Identification and fine mapping of a novel qGR6.2 locus controlling rice salt tolerance during seed germination

The rice growth is strongly affected by salt stress. When exposed to high salt stress, seed germination and seedling establishment are significantly inhibited. Particularly, with the promotion of rice direct-seeding technology in Asia, improving rice salt tolerance during seed germination is of great importance for rice breeding. In this study, an indica rice landrace Wujiaozhan (WJZ) showed significant salt tolerance during seed germination. The dynamic process of seed germination at 300 mM NaCl was observed via the germination rate (GR), seedling percentage (SP), and germination index (GI) from the progeny by crossing WJZ with the salt-sensitive Nipponbare (Nip). By QTL mapping, sixteen quantitative trait loci (QTLs) related to salt-tolerance were detected via a BC1F2 population, including six QTLs for the GR, eight QTLs for the SP, and two QTLs for the GI. Of them, four QTLs, qGR6.1, qGR6.2, qGR10 and qSP10.1, were expressed continuously, especially qGR6.2 and qGR10. Furthermore, twelve significant associated simple sequence repeats (SSR) markers were detected in the F2 population by bulked segregant analysis (BSA). Four SSR markers (RM588, RM190, RM276 and RM5493) were closed to qGR6.1, qGR6.2, qGR6.3, qSP6.1, qSP6.2, qGI6 and qSP8.1 identified by QTL mapping. The major qGR6.2 on chromosome 6 was confirmed via the BC2F2 population, which explained more than 20% phenotype variation of the GR. Fine mapping results displayed that qGR6.2 was narrowed down to a 65.9 kb region between the Z654 and Z619 markers, with eleven candidate genes predicted. Based on the microarray database, six genes (LOC_Os06g10650, LOC_Os06g10660, LOC_Os06g10690, In our study, we identified 16 QTLs and 12 associated SSR markers conferring salt tolerance from indica landrace WJZ during seed germination, and the major qGR6.2 was fine mapped to a 65.9 kb region flanked by marker Z654 and Z619. Our results provide new information on the genetic basis of improving salt tolerance during seed germination by marker-assisted selection.


BSA methods.
Salt tolerance is a polygenic characteristic highly influenced by the environment and multiple genes (Johnson et al. 1992;Wang et al. 2012b), which makes it difficult to identify the QTLs and genes. Up to now, more than hundreds of salt-response QTLs have been reported at different developmental growth stages in rice (Ganie et al. 2019). Salt tolerance during seed germination is not consistently related to other stages, such as vegetative growth or fruit development (Johnson et al. 1992, Shi et al. 2017. So far, few studies aimed at salt tolerance during seed germination in rice. Wang et al. (2011) detected 16 QTLs of rice seed germination ability from the salt tolerance variety Jiucaiqing at 100 mM NaCl. Recently, around 50 salt tolerance loci have been identified using GWAS at the seed germination stage (Cheng et al. 2015;Shi et al. 2017;Yu et al. 2018). Fujino et al. (2008 reported that qLTG3-1 responded to seed germination under salt stress because of tissue vacuolation and weakening. Recently, a potassium transport OsHAK21 (He et al. 2019) was reported to mediate salt tolerance at the seed germination stage by involving in abscisic acid (ABA) metabolism. However, it is still limited genes or QTLs to reveal the genetic mechanisms of salt tolerance during seed germination in rice.
In this study, an indica rice variety Wujiaozhan (WJZ) from 276 indica accessions (Cheng et al. 2015) was identified as high salt tolerance during seed germination. Salt tolerance at the seed germination stage was evaluated by the traits of GR (germination rate), SP (seedling percentage) and GI (germination index). QTL mapping and BSA (bulked segregant analysis) were developed to excavate QTLs by the progeny derived from WJZ and the salt-sensitive japonica rice variety Nipponbare (Nip) at 300 mM NaCl.
Consequently, a major QTL qGR6.2 on the short arm of chromosome 6 could be detected among several generations, and finally fine mapped to a 65.9 kb region between the two markers Z654 and Z619. Our work is beneficial to better understand the genetic and molecular basis of salt tolerance during seed germination.

Salt tolerance of two parents during seed germination
There were no differences in germination rate (GR) and seedling percentage (SP) between salt-sensitive variety Nip and salt-tolerance WJZ at 0 mM NaCl (or H 2 O) imbibition for ten days (Fig. 1a,1b). Under NaCl stress conditions, the GR and SP of WJZ were significantly higher than those of Nip, especially at 300 mM NaCl imbibition (Fig. 1a, 1b). There was significantly higher germination index (GI) of WJZ under H 2 O or NaCl imbibitions (Fig. 1c).
It indicated that WJZ was more salt-tolerant at the seed germination stage. Since there was an irreparable injury for the seedlings at 350 mM NaCl, the SP of WJZ and Nip was 23.33% and 0.00% respectively (Fig. 1b). The seed germination was assessed with 300 mM NaCl for further analyses.
At 300 mM NaCl, the significant differences in GR were observed between WJZ and Nip from 3 to 14 days during seed germination (Fig. 1d, 1e). The WJZ started to germinate after imbibition for 3 days and its GR reached to 90% after 7 days (Fig 1e). After imbibition for 14 days, WJZ presented a higher GR of 95.55%, while Nip only 34.44%.

QTL mapping combined with BSA identify salt tolerance loci during seed germination
To identify QTLs related with salt tolerance during seed germination, a BC 1 F 1 line containing about 37.76% genetic region of Nip was used to generate a BC 1 F 2 population ( Fig. S1). It was analyzed that the GR, SP and GI of the BC 1 F 2 population (181 lines) from from 4 to 14 days (Fig. S2, S3). The dynamic distributions of GR and SP of this population under salt stress could be classified into two categories during seed germination, symmetrical distribution and skewed distribution (Fig. S2, S3). It showed a skewed distribution for GR at the beginning of seed germination (4 ~ 6 days) and the end of seed germination (9 ~ 14 days) ( Fig. S2a-c, f-k), while a symmetrical distribution at the day 7 and (Fig. S2d, e). Besides the two parents, these 181 lines also had a large variation of GR on days 7, 8, 9, ranging from 0 to 100% (Fig. S2d-f). It indicated that this population had a large genetic diversity. Similar results were observed from the SP of this population. On day 11 the symmetrical distribution was observed, whereas from 7 to 10 days a positively skewed distribution was apparent, and from 11 to 14 days a negatively skewed distribution appeared (Fig. S3a-h).
Additionally, the linkage markers of salt tolerance were also identified by bulked segregant analysis (BSA) in this study. Total 73 individual plants were selected and obtained for BSA as they had similar agronomic traits to Nip but shared a great variation of GR, SP and GI (Table S2). Of them, eight high salt-tolerant lines and eight salt-sensitive lines were applied for two extreme phenotype bulks (Fig. S4). A total of 12 SSR markers associated with salt tolerance at the seed germination stage were identified on chromosome 1, 3, 4, 5, 6, and 8 (  Fig. 3), indicating these seven QTLs might be the key QTLs for salt tolerance during seed germination.
Validation and fine mapping of qGR6.2 for controlling salt tolerance during seed germination Based on QTL mapping and BSA, a major qGR6.2 within the interval of Z604 ~ RM276 was further confirmed. A significant peak was obtained between the markers Z604 and RM276 in the BC 2 F 2 population and accounted for 20.96 % and 23.53 % of the GR variation at 12th and 14th day respectively ( Fig. 4; Table S3). The genetic effect of qGR6.2 was validated in different populations (Table 1, 2, S3; Fig. 2-4). These results indicated that qGR6.2 was a major locus controlling salt tolerance during seed germination in rice.
To fine map the major-effective locus qGR6.2, a larger BC 2 F 3 population comprising 1,205 individuals was developed. Total eighty-six recombinants were identified between Z604 and RM276. Eight polymorphic markers were developed within a wider interval between RM190 and RM3183, in case of possible shifting of the target locus (Fig. 5a). The progeny of recombinants was tested using these markers to identify homozygous individuals for the salt response during seed germination. Based on the genotypes, the 86 recombinants were classified into four groups. 18 recombinant events were between Z604 and Z616, 57 recombinant events between Z617 and Z619, and 11 between Z605 and RM276 (Fig. 5b).
The mean GR of homozygous WJZ alleles in group B and D was significantly higher than Nip alleles, while no difference in group A and C. Therefore, the qGR6.2 was delimited between the Z617 and Z619 markers. In the same way, a larger BC 2 F 4 population derived from heterozygous BC 2 F 3 plants in marker Z617 and Z619 was conducted, containing 2,318 individuals. Seventeen recombinants were obtained, including three types of recombination ( Fig. 5c). Based on progenies assay of each recombinant, the qGR6.2 locus was finally narrowed down to a 65.9 kb region between the markers Z654 and Z619 (Fig.   5c).
Meanwhile, the expression profiles of these candidate genes were conducted in various developmental stages incubation of seed imbibition and salt stress based on mRNASeq data and Array database deposited in the GENEVESTIGATOR (Fig. 6). The higher transcript (LOC_Os06g10730)and ORF11 (LOC_Os06g10750) were observed at all of the developmental stages, while almost no expression for ORF4 (LOC_Os06g10680), ORF6 (LOC_Os06g10700) and ORF8 (LOC_Os06g10720) (Fig. 6a). During seed germination, the expression of ORF1 (LOC_Os06g10650) was significantly up-regulated by seed imbibition, and ORF7 (LOC_Os06g10710) was obviously down-regulated (Fig. 6b). Interestingly, only ORF11 (LOC_Os06g10750) showed significant down-regulated expression patterns when exposed to salt stress in root and seedling samples (Fig. 6b).
We subsequently conducted quantitative real-time PCR (RT-qPCR) analysis to detect the expression of these genes in WJZ and Nip during seed germination subjected to salt stress. It was found that the gene ORF1 (LOC_Os06g10650) was dramatically induced by salt stress after imbibition for 24 h, nearly 20-fold change in WJZ and 14-fold in Nip (Fig.   7a). Eight genes, (LOC_Os06g10730), ORF10 (LOC_Os06g10740) and ORF11 (LOC_Os06g10750) showed down-regulation of expression pattern during seed germination (Fig. 7). Comparing the expression level between two parents, we found ORF11 (LOC_Os06g10750) in Nip was nearly 10-fold higher than in WJZ (Fig. 7k).

Discussion
Salinity seriously affects rice seed germination and seedling establishment, especially in the direct-seeding area, leading to a reduction of rice yields (Wang et al. 2012b;Shi et al. 2017). In this study, the indica landrace WJZ from Yunnan province in China showed a strong capability of seed germination and seedling establishment under high salinity stress. Even when exposed to 300 mM NaCl, visible seeds of WJZ could start to germinate after three days of imbibition and established normal seedlings after five days. It suggests that WJZ is an important germplasm source with strong salt tolerance during seed germination, similar to N22-334-3 (Kumar et al. 2013) and Jiucaiqing (He et al. 2019).
However, WJZ was much higher than Nip with 167.62 vs.84.18 cm of plant height (data not shown), and easy to fall-down in field planting. It is of great importance to explore elite genes controlling salt tolerance during seed germination from WJZ, which will be beneficial for improving rice salinity tolerance in the direct-sowing areas.
Evaluating the phenotype of salt tolerance comprehensively and accurately is the most crucial step for QTL mapping (Wang et al. 2012a). To estimate the dynamic characteristics of salt tolerance during seed germination, three germination-related traits (GR SP, and GI) at 300 mM NaCl were recorded in this study. During seed germination, the dynamic distribution characteristics of the BC 1 F 2 population transformed positively skewed distribution (4 to 6 days for GR, 7 to 10 days for SP) into negatively skewed distribution (8 to 14 days for GR, 12 to 14 days for SP). Obviously, seed germination was delayed by salt stress, leading to impaired seedling establishment. Besides, the symmetrical distributions were showed at GR of 7th day, SP of 11th day, where most of the seeds started to germinate and establish seedlings, respectively. These results suggested that there were large variations at different germination stages, and salt tolerance during seed germination is a quantitative trait regulated by multiple genetic mechanisms.
In this study, a total of 16 QTLs and 12 linkage markers controlling salt tolerance during seed germination were identified via QTL mapping and BSA mapping. The comparison revealed that seven QTLs (qGR6.1, qGR6.2, qGR6.3, qSP6.1, qSP6.2, qGI6 and qSP8.2) in the BC 1 F 2 population were co-localized with linkage markers in the F 2 population. The coincidence indicated that these QTLs identified by QTL mapping and BSA might be the key QTLs for salt tolerance during rice seed germination. In addition, nine QTLs were not identified nearby linkage markers via BSA mapping. This might be due to the different genetic background between the F 2 and BC 1 F 2 populations. Simultaneously, these colocalized loci being stably expressed via multiple generations could contribute to increasing salt tolerance during seed germination via the marker-assisted breeding program.
Until now, many studies have been exploring QTLs for salt stress in rice. By comparing chromosomal locations, eleven QTLs in our study were located in the same or adjacent regions with previous reported QTLs. The linkage markers RM128, RM1164, RM3436, and RM480 in the F 2 population were located near the region of qSSIGY1.1, qSSIGY3.3, qSSIGY3.7, qSSIGY5.2 and qSSIGY5.5 for seedling salt tolerance respectively (Tiwari et al. 2016). The region of qGR10, qSP10, and qGI10 in the BC 1 F 2 population were similar to qSKC10 and qRKC10 for seedling salt stress (Wang et al. 2012a). There is a gene OsRR22 located within the region of qGR6.2, qSP6.1, and qGI6, which controlling salt stress at the seedling stage (Takagi et al. 2015). Similar results were found for RM588 and RM190 with OsDREB1C (Dubouzet et al. 2003), and for qGR1 and qSP1 with OsSKC1 (Ren et al. 2005). It indicated these co-localized QTLs at the different developmental stages were the weightily genomic regions for salt tolerance in rice. In addition, the other five mapped QTLs (qSP7, qSP8.1, qSP8.2, qGR8 and qSP10.2) in this study are novel alleles with no previous reports of salt tolerance in rice.
Interestingly, the major qGR6.2 was identified via multiple mapping methods and expressed continuously, with a relatively higher contribution and additive effects at 300 mM NaCl during seed germination. With validation in the BC 2 F 2 population and finemapping strategy, qGR6.2 was finally mapped in a region of 65.9 kb between the two markers Z654 and Z619. A total of 11 candidate genes in the Nipponbare genome were predicted. Of these genes, we found that LOC_Os06g10650 encoded tyrosine phosphatase (TPT) family protein and LOC_Os06g10750 encoded integral membrane protein DUF6 containing protein with differentially expressed by seed imbibition or exposure of salt stress based on mRNASeq data. According to the previous studies, TPT family protein has been reported to regulate signal transduction and control plant growth and development (Stoker, 2005). AtPTP1 encoding a putative protein tyrosine phosphatase (PTPase), the first TPT family gene in plant was upregulated by high salt stress (Xu et at. 1998). An integral membrane protein DUF6 containing protein At01g09380 was involved in salt stress and drought stress (Sun et al. 2009). Furthermore, the different expression patterns of these two genes from RT-qPCR were also observed in WJZ and Nip after seed imbibition at 300 mM NaCl, suggesting the different roles of them in two parents. All these indicated that LOC_Os06g10650 and LOC_Os06g10750 might be the casual candidate genes responding to salt tolerance during seed germination. More experiments should be performed to validate the functions of these candidate genes.
In summary, we fine mapped a novel major QTL qGR6.2 for salt tolerance during seed germination within a region of 65.9 kb and identified two more likely candidate genes, LOC_Os06g10650 and LOC_Os06g10750. Sequence analysis and genetic transformation will be carried out in the future to validate the candidate genes and elucidate the molecular mechanism underlying salt tolerance at the seed germination stage. We believe that the major QTL qGR6.2 is useful for improving salt tolerance by the MAS strategy.

Plant materials
Two rice varieties, the high salt-tolerant indica Wujiaozhan (WJZ) and the salt-sensitive japonica Nipponbare (Nip), and their F 2 , BC 1 F 2~4 , populations were used in this study. The BC 1 F 2 and BC 2 F 2 were developed for QTL mapping and the F 2 population was for BSA (Fig.   S1). All plants were grown in a paddy field at the Jiangpu Experimental Station (Jiangsu Province, China) in 2014, 2015, 2016, 2017and 2018 cm between plants within a row and 33 cm between rows. Seeds of each line were harvested at maturity and dried at 42 ℃ for 7 days to break seed dormancy, and then stored at -20 ℃.

Evaluation of seed tolerance during seed germination
A total of 30 healthy grains of each accession were surface-sterilized with 0.5 % sodium hypochlorite solution for 15 min and then rinsed three times with sterile distilled water.
Seeds were imbibed in 9 cm Petri dishes with 40 mL quartz and 20 mL solution, with normal (sterile distilled water), and salt conditions (150 mM, 200 mM, 250 mM, 300 mM, 350 mM NaCl) at 25 ± 1 ℃for 14 days in a growth chamber under 12 h light/12 h day conditions. Seed germination was defined as the emergence of the radicle (2 mm) through the surrounding tissue, and the seedling establishment was considered when the root length reached the seed length and the shoot length reached half of the seed length (Cheng et al. 2015). Germination rate (GR) and seedling percentage (SP) were calculated on a daily basis. Germination index (GI) was calculated as GI = ∑ (Gt/t), where Gt is the number of germinated seeds on day t (Wang et al. 2011). Three replications of each line were performed.  (Table S1). PCR was performed as described by Chen et al. (1997). The PCR products were separated by electrophoresis through 8 % non-denaturing polyacrylamide gels and visualized by silver staining.

BSA analysis and QTL mapping
A total of 73 plants with similar agronomic characters as Nip were selected from the 3,000 F 2 progeny for BSA (Table S2). Among them, two extreme phenotypes bulks were selected to determine the linkage markers with salt tolerance, including salt-tolerant bulk with simultaneously higher GP, SP, and GI and sensitive bulk with lower GP, GR, and GI (

Confirmation and fine mapping of qGR6.2
To ensure the presence of the major QTL qGR6.2 , a linkage map of 70 individuals from BC 2 F 2 progeny were analyzed with 11 SSR markers on chromosome 6 ( Table. S4). Another ten polymorphic InDel markers were explored and used for fine mapping. In this study, 1205 BC 2 F 3 and 2318 BC 2 F 4 plants were used for fine mapping the target QTL qGR6.2. To determine the location of more precisely, totally seven types of recombinants from the BC 2 F 3~4 segregating populations were identified. Twenty progenies of each recombinant were planted in a two-row plot with 10 individuals in each row, and the homozygous plants were selected for the seed germination test.

Prediction and expression analysis of candidate genes
Open reading frames (ORFs) in the region of markers Z654 and Z619 were predicted by Rice Annotation Project Database (http://rice.plantbiology.msu.edu/). The GENEVESTIGATOR (https://genevestigator.com/gv/) was employed to analyze the expression patterns of eleven candidate genes based on 2836 Affymetrix microarray datasets and 986 mRNASeq data with a significant level P < 0.05 in seed imbibition and salt stress.
Seeds of two parents were sampled after 0 h, 6 h, 12 h, 24 h, 36 h and 48 h imbibition at 300 mM NaCl, and then frozen quickly in liquid nitrogen and stored at -80 ℃ for RNA extraction. Total RNA was isolated from approximately 80∼100 mg powder with Total RNA Kit (BioTeke, http://www.bioteke.com). The first-strand cDNA was synthesized with random

Data analysis
The experimental data were analyzed using the Statistical Analysis System (SAS) software (Cary, NC, USA), and compared with Student's t-test at the 5% and 1 % levels of probability.

Conclusions
In our study, we identified 16 QTLs and 12 associated SSR markers conferring salt tolerance from indica landrace WJZ during seed germination, and the major qGR6.2 was fine mapped to a 65.9 kb region flanked by marker Z654 and Z619. Our results provide new information on the genetic basis of improving salt tolerance during seed germination by marker-assisted selection.    (e) Morphology of two varieties during seed germination after 3, 5, 7 and 9 days at 300 mM NaCl. Scale bar = 1 cm. The arrows in a-c indicate the large variation between WJZ and Nip at 300 mM NaCl.    The mean germination rate of each recombinants group after imbibition for 10 days was indicated in the bar diagram. Orange bars represent the WJZ genotype region, light blue and light green bars represent the Nip genotype and Figure 6 Expression pattern of candidate genes in various developmental stages (a) and response to seed germination and salt stress (b) based on Affymetrix microarray datasets and mRNASeq data from the GENEVESTIGATOR (http://www.genevestigator.com).

Figure 7
Expression levels of candidate genes regulated by WJZ and Nip during seed germination at 300 mM NaCl.

Supplementary Files
This is a list of supplementary files associated with the primary manuscript. Click to download.