Osa-miR159a is Responsive to the Infection by M. oryzae
Previously, the expression of Osa-miR159a was reported to be responsive to M. oryzae or its elicitors (Li et al., 2014; Li, Z-Y et al., 2016; Li et al., 2019b). To confirm this conclusion, we examined its expression pattern in susceptible and resistant accessions of rice after inoculation of M. oryzae at the three-leaf seedling stage. The universally susceptible accession Lijiangxin Tuan Heigu (LTH) showed a severedisease phenotype, whereas the accession (IRBLkm-Ts) that contains the gene Pikm, which confers M. oryzae resistance displayed resistance (Fig. 1a). Compared with mock inoculation, M. oryzae infection resulted in decreased accumulation of Osa-miR159a at 12, 24, and 48 hours post-inoculation (hpi) (Fig. 1b). In contrast, Osa-miR159a was significantly upregulated at 24 hpi; Osa-miR159a showed a significant decrease at 48 hpi in IRBLkm-Ts (Fig. 1b), indicating that the response of Osa-miR159a to M. oryzae infection was different in the susceptible and resistant accessions. Therefore, Osa-miR159a may play a role in rice immunity against M. oryzae.
Osa-miR159a Positively Regulates Rice Resistance Against M. oryzae
To explore how Osa-miR159a acts in the interaction between rice and M. oryzae, we made a construct overexpressing Osa-miR159a (OX159) and introduced the construct into Nipponbare (NPB), generating 24 independent transgenic lines, out of which we chose two lines that showed high Osa-miR159a accumulation for further investigation (Fig. 2a). We made a construct expressing STTM of Osa-miR159a (STTM159) and introduced it into NPB, which mayprevent Osa-miR159a from binding to its target sites (Franco-Zorrilla et al., 2007; Todesco et al., 2010). We also selected two independent transgenic lines that showed a significant reduction of Osa-miR159a accumulation for further investigation (Fig. 2a). Then we conducted blast disease assays by punch- or spray- inoculation of the M. oryzae strain GZ8. We found that OX159 lines generated significantly smaller disease lesions than NPB harboring an empty vector (EV) (Fig. 2b and Fig. S1a). Consistently, the lesions from OX159 lines contained significantly less fungal mass and shorter lesion length than the control at 5 days post-inoculation (Fig. 2c, d and Fig. S1b). In contrast, STTM159 lines generated significantly bigger disease lesions than that of the control (Fig. 2e and Fig. S2a), and the lesions from STTM159 lines contained significantly more fungal mass and longer lesions than control at five days post-inoculation (Fig. 2f, g and Fig. S2b, c). These data indicated that Osa-miR159a positively regulated the resistance of rice to rice M. oryzae.
Next, we exploited the GFP-tagged strain GZ8 to observe the infection process in sheath cells using laser scanning confocal microscopy. Compared with the control, our observation found that the infection progress was delayed in OX159 (Fig. S1c, d), but accelerated in STTM159 (Fig. S2d, e). At 24 hpi and 36 hpi, the percentages of invasive hyphae were much lower in OX159 (Fig. S1c, d) compared with the control; however, the percentages of invasive hyphae was greater in STTM159 (Fig. S2d, e). These results indicated that overexpressing Osa-miR159a delayed infection, whereas blocking Osa-miR159a facilitated M. oryzae infection.
To explain why Osa-miR159a positively regulated resistance to M. oryzae, we used RT-qPCR to examine the expressions of some marker genes, including OsNAC4, OsPR10b (Pathogenesis-Related 1b) and OsJAMYB, acting in immune responses after infection of M. oryzae (Park et al., 2012; Pan et al., 2014). The expression of OsNAC4 was higher in OX159 than in the control at 6 and 12 hpi (Fig. S1e), whereas it was lower in STTM159 than in the control at 6 and 24 hpi (Fig. S2f). The expression of OsPR10b was higher in OX159 than in the control at 6 hpi (Fig. S1f); it was lower in STTM159 than in control at 0, 12, and 24 hpi (Fig. S2g). The expression of OsJAMYB was higher in OX159 than in the control at 6 and 12 hpi (Fig. S1g), while it was lower in STTM159 than in the control at 6 and 12 hpi (Fig. S2h). These data indicated that Osa-miR159a activated defense-related genes, positively regulating rice resistance to M. oryzae.
Alteration of Osa-miR159a Accumulation Leads to Defects in Development
In addition to the resistance conferred by Osa-miR159a in rice against M. oryzae, we found that both OX159 and STTM159 showed some altered development and yield traits. All the OX159 and STTM159 transgenic lines were shorter than the control (Fig. 3a, b and Table 1), with STTM159 lines significantly shorter than OX159 lines and the control (Fig. 3b and Table 1). Both OX159 and STTM159 had a lower yield (Table 1). The OX159 lines were sterile and had only a few filled grains on the panicle, leading to straight panicles at the mature stage (Fig. 3a, c and Table 1). The stamen development was deficient in OX159 lines (Fig. 3e, f). In comparison with the control, which had yellowish anthers containing fertile pollen indicated by starch-staining, anthers from OX159 were pale with sterile pollen lacking starch (Fig. 3g). In addition, grains from OX159 lines lacked starch accumulation, although the ovary grew to a size comparable to that of the control (Fig. 3h, i, l, m and Table 1). However, STTM159 showed smaller panicles than that of the control, but the starch accumulation in the grain was normal (Fig. 3d, j, k). STTM159 was also observed to be less productive than the control (Table 1). The grain width of STTM159 was the same as the control, whereas the grain length was shorter than the control (Fig. 3l, m and Table 1). These results indicated that the alteration of Osa-miR159a expression led to defects in growth and development, particularly in pollen and grain development.
Table 1
Yield traits of the wild type, OX159, and STTM159 lines grown in rice paddies
Materials | Plant Height/cm | No. of Tillers | Panicle Length/cm | No. of Filled Gains Per Panicle | Yield Per Plant/g | 1000-grain weight/g | Grain Length/mm | Grain Width/mm |
EV | 95.83 ± 0.58a | 12.80 ± 1.73a | 19.40 ± 0.25a | 1096.33 ± 23.50a | 28.26 ± 0.94a | 25.76 ± 0.49a | 7.19 ± 0.013b | 3.32 ± 0.02a |
OX159-4 | 85.33 ± 2.08b | 12.67 ± 1.53a | 17.65 ± 0.35b | 146.67 ± 106.00d | 6.79 ± 3.53d | 24.30 ± 0.53b | 7.42 ± 0.082a | 3.32 ± 0.02a |
OX159-19 | 87.00 ± 1.00b | 12.00 ± 2.65a | 17.46 ± 0.30b | 34.00 ± 28.48e | 0.82 ± 0.68e | 24.07 ± 0.13b | 7.46 ± 0.050a | 3.32 ± 0.012a |
STTM159-8 | 58.96 ± 5.63c | 8.20 ± 1.30b | 15.74 ± 0.38c | 670.00 ± 22.30c | 12.70 ± 0.60c | 18.97 ± 0.15d | 6.38 ± 0.15c | 3.34 ± 0.0074a |
STTM159-27 | 66.82 ± 1.71c | 11.00 ± 1.87a | 15.49 ± 0.58c | 875.80 ± 68.56b | 19.15 ± 0.75b | 21.90 ± 0.23c | 6.60 ± 0.08c | 3.34 ± 0.0017a |
Alteration of Osa-miR159a Expression Impacts the Expression of Its Target Genes that are Responsive to M. oryzae
Six Osa-miR159 loci in rice generate five mature isoforms that share 18 central nucleotides (Fig. S3a). Among them, Osa-miR159a/b targeted two confirmed genes, namely, OsGAMYB (LOC_Os01g59660) and OsGAMYBL (LOC_Os06g40330) (Li, H et al., 2016), and one predicated gene, LOC_Os10g05230 (encoding a C3HC4-type domain-containing zinc finger protein, herein designated OsZF) (Khan et al., 2017). The target sites in OsGAMYB and OsGAMYBL were in the codon region, whereas the target site was in a 5' untranslated region (UTR) in OsZF (Fig. S3b). To examine how the expression of these genes was impacted by the alteration of Osa-miR159a expression in OX159 and STTM159, we performed a RT-qPCR analysis. As expected, the expression of all three genes was significantly less in OX159 than in the control (Fig. S3c). In contrast, the expression of all these genes was more in STTM159 than in control (Fig. S3c). These data indicated that the overexpression of Osa-miR159a significantly suppressed the expression of its target genes, and the STTM of miR159 prevented the suppression of Osa-miR159a on the expression of its target genes.
Next, we examined the expression of OsGAMYB, OsGAMYBL, and OsZF in LTH and IRBLkm-Ts after infection with M. oryzae. The expression of OsGAMYB and OsGAMYBL was constitutively higher in IRBLkm-Ts than in LTH (0 hpi in Fig. 4a, b). After M. oryzae infection, OsGAMYB was significantly up-regulated at 12 hpi and 24 hpi; it was down-regulated at 48 hpi in LTH. In IRBLKm-Ts OsGAMYB was up-regulated at 12 hpi, but significantly down-regulated at 24 hpi and 48 hpi (Fig. 4a). The expression of OsGAMYBL was relatively stable in both LTH and IRBLKm-Ts with a significant up-regulation at 24 hpi in IRBLKm-Ts (Fig. 4b). OsZF was significantly upregulated at 12 hpi and 24 hpi, but decreased to the level of background expression at 48 hpi in both LTH and IRBLKm-Ts (Fig. 4c). These expression patterns indicated that they were responsive to the infection of M. oryzae in both susceptible and resistant accessions.
Knocking Out OsGAMYB, OsGAMYBL and OsZF Leads to Compromised Susceptibility to M. oryzae
To investigate the function of OsGAMYB, OsGAMYBL, and OsZF, we obtained mutants using CRISPR/Cas9 DNA editing. We identified two independent mutants for OsGAMYBL, one mutant for OsZF, and one independent mutant for OsGAMYB. Among them, gamybl-1 carried a 1-bp insertion resulting in an early stop codon after aa 325 (Fig. 5a). gamybl-2 carried a 1-bp deletion resulting in an early stop codon after aa 311 (Fig. 5a); zf-4 carried a 1-bp insertion resulting in an early stop codon after aa 42 (Fig. 5b). The zf-8 had a 1-bp deletion resulting in an early stop code after aa 32 (Fig. 5b); gamyb-5 carried a 2-bp deletion resulting in an early stop codon after aa 127 (Fig. 5c). The gamyb-10 had a 1-bp insertion resulting in an early stop codon after aa 128 (Fig. 5c). We conducted a M. oryzae assay via punch-inoculation. All the knockout lines significantly decreased the size of M. oryzae lesions that contained significantly less fungal mass and shorter lesions than that of the control (Fig. 5d-f), indicating greater resistance. These results indicated that OsGAMYB, OsGAMYBL, and OsZF contributed to Osa-miR159a-mediated regulation of rice resistance to M. oryzae.