Comparative analysis of drought responsive transcriptome in Brassica napus genotypes with contrasting drought tolerance under different potassium levels

Drought is a major limiting factor of Brassica napus (rapeseed) and potassium (K) plays important roles in rapeseed drought tolerance. Previous studies have reported that rapeseed cultivars characterized by different K status showed contrasting drought tolerance. However, the molecular mechanism underlying drought tolerance remains unclear. In this study, comparative transcriptome analysis was conducted between drought-tolerant cultivar Youyan57 and drought-sensitive cultivar Chuanyou36 exposed to PEG6000 simulated drought stress with two K levels (1.0 and 0.01 mM K2SO4, referred to NK and LK, respectively). A total of 1689 differentially expressed genes (DEGs) were identified at NK. DEGs involved in photosynthesis, glutathione biosynthesis, IAA signal transduction were up-regulated in Youyan57 at NK. By contrast, the down-regulated DEGs were significantly enriched in biosynthesis of amino acids, cysteine and methionine metabolism and glucosinolate biosynthesis. Transcription profile was affected seriously at LK treatment since only 1050 DEGs were identified. DEGs involved in biosynthesis of amino acids reduced largely. Furthermore, the conspicuous up-regulation of protein phosphatase 2 C in Chuanyou36 could lead to more severe drought stress at LK, which negatively participated in abscisic acid (ABA) signal transduction. Taken together, the comparative transcriptome analysis identified a set of drought-regulated genes involved in several pathways, and provided important information about molecular mechanisms underlying rapeseed drought tolerance.


Introduction
Drought is one of the most common abiotic stresses which is a leading cause of crop failure globally (Lesk et al. 2016). Water deprivation causes decrease of water potential, nutrients uptake and photosynthesis, and induces oxidative damage from reactive oxygen species (ROS) and disturbance of metabolism, finally results in reduced production and yield quality of crops (Wahab et al. 2022). Rapeseed is the third largest vegetable oil crop worldwide in terms of the planting area (Abdallah et al. 2010). As the shortage of fresh water and high variation of seasonal precipitation (Lobell et al. 2011;Franks 2011), rapeseed suffers from drought stress frequently. Drought has deleterious effects during rapeseed vegetative growth and would be more detrimental during reproductive stage (Ghobadi et al. 2006), resulting in growth inhibition, oil content reduction and yield loss (Raza et al. 2017). Thus, it is critical to reveal the molecular mechanism underlying drought tolerance for rapeseed production.
RNA sequencing (RNA-seq) for transcriptome analysis which based on high throughput genotyping technologies, is one of the most important tools for gene expression analysis (Unamba et al. 2015), and enables researchers to investigate plant regulation and identify genes involved in stress tolerance mechanisms. Numerous candidate genes associated with rapeseed drought tolerance have been identified by transcriptome profiling. Liu et al. (2015) identified differentially expression genes relevant to drought tolerance in root and leaf by transcriptome sequencing, and characterized a comparative network related to phytohormone signal transduction and AREB/ABF, AP2/EREBP, NAC, WRKY, and MYC/MYB transcription factors (TFs). Wang et al. (2017) identified 169 highly differentially expressed genes in response to drought stress in rapeseed, including 37 droughtresistant cultivar-related genes, 35 drought-sensitive cultivar-related genes, and 97 cultivar nonspecific genes. Xiong et al. (2018) revealed the molecular mechanism of ABA mimic 1 improves drought resistance in rapeseed by transcriptome analysis. Li et al. (2021b) demonstrated the primary mechanism underlying decrease of rapeseed oil content under drought by transcriptome analysis, including down-regulation of fatty acid biosynthesis-associated genes, up-regulation of fatty acid degradation-associated genes, protein storage relevant genes, and Gly-Asp-Ser-Leu (GDSL) gene.
K is one of the most important nutrient elements in plants and participates in a wide range of physiological processes, such as osmotic adjustment, enzyme activation, ion balance and solute transport, which plays an important role in alleviating biotic and abiotic stresses (Broadley et al. 2004). K is the most important inorganic osmolyte in plants, and more accumulation of K in tissues is beneficial for water acquisition under drought stress (Wang et al. 2013). Enhanced photosynthetic rate, plant growth and yields with increased application of K are observed in many crops including rapeseed under water stress condition (Eyni-Nargeseh et al. 2022;Waraich et al. 2020;Pervez et al. 2004). Wide genotypic difference in K uptake and internal use is observed in rapeseed (Lu et al. 2016), and it is revealed that K uptake and retention is closely associated with rapeseed drought tolerance in our previous study (Zhu et al. 2020). The drought tolerant rapeseed cultivar kept much higher K concentration in tissue, and showed higher antioxidant capacity and less disturbance of metabolism under K deficiency compared to drought sensitive cultivar (Zhu et al. 2020).
Although great advances have been achieved over the past few decades, the general molecular basis of rapeseed drought tolerance is still not well understood. In present study, seedlings of two rapeseed cultivars with contrasting drought tolerance were treated with 0.01 and 1.0 mM K 2 SO 4 nutrient solution supplied with PEG6000 simulated drought stress. Comparative analysis of transcriptome profile between drought tolerant and drought sensitive cultivars was conducted, to identify the drought-regulated genes with different K levels. The objective of this study was to investigate the molecular mechanism of rapeseed drought tolerance with different K status, which would be helpful in unraveling the basic mechanisms of environmental stress tolerance.

Materials and methods
Plant culture and stress treatment Two rapeseed cultivars Youyan57 (drought tolerant) and Chuanyou36 (drought sensitive) were used in this study (Zhu et al. 2020). A hydroponic experiment was carried out in a greenhouse with natural light at Southwest University, Beibei, Chongqing, China. Seeds were germinated on quartz sand in a tray on August and were put into a growth chamber (20/15 ℃, light/dark). Uniform ten-day-old seedlings were transplanted to 1-L pot matching a 5-hole lid for hydroponic culture, being one plant for each hole. The hydroponic solution was prepared according to Zhu et al. (2020). The nutrient solution was renewed every five days.
After 20 d cultivation, the plants were divided into three groups. Two groups of plants were treated with 0.01 and 1.0 mM K 2 SO 4 nutrient solution, respectively, and both supplied with 7% PEG6000 making an osmotic potential of -0.08 MPa to Page 3 of 18 25 Vol.: (0123456789) simulate drought stress (Michel and Kaufmann 1973), being referred to low K (LK) and normal K (NK) treatment. The other group of plants were supplied with the same level of K as NK but without PEG6000 simulated drought stress, being referred to control (CK). After 8 d treatment, two plants with four replications were harvested and dry mass weight (DM) was recorded after oven-dried. Meanwhile, the second leaf from the top to bottom was sampled, and were frozen immediately in liquid nitrogen and stored at -80 ℃ for extraction of endogenous hormone and RNA.

K determination
The oven-dried plants were ground into powder by a grinder, and accurately 0.2 g of sample was dissolved with HNO 3 -H 2 O 2 in an infrared digestion furnace (SKD-20S2, China). The concentration of K was determined using ICP-OES spectrometer (Agilent, America).

Net photosynthesis rate (pn) determination
Pn was measured on second uppermost fully expanded leaf with an LI-6400 portable photosynthesis system (LI-COR, Lincoln, NE, USA) after 7 d drought stress. The photosynthetically active radiation (PAR) in leaf chamber was 1000 µmol m − 2 s − 1 and carbon dioxide concentration was 430 µmol mol − 1 and at an adjusted leaf chamber temperature of 25 °C.

Concentration of phytohormone determination
Concentration of phytohormone were determined by enzyme-linked immunosorbent assay (ELISA). Briefly, accurate 0.2 g fresh leaf was ground in liquid nitrogen. The powders were mixed with 1.6 mL phosphate-buffered saline (pH 7.4) and were centrifuged at 8000 rpm for 10 min. Subsequently, the supernatant was collected to measure the content of indoleacetic acid (IAA), cytokinin (CTK), gibberellin (GA) and ABA using corresponding ELISA Kits (Shanghai Enzyme-linked Biotechnology Co., Ltd. Shanghai, China).

RNA-Seq library preparation, sequencing and data analysis
The total RNA was extracted from the harvested plant tissues using Trizol reagent (Invitrogen, CA, USA). A total of 1 µg RNA was purified by poly-T oligoattached magnetic beads and was used to construct cDNA library. Sequencing libraries were generated using NEBNext UltraTM RNA Library Prep Kit for Illumina (NEB, USA) following manufacturer's recommendations and index codes were added to attribute sequences to each sample. The PCR products were purified (AMPure XP system) and library quality was assessed on the Agilent Bioanalyzer 2100 system. The clustering of the index-coded samples was performed on a cBot Cluster Generation System using TruSeq PE Cluster Kit v4-cBot-HS (Illumia) according to the supplier's instructions. The library was sequenced by using an Illumina HiSeqXten platform with a paired-end sequencing length of 150 bp (PE150) at Biomarker Technologies (Beijing, China). The raw data was deposited at NCBI and accession number was PRJNA901224 (https:// www. ncbi. nlm. nih. gov/ sra/ PRJNA 901224). Clean data was obtained by removing reads containing adapter, reads containing ploy-N and low quality reads from raw data. At the same time, Q-score, GC-content and sequence duplication level of the clean data were calculated. The clean reads were then mapped to the reference Brassica napus genome sequence (https:// www. genos cope. cns. fr/ brass icana pus/ data/) by Hisat2 (Kim et al. 2015). Quantification of gene expression levels were estimated by fragments per kilobase of transcript per million fragments mapped (Robinson et al. 2009). The gene expression meeting the criteria of absolute log2 (fold change) > 1 and false discovery rate (FDR) < 0.05 were determined to be significantly different expression.
Gene ontology (GO) enrichment and Kyoto encyclopedia of genes and genomes (KEGG) analysis GO enrichment analysis of the differentially expressed genes (DEGs) was implemented by the GOseq R packages based Wallenius non-central hyper-geometric distribution (Young et al. 2010). The statistical enrichment of differential expression genes in KEGG (http:// www. genome. jp/ kegg/) pathways was conducted by KOBAS software (Mao 25

Statistical analysis
Significant differences of physiological traits among treatments and genotypes was determined by Fisher's least significant difference test (LSD) and ANOVA analysis was conducted according to the general linear model (GLM) by SAS software. And the difference at P < 0.05 and P < 0.01 was considered as significant and highly significant, respectively.

Genotype difference in physiological index and phytohormone under NK and LK
Rapeseed growth was hindered significantly after 8 d drought stress. Biomass of Chuanyou36 and Youyan57 decreased by 17.8% and 10.8% at NK, while larger loss was observed at LK which decreased by 31.4% and 23.7% compared with those of CK, respectively (Table 1). Youyan57 showed higher biomass and less reduction compared with those of Chuanyou36 both at LK and NK. Accordingly, LK impeded the drought tolerance for both the two rapeseed cultivars, and Youyan57 showed higher drought tolerance compared with Chuanyou36. Large genotypic difference in K concentration was observed between Youyan57 and Chuanyou36 under NK and LK. Tissue K concentration of Youyan57 was 43.63 mg g − 1 and was significantly higher than that of Chuanyou36 recorded 38.09 mg g − 1 at NK treatment. K concentration decreased by 29.1% under LK compared to NK averagely, being 30.40 mg g − 1 for Youyan57 which was significantly higher than that of Chuanyou36 recorded 27.40 mg g − 1 . Photosynthesis decreased significantly under drought stress, and the loss in photosynthesis was more severely under LK compare to NK. Meanwhile, Youyan57 kept stronger photosynthesis under LK, being 1.73-times in Pn to that of Chuanyou36. K levels affected phytohormone significantly, and showed wide genotypic difference between the two cultivars. IAA content decreased significantly under drought stress. Compared with NK, IAA content decreased significantly under LK both for Youyan57 and Chuanyou36, and greater decline was observed in Chuanyou36. Similarly, CTK content decreased significantly under LK compared to that of NK, while Youyan57 kept higher CTK content both under NK and LK. The GA content increased under drought stress, and Chuanyou36 kept higher GA content than that of Youyan57 both under LK and NK. ABA content increased under drought stress. Compared to NK, ABA content increased significantly under LK for both the two cultivars, and no significant difference was observed between the two cultivars.

Identification of DEGs between contrasting cultivars under NK and LK
The paired-end reads were sequenced using the Illumina platform in order to obtain the whole transcriptome profiling of the two varieties under drought stress. A total of 75.23 Gb clean reads were generated from 12 leaf samples by sequencing cDNA libraries (Table S1). The clean data of each sample reached up to 5.95 Gb, and Q30 rate was all above 93.59%, indicating that the quality of sequencing data was sufficient to support further transcriptome analysis. Clean reads of each sample were sequenced with the assigned reference genome, and the sequence alignment efficiency ranged from 87.88 to 90.29%. All mapped reads were categorized into two classes, i.e. multiple mapped and unique mapped reads. Of the 35 ~ 40 million clean reads from each library, 83.64 ~ 85.89% was mapped to unique locations, whereas 4.11 ~ 4.87% was mapped to multiple locations in the genome. Principal component analysis (PCA) was performed to assess the overall relationships of transcriptome in different treatments. The transcription profile between Chuanyou36 and Youyan57 could be clearly divided (Fig. S1), and the transcription profile under NK and LK of Chuan-you36 could be divided by component 1, which suggested high correlation of repeatability among biological replicates. Gene expression profiles of the drought-sensitive and drought-tolerant rapeseed leaves in response to different K levels under drought stress were analyzed. Fragments per kilo-base of transcript per million fragments mapped (FPKM) was used to normalize gene expression levels. Meanwhile, DEGs were identified by comparing the expression levels of Youyan57 to those of Chuanyou36 according to log2 fold change > 1 and false discovery rate (FDR) < 0.05. There were 1689 DEGs between Youyan57 and Chuan-you36 under NK treatment, with 910 up-and 779 down-regulated, respectively. Compared with NK, the number of DEGs decreased by 37.8% recorded 1050 under LK, and with 594 up-and 456 down-regulated, respectively (Fig. 1A). A total of 461 DEGs were commonly expressed and showed the same expression pattern both in NK and LK. There were 1228 and 589 unique DEGs in NK and LK, respectively (Fig. 1B). The findings indicated that there was fundamental differences in gene expression levels within contrast varieties in response to drought with differ K levels.
Gene ontology (GO) enrichment analysis of DEG under NK and LK GO annotation analysis of the DEGs under NK showed that these DEGs could be divided into cellular component, molecular function and biological process 3 main classifications and 53 sub-groups. The top 5 enrichment GO terms in biological process were glucosinolate biosynthetic process, response to light stimulus, regulation of flower development, toxin catabolic process, and response to chitin. The 5 major enrichment GO terms in cellular component were cell part, cell wall, intracellular part, cytosolic small ribosomal subunit, vacuole. And 2-(2'-methylthio) ethylmalate synthase activity, transcription factor activity, ion channel inhibitor activity, 2-isopropylmalate synthase activity, ammonia-lyase activity were the most enriched GO terms within molecular function ( Fig. 2A).
Compared to NK, large difference was observed in GO functional enrichment analysis of DEGs under LK. The top 5 enrichment GO terms in biological process were regulation of flower development, toxin catabolic process, glucosinolate biosynthetic process, response to chitin, carboxylic acid catabolic process. Cell part, intracellular part, cell wall, cytosolic small ribosomal subunit, extracellular region were 5 major categories in cellular component. And 2-(2'-methylthio) ethylmalate synthase activity, transcription factor activity, ion channel inhibitor activity, oxalate oxidase activity, solute: cation symporter activity were the most enrichment terms within molecular function (Fig. 2B).

KEGG analysis of DEGs identified in NK and LK
Totally 427 DEGs were annotated in KEGG pathways under NK. The KEGG pathway enrichment analysis showed that no pathways were significantly (P < 0.05) enriched in up-regulated genes under NK (Fig. 3A), while the down-regulated DEGs were significantly enriched in 2-oxocarboxylic acid metabolism, glucosinolate biosynthesis, cysteine and methionine metabolism, biosynthesis of amino acids, and valine, leucine and isoleucine biosynthesis (Fig. 3B). Meanwhile, biosynthesis of amino acids was the pathway with the largest number of annotated DEGs under NK, being recorded 45 (Fig. 4A). The results indicated that the down-regulated genes was mostly related with biosynthesis of amino acids under NK. There were 225 DEGs annotated in KEGG pathways under LK. No pathways were identified significantly enriched in up-regulated DEGs (Fig. 3C), and 3 pathways were significantly enriched in down-regulated DEGs under LK, including 2-oxocarboxylic acid metabolism, glucosinolate biosynthesis, and valine, leucine and isoleucine biosynthesis (Fig. 3D). There were 6 pathways annotated with more than 10 DEGs, including biosynthesis of amino acids, carbon metabolism, ribosome, 2-oxocarboxylic acid metabolism, cysteine and methionine metabolism, plant hormone signal transduction pathway (Fig. 4B).
DEGs related to biosynthesis of amino acids and carbon metabolism A total of 45 DEGs (13 up-and 32 down-regulated) were identified involved in biosynthesis of amino acids under NK (Fig. 5). Large number of the down-regulated DEGs were identified related to cysteine and methionine metabolism, arginine and proline metabolism. By contrast, only 18 DEGs (8 up-and 10 down-regulated) were identified involved in biosynthesis of amino acids under LK. Besides, there were 27 DEGs (5 up-and 22 down-regulated) and 15 DEGs (3 up-and 12 downregulated) were identified involved in 2-oxocarboxylic acid metabolism under NK and LK, respectively, and most of the DEGs were related to biosynthesis of amino acids and glucosinolate biosynthesis (Fig. S2). A total of 14 DEGs (3 up-and 11 down-regulated) were identified involved in starch and sucrose metabolism under NK (Fig. 6A). Genes encoding β-glucosidase, acid β-fructofuranosidase and sucrose-phosphate synthase were down-regulated under NK. However, no significant difference was observed in these genes under LK. By contrast, 2 starch synthase and 1 alpha amylase were down-regulated under LK. Meanwhile, the genes encoding trehalose-phosphate synthase and trehalose-phosphate phosphatase were up-regulated both under LK and NK.
DEGs associated with glutathione metabolism, glucosinolate biosynthesis, photosynthesis and oxidative phosphorylation There were 16 DEGs (4 up-and 12 down-regulated) and 7 DEGs (4 up-and 3 down-regulated) involved  in glutathione metabolism under NK and LK, respectively (Fig. 6B). Interestingly, the up-regulated genes mainly related to glutathione synthase and 6-phosphogluconate dehydrogenase, while the downregulated genes was mostly related to glutathione S-transferase both under NK and LK. There were 9 and 5 DEGs were identified associated with glucosinolate biosynthesis under NK and LK, and all the DEGs were down-regulated (Fig. 6 C), including genes encoding cytochrome P450 and methylthioalkylmalate synthase. The results indicated that glucosinolate biosynthesis in Chuanyou36 was enhanced than that of Youyan57. There were 7 DEGs (6 upand 1 down-regulated) associated with photosynthesis under NK, including 5 up-regulated chlorophyll a-b binding protein (Fig. 6D). Five DEGs (all upregulated) were identified related to photosynthesis under LK, including 2 oxygen-evolving enhancer protein, 1 ferredoxin, and 1 ferredoxin-NADP reductase. A total of 15 DEGs (6 up-and 9 down-regulated) were identified involved in oxidative phosphorylation under NK (Fig. 6E). The down-regulated DEGs were mostly related to NADH dehydrogenase and cytochrome, while the up-regulated DEGs were mostly related to ATP synthesis. There were 6 DEGs (4 up-and 2 down-regulated) were identified involved in oxidative phosphorylation under LK, and the upregulated DEGs were related to NADH dehydrogenase, cytochrome and ATP synthesis.

DEGs associated with plant hormone signal transduction
Large difference was observed in DEGs associated with plant hormone signal transduction between NK and LK. The total regulated genes reduced under LK, however, DEGs involved in plant hormone signal transduction were increased (Fig. 4). Seven up-regulated DEGs were identified related to plant hormone signal transduction under NK, including 4 auxin responsive genes, 1 abscisic acid receptor PYL8 gene, 1 EIN3-binding F-box protein (Fig. 6 F). Protein phosphatase 2 C and ethylene-insensitive protein 2 were down-regulated under NK. By contrast, 4 up-regulated DEGs were identified related to plant hormone signal transduction under LK, including 2 auxin responsive protein, 1 auxin responsive GH3 gene and 1 EIN3-binding F-box protein 2. And 6 down-regulated DEGs were identified related to plant hormone signal transduction under LK, including 5 Protein phosphatase 2 C. Furthermore, 2 ABC transporter B family member genes were down-regulated under LK (Fig. 6 F).

DEGs encoded transcription factors (TF) and protein kinase
TFs and protein kinase was identified by iTAK according to Zheng et al. (2016). The GO enrichment analysis showed that term of transcription factor activity was enriched significantly both under NK and LK (Fig. 2). A total of 60 DEGs (36 up-and 24-down regulated) and 39 DEGs (24 up-and 15 down-regulated) related to TFs were identified under NK and LK, respectively (Fig. S3). The TFs belonged to 24 families, and mainly involved in MYB, AP2/ERF, M ADS, GRAS, HSF, NAC and B3. Furthermore, large number of DEGs related to MYB and MADS were up-regulated both under NK and LK, suggesting that they could play vital roles in rapeseed drought tolerance. Besides, a total of 58 and 35 differentially expressed protein kinase were identified under NK and LK, respectively, and mainly grouped into receptor-like kinase (RLK). Most of the RLKs were up-regulated under LK, such as 7 RLK _LRR and 3 RLK-_ DLSV. The results indicated that RLKs could play important role in response to drought for rapeseed.

Discussion
The growth of rapeseed was stunted under drought stress and the loss in biomass was more seriously under K deficiency. Youyan57 kept much higher K level in tissue and larger biomass both under NK and LK, showing higher drought tolerance than Chuan-you36. Large difference was observed in transcriptom profile between the two contrast cultivars in response to drought stress, being 1689 and 1050 DEGs identified between Youayn57 and Chuanyou36 under NK and LK, respectively. Furthermore, there were 461 commonly expressed DEGs, and 1228 and 589 unique DEGs were detected in NK and LK, respectively, indicating that K deficiency hindered ability of transcription regulation and induced diverse transcriptome in rapeseed encountering drought stress.
K is the most abundant inorganic cation which can make up 10% of plant's dry weight. It is a primary 25 Page 12 of 18 Vol:. (1234567890) cellular osmoticum and also plays important role in neutralization of negative charges (Karley 2010). Reduced K concentration in plant tissues should lead to disturbance of metabolism, and amino acids are the most important contributor of osmoregulation in addition to K for plants (Morgan 1992). In this study, the drought tolerant cultivar Youyan57 kept higher K level than Chuanyou36, being 14.5% and 10.9% over than that of Chuanyou36 under NK and LK, respectively. By contrast, large number of DEGs associated with biosynthesis of amino acids were downregulated under NK. The results indicated that the drought-sensitive cultivar Chuanyou36 largely relied on biosynthesis of amino acids for osmoregulation rather than K under drought stress. However, the great enhancement of transcription regulation involved in amino acids biosynthesis was eliminated in Chuan-you36, with much less DEGs identified under LK. A previous study found that biosynthesis of amino acids was increased both for Youyan57 and Chuanyou36 under K deficiency and no significant difference was observed in total content of amino acids between the two cultivars under LK (Zhu et al. 2020). We deduced that K deficiency hindered the metabolism processes of Chuanyou36 more seriously and further affected amino acids biosynthesis under drought stress.
Carbohydrates are the ultimate source of carbon skeleton for biosynthesis of amino acids (Stewart et al. 1966), and carbon starvation is observed with increased accumulation of amino acids (Smith and Stitt 2007). Sucrose phosphate synthase is the key rate-limiting enzyme in sucrose synthesis which controls sucrose content in plants (Anur et al. 2020). β-fructofuranosidase or invertase catalyzes the hydrolysis of sucrose into fructose and glucose (Pedezzi et al. 2014). β-glucosidase catalyzes the hydrolysis of cellobiose and cellooligosaccharides containing (1, 4)-β-glycosidic bonds to glucose (Huang et al. 2021). Glucose-1-phosphate uridylyltransferase catalyzes the formation of UDP-glucose from glucose-1-phosphate and UTP (JB and HM 2007). In this study, sucrose phosphate synthase, β-fructofuranosidase, glucose-1-phosphate uridylyltransferase and three β-glucosidases were down regulated under NK, while no significant difference was observed in these genes under LK. Instead, two DEGs encoding starch synthase which related to starch synthesis were down-regulated under LK. The enhanced expression of genes related to production of glucose and sucrose in Chuanyou36 was benefited to biosynthesis of amino acids under NK. Trehalose is a non-reducing disaccharide of glucose which plays an important role in mediating biotic and abiotic stresses in plants, and improved biosynthesis of trehalose by overexpression of trehalose phosphate synthase/phosphatase gene can enhance crops tolerance to drought and saline stresses (Joshi et al. 2020;Lyu et al. 2013). Trehalose-phosphate synthase and trehalose-phosphate phosphatase were up-regulated both under NK and LK, indicating that drought tolerant cultivar Youyan57 could accumulated more trehalose in response to drought stress.
Glutathione (GSH) is an important antioxidant to detoxify reactive oxygen species in plant under stress, and improved expression of glutathione synthetase (GS) enhances the glutathione pool which results in greater tolerance to environmental stresses (Park et al. 2017). Glutathione S-transferases (GST) play major roles in oxidative stress metabolism, which catalyze the conjugation of reduced GSH to electrophilic substrates (Cummins et al. 2011). Drought stress could induce the expression of GST and improved tolerance was observed in GST overexpression plants (Srivastava et al. 2019). Different molecular regulation mechanism was observed in Youyan57 and Chuanyou36 under NK and LK. GS and 6-phosphogluconate dehydrogenase gene which promoted glutathione biosynthesis were both up-regulated under NK and LK, indicating that Youyan57 was capable of biosynthesis of GSH under drought stress. A total of 11 GST were down-regulated under NK, and the expression pattern of GST was affected largely under LK with 3 GST down-regulated. The results indicated that Youyan57 could improve GSH pool in tissues to eliminate the excess reactive oxygen radicals, while Chuanyou36 adapted to drought stress through consuming of GSH by enhanced expression of GST which was sensitive to K deficiency. Glucosinolate is a kind of secondary metabolites found in Brassicaceae that protect plants from herbivory and pathogen attack, and also plays important roles in drought tolerance (Mohammad et al. 2019). Drought-induced accumulation of glucosinolate in leaves directly or indirectly controls stomatal closure to prevent water loss in rapeseed (Seung Hee et al. 2018). Cytochrome P450 plays important role in yield of glucosinolate in plants (Kai et al. 2011). Methylthioalkylmalate synthase catalyzes the committed step in the side chain Page 13 of 18 25 Vol.: (0123456789) elongation of methionine, yielding important precursors for glucosinolate biosynthesis in Brassicaceae species (Kraker and Gershenzon 2011). In this study, 3 cytochrome P450 and 5 methylthioalkylmalate synthase were down-regulated under NK. Meanwhile, 2 cytochrome P450 and 3 methylthioalkylmalate synthase were down-regulated under LK. Coincidently, we found that K deficiency led to massively accumulation of glucosinolate in seeds, and Youyan57 kept much less glucosinolate in seeds than that of Chuanyou36 under drought stress with different K levels in our later study (Zhu et al. 2021). The findings indicated that biosynthesis of glucosinolate was an important mechanism for Chuanyou36 in response to drought stress.
The photosynthesis decreased dramatically under LK compared to NK. Large difference in photosynthesis was observed with two genotypes under NK and LK, and Pn of Youyan57 was 1.73-timefolds to that of Chuanyou36 under LK. Changed molecular regulation mechanism of photosynthesis was observed in response to drought stress under different K status. There were 6 DEGs up-regulated related to photosynthesis under NK, including 5 Chlorophyll a-b binding protein which involved in light harvesting chlorophyll protein complex (Horn et al. 2007). A total of 5 DEGs up-regulated related to photosynthesis under LK, including 1 chlorophyll a-b binding protein, 2 oxygen-evolving enhancer protein involved in photosystem II, 1 ferredoxin and 1 ferredoxin-NADP reductase which involved in photosynthetic electron transportation (Kimata-Ariga et al. 2019). The results indicated that photosynthetic electron transportation was affected seriously in Chuanyou36 under LK as K depletion in tissue. Mitochondrial oxidative phosphorylation (OXPHOS) provides ATP for driving cellular functions, and is composed of 5 protein complex, including NADH dehydrogenase complex (complex I), succinate dehydrogenase complex (complex II), cytochrome c reductase complex (complex III), cytochrome c oxidase complex (complex IV) and ATP synthase complex (complex V). At daytime, OXPHOS takes place in the context of photosynthesis in plants, and NADH produced by photorespiration is the main substrate of respiratory electron transfer chain rather than NADH produced by the TCA cycle (Braun 2020). Besides, plants possesses type II NADH dehydrogenase enzymes bypass the complex I allowing turnover of NADH without translocating protons into OXPHOS, which can avoid electron transfer chain over-reduction and help prevent cellular damage under environmental stress (Rasmusson et al. 2020;Sweetman et al. 2019). In this study, four genes encoding NADH dehydrogenase were down-regulated under NK. By contrast, five up-regulated DEGs were identified related to ATP synthesis under NK, including 2 soluble inoranic pyrophosphatase and 3 ATP synthase. We deduced that enhanced expression of NADH dehydrogenase genes in Chuanyou36 could alleviate photoinhibition under drought stress by sustaining non-phosphorylating pathway of electron transportation. Four out of 6 DEGs associated with oxidative phosphorylation were up-regulated under LK, including NADH dehydrogenase, cytochrome c reductase, soluble inorganic pyrophosphatase, ATP synthase. The results indicated that the drought tolerant cultivar Youyan57 kept higher level of oxidative phosphorylation to biosynthesis of ATP than that of drought-sensitive cultivar both under LK and NK.
The plant hormone IAA regulates many aspects of plant growth and development, including stem elongation, establishment of embryonic polarity, vascular development, and cell expansion. Several major classes of auxin-responsive genes involve in plants quickly sense and respond to changes of auxin levels, including Aux/IAA family, auxin response factor (ARF) family, small auxin upregulated RNA (SAUR ), and the auxin-responsive Gretchen Hagen3 (GH3) family genes (Luo et al. 2018;Spartz et al. 2012). In this study, IAA content decreased under drought stress and no significant difference was observed between the two cultivars under NK. There were 4 up-regulated DEGs related to auxin signal transduction under NK, including 2 auxin-responsive protein IAA9, 1 auxin responsive GH3 family gene, and 1 auxin-responsive protein SAUR gene. Compared to NK, IAA content at LK decreased significantly, and IAA content in Youyan57 was significantly higher than that of Chuanyou36. Furthermore, 3 out of 4 DEGs were up-regulated and involved in auxin signal transduction under LK, including 2 AUX/IAA family genes and 1 GH3 auxin-responsive gene. Thus, the drought tolerant cultivar Youyan57 kept higher level of IAA in tissue and enhanced IAA signal transduction than that of Chuanyou36 under LK. The plant hormone ABA plays a crucial role in response to drought stress, which can induce stomata movement and thus reduce water loss, ROS scavenging enzymes, and proline accumulation (Chen et al. 2021). Compared to NK, ABA content increased significantly under LK, and no significant difference was observed in ABA content between the two cultivars. Instead, large difference was observed in ABA signal transduction between Youyan57 and Chuan-you36. Protein phosphatase 2 C is a negative regulator of ABA signal transduction and functions as a switch at the center of the ABA signaling network (Chen et al. 2021). Protein phosphatase 2 C counteracts SnRK2 kinases by physical interaction, and thereby inhibit activation of the transcription factors that mediate ABA-responsive gene expression. Under osmotic stress conditions, Protein phosphatase 2 C binds to pyrabactin resistance 1 (PYR1)/ PYR1-LIKE (PYL)/ regulatory components of ABA receptors (RCAR) intracellular ABA receptors to capture ABA and releases active SnRK2s, resulting in phosphorylation of ABFs and activation of other ABA response pathways (Jung et al. 2020;Dupeux et al. 2011). In this study, 1 protein phosphatase 2 C was down-regulated under NK. However, 5 protein phosphatase 2 C genes were down-regulated under LK. We deduced that the enhanced expression of protein phosphatase 2 C disrupted ABA signal transduction under K deficiency in Chuanyou36, which induced defective regulation of stomatal closure, resulting in larger water loss under drought stress. Besides, 2 ABC transporters were down-regulated under LK. ABC transporters are responsible for ABA transportation (Kuromori et al. 2011). It was reasonable to speculate that the enhanced expression of ABA transporters in Chuanyou36 could make up some defective in ABA signaling. MADS TFs have been shown to play essential roles in the adaptation to drought stress for plants (Li et al. 2020). MADS functions as a positive regulator in response to drought stress via ABA signaling, and its transcriptional activity increased by interacting with SnRK2 kinase in an ABA-dependent manner (Li et al. 2021a). In this study, a total of 5 and 4 MADS up-regulated under NK and LK in Youyan57, which could enhance ABA signaling and improve drought tolerance. Besides, 7 LRR_RLKs were up-regulated in Youyan57 compared to that of Chuanyou36 under LK, which play key roles in ABA signaling transduction in plants response to drought stress (Nagar et al. 2022;Du et al. 2022), indicating enhanced ABA signaling in Youyan57 under LK. The ethylene signal is perceived by ethylene receptors in plants, such as ethylene insensitive 2 (EIN2), and then induces the accumulation of transcription factor EIN3 which is a key positive switch in ethylene perception (Yanagisawa et al. 2003;Binder 2020). EIN3 is recognized for ubiquitination through the identification of two E3 components, EIN3 Binding F-BOX1 (EBF1) and EBF2 proteins. EBF1 and EBF2 interact directly with EIN3 and subsequently promote degradation of EIN3, which is critical not only for proper ethylene signaling but also for growth in plants ( Gagne et al. 2004). EBF1 and EBF2 fine-tune ethylene responses by repressing signaling in the absence of the hormone, dampening signaling at high hormone concentrations, and promoting a more rapid recovery after ethylene levels dissipate (Binder et al. 2007). In our study, the EIN2 was down-regulated under NK, while EBF2 was up-regulated both under NK and LK, indicating that Youyan57 regulated ethylene signaling properly under drought stress. CTK affects many aspects of plant development including cell division, shoot induction, and vascular development. And histidinecontaining phosphotransfer proteins act as positive regulators of CTK signaling in plants (Hutchison and Kieber 2007). Compared to NK, CTK content decreased under LK, and Youyan57 kept higher level of CTK in tissue than that of Chuanyou36 both under NK and LK. A histidine-containing phosphotransfer protein was down regulated under NK, which could compensate the decrease of CTK in Chuanyou36.

Conclusion
Overall, comparative transcriptomic profiling identified a set of genes differentially expressed in drought tolerant rapeseed cultivar Youyan57 charactered by high K status. The genes involved in photosynthesis, biosynthesis of glutathione, IAA signal transduction, oxidative phosphorylation were highly related to drought stress, and was little affected by K deficiency (Fig. 7). Conversely, drought sensitive cultivar relied heavily on biosynthesis of amino acids instead of K participating in osmoregulation, and showed a diverse transcription profile in response to drought stress under K deficiency. The comparatively enhanced expression of protein phosphatase 2 C induced by K deficiency, which negatively regulated ABA signal transduction, led to more severe inhibition in drought