected a 598‐bp partial UvChs2 coding region (1192‐ to 1789‐bp of its cDNA) and a 527‐bp partial UvChs5 coding region (2043‐ to 2569‐bp of its cDNA), which were highly specific to UvChs2 or UvChs5 with no consecutive nucleotide sequences matching rice and human genomes, to generate RNAi constructs driven by the promoter of U. virens β‐actin gene. The constructs were introduced into U. virens isolate WH17, each two transgenic U. virens strains designating as UvChs2RNAi and UvChs5 RNAi were selected for analysis. UvChs2 and UvChs5 had significantly reduced transcripts, while other UvChs genes had unchanged expressions in the transgenic U. virens strains compared with non‐transgenic WH17 (Figure 1c), suggesting the RNAi constructs are able to effectively and specifically silence the target genes in U. virens. In comparison with WH17, conidiation was reduced in UvChs2RNAi but not in UvChs5RNAi U. virens (Figure 1d). Both UvChs2RNAi and UvChs5RNAi U. virens had significantly reduced conidium germination (Figure 1e), decreased chitin accumulation (Figure 1f) and reduced chitin synthase activity (Figure 1g). Additionally, we assessed their pathogenicity on rice spikelets by inoculation them on susceptible rice WDR64 panicles. The disease symptoms consisting of yellow smut balls were markedly compromised in panicles infected with UvChs2RNAi or UvChs5RNAi U. virens than WH17 (Figure 1h). Thus, UvChs2 and UvChs5 are essential for pathogenicity of U. virens.

To investigate that siRNA molecules from outside can enter into U. virens cells, commercial non‐ U. virens fluorescein (FAM)‐labelled siRNAs (21 nt) were incubated with conidial spores of U. virens and assayed with a fluorescence microscope. The results showed that clearly observable fluorescence in the conidium and mycelium treated with FAM‐siRNA (Figure 1i), indicating that exogenous siRNAs can actively migrate into U. virens cells.

Next, these two RNAi constructs, and a chimeric RNAi construct that simultaneously silence UvChs2 and UvChs5, driven by 35S promoter, were introduced into WDR64 by Agrobacterium‐mediated transformation to generate transgenic plants (HIGSChs2, HIGSChs5 or HIGSChs2/5). After inoculation with U. virens WH17, WDR64 had about 13 RFS balls per panicle, whereas these transgenic lines showed dramatically attenuated disease symptoms with 2–4 RFS balls per panicle (Figure 1j). Consistent with enhanced resistance, transcripts of UvChs2 and UvChs5 were markedly reduced in the transgenic rice than in WDR64 (Figure 1k), suggesting RFS resistance in these transgenic lines was caused by in planta‐derived silencing of the target genes. Notably, these transgenic lines maintained agronomic traits that were indistinguishable from those of wild type (Figure 1l).

Additionally, we transferred the chimeric UvChs2 and UvChs5 RNAi cassette driven by the promoter of OsSWEET11 which is highly expressed in spikelet and dramatically activated by U. virens infection (Chu et al., 2006; Fan et al., 2020), into WDR64 to generate OsSWEET11:HIGSChs2/5 transgenic lines. Similarly, the lines displayed significantly lower transcripts of UvChs2 and UvChs5 and dramatically compromised disease symptoms than in WDR64 after U. virens infection (Figure 1m, n), whereas unchanged agronomic traits (Figure 1l). To confirm that silencing of UvChs2 and UvChs5 in infected U. virens was mediated by homologous siRNAs generated in transgenic rice spikelets, small RNA sequencing was carried out to identify siRNAs specific to the RNAi cassette in OsSWEET11:HIGSChs2/5 transgenic lines. The sequencing data showed that the unique small RNA reads mapping to UvChs2 and UvChs5 genes were 3.25% and 2.07% of the total small RNAs detected respectively. The lengths of siRNAs mapped to UvChs2 and UvChs5 were distributed between 18 and 28 bp, wherein 21 bp was the most abundant (Figure 1o). An ideal accumulation of 21‐nt siRNA in OsSWEET11:HIGSChs2/5 transgenic lines but not in wild type was observed by polyacrylamide gel electrophoresis‐northern assay (Figure 1p). These data indicate that the RNAi constructs were successfully processed into specific siRNA molecules in transgenic rice, and these siRNAs were translocated to fungal cells upon U. virens infection, thereby suppressing the expressions of UvChs2 and UvChs5 to attenuate pathogenicity of U. virens.

In summary, our results suggest that HIGS targeting chitin synthase genes involved in chitin synthesis and pathogenicity of U. virens is effective and can be used as an alternative strategy for developing RFS resistant rice.

Conflict of interest

The authors declare no conflict of interest.

Author contributions

X.L., R.H. and M.Y. conceived the project. X.L., R.H., J.L., G.X. and M.Y. performed the research and analysed the data. M.Y. wrote the manuscript.