OsSRK1, an Atypical S-Receptor-Like Kinase Positively Regulates Leaf Width and Salt Tolerance in Rice

ZHOU Jinjun , JU Peina , ZHANG Fang, ZHENG Chongke BAI Bo LI Yaping ,WANG Haifeng CHEN Fan, XIE Xianzhi

(1Shandong Rice Research Institute, Shandong Academy of Agricultural Sciences, Jinan 250100, China; 2Institute of Genetics and Development Biology, the Chinese Academy of Science, Beijing 100190, China; 3Shangdong Normal University, Jinan 250014,China; #These authors contributed equally to this work)

Abstract: Receptor-like kinases (RLKs) are important for plant growth, development and defense responses. The S-receptor protein kinases (SRKs), which represent an RLK subfamily, control the selfincompatibility among Brassica species. However, little information is available regarding SRK functions in rice. We identified a gene OsSRK1 encoding an atypical SRK. The transcript of OsSRK1 was induced by abscisic acid (ABA), salt and polyethylene glycol. OsSRK1 localized to the plasma membrane and cytoplasm. Leaf width was increased in OsSRK1-overexpression (OsSRK1-OX) transgenic rice plants,likely because of an increase in cell number per leaf. Furthermore, the expression levels of OsCYCA3-1 and OsCYCD2-1, which encode positive regulators of cell division, were up-regulated in leaf primordium of OsSRK1-OX rice plants relative to those in wild type. Meanwhile, the expression level of OsKRP1,which encodes cell cycle inhibitor, was down-regulated in the OsSRK1-OX plants. Therefore, it is deduced that OsSRK1 regulates leaf width by promoting cell division in the leaf primordium. Additionally,OsSRK1-OX plants exhibited enhanced ABA sensitivity and salt tolerance compared with wild type.These results suggest that OsSRK1 plays important roles in leaf development and salt responses in rice.

Key words: rice; S-receptor-like protein kinase; OsSRK1; salt tolerance; leaf width; cell division

Receptor-like kinases (RLKs) reportedly play key roles in regulating plant development, hormone perception, self-recognition and disease resistance(Clark et al, 1997; Gómez-Gómez and Boller, 2000;Takasaki et al, 2000; Kinoshita et al, 2005). These kinases are transmembrane proteins composed of an N-terminal extracellular domain, a transmembrane domain, and C-terminal intracellular kinase domain. The extracellular domain recognizes diverse environmental signals, while the intracellular C-terminal including a conserved serine/threonine kinase domain is involved in signal transductions (Wang Q G et al, 2018). There are more than 610 and 1 131 RLKs in Arabidopsis thaliana and rice, respectively (Shiu and Bleecker,2001). Based on the amino acid sequence identity of extracellular domains, RLKs have been classified into 44 subfamilies, including leucine-rich repeat group,lectin-like domain group and epidermal growth factor repeat receptor group (Becraft, 2002).

S-receptor protein kinases (SRKs), which represent an RLK subfamily containing an S-domain, control self-incompatibility among Brassica species (Becraft,2002). SRKs include an N-terminal extracellular domain,transmembrane domain, and C-terminal kinase domain.The extracellular domain of typical SRKs contains three modules (S-LOCUS GLYCOPROTEIN, B_lectin and PAN domain) (Becraft, 2002; Nasrallah and Nasrallah, 2014). Previous studies revealed that transgenic A. thaliana and Medicago sativa plants overexpressing GsSRK from Glycine soja exhibit enhanced salt tolerance (Sun et al, 2013; Fan et al,2018). Additionally, four SRK genes, including Pi-d2,SDS2 (SPL11 cell-death suppressor 2), OsLSK1 (large spike S-domain receptor like kinase 1) and OsSIK2 have been functionally characterized in rice. Transgenic plants carrying Pi-d2 exhibit race-specific resistance to Magnaporthe grisea (strain ZB15), suggesting Pi-d2 mediates fungal disease resistance in rice (Chen et al, 2006). Moreover, SDS2 regulates programmed cell death and immunity in rice, with implications for defense responses (Fan et al, 2018). Meanwhile,OsLSK1 was reported for the improvement of grain yield components in rice (Zou et al, 2015).Interestingly, plants overexpressing full-length OsLSK1 or lacking OsLSK1 expression (i.e., RNAi plants) are phenotypically the same as the wild type (WT).However, overexpression of a truncated version of OsLSK1 (including the extracellular and transmembrane domain of OsLSK1 without the intracellular kinase domain) increases plant height and improves yield components (Zou et al, 2015). Expression of OsSIK2 can be induced by NaCl, drought, cold and abscisic acid (ABA). Overexpression of OsSIK2 enhances plant tolerance to salt and drought stresses.Interestingly, overexpression of a truncated version of OsSIK2 without most of extracellular region confers higher salt tolerance than the full-length OsSIK2(Chen et al, 2013). OsSIK2 is also associated with delayed dark-induced leaf senescence (Ouyang et al,2010; Chen et al, 2013). These observations suggest that N-terminal extracellular region is important for the perception of environmental signals. However, the N-terminal extracellular domain of some SRKs does not contain all the three previously mentioned modules, with missing of one or two. Additionally,there are currently 28 and 9 known SRKs lacking all the three extracellular modules in rice and A. thaliana,respectively (Xing et al, 2013). These SRKs have not been functionally characterized.

In this study, we identified one of S-receptor-like kinase genes, OsSRK1, by GeneChip analysis. Protein structure analyses indicated that OsSRK1 lacks three N-terminal modules, making it an atypical S-receptorlike kinase. Expression of OsSRK1 was induced by ABA, salt and polyethylene glycol 4000 (PEG).Additionally, OsSRK1 was mainly localized to the cell membrane. Transgenic plants overexpressing OsSRK1 exhibited enhanced salt tolerance and increased leaf width.

MATERIALS AND METHODS

Rice materials and growth conditions

The phytochrome B (phyB) mutant phyB1 was used(Takano et al, 2005). The genetic background of wild type (WT) and phyB1 mutant was Oryza sativa L.subsp geng cv. Nipponbare (Wang W S et al, 2018).

Seeds were surface-sterilized in 70% ethanol for 30 s and in 20% NaClO for 20 min. Then, seeds were washed four times with sterile distilled water and germinated in 0.4% agar supplemented with different ABA (Sigma, Saint Louis, USA) concentrations (ABA was dissolved in alcohol to the stock solution concentration of 10 mmol/L). Nine-day-old seedlings were analyzed.

Seeds were incubated in darkness for 3 d at 37 °C to induce germination and were then grown in Yoshida’s culture solution in a growth chamber under controlled photoperiodic conditions (14 h light, 28 °C/10 h dark,23 °C). For salt, PEG and ABA treatments, seedlings at the 3-leaf stage were transferred to culture solution containing 150 mmol/L NaCl, 15% PEG or ABA.

To examine the expression levels of cell cyclerelated genes in OsSRK1-overexpression (OsSRK1-OX) lines, surface-sterilized WT and phyB1 seeds were incubated in an illumination incubator set at 14 h light (28 °C)/10 h dark (25 °C) for 5 d. White light was supplied at an irradiance of 90 μmol/(m2·s) using FL20W-B white fluorescent tubes (Hitachi, Tokyo,Japan). Stem apical meristem and leaf primordium were harvested for subsequent RNA extraction.

Histological analysis

To measure cell length and the number of epidermal cells, middle portions of the third leaves were harvested from greenhouse-grown WT and OsSRK1-OX lines at 4-leaf stage. These samples were incubated overnight in ethanol : acetic acid (9 : 1)solution and then cleared overnight in Hoyer’s solution (chloral hydrate : glycerol : water = 8 : 1 : 2)(Wilkinson and Tucker, 2017). The cleared leaves were analyzed by a differential interference contrast microscopy (ECLIPSE 80i microscope, Nikon, Tokyo,Japan). The resulting images were used to measure epidermal cell lengths with the Image J program (cell number ＞ 60). Epidermal cell counts in five square areas of 0.068 mm2per leaf were also determined using the ECLIPSE 80i microscope (400×magnification). Epidermal cell counts were calculated from four areas per leaf and one leaf per plant from five individual plants, for a total of 20 measurements.Data were analyzed using the Student’s t-test (n = 20).The third leaves of plants at the 4-leaf stage were fixed with formalin : glacial acetic acid : 70% ethanol(1 : 1 : 18) solution and then dehydrated in a graded ethanol series. Fixed tissues were embedded in paraffin (Sigma, Saint Louis, USA), cut into 10 μm thick sections using a microtome, and then applied to glass slides. The sections were counter-stained under the microscope.

Reverse transcription polymerase chain reaction(RT-PCR)

To compare the expression levels of OsSRK1 gene in OsSRK1-OX lines and WT, the WT and OsSRK1-OX seeds were sown in 0.4% agar and incubated at 28 °C under continuous white light for 8 d, the up-ground parts were harvested and conserved in liquid nitrogen for extraction of total RNA using RNAiso (TaKaRa,Dalian, China). First-strand cDNAs were synthesized from DNaseI-treated total RNA using PrimeScriptRTEnzyme Mix I (TaKaRa, Dalian, China), according to the manufacturer’s instruction. RT-PCR reactions were performed with LA Taq DNA polymerase(TaKaRa, Dalian, China). Each reaction contained 10 μL of 2× GC buffer I, 2 μL dNTPs (2.5 mmol/L), 0.2 μL LA-Taq (5 U), 0.2 μmol/L gene-specific primer pairs (Supplemental Table 1) in a final volume of 20 μL. The PCR thermal cycle used was as follows:denaturation at 94 °C for 4 min and 25 cycles at 95 °C for 30 s, 60 °C for 30 s and 72 °C for 30 s. As an internal control, rice ubiquitin gene (OsUBQ,Os03g13170) was used to quantify the relative transcript level of each target gene. Three biological replications were performed.

Quantitative RT-PCR (qRT-PCR)

Total RNA was extracted from leaves harvested from seedlings treated with ABA, PEG or salt using RNAiso (TaKaRa, Dalian, China). First-strand cDNAs were synthesized from DNaseI-treated total RNA using PrimeScriptRTEnzyme Mix I (TaKaRa,Dalian, China), according to the manufacturer’s instruction. qRT-PCR was performed on an ABI 7900HT Real-time PCR System (Applied Biosystems,Foster, CA, USA) using SYBR Premix Ex Taq TM(TaKaRa, Dalian, China). Each reaction contained 10 μL of 2× SYBR Premix Ex TaqTM(TaKaRa, Dalian,China), 2.0 μL of cDNA samples, and 0.2 μmol/L gene-specific primer pairs (Supplemental Table 1) in a final volume of 20 μL. PCR thermal cycle used was as follows: denaturation at 95 °C for 30 s and 40 cycles at 95 °C for 5 s and at 60 °C for 30 s. As an internal control, OsUBQ was used to quantify the relative transcript level of each target gene. Relative expression levels were calculated using the 2-ΔΔCTmethod (Livak and Schmittgen, 2001). Three biological replications were performed.

Construction of plant expression vector and rice transformation

To construct the OsSRK1-overexpression vector,OsSRK1 open reading frame was amplified by PCR using primers cSRK-F and cSRK-R with cDNA as the template (Supplemental Table 1). The resulting PCR product was digested with SpeI and BamHI, and then subcloned into the pCAMBIA1390-Ubi vector between maize ubiquitin promoter and nos terminator. The plasmid was introduced into Agrobacterium tumefaciens strain EHA105 cells by electroporation. Nipponbare plants were transformed via an A. tumefaciensmediated infection method (Hiei et al, 1994). The transplant seedlings were grew in a paddy field.

Production of OsSRK1-green fluorescent protein(GFP) fusion protein

To assess the subcellular localization of OsSRK1, we constructed a 35S::OsSRK-GFP expression vector.The OsSRK1 open reading frame without the stop codon was amplified by PCR using primers gSRK-F and gSRK-R with rice leaf cDNA as the template(Supplemental Table 1). The PCR product was digested with XbaI and BamHI, and then introduced into the pBI221-GFP vector. The resulting plasmid DNA was transiently introduced into rice protoplasts.

After an overnight incubation, GFP signal was observed with a FluoView? FV1000 confocal microscope (Olympus, Tokyo, Japan).

GeneChip analysis

The OsSRK1-OX and WT lines were grown under continuous white light for 7 d, after which the aerial parts were harvested. Two independent OsSRK1-OX lines (#2 and #8) and duplicate WT lines were used for microarray analysis, which was conducted based on the standard protocol for Affymetrix GeneChip(CapitalBio, Beijing, China). Probe signals were normalized according to the robust multichip analysis(Irizarry et al, 2003). Genes with a call value of P(present) and a signal ratio of OsSRK1-OX/WT ≥ 2(upregulated) or ≤ 0.5 (downregulated) were considered to be differentially expressed.

They were all very anxious to know what had happened, but he rushed off and locked himself into a room, where he ramped66 and raved67 until he was tired

RESULTS

OsSRK1 is an atypical S-receptor-like kinase localized to plasma membrane and cytoplasm

We previously compared gene expression profiles in WT and phyB1 plants at the 4-leaf stage, and detected 236 differentially expressed genes (data not shown).Among these genes, OsSRK1 (accession number AK068992) was expressed approximately 3-fold higher in the phyB1 plants than in the WT plants(Supplemental Fig. 1). A subsequent comparison of OsSRK1 expression levels in the third leaves of WT and phyB1 plants at 4-leaf stage by qRT-PCR assay revealed that OsSRK1 was expressed about 15-fold higher in the phyB1 plants than in the WT plants(Supplemental Fig. 1). Thus, PHYB negatively regulated OsSRK1 expression. In this study, we further characterized the function of OsSRK1 in rice.

OsSRK1 gene encoded an S-receptor-like kinase with 444 amino acids (Fig. 1-A). Several domains were identified in the OsSRK1 protein following an online database search (http://www.ebi.ac.uk/interpro/).An extracellular domain comprising amino acids 1-19 and a transmembrane domain consisting of amino acids 20-46 were detected in the OsSRK1 N-terminal region. Additionally, amino acids 104-400 were predicted to represent serine/threonine protein kinase domain (Fig. 1-A). More specifically, amino acids 114-136 and 232-244 were predicted to form ATP-binding site and serine/threonine protein kinase active site, respectively (data not shown). However,the S-LOCUS GLYCOPROTEIN, B_lectin and PAN modules existing in typical SRKs were not observed at the N-terminal, implying that OsSRK1 was probably an atypical S-receptor-like kinase. Therefore, we compared the identity between OsSRK1 and the four reported rice SRK proteins using the amino acid sequence of the serine/threonine-protein kinase domain. The amino acid sequence alignments revealed the following identities:41.96% between OsSRK1 and OsSIK2, 42.12% between OsSRK1 and Pi_d2, 36.79% between OsSRK1 and OsSLK1, 39.16% between OsSRK1 and SDS2(Supplemental Fig. 2). The relative low identities between OsSRK1 and the other four rice SRKs indicated the possible functional difference among these SRK proteins.

To analyze the subcellular localization of OsSRK1,we generated a construct harboring an OsSRK1-GFP fusion gene under the control of the CaMV35S promoter for the production of OsSRK1 fused to the N-terminal of GFP. This construct was transiently expressed in rice protoplasts. The OsSRK1-GFP signal was clearly observed in the plasma membrane and cytoplasm whereas the control GFP signal was distributed throughout the cell (Fig. 1-B). Localization of OsSRK1 to plasma membrane was consistent with the presence of the predicted transmembrane domain in OsSRK1.

Expression patterns of OsSRK1

To further clarify the effects of OsSRK1 on stress tolerance, we examined the OsSRK1 transcript levels in response to PEG, salt and ABA treatments. The OsSRK1 transcription was induced at 24 h after PEG treatment (Fig. 1-C), while OsSRK1 mRNA levels initially increased in response to NaCl, but then decreased at 24 h (Fig. 1-C). Application of exogenous ABA rapidly induced OsSRK1 expression (Fig. 1-C). Therefore,OsSRK1 is likely involved in responses to abiotic stress and ABA. We also analyzed the OsSRK1 expression in different organs (root, stem, leaf, flag leaf, panicle and embryo), and observed that the transcript levels were the highest in the root followed by the flag leaf (Fig.1-D).

Overexpression of OsSRK1 increased leaf width

To functionally characterize OsSRK1 in rice, we produced transgenic OsSRK1-OX rice plants (Fig. 2-A). The transcript levels of OsSRK1 in the three independent transgenic lines were examined by RT-PCR. The OsSRK1 expression levels were significantly higher in the transgenic lines than in the WT plants (Fig. 2-B).Three independent transgenic lines (#1, #2 and #8)underwent further analyses. The leaf angles of OsSRK1-OX plants were larger than that of WT seedlings (Fig.2-C and -D), which is similar to the phenotype of phyB mutants (Takano et al, 2005), implying that OsSRK1 is probably involved in the phyB-mediated formation of leaf angle. Moreover, the widths of the second and third leaf blades were greater in the OsSRK1-OX lines than in the WT plants (Fig. 2-E and-F). To verify this observation, we measured the widths and lengths of the first, second, third and fourth leaves at the 4-leaf stage as well as the flag leaf at the maturity stage of OsSRK1-OX and WT plants.The leaf widths were significantly greater in all the OsSRK1-OX lines relative to the WT (Fig. 2-H and -J).The lengths of the first, second and third leaves were slightly longer in the OsSRK1-OX lines #1 and #2 but not significantly different with the WT plants at the 4-leaf stage (Fig. 2-G). The lengths of flag leaves in the OsSRK1-OX lines were comparable to that in the WT (Fig. 2-I). These results suggested that OsSRK1 contributed to regulate rice leaf width.

To determine whether the overexpression of OsSRK1 affected cell division in leaf, we statistically compared the number of epidermal cells in the adaxial epidermis per unit leaf area. The cell numbers per unit area for OsSRK1-OX lines #2 and #8 were similar to that of the WT plants (Fig. 3-A). In contrast, there were significantly fewer epidermal cells in OsSRK1-OX line #1 than in the WT plants. Considering the similarities in the cell number per mm2leaf area between the two OsSRK1-OX lines (#2 and #8) and the WT plants, we speculated that the greater number of cells per leaf was responsible for the increased width of OsSRK1-OX leaves. To confirm this speculation,we compared leaf anatomical structures of OsSRK1-OX lines and WT plants at the 4-leaf stage. There were more small vascular bundles between large vascular bundles in the OsSRK1-OX lines than in the WT plants (Fig. 3-B and -C). These results suggested that the overexpression of OsSRK1 affects leaf development by regulating the cell number per leaf and formation of vascular bundles.

Cell division influences cell number (Dewitte and Murray, 2003). To reveal the molecular mechanism underlying the effects of OsSRK1 on cell number per leaf, we compared the expression levels of cell cyclerelated genes in the primordium of developing leaves,including OsCDKB2-1, OsCYCA3-1, OsCYCD2-1,OsCYCD5-1, OsKRP1 and OsKRP4. The OsCYCA3-1 and OsCYCD2-1 expression levels were upregulated in the OsSRK1-OX lines, while the OsKRP1 expression level was downregulated (Fig. 3-D). An earlier investigation concluded that OsCYCA3-1 and OsCYCD2-1 encode positive regulators of cell division, while OsKRP1 encodes cell cycle inhibitor (Morgan, 1995; Sherr and Roberts, 1999; de Veylder et al, 2001). Therefore,OsSRK1 can promote cell division in the leaf primordium, which is probably one of factors contributing to leaf width in the OsSRK1-OX lines.

Overexpression of OsSRK1 enhanced ABA sensitivity and salt tolerance

Because OsSRK1 expression was induced by various stresses, we tested the sensitivity of transgenic plants to different abiotic stresses. Treatments with different ABA concentrations decreased seedling height in the WT plants and three independent OsSRK1-OX lines,but the inhibitory effects were more obvious in the transgenic lines (Fig. 4-A). This result implied that OsSRK1 positively regulates ABA sensitivity.

We further analyzed the regulatory roles of OsSRK1 in rice tolerance to drought stress and high salinity.The OsSRK1-OX lines and WT plants were grown to 3-leaf stage under well-watered conditions before being treated with 150 mmol/L NaCl for 4 d. The OsSRK1-OX lines and WT plants exhibited wilting symptoms, but wilting was less obvious in the OsSRK1-OX plants than in the WT plants (Fig. 4-B).The survival rate showed that 33.0%, 50.0% and 55.6% of OsSRK1-OX lines #1, #2 and #8,respectively, were survived, while only 19.4% of the WT plants grew new leaves after 4 d NaCl treatment(Fig. 4-C). Additionally, more OsSRK1-OX plants grew with fewer dead leaves (Fig. 4-D). These observations suggested that the overexpression of OsSRK1 enhances salt tolerance in rice. We also conducted a similar experiment using PEG, and observed a lack of phenotypic differences between OsSRK1-OX lines and WT plants (data not shown).

Genes regulated by OsSRK1

To examine genes regulated by OsSRK1, we compared the transcriptional profiles of WT and OsSRK1-OX plants using Affymetrix GeneChips. A total of 41 and 132 genes exhibited upregulated and downregulated expressions, respectively, in the OsSRK1-OX lines(Supplemental Table 2). The expression levels of genes related to biotic and abiotic stress responses were upregulated by OsSRK1 (Supplemental Table 2). Among them, OsMyb4, OsDREB1A, ZOS3-22, OsWRKY08 and EL5 have been reported to positively regulate abiotic stress tolerance in plants. We examined the transcript levels of genes mentioned above in the WT plants and OsSRK1-OX lines (#1, #2 and #8). Consistent with the microarray result, these genes were more highly expressed in the OsSRK1-OX lines (#2 and #8)(Fig. 5). Therefore, these genes probably contribute to OsSRK1-regulated responses to ABA or salt.

DISCUSSION

OsSRK1 is not a typical S-receptor-like kinase

SRKs form a subfamily of RLKs consisting of an N-terminal extracellular domain, C-terminal intracellular kinase, and a transmembrane domain. N-terminal of typical SRKs is comprised of S-locus glycoprotein,B_lectin, and PAN modules (Becraft, 2002; Xing et al,2013). However, these three modules are absent in the N-terminal of OsSRK1, suggesting OsSRK1 represents an atypical SRK. An earlier study unveiled 28 rice SRKs lacking three extracellular modules (Xing et al,2013). However, functions of these genes had not been determined. To date, four rice SRK genes (Pi-d2,SDS2, OsLSK1 and OsSIK2) encoding typical SRKs have been functionally characterized. Interestingly,overexpression of a truncated version of OsSIK2 without most of extracellular region confers higher salt tolerance than the full-length OsSIK2 (Chen et al,2013). These results indicate that N-terminal extracellular region is important for perception of environmental signals. Lack of the three typical SRK N-terminal domains in OsSRK1 may be responsible for any functional differences associated with OsSRK1. Of note, we could not exclude the possibility that OsSRK1 may actually have an N-terminal extracellular domain as typical SRKs and that the N-terminal of OsSRK1 may be absent due to the problems from either truncated reverse transcription or sequencing in this study. Additionally, the OsSRK1 amino acid sequence differs considerably from that of the other reported rice SRKs (Supplemental Fig. 2), implying that the effects of OsSRK1 on rice growth and development vary from those of the other rice SRKs.We observed that the fusion protein OsSRK1-GFP was obviously distributed in the plasma membrane and cytoplasm (Fig. 1-B). Considering OsSRK1, as an RLK protein, contains a transmembrane domain (Fig.1-A), it is easy to understand that the fusion protein OsSRK1-GFP should localize in the plasma membrane. Similarly, Pi-d2, OsSIK2 and SDS2 also reportedly localize in the cell membrane (Chen et al,2006; Chen et al, 2013; Fan et al, 2018). However, a portion of OsSRK1-GFP clearly distributed in the cytoplasm, which indicates that OsSRK1 also localize at the endomembrane system such as membrane of endoplasmic reticulum and/or Golgi complex.

Overexpression of OsSRK1 increases leaf width probably by promoting cell cycle in leaf primordium

We observed that the overexpression of OsSRK1 increased the width of leaf blades, suggesting this gene is involved in leaf development (Fig. 2-H and -J).Although relationship between SRK functions and self-incompatibility has been established, it is unclear if SRKs mediate leaf development. Our cytological analysis revealed that OsSRK1 has little effect on cell number per unit area. Therefore, increases in the width of leaf blades in OsSRK1-OX lines are likely due to the greater cell number per leaf in the transgenic lines compared with the WT plants. Consistent with this conclusion, we observed more vascular bundles in the OsSRK1-OX lines than in the WT plants (Fig. 3-B and-C). On the basis of these results, we speculated that OsSRK1 probably affects cell division in the primordium of developing leaves (Fig. 3-A to -C). The expression levels of OsCYCA3-1 and OsCYCD2-1,which promote cell cycles, were upregulated in the OsSRK1-OX lines, while the expression level of OsKRP1, which represses cell cycles, was downregulated in the OsSRK1-OX lines (Fig. 3-D). These results suggest that OsSRK1 regulates leaf development by promoting cell cycle in the leaf primordium.

Overexpression of OsSRK1 enhances salt tolerance probably by regulating stress-related genes and ABA pathway

The rice RLK family is important for plant stress tolerance (Ouyang et al, 2010; Chen et al, 2013; Sun et al, 2018; Zhou et al, 2018). In this study, OsSRK1 mRNA levels were upregulated by ABA, PEG and salt treatments (Fig. 1-C). Additionally, OsSRK1-OX lines were more sensitive to ABA and were more tolerant to salt stress than the WT plants (Fig. 4-A to -D). In contrast, drought stress responses were similar in OsSRK1-OX lines and WT plants. These results imply that OsSRK1 is important for salt stress responses, but not for drought stress responses. Transcriptional profiles revealed that the expression levels of a set of genes related to abiotic stress responses were upregulated by OsSRK1 (Supplemental Table 2). OsMyb4 gene encodes an R2R3 type-Myb transcription factor, and overexpression of this gene can enhance plant tolerance to cold and drought stresses in tobacco, tomato, apple, A.thaliana, and barley plants (Vannini et al, 2004, 2007;Pasquali et al, 2008; Park et al, 2010; Soltész et al,2012). Overexpression of OsDREB1A increases plant tolerance to drought, high-salinity and cold stresses(Dubouzet et al, 2003; Zhang et al, 2009). ZOS3-22,encoding a C2H2-type zinc finger protein, was reported to regulate cadmium stress responses (Tan et al, 2017).Overexpression of OsWRKY08 improves osmotic stress tolerance in Arabidopsis (Song et al, 2009). EL5 encodes a RING-H2finger protein that influences plant defense responses via protein turnover mediated by ubiquitin/proteasome system (Takai et al, 2002).Therefore, we speculated that these genes probably contribute to OsSRK1-regulated responses to salt stress.In addition, the expression of OsSRK1 was induced by ABA (Fig. 1-C), while the OsSRK1-OX lines were more sensitive to ABA than the WT plants (Fig. 4-A).Thus, it is possible that ABA-dependent pathway is also involved in OsSRK1-regulated salt tolerance.

In conclusion, we identified OsSRK1, which encodes an atypical S-receptor-like kinase. We subsequently functionally characterized this gene regarding its effects on salt-stress responses and leaf development. However, future studies need to address two issues. First, OsSRK1 is predicted to be an RLK,but its kinase activity needs to be examined. Second,the mechanism underlying OsSRK1-regulated salt tolerance needs to be clarified.

ACKNOWLEDGEMENTS

This work was supported by grants from Excellent Middle-Aged and Youth Scientist Award Foundation of Shandong Province (Grant No. BS2014SW029),the Shandong Natural Science Foundation (Grant Nos.ZR2016CB17 and ZR2018ZC08N2) and Shandong Major Agricultural Applied Technological Innovation Projects (Grant No. 2017.04-2020.04) in China. We thank Liwenbianji, Edanz Editing China (www.liwenbian ji.cn/ac) for editing the English text of this manuscript.

SUPPLEMENTAL DATA

The following materials are available in the online version of this article at http://www.sciencedirect.com/science/journal/16726308; http://www.ricescience.org.

Supplemental Table 1. Primers used in this study.

Supplemental Table 2. Down-regulated or up-regulated 2-fold genes identified by Gene-chip.

Supplemental Fig. 1. Expression levels of OsSRK1 in wild type (WT) and phyB1.

Supplemental Fig. 2. Identity between OsSRK1 and the four reported SRK proteins based on amino acid sequence of serine/threonine-protein kinase domain.