The seeds of C18 (cold-tolerant) and C20 (cold-sensitive) cultivars were provided by the Oil Crops Research Institute, Chinese Academy of Agricultural Sciences (Wuhan, China). Rapeseed seedlings were grown in small pots having soil in a growth chamber with a 16-/8-h light/dark cycle at 22°C. Then, the 21-day-old seedlings (three-leaf stage) were placed in low-temperature incubators at 4°C for 24 h. Cold stress was imposed at a temperature of 2°C for 1 h, and then the temperature was reduced gradually to 0°C, −2°C for 1 h, respectively. The leaf samples from all the seedlings grew under corresponding controls (22°C when sampled under 2°C, 0°C and −2°C), 2°C, 0°C and −2°C were collected, frozen in liquid nitrogen, and stored at −80°C for RNA extraction and further use. Arabidopsis freezing treatment was carried out as described by Zhao et al. (2016). RNA extraction and purification were performed as described in Yan et al. (2019). Using the above-mentioned samples, the RNA-seq was performed using the Illumina Hiseq 2000 (LC, Science, and Houston, TX) platform, and then the raw reads were generated. The whole experiment was carried out with three biological replications.

Raw data were processed with Fast QC 10.1 software and filtered by removing the unqualified reads. The descriptive statistics for the clean data, such as Q20, Q30, and GC content, were calculated. The clean reads of each library were mapped to the Brassica napus L. genome using Hisat 2.0 software to count the mapping rate.¹ The remaining sequences were considered as clean reads (the corresponding data have been submitted into the database of National Center for Biotechnology Information, Short Read Archive (NCBI, SRV) with accession number PRJNA562064). Fragment per kilobase of exon per million fragments mapped (FPKM) value was used to evaluate the gene expression level.

The DEGseq R package was used to identify differentially expression genes (DEGs). The integration of a p < 0.05 and an absolute value of log2 ratio > 1 were used to identify DEGs. Gene ontology (GO)² and MapMan³ analysis were performed for the functional analysis of DEGs.

The above-mentioned RNA samples were used for qRT-PCR analysis. The primers were designed using Primer 5.0 software and listed in Supplementary Table S5. The Brassica napus actin gene-specific primer (GeneBank: AF111812) was used as a control to normalize the expression data (Wang et al., 2014). The qRT-PCR was carried out with the TransStart Tip Green qPCR SuperMix (TransGen, China) on StepOnePlus real-time PCR system (Applied Biosystems, United States). The condition for PCR was as follows: 95°C for 30s and 40 cycles of 95°C for 10s, followed by 60°C for 30s (extension and signal acquisition). The relative expression levels were calculated using the 2^−∆∆Ct method.

Endogenous inositol and ABA concentration were estimated using the ELISA kit produced by You Xuan Biological Technology Co. Ltd. (Shanghai, China). A total of 0.1 g of leaves was homogenized in chilled pestle and mortar using the 50 mmol L⁻¹ phosphate buffer (pH 7.4). The homogenate was centrifuged at 4,000 rpm for 10 min. Then, the supernatant was collected into another tube to measure the inositol concentration following the kit manufacturer’s instructions.

The 21-day-old seedlings of four cold-sensitive cultivars (Zheyou21, Ningyou18, C06, and C20) were pretreated with different concentrations (0.1, 0.5, 1.0, 2.0, 5.0, and 8.0 g L⁻¹) of inositol solution, and the CK plants were treated with water. The seedlings were divided into two groups. In the first group, the seedlings were treated with freezing stress (−2°C) immediately, and the second group seedlings were treated with freezing stress (−2°C) 2 days after the pretreatment with inositol. After the freezing treatment, the survival rate was evaluated as described in our previous report (Yan et al., 2018).

Net Ca²⁺ fluxes of leaves of 21-day-old seedlings were measured at the YoungerUSA Xuyue (Beijing) BioFunction Institute by using Non-invasive Micro-test Technology (NMT100 Series①, YoungerUSA LLC, Amherst, MA 01002, United States) Xuyue (Beijing) Sci. & Tech. Co., Ltd., Beijing, China, and imFluxes V2.0 (YoungerUSA LLC, Amherst, MA 01002, United States) Software (Ma et al., 2015).

The CDS of BnCBL1-2 was cloned by primers 5′- ATGGGCTGCTTCCACTCAA-3′ and 5′-TCATGTGGCAATCTCATCGAC-3′ and inserted into binary vector pBCXUN (ubiquitin promoter) linearized with Sma I and BamH I enzymes. The recombinant plasmid pBCXUN-BnCBL1-2 (Supplementary Figure S5) was introduced into Agrobacterium tumefaciens strain GV3101 using the heat shock method. Agrobacterium-mediated transformation of Arabidopsis was carried out as described by Clough and Bent (1998).

To check the effect of exogenous inositol on cold-responsive markers gene (CBFs, and CORs) and ABA contents, the 21-day-old seedlings of C20 were sprayed with 0.5 g L⁻¹ of inositol solution water (control). Then, the seedlings were divided into two groups. One group continued to grow under 22°C, and the other was treated at 4°C for 1 h. After 1 h, green leaves were harvested and were stored at −80°C for the qRT-PCR.

Similarly, the 21-day-old seedlings of C20 were used to evaluate the transient effect of exogenous inositol on CBLs genes. The seedlings were subjected to the cold stress as 4°C for 1, 2, 4, 8, 12, and 24 h, 2°C for 1 h, 0°C for 1 h, and −2°C for 1 h. Green leaves were harvested at each time point and were stored at −80°C for the qRT-PCR. Meantime, the seedlings with the pretreatment of inositol and water under 22°C were also harvested as the CK. For durative effect, the cold stress was imposed after cold acclimation of 4°C for 24 h to 21-day-old seedlings of C20. The cold stress treatment was set as 2°C for 1 h, 0°C for 1 h, and −2°C for 1 h. The pretreatment of inositol and the harvest of samples were the same as the explained earlier. The primers used for qRT-PCR are shown in Supplementary Table S5.

Statistical analysis were performed using SPSS statistical package (SPSS Student version 15.0). Significant differences were evaluated using the two-tailed Student’s t-test or one-way ANOVA and Duncan’s test. All test differences at p ≤ 0.05 were considered to be significant. All the error bars were SD (Standard Deviation) value.

C18 and C20 were verified to be cold-tolerant and cold-sensitive cultivars in previous morphological and physiological experiments (Yan et al., 2018, 2019). To identify the molecular mechanism of cold tolerance, the two cultivars were exposed to 2°C, 0°C, −2°C, and 22°C (CK, control) for cold stress treatment and 36 cDNA libraries (two genotypes × six treatments × three biological replicates) were constructed for RNA-seq (Supplementary Figure S1). After a series of processing, such as filtering and quality control (Supplementary Table S1), the sequencing data were used for further analysis. With the criterion of |log₂FC| > 1 and q value < 0.05, a total of 31,392 DEGs were identified between cold stress and corresponding control.

Further, these DEGs were subjected to GO enrichment. The results of GO molecular function enrichment showed oxidoreductase activity, asparagine synthase (glutamine-hydrolyzing) activity, NADP binding, and phosphoric ester hydrolase activity were enriched in S1, S2, S3 (cold-sensitive cultivar C20 under 2°C, 0°C, −2°C compared to 22°C); oxidoreductase activity, asparagine synthase (glutamine-hydrolyzing) activity, catalytic activity, glutamate–ammonia ligase activity, and inositol-3-phosphate synthase activity were enriched in T1, T2, T3 (cold-tolerant cultivar C18 under 2°C, 0°C, −2°C compared to 22°C; Figure 1). The functional enrichment analysis by Mapman indicated that the pathway of inositol phosphate was enriched (Supplementary Figure S2). Inositol biosynthesis and degradation were enriched under cold stress, and 35 related DEGs were identified to be differentially expressed (Figure 2A). Of these 35 DEGs, 28 DEGs are involved in inositol biosynthesis, encoding three enzymes (hexokinase HXK, myo-inositol-1-phosphate synthase MIPS, and inositol monophosphate IMPase), and seven DEGs involved in inositol degradation, encoding myo-inositol oxygenase (MIOX; Figure 2A). Among these 35 DEGs, 14 DEGs were randomly selected for qRT-PCR-based validation. The expression patterns for these DEGs were generally consistent between RNA-seq and qRT-PCR data (Supplementary Figure S3), indicating the reproducibility of the RNA-seq data.

Among different HXK genes, two HXK2 genes were upregulated in response to cold stress in both cultivars, and one HXK3 and HXK4 were upregulated in C20. Seven HXK3 were downregulated in both cultivars (C18 and C20; Figure 2A). MIPS catalyzes the conversion of glucose 6-phosphate to myo-inositol 1-phosphate, which is a rate-limiting step of the inositol biosynthetic pathway in plants. All the eight DEGs encoding MIPS were increased in both cultivars under cold stress; however, the genes in C18 had a stronger increasing trend than that in C20. For IMPase, eight upregulated DEGs were identified in C18 under 2°C, while there were only four upregulated genes identified in C20 (Figure 2A). Notably, the expression trends have varied under 0°C and −2°C. For instance, more MIOXs genes were downregulated in C18 than C20 under cold stress. Especially under 0°C, seven downregulated MIOXs were identified in C18, and only three downregulated MIOXs were found in C20. The different expression patterns of genes responsible for myo-inositol biosynthesis and degradation between C18 and C20 might result in higher endogenous inositol accumulation in C18 than in C20 (Figure 2A).

The endogenous inositol concentration in C18 was 18.9 ng g⁻¹ and 20.1 ng g⁻¹ under 2°C and CK, and there was no significant difference. With the continually dropping temperature, the inositol content in C18 increased significantly (22.4 and 22.2 ng g⁻¹ under 0°C and −2°C treatment) compared to control conditions (16.9 and 17.4 ng g⁻¹; Figure 2B). The inositol content of C20 under 2°C (19.2 ng g⁻¹) was higher than that under CK (16.4 ng g⁻¹). Notably, at 0°C, there were no significant differences (18.7 and 18.0 ng g⁻¹). However, the inositol content (15.9 ng g⁻¹) of C20 under −2°C was decreased compared to control (20.4 ng g⁻¹; Figure 2B). In a nutshell, the results showed that endogenous inositol concentration increased in C18 and decreased in C20 with the gradually dropping temperature.

To investigate the effect of inositol on cold tolerance in rapeseed, four representative sensitive cultivars (Zheyou21, Ningyou18, C06, and C20; Yan et al., 2018) were chosen to be treated with exogenous inositol under cold stress. The results presented that all the sensitive varieties with inositol pretreatment were found to be more cold-tolerant than CK (Figures 2C,D). Interestingly, control seedlings were failed to survive under −2°C. At the same time, the treated cultivars showed better performed. The survival rate of Zheyou21, Ningyou18, C06, and C20 was 79, 46, 17, and 46%, respectively, when pretreated with optimum concentration of inositol 1.0 g L⁻¹, 2.0 g L⁻¹, 0.1 g L⁻¹, and 0.5 g L⁻¹, respectively (Figures 2C,D).

To clarify signal transduction of inositol, the relation was explored between inositol and known signaling pathways involved in cold stress, such as the CBF pathway. Therefore, the expression of several markers genes, including CBFs and CORs were analyzed after exogenous inositol pretreatment. Under 22°C, the expression of BnCBF1, BnCBF2, and BnCBF4 was slightly increased after exogenous inositol application (Figures 3A,B). However, at 4°C, the expression of BnCBF1, BnCBF2, and BnCBF4 was decreased dramatically compared with CK. The result showed that exogenous inositol increased cold tolerance, not by increasing the expression of CBFs. Besides, the expression pattern of CORs after exogenous inositol application was also analyzed. Under 22°C, the expressions of BnCOR6.6, BnCOR15, and BnCOR25 increased significantly after exogenous inositol application (Figure 3C). At 4°C, the expression level of BnCOR6.6 and BnCOR15 was increased in inositol-treated seedlings than CK. However, for the expression of BnCOR25, there was no significant difference between the inositol treatment and CK.

Considering the expression of BnCOR25 is regulated by ABA (Chen et al., 2011), the ABA concentration of seedlings after exogenous inositol application was detected. The results showed that, under 0°C and −2°C, there was no significant difference for ABA concentration between inositol treatment and CK (Figure 3D). In addition, the genes involved in CBF or ABA pathway were explored in the transcriptome data, and there was no difference of expression between C18 and C20 cultivars under cold stress (Supplementary Table S2). All of the above evidence indicated that exogenous inositol increased cold tolerance, not through the ABA pathway, but maybe by regulating the expression levels of cold-responsive marker genes.

After excluding the CBF and ABA pathway, it is speculated that the mechanism of inositol improving cold tolerance may be related to Ca²⁺ flux. Therefore the Ca²⁺ flux value in C20 seedlings treated with water and inositol was evaluated. Under 22°C, the initial Ca²⁺ flux value of inositol and water treated seedlings was 370.1 and 60.3 pmol cm⁻² s⁻¹, respectively (the positive value represents Ca²⁺ efflux, and a negative value represents Ca²⁺ influx). Under 4°C, the Ca²⁺ influx of inositol and water treated seedlings increased by 195.5 and 108.7 pmol cm⁻² s⁻¹, respectively, and the Ca²⁺ flux value finally reached 175.6 and − 48.4 pmol cm⁻² s⁻¹, respectively (Figures 4A,B). Compared with the control, inositol treatment led to a stronger calcium influx. The results suggested that inositol increased Ca²⁺ influx under cold stress, resulting in enhanced cold tolerance in rapeseed.

In our study, inositol was proved to increase Ca²⁺ influx under cold stress which resulted in enhanced cold tolerance in rapeseed. In other studies, the combination of different concentrations of inositol and Ca²⁺ significantly promoted the growth of Chinese cabbage and pepper seedlings (Liu et al., 2017; Yang et al., 2017).

Then, we identified all the Ca²⁺ related gens in both the transcriptome of C18 and C20, and 351 DEGs were obtained to be different expressed in C18 and C20 (Supplementary Table S3). It has also been reported that some calcium sensors, such as CBL1, act as a rate-limiting factor to response to stress signals (Cheong et al., 2003). Of these 351 DEGs, there were four CBL1s. The expression of all the four BnCBL1s was increased in C20 under cold stress. There were no significant differences of expressions of BnCBL1-3 and BnCBL1-4 in C18 under cold stress. The expressions of BnCBL1-1 and BnCBL1-2 were decreased in C18 under cold stress (Supplementary Table S4). The result indicated that BnCBL1 might be involved in signaling in cold tolerance. With the pretreatment of exogenous inositol, the expression of BnCBL1-2 decreased in C20 under freezing stress (Supplementary Figure S4). BnCBL1-2 was overexpressed in Arabidopsis, and two lines, BnCBL1-2-1 and BnCBL1-2-8, were selected. After the cold treatment at −4°C for 3 h, approximately 41% of WT and 7–18% of CBL1-overexpressing plants recovered from the freezing treatment The survival rate of transgenic lines was significantly lower than WT plants (Figures 5C,D). The results indicated that overexpression of BnCBL1-2 reduced cold tolerance.

The further experiment showed that overexpression of BnCBL1-2 reduced cold tolerance and Ca²⁺ flux (Figures 5E,F). This research also revealed that inositol inhibited the expression of BnCBL1-2 (BnaA01g08510D) under cold stress. Therefore, it is proposed there is a mechanism that inositol regulates the Ca²⁺ signaling and the expression of CBL1 (Figure 6), which is a brand-new signal pathway under cold stress in plants.

The Ca²⁺ flux was measured in WT and transgenic Arabidopsis with overexpressing BnCBL1-2. Under normal condition (22°C), in mesophyll cells, the initial Ca²⁺ flux values were 94.7, 35.6, and 61.7 pmol cm⁻² s⁻¹, respectively in WT, and overexpressing (BnCBL1-2\u20131 and BnCBL1-2\u20138) transgenic plants. Under cold stress (4°C), the Ca²⁺ flux of WT, overexpressing (BnCBL1-2\u20131 and BnCBL1-2\u20138) plants in mesophyll cells were − 117.2, −101.8, and − 41.1 pmol cm⁻² s⁻¹, respectively (Figures 5E,F). Compared with overexpressing BnCBL1-2 plants, WT plants showed more Ca²⁺ efflux under normal conditions and more Ca²⁺ influx under cold stress, suggesting a stronger ability in Ca²⁺ regulation.

The best understood cold signaling pathway is the ICE-CBF transcriptional cascade, and the CBF genes play crucial roles in this cascade. In this study, three CBF related DEGs, BnaA03g13620D (homologous to AtCBF1), BnaAnng34260D (homologous to AtCBF1), and BnaA10g07630D (homologous to AtCBF4) were identified (Supplementary Table S2). The BnCBF4 (BnaA10g07630D) was overexpressed in Arabidopsis, and the plants had a retarded flowering period (data not shown), which was similar to the previous studies (Jaglo-Ottosen et al., 1998; Gilmour et al., 2004). Overexpression of CBFs not only improved freezing tolerance but also led to dwarfism and late flowering (Jaglo-Ottosen et al., 1998; Gilmour et al., 2004). Therefore, the CBFs are not ideal target genes in transgenic breeding due to their negative effects. Previous transcriptome analysis showed that only 12% of the cold-responsive genes were controlled by CBFs (Fowler and Thomashow, 2002), which indicated there were other unexplored important pathways responding to cold stress.

Inositol, a cyclic polyol, is involved in various physiological processes and participates in programmed cell death, pathogen resistance, and stress adaptation in plants (Bohnert et al., 1995; Chaouch and Noctor, 2010; Bruggeman et al., 2014). In Medicago falcata, the concentration of inositol increased significantly under low temperature, and the inositol concentration in Medicago falcata (cold-tolerant) was higher than Medicago sativa (cold-sensitive) at 5°C (Tan et al., 2013). In animals, inositol was also reported to be involved in cold tolerance. For instance, the inositol concentrations increased up to 400-fold and peaked at 147 nmol mg⁻¹ fresh mass in overwintering flies during the winter (Vesala et al., 2012). Previous investigations also reported that inositol alleviated the damage of cold stress on maize and rice seedlings (You et al., 2000; Guo et al., 2014; Yao and Xia, 2014). When inositol was used as a seed coating, it significantly increased the germination rate and fresh weight in maize under cold stress (Yu et al., 2009). Notably, the combination of different concentrations of inositol and Ca²⁺ significantly promoted the growth of Chinese cabbage and pepper seedlings (Liu et al., 2017; Yang et al., 2017). In this study, inositol-3-phosphate synthase activity was enriched in T1, T2, T3, and S2 (Figure 1), and C18 showed higher inositol biosynthesis gene expression and lower degradation gene expression than in C20 (Figure 2A). Further, the endogenous inositol concentration was increased in C18, and decreased in C20 with the continually dropping temperature (Figure 2B), suggesting the crucial role of inositol in cold tolerance. Besides, exogenous application of inositol increases cold tolerance of all four sensitive rapeseed cultivars (Figures 2C,D). Briefly, these results indicated that inositol played an important role in cold tolerance in rapeseed.

As a secondary messenger, Ca²⁺ are participate in many biological processes and plays a critical role in signaling pathways under stress conditions in plants. Under stress, stress signals typically boost the Ca²⁺ levels over the threshold and activate calcium sensors. In this study, inositol positively regulated cold tolerance by increasing Ca²⁺ influx in rapeseed. Inositol is an important precursor of inositol phospholipids (Downes and Macphee, 1990). Phosphatidylinositol (4, 5) P2 are hydrolyzed by phospholipase C (PLC) to InsP3 and diacylglycerol (DAG). In animal cells, InsP3 activate InsP3 receptor (Ca²⁺ channel) and release Ca²⁺ (Valluru and Van den Ende, 2011). These outcomes suggest that there is a crosstalk between the signal pathway of inositol and Ca²⁺ in improving cold tress tolerance.

Some calcium sensors, such as CBL1, act as a rate-limiting factor in response to stress signals (Cheong et al., 2003). In Arabidopsis, CBL1 improved the salt and drought stress tolerance but decreased cold stress tolerance (Cheong et al., 2003). Overexpression of BnCBL1 improved tolerance to high salinity and low phosphate conditions in rapeseed (Chen et al., 2012). In this study, inositol increased Ca²⁺ influx under cold stress (Figures 4A,B), 351 Ca²⁺ DEGs and found CBL1s had different expression trend in both cultivars. Among different CBL1s, there were no significant differences in the expression levels of BnCBL1-3 and BnCBL1-4. On the other hand, the expression levels of BnCBL1-1 and BnCBL1-2 were decreased in C18 under cold stress, whereas the expressions of four BnCBL1s increased in C20 under cold stress (Supplementary Table S4). Moreover, exogenous inositol decreases the expression of BnCBL1-2 in C20 under freezing stress (Supplementary Figure S4). Overexpression of BnCBL1-2 reduced cold tolerance and Ca²⁺ flux (Figures 5E,F). It was assumed that overexpression of BnCBL1-2 may disturb the Ca²⁺ homeostasis in plants, leading to the failure in sending cold signals to the downstream signaling pathway. This research also revealed that inositol inhibited the expression of BnCBL1-2 (BnaA01g08510D) under cold stress. Therefore, it is proposed there is a mechanism that inositol regulates the Ca²⁺ signaling and the expression of CBL1 (Figure 6), which is a brand-new signal pathway under cold stress in plants.