Two set of synthetic allohexaploid wheat (SCAUP/SQ523, DM4/Y199) and their tetraploid (SCAUP, DM4) and diploid (SQ523, Y199) progenitors, 48 natural wheat accessions with different ploidy levels ( Table S1 ) were used to examine the difference of salt tolerance between different ploidy wheats and to explore the expression pattern of TaSOS1 before and after salt treatment. wheat plants were grown under 16/8 h, light/dark conditions at a temperature of 22°C. wheat variety Nongda3338 was used to clone the three CDS and promoter sequences of TaSOS1 homoeologous genes and analyze the expression pattern of three TaSOS1 homoeologous genes under normal or salt condition.

Arabidopsis (Arabidopsis thaliana) wild-type Columbia was used as the recipient of the overexpression and promoter analysis of pSOS1:uidA reporters transformation. Sterilized seeds were incubated at 4°C for 3 d then sowed on MS plates containing 1% (w/v) Suc and 0.8% (w/v) agar. Seedlings were grown under a 16-h-light/8-h-dark cycle at 22°C in a growth room.

For salt tolerance comparisons, seedlings of the synthetic allohexaploid wheat genotype (SCAUP/SQ523), and their respective tetraploid and diploid parents, 48 natural wheat accessions with different ploidy levels were grown in vermiculite and watered with half strength Hoagland’s liquid medium (pH 6.0). The seedlings were treated with 0 or 200 mM NaCl at three-leaf stage for 10 d. Then the chlorophyll content, shoot fresh weight, root length and root fresh weight were quantified and photos were taken. Root length was quantified by measurement with a ruler and the fresh mass of the shoots and roots were measured using an automated electronic scale. Chlorophyll were measured by aqueous acetone (80%) described earlier.

For Arabidopsis germination rate analysis, Seeds were sown on MS medium supplemented with 0, 100, 200 mM NaCl, and germination rates were scored every 24 hours. For the salt tolerance assay of plants grown in soil, 5-d-old seedlings were transferred from MS medium to soil. The seedling of wild type and transgenic lines plants grown in soil pots were irrigated with 250 mM NaCl every 5 d, then the chlorophyll content, shoot fresh weight and the leaves electrolyte leakage of each genotype were evaluated after 15 d. For whole growth duration salt tolerance assay, seedlings plants were irrigated every 5 d with or without the addition of 200 mM NaCl until seed maturation.

The genomic DNA of wheat genotype Nongda3338 was extracted using the CTAB protocol. Total RNA was extracted with TRIzol reagent (Invitrogen), and purified RNA was treated with DNase I. Subsequently, First-strand cDNA synthesis was performed using HiScipt II One Step RT-PCR Kit (Vazyme Biotech, Nanjing, China).

Real-time qRT-PCR analysis was conducted using SYBR Premix EX TagTM (TaKaRa, Dalian, China) in a volume of 20μl in a Bio-Rad CFX96 real-time PCR detection system3. The PCR parameters were as follows: 94°C for 3 min, followed by 40 cycles of 94°C for 15 s, 60°C for 20 s, and 72°C for 20 s. Each sample was quantified at least in triplicate and normalized using Actin as an internal control.

Specific primer sets were designed to obtain the CDS and promoter sequences of TaSOS1-A, TaSOS1-B and TaSOS1-D. The PCR was performed using the Tks Gflex DNA Polymerase (TaRaKa, Beijing, China) and the PCR products were recovered by TIANgel Midi Purification Kit (TIANGEN BIOTECH, Beijing, China) then cloned into the pEASY-Blunt Simple Cloning vector (TransGen Biotech, Beijing, China) and sequenced. Amplified fragments of the full-length coding sequence of TaSOS1-A, TaSOS1-B and TaSOS1-D were cloned into the binary expression vector pCAMBIA1300 (driven by the CaMV 35S promoter). The promoter region of target genes that contains about 2-kb fragment upstream of the coding sequence was fused to the reporter gene encoding GUS and was then cloned into the binary expression vector pCAMBIA1300 to generate reporters. All the constructs were constructed by In-Fusion PCR Cloning Kits (TaRaKa, Beijing, China) and were confirmed by sequencing. These vectors were transferred into Agrobacterium tumefaciens strain GV3101 and transgenic plants were identified on Murashige and Skoog medium containing hygromycin. The primers employed for plasmid construction are listed in Table S2 .

Two-week-old transgenic Arabidopsis seedlings were transferred to MS medium (1% Suc) with 0 or 200 mM NaCl, and after 24h whole seedlings or different tissues expressing the GUS gene were immersed in staining buffer ((100 mM NaPO₄, 0.5 mM K₃[Fe(CN)₆, 0.5 mM K₄[Fe(CN)₆], 10 mM EDTA, 1 mg/ml 5-bromo-4-chloro-3-indolyl-b-D-glucuronide (X-gluc)) at 37°C in the dark for overnight. Subsequently, samples were washed in 75% ethanol and the intensity of color development in different tissues was monitored and photographed.

For measurement of Na⁺ and K⁺ contents in the seedling leaves, 10-day-old uniform seedlings were transferred to MS plates with 100 mM NaCl solution. 2 days later, the samples were harvested and dried, then dissolved in HNO₃ and analyzed by inductively coupled plasma optical emission spectrometry (FP6410, INESA).

In order to identify TaSOS1 homoeologous genes in wheat, we conducted a BLAST search against the wheat genome database (IWGSC RefSeq v1.0) using rice and barley SOS1 sequences as queries. After the redundancy elimination and domain confirmation, the three sets of wheat TaSOS1 homoeologs on chromosome 3AS, 3BS and 3DS were obtained and these homologous alleles were designated as TaSOS1-A, TaSOS1-B, or TaSOS1-D according to their chromosome location ( Figure 1A ). A sequence comparison between cDNA and genomic DNA revealed that TaSOS1 has 23 exons and 22 introns ( Figure S1 ). The coding sequences (CDS) of the three TaSOS1 homoeologs were comparatively conserved, and can be distinguished from one another by series of single nucleotide polymorphisms (SNPs). In contrast with the CDS, differences in molecular size were found in introns and in upstream/downstream regions. Deduced amino acid sequence alignments showed that TaSOS1-A, TaSOS1-B and TaSOS1-D encode 1142 amino acid residues and three TaSOS1 homoeologs shared 98.16% sequence similarity. The amino acid sequences differed from one another at 49 positions ( Figure 1B ). In addition, similar to its orthologues from Arabidopsis, the predicted proteins of TaSOS1 homoeologs contain 12 transmembrane domains (TM) in N-terminus and all of them are conserved in three TaSOS1 homoeologs, except that a single amino acid substitution in the first predicted membrane-spanning region of TaSOS1-D (Ala-Gly). Previously analysis in Arabidopsis showed that AtSOS1 protein contains three pivotal and highly conserved consensus motifs, including CNBD activation domain in N-terminus, auto-inhibitory domain and phosphorylation site in C-terminus. The alignments revealed the phosphorylation site and the auto-inhibitory domain highly conserved at the C-terminus of three TaSOS1 homoeologs ( Figure 1B ). However, the sequence of CNBD activation domain, a conserved domain interacts with auto-inhibitory domain to keep the SOS1 protein with basal activity under control condition, show differences between TaSOS1-D and TaSOS1-A, TaSOS1-B ( Figure 1B ).

The spatial and temporal expression patterns of TaSOS1 under normal and salt stress condition were first analyzed in a wheat variety Nongda3338 by quantitative real-time PCR (qRT-PCR). The results indicated the transcription abundance of TaSOS1 were improved after salt treatment but the transcription patterns were different in seedling shoot and root. In root, the expression of TaSOS1 was induced quickly and peaked at 9 h after salt treatment then the level of mRNA accumulation gradually declined ( Figure 2A ). However, the mRNA accumulation remains almost unchanged in shoot until 48h after salt condition ( Figure 2B ).

Hexaploid bread wheat has a greater discriminating capacity between Na⁺ and K⁺ than durum wheat under salt stress and is generally more salt tolerant. In order to assess the salt tolerance variation after the formation of allopolyploid wheat, two set of synthetic hexaploids (SCAUP/SQ523 and DM4/Y199, BBAADD) with their tetraploid (SCAUP and DM4, BBAA) and diploid Ae. tauschii (SQ523 and Y199, DD) parents were used for salt tolerance assessment. The root length, root and shoot fresh weight for each of the three genotypes under normal and salt stress conditions were recorded and quantified. Consistent with previous study (Zheng et al., 2021), the results indicated the synthetic hexaploid wheats and diploid Ae. tauschii exhibited superior growth vigour relative to the tetraploid lines ( Figures 3A-E ; Figure S2 ). We further analyzed the expression patterns of TaSOS1 gene in two synthetic hexaploids with their tetraploid and diploid parents using qRT-PCR. Consistent with our expectation, the TaSOS1 genes were up-regulated in synthetic hexaploid wheats (SCAUP/SQ523, DM4/Y199) compared with their tetraploid parents (SCAUP, DM4) both in seedling roots and shoots upon NaCl treatment. Notably, the expression of TaSOS1 was upregulated more significantly in diploid (SQ523, Y199) parents ( Figures 3F-I ). Thus, we examined whether the expression pattern of TaSOS1 is correlative to the variation in the salt tolerance (scored root length as a growth indicator and compared the changes before and after salt treatment) in 48 natural wheat accessions with varying ploidy, including 16 allohexaploid wheats, 16 allotetraploid and 16 diploid species Ae. tauschiis ( Figures 4A, B ; Table S1 ). Across the three different ploidy levels, the fold change of TaSOS1 upregulation expression was positively correlated with the root length under salt treatment of the natural wheat accessions (BBAADD, BBAA and DD; R² = 0.613, P < 0.01; Figure 4C ). In addition, we also performed the correlation analysis between the TaSOS1 expression pattern with the change of root and shoot fresh weight before and after salt treatment in 48 natural wheat accessions, these results also indicated the fold change of TaSOS1 up-regulation expression was positively correlated with salt tolerance in wheat accessions of different ploidy ( Figures S2A-E ).

To further investigate how the expressions of TaSOS1 homoeologous genes were regulated in allopolyploid wheat under salt stress, gene-specific PCR primers were designed and the amplification efficiency of the primers was confirmed by homoeolog-specific genomic PCR ( Figure S3A ). Considering the expression of TaSOS1 peaked at 9 h in root and 2 d after salt stress in shoot, we coupled quantitative RT-PCR to quantity the TaSOS1-A, TaSOS1-B and TaSOS1-D subgenome specific expression at these specific time points after salt stress in root and shoot, respectively. Interestingly, different expression patterns were found between the three TaSOS1 homoeologous genes. In root, NaCl treatment increased the transcription level by more than 12 times of TaSOS1-D and 2.7-, 6.4-fold enhancements were observed of the amount of TaSOS1-A and TaSOS1-B ( Figure 5A ). However, in shoot, the amount of TaSOS1-B mRNA was much higher than TaSOS1-A and TaSOS1-D, while the transcription level of TaSOS1-D was the lowest in both control and salt condition ( Figure 5B ). Taken together, these results showed that these homoeologs may play versatile roles in different organs under salt stress condition.

To gain further insight into the transcriptional divergence of three TaSOS1 homoeologous genes in normal and salt treatment conditions, we generated transgenic Arabidopsis plants expressing the β-glucuronidase (GUS) gene (uidA) under the control of the TaSOS1-A, TaSOS1-B and TaSOS1-D promoter, respectively. About 2 kb genomic DNA fragments which included the sequence upstream of transcription initiation site of three TaSOS1 homoeologs were amplified, respectively. The reporter constructs using uidA fused to the promoter sequence of TaSOS1-A, TaSOS1-B and TaSOS1-D (pTaSOS1-A:GUS, pTaSOS1-B:GUS, pTaSOS1-D:GUS) were transformed into the wild type Arabidopsis (Columbia ecotype) ( Figure S3B ). The construct was introduced via Agrobacterium-mediated transformation by the floral-dip method. For each transgenic assay, three independent T₃ homozygous lines carrying the construct for a single insertion were selected and used for further characterization ( Figure S3C ). The GUS activity of transgenic Arabidopsis plants carrying pTaSOS1-A:GUS, pTaSOS1-B:GUS and pTaSOS1-D:GUS were examined under normal or salt conditions by histochemical staining ( Figures 5C, D ). In normal conditions, the GUS activity can be detected in roots, stems and leaves of pTaSOS1-A:GUS and pTaSOS1-B:GUS seedling plants, but undetectable in leaves and slightly in root of pTaSOS1-D:GUS lines ( Figures 5C, D ). These differential expression pattern between the three promoters are consistent with our qRT-PCR results in wheat. To assess the expression of TaSOS1s under salt stress conditions, transgenic Arabidopsis seedlings were subjected to 200 mM NaCl treatment for 1 d, after which the seedlings were stained. The results showed that NaCl induced deeper blue staining in root of pTaSOS1-B:GUS than pTaSOS1-A:GUS and pTaSOS1-D:GUS lines, but the improvement ratio of GUS activity of TaSOS1-D in root was more than TaSOS1-A and TaSOS1-B ( Figure 5C ). Meanwhile, the histochemical staining showed that the activities of the three wheat TaSOS1 promoters in leaves under salt stress almost unchanged or a bit higher than untreated controls ( Figure 5D ). Collectively, our results, on the one hand, were consistent with our qRT-PCR results showing that the transcription patterns of TaSOS1-A, TaSOS1-B and TaSOS1-D existed discrepancies under normal and salt treatment. On the other hand, the results indicated that TaSOS1-A and TaSOS1-B showed high basal expression in roots and leaves in normal condition and they can further up-regulation in roots under salt stress. While TaSOS1-D showed markedly lower expression in both roots and leaves under normal conditions, but showed significant up-regulation in roots but not leaves under salt stress.

To explore the molecular basis of the differential expression pattern of three TaSOS1 homoeologous genes, cis-acting elements analysis for promoters of three TaSOS1 homoeologs was carried out through web-based PLANTCARE and PLACE databases (Lescot et al., 2002). Results showed that there are 23 cis-acting elements owned by all the three TaSOS1 promoters and several of them involved in the activation of defense-related genes, including predicted ABREs, which are known to be involved in abscisic acid responsiveness. TGACG-motif, CGTCA-motif, which are involved in jasmonic acid methyl ester (MeJA) responses. AuxRR-core, which is associate with auxin responsiveness. Further, several elements involved in light responsiveness such as G-box, ATC-motif, and CAT-box, which is related to meristem expression were also present in the promoters of three TaSOS1 homoeologs ( Figure S4 ). Moreover, there are several types of cis-acting elements were detected only in one or two promoter regions of three TaSOS1 genes, such as EIRE and O2-site motif only presented in TaSOS1-A; while ACE, ATCT-motif, CATT-motif, HSE and TC-rich repeats are presented both in promoters of TaSOS1-A and TaSOS1-B, but absent in TaSOS1-D. Interestingly, we also found the P-box, AP-2-like and WUN-motif are specifically exhibited in promoter of TaSOS1-D. These results suggest there are genetic variations in cis-regulatory elements or regulatory regions in the three TaSOS1 homoeologs and may be involved in gene expression changes that impact phenotype.

To further understand the functions of three TaSOS1 homoeologs in wheat, ectopic expression of TaSOS1-A, TaSOS1-B and TaSOS1-D were performed using Arabidopsis transgenic system. For each transgenic line, three independent T₃ homozygous overexpression lines carrying the construct for a single insertion were selected and used for further characterization. RT-PCR analysis showed that the TaSOS1 were expressed in all the transgenic Arabidopsis plants, respectively, but was absent in the wild-type ( Figure S5A ). We investigated the effects of three TaSOS1 homoeologs overexpression on the morphology and growth of transgenic plants under normal and salt conditions. Germination rates of TaSOS1-overexpressing and wild type Arabidopsis seeds were first compared. In normal conditions, the transgenic lines of 35S:TaSOS1-A, 35S:TaSOS1-B and 35S:TaSOS1-D and wild type plants showed similar germination rates ( Figure 6A ). In MS medium with 100 mM NaCl, three TaSOS1 homoeologs transgenic lines exhibited higher germination rates than wild type plants ( Figure 6A ). Notably, the transgenic lines of 35S:TaSOS1-D displayed more salt tolerant phenotype than 35S:TaSOS1-A and 35S:TaSOS1-B plants. The germination rate for 35S:TaSOS1-D lines were nearly 60% in 48h under salt stress treatment, but only 20% for 35S:TaSOS1-A and 41% for 35S:TaSOS1-B lines during the same time period ( Figure 6B ). More significantly, when sowed in MS medium with 200 mM NaCl, the seed germination of 35S:TaSOS1-D transgenic lines were significantly higher than that of the 35S:TaSOS1-A, 35S:TaSOS1-B and wild type plants, with more than 30% of the seeds of the 35S:TaSOS1-D transgenic lines germinating in the 84h after sowing, while less 6% of the 35S:TaSOS1-A, 35S:TaSOS1-B transgenic lines and wild type seeds germinated at the same treatment ( Figures S5B, C ). We further determined the seedlings growth of the wild type and the three TaSOS1 homoeologs transgenic plants and found that the growth performance of transgenic lines was similar to that of wild type plants when grown in soil pots in non-saline conditions ( Figure 6C ). When irrigated with 200 mM NaCl for 15 d, three TaSOS1 homoeologs transgenic lines exhibited a salt tolerant phenotype compared with wild type plants ( Figure 6C ). We scored the shoot fresh weight and compared the changes between transgenic and wild type plants before and after salt treatment. Overall, NaCl had a lower inhibitory effect on TaSOS1 overexpression lines than it had on the wild type plants. The shoot fresh weight of wild type was 38% of the control weight, while the shoot fresh weight of TaSOS1-B and TaSOS1-D overexpression lines was more than 70% of the control weight ( Figure 6D ). We also determined the relative electrolyte leakage (REL) of wild type, 35S:TaSOS1-A, 35S:TaSOS1-B and 35S:TaSOS1-D plants after salt stress and found that the REL of the TaSOS1-overexpression lines was significantly lower than that of the wild type ( Figure 6E ). Since low Na⁺ content and high K⁺/Na⁺ ratio in cells are crucial for salt tolerance, we further compared the Na⁺, K⁺ content and K⁺/Na⁺ ratio in leaves of three TaSOS1 homoeologs transgenic lines and wild type under salt stress conditions. As shown in Figures 6F-H , generally, the TaSOS1 overexpression lines keep less Na⁺, more K⁺ content and higher K⁺/Na⁺ ratio than wild type and among them, 35S:TaSOS1-D showed lower Na⁺ content and higher K⁺/Na⁺ than the 35S:TaSOS1-A, 35S:TaSOS1-B transgenic lines. Furthermore, all TaSOS1 homoeologs transgenic lines exhibited a salt tolerant phenotype compared with wild type plants in the reproductive growth stage. Among them, 35S:TaSOS1-B and 35S:TaSOS1-D showed much less growth retardation than the 35S:TaSOS1-A transgenic lines and wild type plants, including fresh weight, silique number and plant height ( Figures 7A-C ; Figure S5D ). Thus, the results demonstrate that three TaSOS1 homoeologous genes display different degree of contribution to salt tolerance at different developmental stages and among of them, TaSOS1-A contributes the least to salt tolerance in all developmental stages, while TaSOS1-D plays the prominent role in germination and seedling stage and both TaSOS1-B and TaSOS1-D exhibit a higher contribution in adult stage under salt stress.