Identification and expression analysis of the sucrose synthase gene family in pomegranate (Punica granatum L.)

Obtaining genome sequences and identifying PgSUS family members in pomegranate

The genome sequences and annotation data of pomegranate cv. Tunisia were obtained from the NCBI genome database (https://www.ncbi.nlm.nih.gov/genome/13946?genome_assembly_id=720008) (Luo et al., 2020). Six known AtSUS proteins sequences were downloaded from TAIR database (http://www.arabidopsis.org/) and were used as a query to search against the pomegranate protein database (e-value <1 × 10⁻⁵, identify >50%). The search used a local BLAST alignment in order to identify potential members of SUS gene family in pomegranate. The hidden Markov model (HMM) profiles of the sucrose synthase domain (PF00862) and glycosyl transferases domain (PF00534) collected from the Pfam website (http://pfam.xfam.org/) were used as queries to search the candidate PgSUS from pomegranate proteins using HMMER 3.1 (e-value < 1 × 10⁻⁵) (Finn et al., 2015). The sucrose phosphate synthase (SPS) gene family with a sucrose-phosphatase domain (PF05116) in the N-terminal was also found to contain SUS protein conserved domains (PF00862 and PF00534). The resulting putative proteins were further examined by using the SMART and NCBI conserved domain database (CDD) (Letunic & Bork, 2018; Lu et al., 2020). We filtered out the candidates with a sucrose-phosphatase domain and those that lacked the sucrose synthase and glycosyl transferases domains.

The information on the pomegranate SUS chromosomal positions was obtained from the genome annotation data. The ExPasy website (http://web.expasy.org/protparam/) was used to evaluate the molecular weight (MW), isoelectric point (pI), instability index, aliphatic index, and grand average of hydropathicity (GRAVY). The NetPhos 3.1 server (http://www.cbs.dtu.dk/services/NetPhos/) was used to predicted the PgSUS proteins phosphorylation sites (Blom et al., 2004).

Nucleotide and amino acid sequences alignment of SUS genes from eight species

The nucleotide and proteins sequences of 68 SUS genes were collected from Arabidopsis thaliana (6), Oryza sativa (6), Glycine max (12), Malus domestica (11), Pyrus bretschneideri (17), Prunus persica (6), Vitis vinifera (5), and Punica granatum (5), respectively. Multiple SUS genes sequence alignments were performed using the CLUSTAL_X program (http://www.clustal.org/).

Phylogenetic analysis and classification of SUS gene family

A phylogenetic tree of 68 SUS proteins from eight species was generated using MEGA X software (http://www.megasoftware.net/). The tree was based on the maximum-likelihood (ML) method with the substitution model JTT+G+I and 1,000 bootstrap replications. PgSUS proteins were further categorized into different subfamilies according to the classification records of subfamily members of other species. The proteins sequences used in the phylogenetic analysis are listed in Data S1.

Gene structure construction, conserved motif, domain, and protein structure analysis

The information on gene structure for each of the 68 SUS genes was extracted from their GFF3 files. This data included sequence length, number, and arrangement of exons and introns. The conserved motif type and sequence of the SUS family were analyzed by MEME (http://meme-suite.org/tools/meme). The phmmer protein database (https://www.ebi.ac.uk/Tools/hmmer/search/phmmer) was used to annotate the MEME motifs. The conserved domains of the SUS proteins were determined using SMART (http://smart.embl-heidelberg.de/). The gene structure, MEME, and conserved domain results were plotted with TBtools (Chen et al., 2020). Secondary and tertiary structures of PgSUS proteins were predicted using NPS@: SOPMA (https://www.predictprotein.org/signin) and the ExPaSy Swiss-Model online software (http://swissmodel.expasy.org), respectively.

Syntenic analysis with four other species

MCScanX was used to obtain the syntenic relationships of five species: Arabidopsis thaliana, Malus domestica, Pyrus bretschneideri, Vitis vinifera, and Punica granatum (Wang et al., 2012). The results were presented with TBtools (Chen et al., 2020).

Cis-acting element analysis of PgSUS genes promoter regions

We extracted 2,000 bp gene sequences of genomic DNA sequences upstream of the initiation codon (ATG). These were used to predict the putative cis-acting elements using PlantCARE online software (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/).

Expression pattern analysis of candidate PgSUS genes in pomegranate

Two published sets of transcriptome data were used to investigate the expression characteristics of the PgSUS genes. The abundance of the PgSUS transcripts of 12 samples, including root, leaf, flower, and three different development stages of the pericarp, inner, and outer seed coats (50, 95, and 140 days after flowering, DAF), were collected from the NCBI Sequence Read Archive database (accession number SRP100581) (Qin et al., 2017). The expression profiles of the PgSUS genes were analyzed at different developmental stages of the seed coats in pomegranate cultivars ‘Dabenzi’ and ‘Tunisia’. These were collected at 50, 95, and 140 DAF and three biological replicates were collected per sample for RNA sequencing (accession number SRP212814, Qin et al., 2020). Clean reads of each sample were aligned to the pomegranate reference genome by HISAT2, using default parameter settings (Kim et al., 2019) after conducting a quality assessment of the filtered reads using Trimmomatic (Bolger, Lohse & Usadel, 2014). The mapped reads assembly of each sample was completed using StringTie (Pertea et al., 2015). The different gene expression levels were calculated according to transcripts per kilobase of exon model per million mapped reads (TPM). The TPM value was transformed into log₂ (TPM + 1). The heatmap of the PgSUS genes expression was plotted using TBtools (Chen et al., 2020).

Plant material

Samples were collected from three-year-old ‘Tunisia’ pomegranate trees at 26 °C under long-day conditions (14-h light/10-h dark) at approximately 60–70% humidity conditions. The trees were cultivated at the horticultural experimental station of Huaibei Normal University. We collected young root, mature leaves, and flowers. Healthy, uniform fruits were randomly collected at 45, 75, 115, and 150 DAF, respectively. Three replicates were prepared for each stage and each replicate contained 15 fruits. The fruit pericarp and seed coat were separated by hand. All samples were collected and immediately frozen in liquid nitrogen and stored at −80 °C.

Total RNA isolation and quantitative PCR expression assay

Approximately 1 µg of high quality RNA per sample was extracted using plant RNA extraction kits (TIANGEN, China). The first strand of cDNA synthesis was performed using the TIANScript II RT kit (TIANGEN, China). We diluted 20 µL of cDNA from each sample to a total volume of 200 µL using DEPC water. These were used as qPCR templates. The reaction mixture contained1 µL cDNA, 0.5 µL each of the forward- and reverse-specific primer, 10 µL chamQ SYBR qPCR Master Mix (Vazyme, China), and 8 µL DEPC water, for a total volume 20 µL. The qPCR reaction was conducted in an ABI 7300 Real-Time PCR system with the following amplification program: 95 °C for 5 min, 40 cycles of 95 °C for 5 s, and 60 °C for 35 s. The pomegranate PgActin gene served as the reference gene, and the relative expressions levels of the genes were calculated according to Livak & Schmitten (2001). Each sample was quantified in triplicate. Data were analyzed using SPSS software (22.0, USA) and Excel. All primers used for qPCR assay are shown in Data S2.

Identification of PgSUS genes

Two searches were performed to identify all possible SUS family members in the pomegranate genome. We obtained 14 putative PgSUS candidates by local Blast alignment according to query sequences of six Arabidopsis SUS proteins. Then, 19 PgSUS candidates were scanned from the pomegranate genome database based on the HMMER search. These two methods identified a total of 14 PgSUS candidates without a sucrose-phosphatase domain, which were verified with SMART and NCBI CDD databases. We found that 14 PgSUS candidates belonged to five genes and determined that each gene two-to-four transcripts after extraction and comparing the generic feature formats of these candidates. Finally, the five longest transcripts were isolated as the representative genes and were named PgSUS1 to PgSUS5, according to their chromosomal information (Table 1).

Five PgSUS were dispersed on four chromosomes (Table 1). Among them, one single SUS gene originated from Chr2, 4, and 6 respectively, while the rest two were located in Chr8. cDNA length analysis of five PgSUS genes revealed variations from 4,249 bp (PgSUS2) to 7,426 bp (PgSUS3). However, the coding DNA sequence (CDS) lengths were similar, and ranged from approximately 2,418 bp (PgSUS4) to 2,706 bp (PgSUS5). Their proteins were composed of 805-901 amino acids, the putative molecular weights (MW) ranged from 92.26 kDa to 102.58 kDa, and the theoretical isoelectric points (pI) were approximately 5.99 to 8.19. The instability index of the five PgSUS proteins ranged between 32.35 and 42.23. The aliphatic index (A.i) were between 81.60 and 92.87 and all of the PgSUS proteins were hydrophilic (Table 1). Our prediction of the phosphorylation sites in PgSUSs showed that serine was the most common site for phosphorylation Tow typically serine phosphorylation sites were observed in all PgSUS proteins (Data S3).

The ClustalW2 program was used to align the nucleotide/amino acid sequences of five pomegranate SUS and 63 SUS members with seven other species. The comparison results showed that these 68 genes shared a high sequence homology at the nucleotide level (average 65.93% identity) as well as the protein level (average 65.16% identity) (Data S4). Among the five PgSUS genes, the nucleotide and amino acid sequences of PgSUS1 were more similar to PgSUS4 and their identity scores reached 80.98% and 89.94%, respectively. PgSUS1 also showed similarity with PgSUS3 with the sequence comparison scores of nucleotide and amino acid sequences of 68.03% and 69.44%, respectively. A pair of PgSUS genes (PgSUS2-PgSUS5) were also observed to be closely related (68.51% and 70.96% respectively) (Data S4).

Phylogenetic analysis of SUS family members

We used five PgSUS from pomegranate, six AtSUS from Arabidopsis, and 57 other SUS proteins to construct the phylogenetic tree in order to clarify the evolutionary relationships. A total of 68 SUS proteins results from phylogenetic analysis were classified into three distinct subgroups categorized as SUS I, SUS II, and SUS III (Fig. 1). Corresponding to the nucleotide/amino acid sequence identity (Data S4), PgSUS1 was clustered with PgSUS4 to form the SUS I branch, which contained well-characterized SUS genes including AtSUS1/2, PpSUS1/2/15, and VvSS4. PgSUS3 belonged to SUS II, which included MdSUS2.1 and VvSS3. Compared with the SUS I and SUS II subgroups, the genes clustered in the SUS III subgroup typically contained the proteins with a C-terminal extension, such as PgSUS2/5, AtSUS5/6 and MdSUS3.1/3.2/3.3 (Data S5). The results showed that although these SUS family genes shared high sequence similarities, including five pomegranate SUS genes, diversification was identified in this family through phylogenetic analysis.

Gene structure, conserved motif, and domain analysis of SUS family genes

We further investigated the exons/introns exon/intron structure of all SUS genes to better understand the molecular evolution mechanism. These included five in pomegranate and 63 in other seven species according to the gene annotation files (Fig. 2A). SUS family gene sequences were split into approximately 15 exons in SUS I, and 14 exons in SUS II and SUS III genes, respectively, after taking introns loss into account (Fig. 2A; Data S6) (Xu et al., 2019). The nucleotide sequences of 68 SUS genes showed high similarity (Data S4), therefore, high conservation of these gene structures was expected. Exons with lengths of 152/155, 193, 177/174/129, 117, 167 and 225, were highly conserved and arranged in same order in the CDS regions of all three SUS subgroups (Data S6). For each SUS subgroup, the gene structure also showed unique features: compared with SUS II genes, intron loss was a common phenomenon in SUS I and SUS III genes (Data S6). In the SUS I subgroup, exons with lengths of 336, 432, and 564, were split into two (119 and 217), three (119, 217 and 96) and two (322 and 245) exons in the SUS II subgroup, respectively. In the SUS III subgroup, exons with lengths of 567, were split into two exons (322 and 245) in the SUS II subgroup (Data S6). The exon sizes and splitting varied among the 3′ end of the genes of the SUS III subgroup. This was associated with the 3′ extension of SUS III proteins (Data S5; Data S6). In the SUS genes of pomegranate, the exons with lengths of 336 (or spilt into 119 and 217), 96, and 139 were typically conserved in PgSUSs (Data S4; Data S7). Moreover, in the same group, PgSUS genes showed a similar exon number, arrangement, and length with SUS genes from Arabidopsis (Baud, Vaultier & Rochat, 2004), apple (Tong et al., 2018), grape (Zhu et al., 2017), peach (Zhang et al., 2015), pear (Lv et al., 2018), soybean (Xu et al., 2019), and rice (Hirose, Scofield & Terao, 2008) (Fig. 2A; Data S6; Data S7).

We used the MEME online server to predict 15 motifs in the SUS gene family (Fig. 2B). Detailed information of these motifs is shown in Data S8. Among these motifs, the motifs 1, 3, 5, 6, 9, 10, 11, 12, and 13 represented the sucrose synthase domain; motifs 2 and 7 corresponded to the glycosyl-transferase domain, and the motif feature of motifs 4, 8, and 15 were unknown (Data S8). The majority of the SUS proteins from eight species contained the 14 predicted motifs, except motif 14, and showed a consistent array (Fig. 2B). Motifs 2 and 7, as the elements of the glycosyl-transferase domain signature, were highly conserved, suggested that these motifs are essential for enzyme function of sucrose synthase. However, several motifs which corresponded to the sucrose synthase domain were missing and the motif composition of some members were found to be variants in apple, pear, peach, and soybean (Fig. 2B). The five PgSUS members also shared common conserved motif compositions and had consistent arrangement (motifs 12, 9, 10, 11, 6, 3, 13, 1, 5, 7, 4, 2, 8 and 15) (Fig. 2B).

Tow typically conserved domains corresponding to the motifs features (sucrose synthase domain and the glycosyl-transferase domain) were screened in each member of 68 SUS proteins by matching with SMART and NCBI CDD (Fig. 2C). These two conserved domains were located at the N and C-terminal ends, respectively. This was consistent with the motif arrangement (Figs. 2B, 2C). In pomegranate, the length and distribution of two conserved domains of five SUS protein showed high consistency and conservation (Fig. 2C), indicating that they are critical for the function of PgSUS proteins.

Prediction of protein structure of pomegranate SUS proteins

The secondary structures analysis showed that five pomegranate SUS proteins were composed of α-helices, extended β strands, β-turns, and random coils (Table 2; Data S9). The α-helix was the major secondary structures among the five PgSUS proteins, accounting for 49.72–53.97%, followed by random coils (25.73–32.74%) and extended β strands (12.26–13.21%) (Table 2). These secondary structure distributions were also highly conserved in five PgSUS polypeptide chains (Data S9).

We predicted tertiary structures of the five pomegranate SUS proteins using the Swiss-model online software. The three-dimensional models of the PgSUSs proteins were based on templates 3s27 (Sucrose synthase) and 4rbn (Glycosyl transferases group 1). The results showed that the tertiary structure for PgSUS1 to PgSUS5 had two symmetric tetramers and comprised four main polypeptide chains. These were similar to PpSus1 to PpSus4 in peach (Data S10; Zhang et al., 2015). The 3D structure of PgSUS1 was quite similar with PgSUS4 among PgSUS1 to PgSUS5 (Data S10).

Syntenic analysis of five species SUS genes

We analyzed the syntenic relationships between pomegranate and four other species, including A. thaliana, M. domestica, P. bretschneideri, and V. vinifera to explore the evolutionary process of pomegranate SUS genes. Four SUS genes were found to have ten orthologous syntenic gene pairs in another four species (Fig. 3). PgSUS1 was found to be syntenic with four genes from apple (MdSUS1.1 and MdSUS1.4), pear (PbrSUS17), and grape (VvSS4). Three genes (AtSUS6, MdSUS3.1 and PbrSUS12) showed synteny with PgSUS5, two genes (PbrSUS17 and VvSS4) were syntenic with PgSUS4, and PgSUS3 was syntenic only with VvSS3 (Fig. 3). The syntenic relationships of these SUS orthologous gene pairs were consistent with their phylogenetic relationship (Fig. 1). However, PgSUS2 in pomegranate was found to have no syntenic counterpart in the other four species. These results help to better understanding the possible roles of SUS gene family members in pomegranate.

Cis-acting element analysis of PgSUS genes promoters

The cis-acting elements are crucial in the spatial–temporal and tissue-specific expression of genes. The cis-acting elements of the PgSUS genes were classified into five categories using the PlantCARE database The categories were: hormone responsive elements (HRE), tissue specific elements (TSE), light responsive elements (LRE), stress responsive elements (SRE), and others responsive elements (ORE) (Fig. 4). Detailed information of cis-acting elements in five PgSUS promoter regions is provided in Data S11. The number of LREs was the largest group (49%), followed by HREs (24%), SREs (15%), OREs (7%) and TSEs (6%) (Fig. 4A; Data S11). Among these, the presence of LREs was universal in five PgSUS genes promoters. The PgSUS3 promoter contained 22 LREs, which was almost two times that of other PgSUS genes These results imply that PgSUS3 may respond to light induction. Other PgSUS genes promoters contained several HREs, with the exception of PgSUS2. These hormones include abscisic acid (ABA), auxin, gibberellin (GA), methyl jasmonate (MeJA), and salicylic acid (SA). MeJA and ABA responsive elements were prevalent in the promoter regions of those four genes. Moreover, each PgSUS promoter contained tow-to-eight SREs, and were responsive to stresses including anoxic environments, low-temperatures, and drought (Fig. 4B; Data S11). In addition, PgSUS genes promoters also contained several OREs, such as circadian control, cell cycle regulation, and MYB binding sites, implying that PgSUS family genes may play multiple roles in plant growth and development.

Expression profile of pomegranate SUS family genes, assessed with RNA-seq and qPCR

In order to analyze the molecular functions of the SUS genes in pomegranate, we studied the transcript characteristics of the PgSUS genes using RNA-seq data downloaded from the NCBI SRA database (Fig. 5). For transcriptome analysis, a total of 300.88 Gb clean data with an average of 94.49% bases scoring Q30 were obtained from 42 RNA-seq libraries. The GC content of all samples ranged from 49.50 to 52.80%. It was found that more than 96% of the reads aligned with the pomegranate genome sequence, indicating a high sequencing quality and that the resulting data was reliable for subsequent analyses (Data S12).

PgSUS genes exhibited an obvious tissue specific expression pattern (Fig. 5A). The PgSUS family members of the SUS I and SUS II subgroups were predominantly expressed in sink organs, particularly in fruit tissues. With the fruit development, PgSUS1, PgSUS3, and PgSUS4 transcripts displayed different expression characteristics. PgSUS1 was mainly expressed in the inner and outer seed coats, and reached its peak at 95 DAF in the outer seed coat. The PgSUS3 transcript was expressed at higher levels in the seed coat and pericarp (Fig. 5A). As the pericarp developed from 50 DAF to 95 DAF, PgSUS3 expression gradually increased to the highest level, but slightly declined at fruit harvest (Fig. 5A). Interestingly, PgSUS4 was strongly expressed in the sink organ, including the outer seed coat, root, and flower. PgSUS4 showed a similar expression trend with PgSUS1 as the outer seed coat developed from 50 DAF to 140 DAF. Its abundance rapidly increased on the 95 DAF (Fig. 5A). However, PgSUS2 and PgSUS5 of the SUS III subgroup were slightly expressed in the root, leaf, and flower, but was almost undetectable in fruits tissues (Fig. 5A). Furthermore, similar expression trends of PgSUS genes were also observed during the fruit development in the ‘Dabenzi’ and ‘Tunisia’ pomegranate cultivars (Fig. 5B).

QPCR was used to analyze the expression patterns of PgSUS genes. The relative expression level of each gene in different organs or tissues was standardized with their expression level in the leaf (Fig. 6). All five genes were up-regulated in root and flower compared with their expressions in the leaf. The relative expression level of PgSUS4 increased more significantly than the other PgSUSs genes in root and flower. During fruit development, expressions of PgSUS1 and PgSUS4 rapidly increased with a tendency toward to a gradual decrease as the seed coat developed from 45 DAF to 150 DAF, which peaked at 75 DAF. These results suggest that isozymes encoded by these two genes may be involved in catalyzing key aspects of sucrose metabolism in the fruit seed coat during the early- and middle- developmental stages. PgSUS3 showed stable expression levels when the fruit seed coat developed from 45 DAF to 115 DAF. Additionally, PgSUS3 showed higher transcript levels in the pericarp than other genes, indicating that PgSUS3 may play an important role in sucrose metabolism during the development of the pomegranate fruit pericarp. However, the transcripts levels of PgSUS2 and PgSUS5 were slightly or not-at-all expressed during fruit development. Our results show that three SUS genes (PgSUS1, PgSUS3 and PgSUS4) may contribute to the sucrose metabolism and fruit development.