3.1. Phenotypic evaluation of salt tolerance in sitl3
A total of 150 TILLING rice lines were screened to select a salt-insensitive rice mutant, (Cho et al., 2010; Lim et al., 2022; Seo et al., 2022). Among them, one salt tolerance line was additionally selected and then named salt-insensitive TILLING line 3 (sitl3), according to nomenclature of previous studies (Lim et al., 2022, Seo et al., 2022).
Wild-type and sitl3 seedlings were grown in a plastic box for hydroponic cultivation (Kimura B nutrient solution) for 21 days after germination (DAG). No significant differences were observed for either shoot or root length, or fresh weight between WT and sitl3 seedlings at 2, 5, 7, 14, or 21 DAG (Fig. S1A-C). When the agronomic traits were investigated at harvest in untreated seedlings, there were no significant differences among the eight categories; neither was there any difference for yield between the two types of plant under normal conditions (Fig. S2E).
First, salt insensitivity was evaluated at the seedling stage. Seven-day-old WT and sitl3 plants were treated with 150 mM NaCl for 7 days (Fig. 1A), and shoot and root length and fresh weight, chlorophyll and H2O2 contents were measured. When measuring shoot and root length we confirmed that the shoot length of sitl3 was greater than that of the WT in the 150 mM NaCl treatment group but there was no significant difference in root length between the two (Fig. 1B). Similarly, shoot weight of sitl3 seedlings was greater than that of WT seedlings under 150 mM NaCl, but there was no significant difference between them in root weight (Fig. 1C). Chlorophyll content was measured using chlorophyll a, chlorophyll b, and total chlorophyll (a + b). We found no significant differences between WT and sitl3 plants before NaCl treatment. In contrast, under 150 mM NaCl, chlorophyll content, including chlorophyll a and chlorophyll b, was higher in sitl3 plants than in WT plants (Fig. 1D). In turn, no significant difference in H2O2 accumulation was found in the leaf blades, leaf sheaths, and roots between untreated WT and sitl3 seedlings under 150 mM NaCl. In contrast, the same three tissues of silt3 accumulated less H2O2 than those of WT seedlings (Fig. 1E). Additionally, salt insensitivity of the WT and sitl3 plants during the entire growth period in the soil was evaluated (Fig S2). There were no significant differences in shoot height, tiller number, or chlorophyll content between untreated WT and sitl3 plants during vegetative growth (Fig. S2A-C). However, when grown under 0.3 dS/m NaCl treatment, sitl3 plants showed a significant difference in chlorophyll content (Fig. S2C) but not in shoot height or tiller number (Fig. 2B). In turn, differences in culm length, panicle length, number of panicles per hill, number of spikelets per panicle, total number of spikelets, maturity rate, total grain weight and 1,000 grain weight were measured between WT and sitl3 under 0.3 dS/m NaCl at the reproductive and harvest stages. We found that the 1,000 grain weight and total grain weight of sitl3 were significantly greater than those of the WT (Fig. S2D, F).
Germination rates were measured for 5 d under NaCl and ABA treatment conditions. In the NaCl treatment, sitl3 seeds exhibited significantly higher germination rates than the WT under 100 to 200 mM NaCl (Fig. S3A). Interestingly, the seed germination rate of sitl3 seeds was significantly lower at 5 d after treatment with 3 µM ABA and at 2 d after treatment with 5 µM ABA; whereas no significant difference was found for germination rate between the two when treated at 2 µM ABA (Fig. S3B).
Ionomic analysis was performed to investigate the accumulation of Na+ and K+ under salt stress. The Na+ contents in WT and sitl3 were not significantly different between shoots and roots before treatment. However, after 150 mM NaCl treatment, the accumulated Na+ in sitl3 was 0.65-fold in shoots and 0.84-fold in roots relative to WT (Fig. 1F). However, there was no difference in K+ content between control and treatment conditions (Fig. 1G).
K, Na, P, and Ca contents were measured in leaf blades, leaf sheaths, and roots of seven-day-old WT and sitl3 plants for 7 d in untreated and 150 mM NaCl-treated plants (Fig. S4). K contents in the three tissues did not significantly differ between WT and sitl3 plants under control or saline conditions (Fig. S4A). In contrast, Na contents were higher in the three tissues of silt3 than in those of the WT (Fig. S4B). In contrast, although there was no difference in P between WT and sitl3 before treatment, it was significantly lower in sitl3 than in WT roots and leaf bladesࣧbut not in leaf sheathsࣧunder salt stress (Fig. S4C). In turn, Ca contents did not differ in untreated nor in salt-treated groups between WT and sitl3 plants, except for a lower content in sitl3 leaf blades than that in WT plants under salt stress (Fig. S4D).
3.2. Whole genome re-sequencing and selection of candidate gene corresponding to salinity tolerance
In a previous study, whole-genome resequencing was performed to identify mutations in 150 core TILLING lines, including sitl3 (Hwang et al., 2020). Genetic mutations were detected in sitl3 to compare genetic polymorphisms with those in the WT genome. In all, 167,646,640 of the 196,179,185 WT reads (84.46%) and 37,158,152 of 43,603,544 sitl3 reads (85.22%) were mapped to the reference genome ( i.e., Oryza sativa cv. Nipponbare) (Table. S3). Genetic variants of 82 SNPs and 53 indels were detected in the sitl3 genome after manual evaluation of each variant individually (Fig. 2A). Further, SNPs and indels were distributed over 12 chromosomes and homozygous variants were mapped to 12 chromosomes (Fig. 2B). Mutations were classified according to their location on the genome, and the degree of their impact on the genome was predicted and evaluated. Among 82 SNPs and 53 indels, 17 SNPs (20.73%) and 7 indels (13.20%) were observed in genic regions (Table S4-5). Additionally, in the genic region, 10 of the 17 SNPs (58.82%) and one of the 7 indels (14.28%) occurred in the CDS region. SNPs had eight missense variants, one synonymous variant, and one stop gain (Fig. 2C), whereas indels had one frameshift variant (Table. S5). Both mutations had a high impact on genetic variation, with SNP and indels. One mutated gene with the SNP encoded a PWWP domain-containing protein (LOC_Os07g46180), and the other with indel was annotated retrotransposons (Fig. 2C and Table. S5). Therefore, the gene encoding the PWWP domain-containing protein was named Oryza sativa PWWP4 (OsPWWP4), in accordance to the nomenclature of a previous report (Guo et al., 2022), and was selected as a candidate for the salt tolerance phenotype of sitl3 for further study.
3.3. Molecular characterization of OsPWWP4 and OsPWWP4p.Gly462*
Further studies were performed to confirm the SNP of OsPWWP4 in sitl3. First, the reference genomic DNA and CDS sequences were retrieved from an internet website (phytozome 13, https://phytozome-next.jgi.doe.gov/). OsPWWP4 is located on chromosome 7 and consists of six exons and five introns with 11,146 nucleotides long of genomic DNA and 4,344 nucleotides long of CDS (Fig. 3A). Sanger sequencing confirmed that the WT gene had the same sequences as the reference gene from gDNA and CDS. In contrast, guanine in the mutant gene of gDNA and CDS at positions 4464th and 1384th was changed to thymine, resulting in premature termination (Fig. 3B-C). Therefore, the mutant gene with only one PWWP domain was named OsPWWP4p.Gly462* according to its nomenclature, although the WT gene consisted of two PWWP domains (Fig. 3B).
3.4. Interaction of OsPWWP4 or OsPWWP4p.Gly462* with OsEUL families
To identify partner proteins (s) of OsPWWP4, we performed yeast two-hybrid screening using a salt-treated rice cDNA library. As a result of this screening, we found 10 blue colonies on DDO/X/A media, followed by patching on QDO/X/A, and finally selected one colony (Fig. S5A). The colony was BLASTed, and five candidate genes belonging to the EUL family (OsEULS2, OsEULS3, OsEULD1a, OsEULD1b, and OsEULD2) were highly matched (Fig. S5B, C). The EUL family consists of five genes in the rice genome and is divided according to its structure, including the genuine EUL domain and inter-domain linker (Fouquaert et al., 2012) (Fig. 4A). In addition, the phylogenetic tree based on the amino acid sequences revealed three subgroups: OsEULD1a-OsEULD2, OsEULS2-OsEULS3, and OsEULD1b (Fig. 4B). Each gene of the EUL family was cloned into the pGADT7-AD vector, and a co-transformation experiment was performed with OsPWWP4 or OsPWWP4p.Gly462*. Five proteins co-transformations of OsPWWP4 with each of the three proteins (OsEULS2, OsEULS3, and OsEULD2) grew well in QDO/X/A medium. In contrast, no co-transformation of OsPWWP4p.Gly462* with each of the five proteins was observed (Fig. 4C). The positive control (P53 + T) but not the negative control (Lam + T) grew well (Fig. 4C).
3.5. Subcellular localization and bimolecular fluorescence complementation assays
To investigate the subcellular localization of OsPWWP4, OsPWWP4p.Gly462* and their partner proteins (OsEULS2, OsEULS3, and OsEULD2), these genes were expressed in rice protoplasts. The positive control, 35S:: sGFP, was highly expressed in the cytosol. In turn, 35S::sGFP-OsPWWP4 and 35S::sGFP-OsPWWP4p.Gly462*, and OsMeCP-DsRed2, were highly expressed in the nucleus (Kim et al., 2019). The partner genes 35S::sGFP-OsEULS2, 35S::sGFP-OsEULS3, and 35S::sGFP-OsEULD2 were expressed in the nucleus and cytosol (Fig. 5A). Rice protoplasts were co-infected with OsPWWP4, OsEULS2, OsEULS3, and OsEULD2), and with OsPWWP4p.Gly462*, EULS2, EULS3, and EULD2 for the BiFC assay. When 35S::YFPn + 35S::YFPc co-infection was used as the negative control, no signal was obtained. Interestingly, all six combinations of OsPWWP4 or OsPWWP4p.Gly462*, and each of the three partner proteins, showed strong yellow fluorescence signals in the nucleus (Fig. 5B).
3.5. In vitro pull-down assay and methyltransferase activity of OsPWWP4 and the mutant protein
To verify interactions of OsPWWP4 or OsPWWP4p.Gly462* with each of the partners (OsEULS2, OsEULS3, and OsEULD2), we constructed each of the His-tagged OsPWWP4 or OsPWWP4p.Gly462* protein followed by constructing each of the MBP-tagged OsEULS2, OsEULS3, or OsEULD2 protein. In vitro pull-down assays showed strong interactions between OsPWWP4 and OsEULS2, and between OsEULS3, and OsEULD2, resulting in the detection of MBP-tagged OsEULS2, OsEULS3, and OsEULD2 proteins pulled down by His-tagged OsPWWP4 using HisPur™ Cobalt resin (Fig. 6A). Interestingly, another in vitro pull-down assay of OsPWWP4p.Gly462* and each of the three partner proteins confirmed all the interactions between the mutant protein and each protein (Fig. 6B).
The PWWP domain was expected to possess methyltransferase activity (Qiu et al., 2002; Sankaran et al., 2016). To confirm the presence of OsPWWP4 and OsPWWP4p.Gly462* with methyltransferase activity towards OsEULS2, OsEULS3, and OsEULD2, the in vitro methyltransferase assay was performed. The expected sizes of the OsPWWP4 and OsPWWP4p.Gly462* recombinants were 163 and 55 kDa including 5 kDa of the pET-28a (+) vector using a prediction program (https://www.bioinformatics.org/), respectively. Additionally, the three partner proteins OsEULS2, OsEULS3, and OsEULD2 were expected to be 66, 76, and 88 kDa, respectively, including the 43 kDa of the pMAL vector. Expressed sizes of purified OsPWWP4 and OsPWWP4p.Gly462*, and OsEULS2, OsEULS3, and OsEULD2 proteins were confirmed using western blotting analysis, finding all same sizes, as expected (Fig. S6A).
An in vitro methyltransferase assay was performed using the purified protein (Fig. S6B-C). The protein combinations were OsPWWP4 and each of OsEULS2, OsEULS3, OsEULD2, and OsPWWP4p.Gly462* and EULS2, EULS3, and EULD2, including positive and negative controls. Unfortunately, despite the use of several vectors and expression systems, the high-molecular-weight protein OsPWWP4 (163 kDa), including the 5 kDa protein of pET-28a(+), showed severe fragmentation during protein purification. The other proteins, including the mutant protein, were expressed and purified without fragmentation (Fig. S6A). Optical density (OD) was measured using a spectrometer at 30 s intervals over 45 min after incubation. The OD value gradually increased until 45 min and, finally, OD values of OsPWWP4 and each of OsEULS2, OsEULS3, and OsEULD2 values were 0.288, 0.261, and 0.295, and values of OsPWWP4p.Gly462* and each of EULS2, EULS3, and EULD2 were 0.279, 0.297, and 0.292, respectively, resulting in no significant difference between OsPWWP4 and the mutant protein (Fig. S6B-C). The OD value of the negative control was 0.156, while that of the positive control was 0.222 (Fig. S6B-C).
3.7. Relative expression level of OsPWWP4, and OsEUL genes
Next, we examined the relative expression levels of OsPWWP4 as well as the mutant gene and partner genes in WT and sitl3 plants under salt stress and ABA treatment. Then, expression levels of the genes were measured at 0, 1, 3, 6, 12, and 24 h under 150 mM NaCl (Fig. S7A) or 100 µM ABA treatment (Fig. S7B) in WT and sitl3. Under saline conditions, the expression levels of OsPWWP4 and its mutant genes increased in the shoots during 6 h and then decreased after 12 h in both WT and sitl3 plants. Interestingly, the expression levels were significantly lower in shoots of sitl3 than in those of WT plants at 6–24 h after treatment, although there was no difference between non-saline and early salt treatments. In the roots, the expression levels of the sitl3, OsPWWP4, mutant gene were significantly lower than those of the WT from 6 to 24 h after salt treatment, under which, the relative expression of the EUL family genes, such as OsEULS2, OsEULS3, and OsEULD2, was significantly higher in the shoots and roots of sitl3 than in those of WT plants, at least one or more sampling time points, particularly at 6 to 12 h after treatment.
Additionally, under the ABA treatment, the relative expression of OsPWWP4 gradually increased after 24 h. However, the OsPWWP4 mutant gene of sitl3 showed a lower level of expression than OsPWWP4 in the WT from 12 to 24 h after salt treatment. In contrast, partner genes showed higher expression in sitl3 than in WT from 3 to 24 h (OsEULS2 and OsEULS3) or from 1 to 24 h (OsEULD2) after salt treatment.
3.8. Relative expression level of transcriptional factors related to EUL family genes
To further verify the molecular regulation of OsPWWP4 and EUL family gene expression during the plant response to salinity, the expression of transcription factors related to partner genes was measured in both WT and sitl3 plants under salt stress. First, upstream sequences (2 kb long) of the three partner genes were scanned with a website program, PlantPAN4.0 (Chow et al., 2019) to detect transcription binding sites and the corresponding transcription factors. We retrieved overlapping transcription factors that were predicted to exist in the promoters of all three partner genes (Fig. S8, Table S6-S8). Among them, five WRKYs, four NACs, and two bZIPs transcription factors related to salt stress were randomly selected based on previous reports, and their expression levels were evaluated in leaf blades, leaf sheaths, and roots of both one-week-old plants under non-treatment and 150 mM NaCl for 7 d (Fig. 7). Five WRKY transcription factors, OsWRKY8, OsWRKY23, OsWRKY45, OsWRK72, and OsWRKY89 were selected, and their expression was significantly increased in leaf blades and roots or in all three all tissues of sitl3 relative to those of WT plants under salinity. However, there were no significant differences in the expression of the genes tested in any of the three tissues between the WT and sitl3 untreated controls, except for OsWRKY8 in the leaf blades and OsWRKY45 in the roots of sitl3 plants.
Expression analysis of four NAC genes, OsNAC4, OsNAC5, OsNAC10, and OsNAC58, revealed significant differences in expression levels in at least one or two tissues between WT and sitl3 plants, resulting in higher levels of OsNAC4, OsNAC5, and OsNAC58 in sitl3 than in WT plants under salt stress; however, the expression of OsNAC10 was higher in the roots of WT plants. No significant difference was found between the tissues of both plant types under control conditions, except for OsNAC4 in the leaf blades of sitl3. Among the two bZIP transcription factors, the expression of OsbZIP23 was higher in roots and leaf blades but not in leaf sheaths under salt stress, while no significant changes were observed in any tissue under control conditions. Lastly, there was no difference in expression of OsbZIP72 between WT and sitl3 plants in any of the three tissues studied under control or salinity conditions.
The expression of OsPWWP4 (or the mutant gene) and its three partner genes were also examined in the same tissues. Expression of OsPWWP4 was higher in the three tissues of WT than those in the mutant gene in sitl3 under salinity conditions, whereas no difference was observed in any of the three tissues under control conditions. Interestingly, the expression of the three partner genes was significantly higher in sitl3 than in WT plants under salt stress; however, no difference was found between the two plant types under control conditions, except in the roots of OsEULS2.
3.9. Relative expression level of salt stress-related genes
In addition, the expression levels of genes related to the plant responses to salinity were evaluated using the same samples (Fig. 8). There was no significant difference in the expression levels of three OsSOS1-3 genes (Xiang et al., 2007) between WT and sitl3 plants before treatment, except in the roots of OsSOS3. However, the expression levels after treatment increased and were higher in sitl3 than in the WT plant tissues, except for the roots of OsSOS2 and OsSOS3 under salinity. The expression pattern of the OsCIPK15 gene, a protein kinase that interacts with OsSOS3 (Kanwar et al., 2014) was similar to that of OsSOS2.
As for OsTPC1, which imports calcium into the cytosol (Kurusu et al., 2012), it showed higher expression in the leaf sheaths and roots of sitl3 than in those of the WT, before treatment; however, no difference was found between the two plant types under salt stress. The expression of OsMSR2, which responds to Ca2+ signaling and the activity OsNHX1 Na+ transport (Xu et al., 2013), was higher in the three tissues of sitl3 than in those of the WT, whereas no significant difference was detected under control conditions, except in the roots. Expression of OsNHX1 (Chen et al., 2007) showed no significant differences in the leaf blades and sheaths before treatment; however, higher expression was observed in both tissues in sitl3 than in the WT. OsCAX1, which exchanges Ca2+ and H+ in the tonoplast (Kamiya 2006), showed no difference between both plants in expression, in any of the three tissues studied, except for leaf blades under saline conditions.