CRISPR/Cas9‐mediated OsFd1 editing enhances rice broad‐spectrum resistance without growth and yield penalty

Ferredoxins (Fds), a category of small iron-sulphur [2Fe-2S] cluster-containing proteins, localize in plastids and are required for distributing electrons from photosystem I (PSI) to downstream metabolic reactions (Hanke and Mulo, 2013). Based on their expression pattern and redox potential, Fds in higher plants are classified into leaf (photosynthetic) and root (non-photosynthetic) types. In rice, five typical Fd genes have been identified, among which OsFd1 encodes the primary photosynthetic Fd. Knockout of OsFd1 caused rice lethal at seedling stage (He et al., 2020), indicating an essential role of OsFd1 in rice photosynthetic electron transport. We recently reported that knockout of OsFd4, the major rice non-photosynthetic type Fd, increased rice resistance against the blight bacteria Xanthomonas oryzae pv. oryzae (Xoo) (Lu et al., 2023). To determine the immune function of OsFd1 and the possibility of OsFd1 to be a target for genomic modification to enhance rice resistance, we performed CRISPR/Cas9-mediated OsFd1 editing in Zhonghua 11 (ZH11) and obtained two loss-of-function alleles Osfd1-1 and Osfd1-2 carrying a 5-bp deletion and 1-bp insertion, respectively, in the coding region (Figure 1a). Consistent with the previous report (He et al., 2020), both alleles were lethal at young seedling stage under the 12-h light/dark cycle condition (Figure 1b). However, when grown under constant dark, the etiolated seedlings of Osfd1-1 and Osfd1-2 grew similarly as ZH11 (Figure 1b), indicating that the lethality of Osfd1 is light-dependent. We also found that OsFd1 transcript levels and OsFd1 protein abundance were significantly increased under light (Figure S1). When the leaves detached from 10-day-old ZH11 and Osfd1-1 seedlings grown under light cycle were stained with H2DCFDA, a visible cellular indicator for reactive oxygen species (ROS), clear fluorescent signals were observed in the chloroplasts of Osfd1-1, but not in those of ZH11 (Figure 1c), indicating that OsFd1 deletion leads to constitutive ROS accumulation in chloroplasts. Similar to Arabidopsis Fd2-knockout mutant, both Osfd1-1 and Osfd1-2 accumulated significantly higher basal levels of jasmonic acid (JA) and JA-Ile than ZH11 (Figure 1d and Figure S2). ROS production and JA/JA-Ile accumulation contribute to rice immunity (Liu and Zhang, 2022; Ma et al., 2022). To investigate OsFd1's function in rice defense, we transformed OsFd1-overexpression (OE) construct into the callus of heterozygous Osfd1-1 and obtained two independent OEOsFd1 transgenic lines in homozygous Osfd1-1 background (Osfd1-1 OEOsFd1, Figure S3a). OEOsFd1 completely rescued the seedling lethal phenotype of Osfd1-1 (Figure S3b,c). When inoculated with the rice blast fungus Magnaporthe oryzae (M. oryzae), the Osfd1-1 OEOsFd1 lines supported significantly more M. oryzae growth than ZH11 (Figure S3d). We then challenged the Osfd1-1 OEOsFd1 lines with Xoo and found that both lines displayed significantly longer blight lesions (Figure S3e,f) and supported more Xoo growth (Figure S3g), indicating that OEOsFd1 compromised rice defence against the pathogens. Moreover, chitin- and flg22-induced ROS burst were severely reduced in the OEOsFd1 lines compared with ZH11 (Figure S4). Taken together, our data showed that OsFd1 plays a negative role in rice defence against the pathogens. The seedling lethality caused by loss function of OsFd1 limits its application in rice disease resistance breeding. Therefore, we sought to identify weak alleles of Osfd1 that can confer robust resistance without growth penalty. By further screening the CRISPR/Cas9-mediated editing progenies, we identified another two homozygous Osfd1 alleles Osfd1-3 and Osfd1-4 containing 3-bp and 15-bp in-frame deletions. The mutation forms are named as OsFd1Δ3bp and OsFd1Δ15bp, which result in 1 and 5 amino acids deletions, respectively (Figure 1e). Notably, both mutants exhibited WT-like growth and no obvious defects in the agronomic traits (Figure 1f and Figure S5). When challenged with M. oryzae and Xoo strains, Osfd1-4, rather than Osfd1-3, displayed significantly enhanced resistance compared with ZH11 (Figure 1g–k, Figure S6). Consistently, the increased disease resistance was accompanied by enhanced chitin- and flg22-induced ROS burst (Figure 1l). Both the OsFd1 variants, OsFd1Δ1aa and OsFd1Δ5aa, carry the amino acids deletion between the chloroplast-localization signal peptides (CSP) and the conserved [2Fe-2S] cluster (Figure 1e), causing no disruption of their chloroplast localization (Figure S7). Fds are shown to form functional dimers to facilitate electron carrying and delivering (Hasan et al., 2002; Iwasaki et al., 2011; Lu et al., 2023). Our yeast two-hybrid assays indicated a strong self-interaction of OsFd1 (Figure S8). Interestingly, OsFd1Δ5aa, but not OsFd1Δ1aa, showed notably decreased self-association (Figure 1m), suggesting that the five-amino acid deletion in OsFd1Δ5aa may compromise OsFd1 dimerization and decrease the efficiency in electron transfer. Consistently, we observed a moderate ROS accumulation in the chloroplasts of Osfd1-4, rather than Osfd1-3 (Figure 1n). In an attempt to investigate the potential of OsFd1Δ15bp in rice-resistant breeding, we crossed NG9108 (as female parent), a commercial conventional japonica cultivar, with Osfd1-4 (as male parent) and obtained the F2 population. All the tested F2 plants grown normally (Figure S9a), indicating that the OsFd1Δ15bp mutation caused no penalty on growth in different genetic background. When the F2 progenies were inoculated with YN-5, an M. oryzae strain with similar virulence to NG9108 and ZH11, the plants carrying homozygous OsFd1Δ15bp showed increased resistance, compared with those carrying wild-type OsFd1 or heterozygous OsFd1/OsFd1Δ15bp (Figure S9b), indicating that the resistance co-segregates with OsFd1Δ15bp mutation. Collectively, our results revealed OsFd1's critical functions in rice growth and defence. Notably, a specific truncated form of OsFd1 was characterized to confer rice broad-spectrum resistance without yield penalty. These findings provide a potentiality of utilizing OsFd1 gene-editing in resistance breeding to balance rice growth and defense. This research was supported by the National Key R&D Program of China (2023YFD1200800), the Special Support Project of Yunnan Xingdian Young Talent (XDYC-QNRC-2023-0417), the National Natural Science Foundation of China (31701777 and 32360675), the Yunnan Fundamental Research Projects (202301BD070001-145), the Fujian Provincial Science and Technology Key Project (2022NZ030014) and the Yunnan Chen Xuewei Expert Workstation Foundation (202305AF150124). The authors declare no conflict of interest. M.W. designed and supervised the project; H.S., J.C., M.L., W.L., W.D., P.K. and X.Z. performed the experiments; M.W., H.S., J.C., M.L. and Q.L. discussed and interpreted the data; H.S. and M.W. wrote the manuscript. Figure S1 OsFd1 transcript levels and OsFd1 protein abundance were induced under light. Figure S2 JA and JA-Ile levels in Osfd1-2 are significantly higher than those in ZH11. Figure S3 Overexpression of OsFd1 compromised rice resistance against M. oryzae and Xoo. Figure S4 OsFd1 overexpression compromised chitin- and flg22-induced ROS burst. Figure S5 The agronomic traits of Osfd1-3 and Osfd1-4 mutants were similar to those of wild type. Figure S6 Osfd1-4 exhibited increased resistance to M. oryzae isolate Zhong1 and Xoo strain Pxo99. Figure S7 Subcellular localization of OsFd1, OsFd1Δ1aa and OsFd1Δ5aa. Figure S8 The Y2H assay showed self-interaction of OsFd1. Figure S9 Investigating growth and resistance against M. oryzae of NG9108×Osfd1-4 F2 population. Table S1 Primers used in this study. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.