Genome-wide identification and characterization of the 14–3-3 family in Vitis vinifera L. during berry development and cold- and heat-stress response

The 14–3-3 family of ubiquitous proteins in eukaryotes plays important roles in the regulation of various plant biological processes. However, less information is known about this family in grape fruit. To investigate the characteristics and functions of 14–3-3 in grape, a total of 11 14–3-3 proteins were identified. Phylogenetic analysis of 14–3-3 proteins in grape (VviGRFs) with homologous proteins in Arabidopsis showed that these proteins were classified into two groups, namely, epsilon and non-epsilon groups. Epsilon group members commonly contained more introns and motifs than non-epsilon group, and some intron positions were found to be conserved between Vitis and Arabidopsis 14–3-3 genes. RNA-seq and qRT-PCR results indicated that VviGRF genes may be involved in the regulation of grape development and berry ripening. Moreover, six VviGRFs exhibited significantly up- or down-regulated expression in response to cold and heat stresses, thereby revealing their potential roles in the regulation of abiotic stress responses. This work provides fundamental knowledge for further studies about the biological roles of VviGRFs in grape development and abiotic stress response. The present result will also be beneficial for understanding their molecular mechanisms and improving grape agricultural traits in the future.


Background
The 14-3-3 proteins were first found in bovine brain and named according to their elution and migration pattern on 2D DEAE-cellulose chromatography and starch gel electrophoresis [1]. This proteins, as highly conserved regulatory proteins, are ubiquitous in eukaryotes, including plants, insects, mammals, amphibians and yeasts [2][3][4][5]. Generally, these proteins exist as homo-or hetero-isoform dimers, which form dimer groove structures [6]. The structures provide binding sites for 14-3-3 proteins to interact with their targets, such as enzymes involved in primary biosynthetic and energy metabolism (e.g. nitrate reductase, brassinosteroid receptor kinase and H + -ATPase) and signal proteins (e.g. lipoxygenase and protein kinases) [7][8][9]. With the interaction, 14-3-3 proteins function as scaffolding proteins that are involved in the regulation of diverse biological processes, including carbon and nitrogen metabolisms, tricarboxylic acid cycle and the shikimate pathway [8,10].
Grape (Vitis vinifera L.) is one of the most important fruit crops worldwide; this crop has been widely cultivated because of its nutritional value and wide application [21]. The grape production is often threatened with abiotic stresses, such as heat or low temperature, drought and salt stresses. Currently, the14-3-3 family in grape has not been studied systematically, except for two 14-3-3 genes (Vv1CS and Vv2CS) that are cloned from V. vinifera cv. 'Cabernet Sauvignon'; nevertheless, the functions of these genes are still unclear [22]. Several unknown grape 14-3-3-like proteins are also involved in biotic and abiotic stress responses [23]. Thus, 14-3-3 gene family should be systematically analysed to characterise the expression during development and response to abiotic stress in grape.
In the present study, the identification, phylogeny, structure and gene location and duplication were performed on the 14-3-3 family in grape. The expression levels of 14-3-3 s in various tissues and organs of grape during different developmental stages and responses to heat and cold stresses were systematically studied in detail. The present results will provide important information for further study of the regulation mechanism of 14-3-3 s in grape berry development and under abiotic stress.

Multiple sequence alignment and phylogenetic analysis
Identified 14-3-3 proteins in grape were aligned with AtGRFs using ClustalX2 program [25]. The secondary structures of VviGRF proteins were analysed by searching in Self-Optimized Prediction Methods with Alignment (SOPMA, https://npsa-prabi.ibcp.fr/cgi-bin/ npsa_automat.pl?page=npsa_sopma.html) [26]. With the combined aligned results, typical α helices in 14-3-3 proteins were marked on their corresponding sequences ( Fig. 1). Phylogenetic analysis was conducted by MEGA7 based on the aligned results [27]. Neighbour-joining method was used with bootstrap replications of 1000 (Figs. 2 and 3a). 11 VviGRF proteins were named based on their domains and their phylogenetic relationships with AtGRFs [28]. When a one-to-one orthology was present in the Arabidopsis 14-3-3 family, the grape 14-3-3 members were given the corresponding Arabidopsis-like name (e.g., AtGRF11 and VviGRF11). Two grape genes had the same phylogenetic distance from a single homologue in Arabidopsis were differentiated by a number (e.g., AtGRF9, VviGRF9a and VviGRF9b). As for remaining 14-3-3 members, there are no Arabidopsis genes homologous to them, thus we named them on the basis of domain analysis: both VIT_00s0199g00140 and VIT_02s0033g00780 have incomplete 14-3-3 domains, thus these two protein were named as VviGRF-like1 and -like2. Since there are only 13 AtGRFs in Arabidopsis 14-3-3 family, other five members in several adjacent subclasses share the same 14-3-3 domain were named from VviGRF14 to 15, 16, 17 and 18. The isoelectric point (PI) and molecular weight (Mol wt) of VviGRF proteins were examined on Swiss-Prot website (http://web.expasy.org/ compute_pi/, Table 1).

Gene location and duplication
The location data of VviGRFs were also obtained from the genome annotation files (http://genomes.cribi.utn ipd.it/DATA/). The gene location map was constructed using MapChart ( Fig. 4) [31]. Moreover, the gene duplication landscape was obtained using the MCScanX [32]. Each duplicate segment with VviGRF genes was selected, and the syntonic map was generated using CIRCOS [33]. The putative duplicated genes were linked by the connection lines (Fig. 5).

Transcriptomic resources
In this work, published data GSE36128, GSE62744, SRP018199 and SRP091989 were used for the analysis of expression profiles of VviGRF genes in 54 tissues and organs, during berry ripening, and in cold and heat stress responses, respectively [34][35][36][37]. To survey the temporal and tissue-specific expression of VviGRFs, 54 tissues and organs in V. vinifera cv Corvina were collected in GSE36128 (Additional file 2, Fig. 6). To search the expression pattern of VviGRFs in different varieties during berry development, berries at four growth periods (Pea, Touch, Soft, Harvest) in five grape varieties, namely, Fig. 1 Multiple sequence alignment of VviGRFs. VviGRF proteins were aligned using ClustalX2 program. ɑ helices of VviGRF proteins were analysed by searching SOPMA website, and were marked on their corresponding sequences. VviGRF-like1 and VviGRF-like2 respectively are the longest and the shortest amino acid sequences among VviGRFs, and they show very different motif and intron-exon structures with other 14-3-3 members (Fig. 3b). In order to present a more visualized secondary structure of VviGRFs, nine VviGRF proteins were used except for VviGRF-like1 and VviGRF-like2 Sangiovese, Barbera, Negro amaro, Refosco and Primitivo were used in GSE62744 (Fig. 7a). In SRP018199, the V. vinifera cv. Muscat Hamburg seedlings were used under the gradual cooling treatment (decreased at 5°C per hour from 24°C to 4°C and then 4°C for additional 4 h) to analyse the expression profile of VviGRFs in response to cold treatment. Besides, to survey their response to heat treatment, detached leaves in V. vinifera cv 'Jingxiangyu' plantlets were treated under 25°C, 35°C, 40°C and 45°C.

Plant material
To verify exon-intron structure, specific cDNA and DNA sequences were obtained from three biological replicates of green stem in V. vinifera cv. 'Cabernet Sauvignon'. To analyse the 14-3-3 expression pattern in grape during berry ripening, three biological replicates represented by 15-20 grape berries from V. vinifera cv. 'Cabernet Sauvignon' were sampled at three developmental stages (EL33, EL35 and EL37 [38]). The grapevines were planted in south-to-north oriented rows by similar management conditions at the Germplasm Repository for Grapevines, Institute of Botany, the Chinese Academy of Sciences, Beijing, China (39°54' N, 116°23′ E).
To analyse the 14-3-3 expression in grape under heat stress, 30-day-old leaves taken from V. davidii were heated in a water bath (25°C for 2 h and 47°C for 40 min) with three biological replicates. V. davidii were planted in the same condition as V. vinifera cv. 'Cabernet Sauvignon' described above. Heat treatment was processed according to a previous study [39]: leaf discs (5.5 cm in diameter) cut from the detached leaves were wrapped in a wet paper and placed in a small vessel made of aluminium foil. Leaf discs were sampled when vessels floated in a temperature-controlled water bath at 25°C for 2 h and subsequently at 47°C for 40 min. The control plants were treated in the same condition as that in heat treatment, except for temperature controlled at 25°C.
To study the expression profile of VviGRFs in response to cold stress, six-week-old Vitis amurensis plantlets grown on 1/2 Murashige and Skoog (pH 5.8, 0.7% agar and 1% sucrose in 120 mL conical flasks) solid medium were placed in 4°C (cold treatment) and 24°C (control) chambers with a 16 h light/8 h dark Fig. 2 Phylogenetic analysis of 14-3-3 members in V. vinifera and Arabidopsis. Phylogenetic tree was constructed by MEGA7 with Neighbour-Joining method and bootstrap of 1000 replications. Grape 14-3-3 s were named by reference to their domains and the grapevine gene nomenclature system [28]. Values less than 40 were cut off

RNA extraction and qRT-PCR analysis
All samples were collected then preserved in liquid nitrogen immediately, and stored at − 80°C for RNA and DNA extraction. Total RNA and DNA were extracted using the RNAprep Pure Plant Kit (DP432, Tiangen Biotech, Beijing, China) and Plant Genomic DNA Kit (DP305-02, Tiangen Biotech, Beijing, China) respectively, according to the manufacturer's instructions. The cDNA used for verifying exon-intron structure was synthesised using HiScript® II 1st Strand cDNA Synthesis Kit (R211-01, Vazyme Biotech Co., Nanjing, China). To amplify the complete cDNA and DNA fragments corresponding to the VviGRF genes, specific primers for each gene were designed through NCBI Primer-BLAST (http://www.ncbi.nlm.nih.gov/tools/primer-blast/, Additional file 3), and the PCR reactions were performed using PrimeSTAR Max DNA Polymerase (R045A, Takara Biotechnology Co., Ltd., Dalian, China) with the following thermal cycling profile: 98°C for 3 min, 34 cycles of 98°C for 10 s and 60°C for 5 s and 72°C for 50 s, 72°C for 5 min, 12°C for ∞. PCR products were submit to Beijing Majorbio Company, and the sequencing results can be seen in Additional file 4. The cDNA used for qRT-PCR analysis was synthesised through reverse transcription of obtained total RNA using HiScript® II Reverse Transcriptase (R223-01, Vazyme Biotech Co., Nanjing, China). The synthesised cDNA was subjected to qRT-PCR with an Opticon thermocycler (CFX Connect Real-Time System; Bio-Rad, Hercules, CA) using SYBR Green PCR master mix (Vazyme, Nanjing, China) according to the manufacturer's instructions. In qRT-PCR, VviActin (Accession: EC969944) was Chromosomes were drawn in different colours. The approximate location of VviGRF genes is shown by short black lines on the circle. Red curves linking VviGRF genes represent the duplication events occurred in grape 14-3-3 gene family used as reference gene, and specific primers for VviGRF genes were designed by using Primer-BLAST (Additional file 5). The specificity of designed primers was verified through gel electrophoresis and sequencing. PCR reactions were as follows: 95°C for 10 min, 40 cycles of 95°C for 10 s and 60°C for 30 s. Each sample was prepared in three biological and technical replicates. The relative expression levels of the VviGRFs were calculated using the 2 −ΔΔCt method. Statistical difference was determined by t-test (**P < 0.01, *P < 0.05, n = 3) using R program [40].

Results
Identification, multiple sequences alignment, and phylogenetic analysis of grape 14-3-3 family Nineteen 14-3-3 proteins were obtained by using HMM and BLASTP. A total of 11 proteins containing 14-3-3 domain were identified in grape after searching in NCBI-CDD. The results of sequence alignment showed that VviGRFs contained several conserved domains in their aligned sequences, and nine α helices were identified in their secondary structures (Fig. 1). Based on phylogenetic results, these proteins were classified into two groups, namely, epsilon (VviGRF9a/9b/11/12) and non-epsilon groups (VviGRF14/15/16/17/18/−like1/ −like2) (Figs. 2 and 3a). VviGRF proteins contained 64 (VviGRF-like2) to 582 (VviGRF-like1) amino acid residues, their Mol wts varied from 6.91 kDa (VviGRF-like2) to 66.03 kDa (VviGRF-like1), and their PIs ranged from 4.74 (VviGRF16) to 7.57 (VviGRF-like2) ( Table 1). Figure 3b illustrates the predicted motif and gene structure of VviGRF genes. And the exon-intron structure of VviGRF genes have been verified by sequencing, which was accordant with our predicted analysis (Additional file 4). The number of exon in VviGRFs varied from one (VviGRF-like2) to seven (VviGRF9b/ 11/ 12), and nine out of 11 VviGRFs showed not less than four exons (except for VviGRF-like1 and VviGRF-like2). A search on MEME program identified seven motifs in VviGRF proteins. The number of motifs in VviGRFs ranged from 1 to 7, and the length of motifs varied from 8 (motif 6) to 50 (motifs 1-3) amino acids. Most VviGRFs contained all seven motifs, except for VviGRF-like1 and VviGRF-like2. VviGRF-like1 contained five motifs without motifs 2 and 5. VviGRF-like2 only had motif 7. Moreover, NCBI-CDD results showed that the 14-3-3 domain of VviGRF-like2 was incomplete in N-terminal, VviGRF-like1 was incomplete in C-terminal. Furthermore, VviGRF-like1 contained a 14-3-3 domain, a domain of unknown function (DUF4283) and a zinc-binding motif (Additional file 6). Associated with gene and motif structure, the position of introns was conserved in grape 14-3-3 family (Fig. 3). In both groups, motif 1 and 6 were translated by a single exon, motif 2 and motif 5 were translated from sequences modelled by "exon-intron-exon". In epsilon group, motifs 3, 4 and 7 in all four members were translated by "exon-intron-exon" sequences. In non-epsilon group, the sequence of 6 members (except for VviGRF-like2) Fig. 6 The temporal and tissue-specific expression of VviGRF genes based on the GSE36128 [35]. Expression data were processed with log2 normalization. The horizontal axis shows fifty-four tissues and organs, which can be seen in Additional file 2. The colour scale represents relative expression levels encoding motif 7 also shown in this model. The detail information of introns position in "exon-intron-exon" sequences was shown in Additional file 7. Between epsilon group and non-epsilon group, the position of introns was absolutely different. But in the same group, the introns were present at the same position within their encoding sequences.

Tissue-specific expression of VviGRF genes
To analyse the temporal and tissue-specific expression of VviGRFs in grapevine, we assessed their expression ). Berries were collected at four phenological stages, as follows: pea-sized berries at 20 days after flowering (Pea), berries beginning to touch just prior to veraison (Touch), softening berries at the end of veraison (Soft), and berries ripe for harvest (Harv) [36]. The obtained data were processed using log2 normalization. The colour scale represents relative expression levels. (b) qRT-PCR results of eight VviGRF genes during berry ripening. As previous study indicated [38], EL33, EL35 and EL37 represent three important stages during berry ripening, respectively. EL33 means the stage when berries are still hard and green, EL35 means the veraison when berries begin to colour and enlarge, and EL37 presents the period when berries will mature soon but not quite ripe. Pea, touch, soft and harvest stages in a correspond to EL31, EL34, EL37 and EL38, respectively. Period before veraison is named the early stage in this study, and the period after veraion is the late stage. Thus EL33, pea, touch belong to the early stages, and EL37, soft, harvest belong to the late stage. Expression data was normalized to VviActin gene expression level, and every VviGRFs at EL33 stage was normalized as "1". The mean expression value was calculated from three replications. Vertical bars indicate the standard error of mean. **P < 0.01 and *P < 0.05 compared with expression level at EL33 stage data in 54 tissues and organs obtained from the published GSE36128 (Fig. 6). Most VviGRFs exhibited a quite moderate expression in different tissues and organs, but some genes showed very high or very low expression in certain tissues. For instance, VviGRF12 was normally lowly expressed in most tissues and organs, but not in some floral organs such as stamen, bud, flower and pollen. And in normal conditions, VviGRF15 showed a high expression level in most tissues and organs, except for leaf, petal and pollen. VviGRF-like1 was generally lowly expressed in most tissues and organs, except for senescencing leaf, bud, berry pericarp and berry skin. VviGRF-like2 exhibited a high expression level in most tissues, but lowly expressed in well-developed inflorescence, young leaf, and woody stem.

Expression pattern of VviGRFs during berry development
To determine the VviGRF expression in different grape genotypes, we analysed their expression levels in five grape varieties during berry ripening (GSE62744, Fig. 7a). Except that VviGRFs in Sangiovese showed similar expression at soft and harvest stages, the expression patterns of VviGRFs in five varieties were similar at the early (pea and touch) and late (soft and harvest) stages of ripening. VviGRF9b, VviGRF11, VviGRF14 and VviGRF16 exhibited low expression level at the early stages and high expression level at the late stages. By contrast, VviGRF9a and VviGRF15 showed high and low expression levels at the early and late stages, respectively. The expression level of VviGRF12, VviGRF17 and VviGRF18 was constantly changed during berry ripening. At early stages, these three genes expressed highly at pea stage and lowly at touch stage, whereas at late stages, they expressed highly at soft stage and lowly at harvest stage. To study their expression patterns during grape berry development (Fig. 7b), eight VviGRFs in V. vinifera cv. 'Cabernet Sauvignon' were selected to qRT-PCR. Five out of 8 VviGRFs showed no significant expression level at three key stages during berry ripening (EL33, EL35 and EL37). Only VviGRF18 expression was up-regulated at both EL35 and EL37 stages. VviGRF14 and VviGRF15 were only up-regulated at EL37.

Response to cold and heat stresses
RNA-seq GSE89113 was used to investigate the response of VviGRFs to heat stress (Fig. 8a). Four VviGRFs   [37]. Log2 change fold between two treatment temperatures was used to present changed expression level under heat treatment by R software. For example, the expression level of 25 vs 35 was calculated from the formula: log2 (the expression at 35°C / the expression at 25°C). The colour scale represents relative expression levels. (b) qRT-PCR profiles of VviGRF genes by heat treatment. CK means control, H means heat treatment. Expression level of heat stress was normalized to VviActin gene expression level, and every VviGRFs at CK was normalized as "1". The mean expression value was calculated from three replications. Vertical bars indicate the standard error of mean. **P < 0.01 compared with expression level at CK Among them, VviGRF12 showed significantly decreased expression, but VviGRF15 significantly increased (|log 2 fold change| > 1). Under 45°C, all of nine VviGRFs showed an increase in expression compared with that at 40°C. According to the data published in SRP018199, GSVIVT01012207001 and GSVIVT01009141001 (corresponding VviGRF12 and VviGRF14 in 12X V1), the V. vinifera cv. Muscat Hamburg seedlings showed up-regulated expression in response to the gradual cooling treatment (decreased at 5°C per hour from 24°C to 4°C and subsequently 4°C for additional 4 h, Additional file 8).

Discussion
In this work, we identified and characterized the 14-3-3 gene family in V. vinifera. It is composed by 11 members, a number higher than in rice [11], but lower respect to Arabidopsis [16], tomato [12], cotton [13], banana [14], switchgrass [41], and soybean [15]. Based on amino acid sequences, gene structure and phylogenetic relationship with A. thaliana, the 11 VviGRF proteins were classified into two groups (Table 1 and Fig. 2,  Fig. 3a). This result is also in accordance with previous studies on other species [11,12,42]. For example, five and eight AtGRF proteins were contained in epsilon and non-epsilon groups in Arabidopsis, respectively.
A total of seven motifs were found in the grape 14-3-3 family. Epsilon group members contained all seven motifs, whereas non-epsilon group members commonly possessed less motifs than those of epsilon group Fig. 9 qRT-PCR profiles of VviGRF genes under cold treatment. CK means control, C8 and C12 represent 8 h and 12 h 4°C treatment, respectively. Expression data of cold stress was normalized to VviActin gene expression level, and every VviGRFs at CK was normalized as "1". The mean expression value was calculated from three replications. Vertical bars indicate the standard error of mean. **P < 0.01 and *P < 0.05 compared with expression level at CK members (e.g., VviGRF-like1 and VviGRF-like2). Similar to the motif analysis, epsilon group members commonly possessed more exons or introns than those of non-epsilon group, which also verified by RT-PCR (Table 1, Fig. 3b, and Additional file 4). Generally, only epsilon group members contained introns within sequences encoding motifs 3 and 4.This observation revealed that different groups of phylogenetic tree presented different motifs and exon-intron structures. This finding can also be observed in other species, such as Arabidopsis and rice; epsilon group members often present two extra introns in the N-or C-terminal compared with those of non-epsilon group [11,42]. Take Arabidopsis for example (Additional file 1), there were a total of six motifs, and both their sequences and distribution were similar with those in grapevine. And as found in grapevine, sequences encoding motifs in Arabidopsis 14-3-3 family also possessed "exon-intron-exon" model. Moreover, Additional file 9 showed that the position of introns in this model was conserved in the same phylogenetic group, which was in accordance with grapevine. And between Vitis and Arabidopsis, introns within motifs 2 and 4 located in the same position. These similarities suggested different distribution of "exon-intron-exon" sequences encoding motifs between epsilon and non-epsilon groups, and also revealed the conservation of the exon-intron and motif structure in 14-3-3 family.
Gene duplication events are crucial in genomic rearrangement and often result in the expansion of gene family, which includes tandem, segment and transposition duplication [43]. In grape 14-3-3 family, three pairs of genes evolved from segment duplication, but no tandem duplication event occurred (Fig. 5), indicating that the segment duplication may be the dominant gene duplication on the expansion of this family. And as for the remaining five VviGRFs, they may evolve in an early divergence time or be obtained from gene transposition.
To further understand the putative function of VviGRFs, the expression pattern of VviGRFs was searched for reference in the Vespucci database [44], the RNA-seq and expression-array data were also analysed. Fig. 6 reveals that VviGRFs were expressed in all investigated tissues and organs, and several VviGRFs showed tissue-specific expression in different tissues (e.g., VviGRF12/15/−like1/−like2). This phenomenon is also supported by some 14-3-3 genes in switchgrass, tomato, Arabidopsis and barley microspore embryogenesis [41,[45][46][47].
During berry ripening, VviGRFs in Sangiovese expressed similar pattern at soft and harvest stages, which may be caused by the varietal character or close sampling time at these two stages. Except this, VviGRFs in different varieties exhibited similar expression patterns between the early and late stages of berry ripening. In accordance with the results, VviGRF14, VviGRF15 and VviGRF18 displayed significant change in expression levels during EL33 (early stage), EL35, and EL37 stages (late stage). However, the up-regulated or down-regulated tendency of VviGRF15 differed in Fig. 7a and b, which may be impacted of variety difference. Taken together, some VviGRF genes presented changed expression during berry ripening, and they may be involved in the regulation of grape development. Accordingly, previous studies also demonstrated that plant 14-3-3 proteins are involved in fruit development and ripening processes. For example, most MaGRF transcriptions are significantly accumulated during fruit development and postharvest ripening in banana [14]. Transcription profiles of switchgrass suggested that four PvGRFs may be involved in regulating lignin metabolism, and PvGRFr may participate in flower development [41]. In Arabidopsis, RNA interference revealed the fundamental role of epsilon group members in regulating PIN polarity and plant development [48].
In the process of heating-up temperature (RNA-seq SRP091989, Fig. 8a), the expression of VdGRFs changed significantly, and nine genes presented up-regulated expression at 45°C. In accordance with this result, six VdGRFs exhibited up-regulated expression in response to heat stress (Fig. 8b). Similar to this result, most 14-3-3 genes also display up-regulated expression under heat treatment in rice and Brachypodium distachyon L. [11,49]. In V. amurensis, six VaGRF expression levels significantly changed in response to cold treatment, and five of six VaGRFs up-regulated or down-regulated after 8 h treatment, thereby revealing their potential roles in cold stress response (Fig. 9). In accordance with previous studies [11,50,51], two AtGRFs, five OsGRFs and two TaGRFs also indicated that cold stress induced significant expression changes in Arabidopsis, rice and wheat, respectively. VviGRF15, VviGRF-like2 and VviGRF17 showed significantly change in expression level under both cold and heat treatments, which indicated their potential roles in multiple-abiotic-stress