Background

Plants must sense and respond to various perturbations or stresses by properly altering their cellular and physiological status to maintain normal growth and development in such adverse environments. As is well known, this adaptation process is associated with two critical cellular components, calcium ion (Ca2+) and Ca2+-binding proteins. This external stimulus is transduced into a ubiquitous and distinct Ca2+ signature through regulating to specific temporal and spatial changes in cytosolic calcium concentration [(Ca2+)cyt] in plant cells (Evans et al. 2001; Sanders et al. 2002). Although the Ca2+ signature plays a vital role in stimulus–response coupling, Ca2+-binding proteins decode and relay the information encoded by the Ca2+ signature, which serves as a Ca2+ sensor, and perceives the specific Ca2+ signature and transmits these signals to downstream pathways to response-specific stress (Luan et al. 2002; Weinl and Kudla 2009).

The Ca2+ sensor harboring the canonical elongation factor hand (EF-hand) Ca2+-binding motifs can be classified into four families: calmodulin (CaM), CaM-like proteins (CMLs), calcineurin B-like (CBL) proteins and Ca2+-dependent protein kinases (CDPKs)(Luan 2009; McCormack et al. 2005; Schulz et al. 2013). Among these, CDPKs possess a kinase domain modulated by an EF-hand domain, thereby forming ‘sensor-response’ modules (Harper and Harmon 2005; Sanders et al. 2002). By contrast, the rest of the family and their interacting kinases are divided into two modules: Ca2+-binding function and kinase activity (Burstenbinder et al. 2017). Therefore, CBL proteins that were activated by binding the calcium signature can physically interact with the NAF (Asn-Ala-Phe) domain of the CBL-interacting protein kinases (CIPKs) (Chaves-Sanjuan et al. 2014). As a calcium sensor protein for Ca2+ binding, CBL in plants are similar to neuronal calcium sensors and calcineurin B (CNB) from animals (Kudla et al. 1999), harboring four EF-hand domains that are characterized by a conserved Asp (D) and Glu (E) residue with a completely constant spacing (Batistic and Kudla 2009; Kolukisaoglu et al. 2004). In addition, the short N-terminal of CBL proteins (CBL1, − 4, − 5, − 8 and − 9) contain conserved MGCXXS/T motifs that contributed to the CBLs to anchor in the membrane to transduce Ca2+ signal after lipid modification of myristoylation (Batistic et al. 2008; Weinl and Kudla 2009). While the extended N-terminus of CBL proteins (vacuolar-targeted CBL2 and -6) had another recognizable lipid modification motif (S-acylation) (Batistic et al. 2012; Zhang et al. 2017). Moreover, the C-terminal of CBL proteins harbor a conserved FPSF domain (because of the presence of conserved P, M, L, F, P and F residues) that are the target of phosphorylation by CIPK, which might generate additional specificity to the CBL–CIPK interaction (Mohanta et al. 2015; Sanyal et al. 2015).

To date, the CBL protein was first identified in Arabidopsis thaliana (Kolukisaoglu et al. 2004), which was followed by identification in Populus (Zhang et al. 2008), rice (Gu et al. 2008), cotton (Lu et al. 2017), grapevine (Xi et al. 2017), canola (Zhang et al. 2014), eggplant (Li et al. 2016), Physcomitrella patens (Weinl and Kudla 2009) and Selaginella moellendorffii (Weinl and Kudla 2009). The ten CBLs encoded in the Arabidopsis and rice genome whose four introns are absolutely conserved in phase and position (Kolukisaoglu et al. 2004). CBL1 mediates cold and low K+ stresses response via CIPK-coupled signaling in Arabidopsis (Huang et al. 2011; Li et al. 2006). Moreover, CBL1 also contributes to rescue the salt and osmotic stress in Sedirea japonica (Cho et al. 2018). CBL1 and CBL9 play an important role in regulating K+ homeostasis in root and stomata through CIPK23 (Cheong et al. 2007) and also affects pollen development (Mahs et al. 2013). It is noteworthy that CBL10 modulates K+ homeostasis by directly interacting with Arabidopsis K+ transporter 1 (AKT1) (Ren et al. 2013). In addition, OsCBL10 confers flooding tolerance during seed germination with enhanced Ca2+ flow (Ye et al. 2018). CBL1 and CBL4 specifically mediate salt stress signaling and membrane H+ transport (Liu and Zhu 1998). CBL2 and CBL3 not only controls ion homeostasis (Tang et al. 2012), but also modulates plant seed size and embryonic development (Eckert et al. 2014). CBL5 can increase plant’s drought stress tolerance (Cheong et al., 2010) and CBL7 can affect plant responses to low nitrate in Arabidopsis (Ma et al. 2015).

The grass family consists of large and nearly ubiquitous monocotyledonous plants that can be used as forage, building materials, fuel, and food. A great deal of effort has been made to investigate the role of the CBL gene family in some plant species, but a more detailed evolution and phylogeny of CBL protein in monocot, including grasses, has not been described yet. Here, we investigated the phylogenetic relationships and adaptive evolution of the 152 CBL genes from 15 grass species at the genome-wide level. Additionally, the expression patterns and the function in response to salt and drought stresses of these genes are examined in rice.

Materials and Methods

Identification of CBL Gene Family Members in Grasses

The ten CBL gene information from the model plant Oryza sativa were searched from the TIGR rice Genome Annotation Resources database (https://rice.plantbiology.msu.edu/) (Ouyang et al. 2007). To identify CBL genes in other grass species, BLASTP searches were performed against orthologous protein sequences using rice CBLs as independent probes (Altschul et al. 1997) in phytozome (https://www.phytozome.net/) (Goodstein et al. 2012) and public PLAZA (https://bioinformatics.psb.ugent.be/plaza/) (Van Bel et al. 2018). The CBL gene was only considered as a candidate when it harbored four EF-hand domains, which was subjected to domain analyses by scanning in InterPro software (de Castro et al. 2006). Moreover, all information was examined for redundancy and no alternative splice variants were selected. All retrieved non-redundant coding and genomic sequences were collected in 15 grass species.

Sequence Alignment, Protein Motif, Phylogenetic Tree and Molecular Evolution Analyses

Multiple sequence alignments of the protein or DNA sequences of all CBL genes were performed by the Clustal Omega (https://www.ebi.ac.uk/Tools/msa/clustalo/) with default settings. The alignment logos of the protein conserved domain were viewed by WebLogo (https://weblogo.threeplusone.com/). The domain analyses were executed using InterProScan against the protein database (https://www.ebi.ac.uk/interpro/). The theoretical pI (isoelectric point) and Mw (molecular weight) of CBLs were measured using Compute pI/Mw tool online (https://web.expasy.org/compute_pi/). Phylogenetic analyses were performed by the maximum-likelihood (ML) method with a Jones–Taylor–Thornton (JTT) model and the 2000 iterations bootstrap test in the MEGA 6.0 software (Tamura et al. 2013). To insure the topology of the ML tree with more divergent domains, all positions with 95% site coverage were eliminated. The aligned genomic sequences were applied to assess the value of Ka and Ks as well as their ratios by the DNASP v5.10 software (Librado and Rozas 2009). To study the genetic divergence between each group, we counted the genetic distances based on the amino acid sequences with the JTT model in the MEGA 6.0 (Tamura et al. 2013). Their overall mean distances were also measured according to the related MEGA file of phylogenetic tree of grass CBLs.

Plant Growth, Treatments and Expression Analysis

Gene expression analyses of rice cultivar Zhonghua 11 (ZH11) were investigated in different tissues and response to abiotic stresses and hormone treatments. It is important that high-frequency genetic transformation is established for the generation of transgenic lines and phenotype analyses in our laboratory. Pre-germinated seeds were placed in water for 2 days that were then transferred into a soil mix. All rice plants have a growth condition with a cycle of 26 °C and 14 h light (> 3000 lx)/22 °C and 10 h dark. For real-time quantitative PCR (RT-qPCR) analyses, 2-week-old rice plants were used for harvesting root, shoot, leaf blade, leaf sheath, shoot apical meristem (SAM) at vegetative or reproductive stage and nine different flower stages. For phytohormone analysis, 2-week-old seedlings were treated in MS liquid medium containing 50 μM TIBA, 100 µM SA, 50 µM IAA, 100 µM ABA and 50 µM GA for 0.5 h, 1 h, 3 h, 6 h, 12 h and 24 h, respectively. For abiotic stress treatment, 2-week-old seedlings were treated in MS liquid medium containing 50 µM CuSO4, 25 µM K2Cr2O7, 20% PEG 6000 and 300 mM NaCl for 0.5 h, 1 h, 3 h, 6 h, 12 h and 24 h, respectively. In addition, seedlings were transferred to a growth chamber at 4 °C for cold treatment. Untreated plants were used as control samples. The seedlings were sampled at different time points after treatment and stored at − 80 °C until use. Total RNA was extracted using TRIzol reagent and reverse-transcribed into cDNA using PrimeScript RT Master Mix Perfect Real Time (TaKaRa) followed the manufacturer’s instructions. The total RNA was quantified using Nanodrop1000 and its integrity was assessed by electrophoresis in 1.5% (w/v) agarose gel. The RT-qPCR was performed in a 10 µl reaction with 5–50 ng of first-stand cDNA products (4 µl), 5 pmol of each primer (0.4 µl), 5 μl SYBR Green Master Mix (2X) and 0.2 µl ROX as a passive reference standard to normalize the SYBR fluorescent signal. The thermal profile for RT-qPCR was: initial activation at 95 °C for 10 min, followed by 40 cycles of 95 °C for 30 s, 60 °C for 30 s, and 72 °C for 30 s. The rice actin gene was used as an internal reference for all RT-qPCR analyses. Each treatment was conducted in triplicate independently using three biological replicates. The relative expression of OsCBL genes was calculated using the 2−ΔΔCt method. The primer sets are listed in Table S1.

Binary Vector Construction, Rice Transformation and Characterization of the Transgenic Lines

To make OsCBLs constitutive overexpression, the modified pCAMBIA1301 vector P1301-UbiNos under the control of the maize ubiquitin promoter was constructed by inserting the coding sequence of OsCBLs which was amplified with primers of OsCBLsOE-F and OsCBLsOE-R (Table S2). Screening of transgenic plants was attained by planting on 1/2 Murashige and Skoog (MS) medium containing 25 µg/L hygromycin. The presence of the transgenic insertion was confirmed through PCR in the transgenic plants. Moreover, seeds were collected and sown for phenotyping from T1 transgenic rice plants. More than 20 independent plant lines from each transformation were picked for morphologic observation, and ZH11 was used as control.

Phenotype Analyses for Drought and Salt Tolerance

For drought stress assay, 2-week-old OsCBLs overexpression (OsCBLs-OE) transgenic plants and wild-type (WT) plants located in same growth condition were subjected to drought stress by 20% PEG 6000 for 10 days. The length of of the shoot and root of at least 20 seedlings of each line were measured at 10 days after treatment, respectively. Likewise, 2-week-old transgenic and non-transgenic plants grown in the same environment were transferred into 1/2 MS containing 300 mM NaCl. At 10 days after treatment, the length of the root and shoot were measured, respectively.

Results

Identification of CBL Protein Sequences in 15 Grass Species Based on the Conserved EF-Hand Domain

To survey the evolution of CBL protein architecture across grass plants, 15 publicly available genomes of Poales were selected for CBL gene family identification analyses (Table 1). We found that the numbers of CBLs varied from species to species across the grass plants. Finally, 152 non-redundant CBL sequences containing the conserved EF-hand domain were retrieved in total (Table 1, Table S3). Individual CBLs were assigned names using the ortholog based on the evolutionary relationship with ten Arabidopsis CBLs. The hexaploid wheat genome contained the largest number CBL genes, whereas Zoysia japonica only contained six, but most of them were around ten (Table 1, Table S3).

Table 1 Table representing genome size of different grass species and number of CBL genes present per genome (species)
Full size table

Phylogenetic Relationships of Grass CBL Genes

To investigate the evolution history among grass CBL gene families, the ML phylogenetic tree was reconstructed based on the full-length amino acid of the 152 CBL protein sequences (Fig. 1). In the phylogenetic tree, the grass CBLs can be divided into four main clades and designated as group A, B, C and D, respectively (Fig. 1, Table S3), which is in accordance with the Arabidopsis and rice reported CBLs (Albrecht et al. 2001; Kolukisaoglu et al. 2004). The groups A, B, C and D contain CBL9/10, CBL2/3/6, CBL1 and CBL4/5/7/8, respectively (Fig. 1). In addition, to assess the difference between whole sequences and functional motifs, we also reconstructed the phylogenetic tree using the EF-hand motif sequences (Fig. S1). As expected, the phylogenetic tree using EF-hand motifs of CBL protein displayed similar phylogenetic relationship compared with whole sequences, indicating that the evolutionary relationship of the CBL proteins was mainly determined by functional motifs, especially their characteristic EF-hand motif.

Fig. 1
figure 1

Maximum Likelihood phylogenetic trees of 15 grass CBL proteins. Phylogenetic analysis was carried with protein sequences for 152 CBL proteins from 15 grass species identified in this study

Full size image

Common Conserved Domain Compositions and Genomic Analysis of CBL Proteins in Grass

To better obtain an overview of the characteristics among grass CBLs, their common conserved domains were further analyzed. As expected, three main common motifs of CBL proteins were subjected to the domain analyses using InterProScan database, suggestive of functional similarities within gene family (Fig. 2). As calcium sensors, CBL proteins harbor four typical EF-hands (IPR002048) as a structural basis for calcium binding (Albrecht et al. 2001) (Fig. 2a). Generally, each EF-hand is composed of 12 amino acid residues, in which 6 amino acids at positions 1(X), 3(Y), 5(Z), 7(− X), 9(− Y) and 12(-Z) are highly conserved and involved in binding of calcium ion (Fig. 2b). It is notable that the EF1 loop contains 14 amino acid residues caused by an insertion between position 1 and 3, and AtCBL6 lacked four amino acids and AtCBL7 had an insertion of five additional amino acids (Kolukisaoglu et al. 2004). Surprisingly, CBL4 can bind up to four Ca2+ ions; however, only two EF-hands bind Ca2+ in CBL4-CIPK24 complex (Fig. 2a) (Sanchez-Barrena et al. 2013). Interestingly, some N-terminus of CBL proteins possessed a highly conserved N-myristoylation motif that favored the attachment of CBL to the plasma membrane and transmitted the calcium signal, while the second glycine residue of the consensus sequence MGCXXSK/T is involved in protein myristoylation (Fig. 2a, c). Moreover, all CBLs similarly harbor a highly conserved C-FPSF domain (Fig. 2d).

Fig. 2
figure 2

Analyses of motif structures and compositions of all grass CBLs identified in this study. a Schematic depiction of the typical domain structure of grass CBLs. The overall structure of CBLs consists of four EF-hands (boxes with numbers), which the first and fourth bind two Ca2+ ions. In all CBL proteins, the first EF-hand has an unconventional structure, encompassing 14 amino acids instead of the 12 in a canonical EF hand (black box) flanked by two alpha helices and the amino acids at position 1(X), 3(Y), 5(Z), 7(− X), 9(− Y) and 12(− Z) of the loop co-ordinate and bind the Ca2+ ion. The position of lipid modification sites (N-myristoylation and S-acylation) and the C-terminal FPSF domain was shown by jagged lines and red box, respectively. b Sequence features shown in the form of web logos representing the EF hands of all CBL sequences. Spacing of EF hands in all CBLs is invariable. c Detailed comparisons of N-myristoylation motif sequences of grass CBLs. d Detailed comparisons of C-terminal FPSF domain sequences of grass CBLs

Full size image

In addition, we also investigated the exon/intron organizations of different CBL genes according to the genomic information received. The majority of CBL genes contain seven introns occupying the main points (55%), which mean that the exon/intron organizations of grass CBL genes have a conserved model in their evolutionary process, while only one CBL gene contains a maximum of 20 introns (PeCBL1-4) or a minimum of being intronless (HvCBL6-2) (Table S3). Group A CBLs possess introns ranging from 5 (HvCBL10) to 11 (AcCBL9) and dominated by 8 (64%), while the numbers of introns in group B CBLs ranged from 0 (HvCBL6-2) to 9 (AcCBL6) and 7 occupied the main points (61%; Table S3). Remarkably, most groups C and D CBLs contain seven introns (Table S3). Afterward, we further determined the Mw and pI of different CBL proteins using the online version of Compute pI/Mw tool. The Mw of CBL proteins ranged from 12.3 (SiCBL7) to 80.115 (PeCBL1-4) kDa and the pI varied from 4.26 (SiCBL7) to 10.13 (SiCBL1) (Table S3). It is noteworthy that the pI of other CBLs removed SiCBL1were in the slightly acidic ranges. The average amino acid composition of CBL proteins ranged from 0.53 (tryptophan) to 10.83 (leucine) (Table S4). Remarkably, the average abundance of the most important amino acids aspartic and glutamic (EF-hand) which were responsible for the Ca2+-binding function of CBLs were 8.31 and 8.87, respectively. Moreover, the average abundance of glycine (N-myristoylation motif) that is required for protein myristoylation was 4.08 (Table S4). Furthermore, the average abundance of the hydrophobic amino acids in CBLs was relatively higher than ones of other amino acids such as alanine (6.63), isoleucine (5.68), leucine (10.83) and valine (6.27) (Table S4).

Previous research have reported that several consensus sequences play vital roles in their structure or function, including the EF-hand, N-myristoylation motif and the FPSF domain in C-terminal region. In particular, the aspartate and glutamate in EF-hand is the structural basis of function for the calcium sensor. To better illustrate consensus sequences that might be characteristic of the phylogenetic group, we then executed sequence logos motif analyses using WebLogo 3 online tool for each group. For example, Fig. 2b displayed the consensus sequence of four EF-hand from all group CBLs. The residues marked stars were required for the calcium binding function (Fig. S2). The CBLs contained group-specific conserved first non-canonical EF-hand motif, S–C–S-I-I-D/N-D-G-L-I-H–K-E-E (group A), S–S-A-V-I/V-D-D-G-L-I-N-K-E-E (group B), S-G-S-V-I-D-D-G-L-I-N-K-E-E (group C) and S-X-S-I-X-K-D-G-L-I-H–K-E-E (group D), respectively (Fig. S2). In addition, grass CBLs also harbored a group-specific conserved FPSF domain in the C-terminal,, F-P-S-F-V/I-F-N-T-X-V-E-D (group A), F-P-S-F-V-F–H/N-S-Q–V-D/E-D (group B), F-P-S-F-V-F–H/N-S-E-V-D-D (group C) and F-P-S-F-V/I-X-X-S-E/G-X-X-D/E (group D), respectively (Fig.S3). The presence of group-specific EF-hand and/or FPSF domain in CBLs indicated that group-specific CBL proteins may show distinct affinities to calcium ion or substrates (CIPKs).

Molecular Evolutionary Analyses

The genetic distance between groups A and D CBLs (0.279) was the lowest compared with other counterparts, suggesting that their sequence similarity was higher between the group A and D CBLs (Table S5), which was in accordance with their evolutionary relationship (Fig. 1). On the contrary, the highest genetic distance was between group B and C CBLs, indicating that the sequence of two group CBLs showed greater difference than other group CBLs. In addition, if there is no significant difference in genetic distance between each group, it implies that there is inconspicuous sequence divergence between each group CBLs. Furthermore, the average overall mean distance of CBL was 0.662 (standard error 0.343).

To elucidate the evolutionary basis of each group CBLs, we assessed the nonsynonymous-to-synonymous rate ratio using full-length DNA sequences (ώ = dN/dS) (Table 2) under different codon substitution-based evolutionary models. The mean ώ values were less than 1 for group A (ώ = 0.403), group B (ώ = 0.710) and group C (ώ = 0.674) CBLs, suggesting strong purifying selection during the CBLs evolution (Fig. 1). However, the average ώ values of group D CBLs was 1.293 that is greater than 1, indicating that these genes show positive selection during their evolution process (Table 2). Moreover, to compare the difference between whole sequences and functional motifs, we evaluated the ώ values employing the motif sequences under the same evolutionary models (Table S6). As expected, there were equally low ώ values in group A, B and C CBLs when the sequences of EF1 were performed, suggesting that EF1 are well conserved during the evolution process, which was also in accordance with the result for four EF-hand motifs together, as supported by a previous study (Table S6). Interestingly, there were diverse results in the analyses of the FPSF domain. The mean ώ values were 2.028 in group B CBLs, while it were 0.306 in group D (Table S6), indicating that the FPSF domain has more variations compared to EF-hand, which may be associate with their roles in the calcium signal transduction process.

Table 2 Molecular evolutionary analysis of the CBL genes using their whole cDNA sequences
Full size table

Expression Patterns of OsCBLs

Expression patterns would provide evidence of their functional divergence among all members of a gene family (Whittle and Krochko 2009). In this analysis, the expression patterns of OsCBL genes in various stages of flower development were investigated using RT-qPCR. OsCBL1, OsCBL2, OsCBL3 and OsCBL6 revealed higher expression in all the tissues tested. OsCBL4 showed a very high expression only in leaf blade, but low expression in other organs. OsCBL9 is moderately expressed in root, shoot, leaf blade and leaf sheath, while it shows high expression in other tissues (Fig. 3). In addition, OsCBL5, OsCBL7, OsCBL8 and OsCBL10 showed a relatively lower expression in all organs other than roots, but it was a little different. OsCBL5 and OsCBL8 displayed relatively low expression in all organs detected and OsCBL10 moderate expression in whole flower developmental stages, while OsCBL7 showed moderate expression in the leaf sheath (Fig. 3). These results indicated that these genes may have distinct functions in the corresponding tissues.

Fig. 3
figure 3

Expression profiles of rice CBL genes in different tissues and flower developmental stages of ZH11 plants. Clustering of OsCBL genes according to their expression profiles in ZH11 plants at different tissues and developmental stages was performed

Full size image

We also had special concerns about the rice group B CBL members on account of them lacking myristoylation motifs which are involved in protein anchoring to the membrane to perceive and transduce the calcium signal, especially OsCBL3 and OsCBL6. To explore their roles in response to hormones, the expression profiles in rice seedlings were examined by RT-qPCR under TIBA, SA, IAA, GA and ABA (Fig. 4a). OsCBL3 and OsCBL6 exhibited increases after TIBA and SA treatments and peaked at 1 h, then remained at relatively high levels under TIBA, and recovered to near normal levels at 3 h after SA treatment. Moreover, OsCBL3 and OsCBL6 were induced by ABA within 1 h and then suppressed at 3 h after treatment. Furthermore, OsCBL3 and OsCBL6 were not significantly affected after IAA and GA treatments. We also analyzed their expression profiles under PEG 6000, salt, cold, CuSO4 and K2Cr2O7 applications (Fig. 4b). OsCBL3 was up-regulated at 1 h after K2Cr2O7 treatment, and then recovered to near normal levels, while OsCBL6 showed no obvious change. Likewise, OsCBL3 and OsCBL6 peaked at 3 h after CuSO4 application and then OsCBL3 maintained relatively high expression level, while OsCBL6 recovered to normal condition. Cold treatment had no apparent influence on OsCBL3 expression, but strongly induced OsCBL6 expression up to 12 h. Salt treatment slightly suppressed OsCBL3 expression, while OsCBL6 had mild up-regulated expression. Drought treatment suppressed OsCBL3 and OsCBL6 expression from 3 h, staying up to 12 h and then had a recovery at 24 h up to normal levels.

Fig. 4
figure 4

Time course quantitative gene expression profiles of OsCBL3 and OsCBL6 under hormone treatments and abiotic stresses. a Clustering of OsCBL3 and OsCBL6 according to their expression profiles in the 2-week-old seedling of ZH11 plants after different phytohormone treatments. b Clustering of OsCBL3 and OsCBL6 according to their expression profiles in the 2-week-old seedling of ZH11 plants under abiotic stresses

Full size image

Tolerance Analyses of OsCBLs-OE Plants Under Salt and Drought Condition

To elucidate the putative roles of OsCBLs in plant, their overexpression transgenic rice plants were generated. To confirm whether the CBL genes were integrated into the genomic DNA, PCR amplification with their specific primers to produce whole coding sequence was performed (Fig. S4). In addition, GUS staining assays in the root of transgenic rice plants were also executed (Fig. S5). Moreover, we also detected their expression levels in transgenic rice plants using RT-qPCR methods (Fig. S6). These results showed that the mRNA of OsCBL genes was increased more than in WT plants (Fig. S6). We further tested the tolerance of OsCBL-OE plants under salt and drought stresses. For salt tolerance analyses, the growth status of WT and OsCBL-OE plants was compared on pre- and post-treatment (Fig. 5a). The root lengths of OsCBL3-OE and OsCBL8-OE plants were significantly higher than those of WT seedlings, while other OsCBLs-OE plants had no apparent change (Fig. 5b). Moreover, the shoot growth of OsCBL4-OE and OsCBL6-OE plants was obviously suppressed by salt stress, while other OsCBLs-OE plants had no effect (Fig. 5b). The result indicated that OsCBL3 and OsCBL8 might improve the tolerance to high-salt stress in rice, while OsCBL4 and OsCBL6 were likely hypersensitive to it.

Fig. 5
figure 5

Characterization of the OsCBLs-overexpressing transgenic rice lines under salt stresses. a Growth performance of the OsCBLs-OE and WT seedlings grown on 1/2MS medium supplemented with 300 mM NaCl. Scale bar = 1 cm. b Root and shoot length of the OsCBLs-OE and WT seedlings grown on 1/2MS medium supplemented with or without 300 mM NaCl. Values are expressed as mean ± SD (n = 3 experiments). Asterisk indicate significant difference at p < 0.05 level between WT and OsCBLs-OE plants

Full size image

We next examined the drought tolerance of OsCBLs-OE and WT plants with 20% PEG 6000 1/2 MS medium as mimetic drought environment. The morphologic observation of WT and OsCBLs-OE plants was executed on pre- and post-treatment (Fig. 6a). The root lengths of OsCBL5-OE, OsCBL6-OE and OsCBL7-OE plants were significantly higher than those of WT after treatment, while those of OsCBL8-OE, OsCBL9-OE and OsCBL10-OE plants were apparently lower (Fig. 6b). However, the shoot length of all OsCBLs-OE plants had no evident influence except for OsCBL8-OE, OsCBL9-OE and OsCBL10-OE plants which were suppressed by drought stress (Fig. 6b). The data suggested that OsCBL5, OsCBL6 and OsCBL7 might confer resistance to drought, while OsCBL8, OsCBL9 and OsCBL10 had the opposite effect.

Fig. 6
figure 6

Characterization of the OsCBLs-overexpressing transgenic rice lines under drought stresses. a Growth performance of the OsCBLs-OE and WT seedlings grown on 1/2MS medium supplemented with 20℅ PEG6000. Scale bar = 1 cm. b Root and shoot length of the OsCBLs-OE and WT seedlings grown on 1/2MS medium supplemented with or without 20℅ PEG6000. Values are expressed as mean ± SD (n = 3 experiments). Asterisk indicate significant difference at p < 0.05 level between WT and OsCBLs-OE plants

Full size image

Discussion

Phylogeny and Evolution of CBL Gene Family in Grass

In this study, the evolution of CBL gene family in grass was addressed. For this purpose, we first identified 152 CBLs from 15 Poales and then reconstructed their ML phylogenetic tree (Fig. 1). Strikingly, all CBLs except Z. japonica can be divided into four groups, which might be involved in the incomplete annotation of the genome sequence (Table S1). In addition, the size of the CBL gene family in different grass species had no apparent difference and was not associated with genome size and mostly had near ten CBL genes (Table 1). For instance, O. sativa, Setaria italica and Hordeum vulgare all had 10 CBL genes, while their genome size is 372 Mb, 441.7 Mb and 5100 Mb, respectively. Nevertheless, the gene expansion might be involved with polyploidy or ancient polyploidization events (Jiao et al. 2011). For example, 21 CBL genes, rather than 10, had been identified in hexaploid bread wheat genome which arose in two polyploidization events (Mayer et al. 2014). Therefore, we can speculate that whole genome duplication contributed to gene expansion of CBLs, which is similar to the patterns of MKK and rhomboid gene families (Jiang and Chu 2018; Li et al. 2015). Moreover, except group D CBLs, the mean ώ values were all less than 1 and showed purifying selection during their evolution process and suggested that duplicated or lost genes were easily affected by positive selection (Table 2). Furthermore, domain analyses revealed that CBL genes were highly conserved in the study investigated including EF-hand motif, N-myristoylation motif and the FPSF domain (Fig. 2). As the basis of calcium binding, the average abundance of amino acid residues D and E contained a relatively high proportion (Table S4). Compared to canonical sequences in CaMs, EF-hands of each group CBLs have variable degrees of conservation (Fig.S2), implying that specific signal transduction is likely to depend on different Ca2+ affinities that is not uniquely governed by the EF-hand sequence (Nagae et al. 2003; Sanyal et al. 2015). Above all, these results suggested that the CBL gene family is highly conserved in grasses.

Differential Expression of OsCBLs Genes

So far, numerous CBL genes have been characterized from the higher plants. However, tissue-specific distribution patterns may allow CBL proteins to have corresponding functions in different pathways. For instance, AtCBL4 has specific expression in roots (Guo et al. 2001), while the expression of AtCBL10 is limited to shoot tissues (Kim et al. 2007). These expression patterns enabled CBL4-CIPK24 to effectively mediate Na+ extrusion from roots, and CBL10-CIPK24 in shoot tissues modulated sequestration of Na+ into vacuoles (Weinl and Kudla 2009). Therefore, we investigated the expression profiles in different rice organs. OsCBL4 was highly expressed in leaf, while OsCBL5, OsCBL7 and OsCBL8 were mainly expressed in roots (Fig. 3), implying that rice might have a different signaling pathway from Arabidopsis CBL4-CIPK24 that contributed to response to salt stress. Moreover, OsCBL3, OsCBL6 and OsCBL9 had a high expression in various flower stages (Fig. 3). Hence, we have reason to speculate that it exist a distinct Ca2+ signaling pathway in flower developmental stages.

Function of Rice CBLs in Salt and Drought Stress Responses

In previous studies, expression of AtCBL1 was induced by various stimuli such as salt, drought, cold, wounding and ABA exposure (Fotster et al. 2019; Kudla et al. 1999). AtCBL9 negatively modulates low-temperature stress (Gao and Zhang 2019) and renders plants hypersensitive to ABA in Arabidopsis (Pandey et al. 2004). In addition, overexpression of SpCBL6 improved cold tolerance and reduced drought tolerance (Zhou et al. 2016), while turnip BrrCBL9.2 enhanced salt tolerance in transgenic Arabidopsis plants (Yin et al. 2017). Interestingly, the expression of OsCBL3 was slightly down-regulated under salt condition (Fig. 4), while their overexpression plants enhanced the tolerance of salt stress (Fig. 5). The result indicated that OsCBL3 positively mediated salt tolerance. Likewise, OsCBL6 expression was suppressed (Fig. 4) and their overexpression in plants conferred tolerance to drought stress (Fig. 6). In addition, OsCBL6 was induced by salt and was hypersensitive to it (Fig. 5). These results suggested that OsCBL6 positively improved drought tolerance and negatively regulated salt stress in rice. Unlike CBL4, both OsCBL3 and OsCBL6 lack N-myristoylation motif which favors carrying out the function in membrane targeting of the CBL–CIPK complex (Ishitani et al. 2000). Their structural difference may be the reasons for distinct function in response to salt stress, CBL4 positively improves salt tolerance, while OsCBL6 may have opposite functions.

Conclusion

This study showed that a total of 152 CBL genes were identified among 15 grass plant species based on the presence of a conserved EF-hand domain. Based on the phylogenetic and domain analyses, the CBL gene family can be divided into four groups. It was noteworthy that only Group D CBLs had positive selection during their evolution process. Expression analysis showed that OsCBL3, OsCBL4, OsCBL6 and OsCBL9 had apparent modulated during various tissues in O. sativa. Tolerance analyses revealed that OsCBL3 and OsCBL8 enhanced salt tolerance in rice, while OsCBL4 and OsCBL6 negatively regulated to it. In addition, OsCBL5, OsCBL6 and OsCBL7 might positively modulate drought stress, while OsCBL8, OsCBL9 and OsCBL10 were hypersensitive to it.