Introduction

Plants comprise various fatty acids and derivatives, including plant lipids, sterols, glycerol lipids, and lipid polyesters. Lipids, the fundamental components of extracellular protective layers and signaling molecules, help construct the cell membrane and provide the energy needed for cell metabolism. Moreover, it plays a significant role in various physiological and metabolic processes, such as plant growth, development, and stress tolerance (Gidda et al. 2009; Okazaki and Saito 2014; Nakamura 2017). Several genes are involved in lipid synthesis (Guschina et al. 2014), and one of these genes, sn Glycerol 3 phosphate O acyltransferase (GPAT; [EC] 2.3.1.15), is required for glycerolipid production. It is also involved in a variety of metabolic and physiological processes. The Kennedy pathway is known as the synthesis pathway of glycerolipids (Chen et al. 2022; Zou et al. 2023). GPAT catalyzes the initial step in nearly all phospholipid production. To generate lysophosphatidic acid, several GPATs transfer an acyl group from acyl-CoA or acyl-ACP (acyl-acyl carrier protein) to the sn-1 position of glycerol-3-phosphate (G3P) (LPA). LPA is important for synthesizing and storing membrane lipids and extracellular lipid polyesters. These GPATs are sn-1 in nature. Other GPATs exhibit additional phosphatase activity, which results in the major product sn-2 monoacylglycerol (2-MAG). In nature, these GPATs are of the sn-2 type, and they play a significant role in the synthesis of cutin and suberin in terrestrial plants (Li et al. 2007, 2023; Beisson et al. 2012; Yang et al.2012).

More than 60 years ago, GPATs were discovered biochemically in plant and animal tissues (Weiss et al. 1939; Kornberg and Pricer 1987). Plant GPAT genes active in three distinct organelles, namely the chloroplast, mitochondria, and endoplasmic reticulum (ER), have been identified (Gidda et al. 2009). Membrane-bound GPATs in mitochondria and the ER prefer the acyl donors acyl-CoA and acyl-ACP. Other GPATs prefer acyl-ACP as an acyl donor in chloroplasts and are soluble (Zheng et al. 2003). When GPAT amino acid sequences from many species were aligned, it was shown that these enzymes had at least four highly conserved motifs (Block I, II, III, and IV) that are imperative for both acyltransferase activity and glycerol-3-phosphate substrate binding (Lewin et al. 1999). There are 10 GPAT genes in Arabidopsis, two of which are of the sn-1 type (AtGPAT1-9 and ATS1). While the ATS1 gene in Arabidopsis chloroplasts is involved in the synthesis of phosphatidylglycerol via the use of acyl groups in acyl-ACP (Nishida et al. 1993; Xu et al. 2006), AtGPAT9 is involved in pollen grain development and is responsible for plant membrane oil and lipid biosynthesis (Shockey et al. 2016). Moreover, multiple GPAT genes have been identified as important in pollen and anther development. AtGPAT7 and AtGPAT5, for example, are acyltransferases with strong acyl-CoA specificity and are required for suberin synthesis (Yang et al. 2012). AtGPAT1 is required for nutrient secretion and band differentiation. (Zheng et al. 2003). AtGPAT6 results in elongation of the pollen outer shell and tapetum and stamen (Zheng et al. 2003; Li et al. 2012). Many GPAT genes are required for male fertility in tomato, maize, and rice. Pollen production was impaired in the tomato gpat6 mutant (Petit et al. 2016). OsGPAT3 has been linked to the development of anther cuticles and pollen outer shells (Men et al. 2017). The GPAT gene family was examined in Brassica napus (Wang et al. 2020) Zea mays (Xu et al. 2019), Oryza sativa (Safder et al. 2021), Helianthus annuus (Paya-Milans et al. 2016), and Solanum lycopersicum (Petit et al. 2016; Fawke et al. 2019).

Any change that characterizes or disturbs plants' metabolic and physiological balance in their living environment is classified as stress (Shulaeva et al. 2008). Drought and salt stress are the most detrimental abiotic stresses affecting plant development and growth (Farooq et al. 2009). A previous study reported that the GPAT gene is associated with stress tolerance (Paya-Milans et al. 2016). In tomato leaves, the expression of the soluble GPAT gene (LeGPAT) increases at low temperatures and decreases at high temperatures. LeGPAT overexpression enhanced the unsaturated fatty acid content of phosphatidylglycerol, improving freezing tolerance (Sui et al. 2007b). Similarly, in tomato plants, the overexpression of soluble LeGPAT promoted freezing (Sui et al. 2007a) and salt tolerance (Sun et al. 2010), while in Arabidopsis, the overexpression of the Suaeda salsa GPAT gene (SsGPAT) improved salt tolerance (Sui et al. 2017).

Melatonin has several functions in plants, the most important of which is to protect the plant from different stress effects by acting as an antioxidant (Tan et al. 2012). Melatonin also influences the expression of critical transcription factors and enzymes involved in defense (Weeda et al. 2014). Melatonin has been demonstrated to contribute to plant growth and development, yield increase, and stress tolerance after research on various plant species, including wheat and barley (Hernandez-Ruiz et al. 2005). According to related research, melatonin promotes resistance to salt and drought conditions in soybean plants (Wei et al. 2015).

The common bean (Phaseolus vulgaris L.) is a diploid grain with a modest genome (633 Mbp (0.66 pg)/IC) and two pairs of chromosomes (2n = 22) (Arumuganathan and Earle 1991). Directly consumed beans are the most critical food grain for humans worldwide (Singh et al. 1991). This study presents a phylogenetic tree, chromosomal location, gene structure comparison, and gene expression models of GPAT genes. This study also provides information on the diversity and orthologous-paralogous associations of GPAT genes in A. thaliana, Glycine max, and P. vulgaris plants. qRT-PCR revealed changes in the expression of PvGPAT genes due to stress and melatonin treatment. These findings will lay the groundwork for future functional research. Additionally, this study aimed to better understand the function of GPAT genes in beans and to fill a gap in the literature on the subject.

Materials and methods

Search and identification of GPAT genes in Phaseolus vulgaris, Glycine max, and Arabidopsis thaliana

The Pfam database was utilized to obtain the Pfam Accession Number (PF01553) needed to obtain the protein sequences of the GPAT gene family members in the genomes of common bean (Schmutz et al. 2014), soybean (Valliyodan et al. 2019), Arabidopsis (Lamesch et al. 2011), Vitis vinifera, Cicer arietinum and Lens culinaris. To find all probable GPAT protein sequences with this accession number, BLASTP and Hidden Markov Model (HMM) screening were performed with the Phytozome v13 database. In addition, the CDS (coding sequence), genomic, transcript, and promoter sequences of the PvGPAT genes were obtained from the Phytozome v13 database.

Analyses of the structure, conserved motifs, cis-regulatory elements, chromosomal location, and gene duplication status of the GPAT genes

The exon and intron organization of the GPAT genes of P. vulgaris were determined using genomic DNA sequences stored in databases as well as by understanding the principles and potential effects of gene organization and structure on protein evolution and functional differences between species (Wang et al. 2013). The annotation-gene-transcript (gff3) file, which contains information about the start and end points of the exon–intron regions of PvGPAT genes, was obtained from the Phytozome v13 database. Additional conserved motifs of PvGPAT amino acid sequences were obtained using the program ‘’Multiple EM for Motif Elimination (MEME)’’ (Bailey et al. 2006). The maximum number of motifs was set at 10, while the range of motif widths was specified at 2 to 50. The theme zones vary from 6 to 300. Any number of repetitions was used to establish the area distribution. The default settings of the InterPro database were used to search for recognized motifs (Quevillon et al. 2005). Conserved motif information for the PvGPAT genes was obtained from the ‘’NCBI Conserved Domain Search’’ browser. The obtained data were visualized with the TbTools program (Chen et al. 2020). Additionally, for conserved region sequence analysis in beans, sequence logo analyses of the BLOCKI-II-III and -IV domains of PvGPAT proteins were performed using the WEBLOGO online web tool (Crooks et al. 2004). The sizes and placements of the PvGPAT genes on the bean chromosome were determined using the Phytozome v13 database. The PvGPAT gene locations were noted on all the bean chromosomes and plotted using the TbTools program. The nonsynonymous ratios (Ka), synonymous ratios (Ks), and evolutionary compulsions (Ka/Ks) between binary gene pairs of PvGPAT genes were defined using the PAML interface (Yang 2007) and the PAL2NAL browser (Suyama et al. 2006). Using PlantCARE (Lescot et al. 2002), cis-element analysis was applied to 5 upstream regions for all members of the PvGPAT gene family, comprising almost 2 kb of DNA. The acquired data were analyzed using the TbTools tool.

Physiological parameters and subcellular localization of PvGPAT proteins

The amino acid number, theoretical isoelectric point (pI), molecular weight, instability index, and stabilization state of the GPAT proteins in the beans were determined using ProtParam. The subcellular localizations of the bean GPAT protein sequences were determined using WoLF PSORT (Horton et al. 2007).

Analyses of the phylogenetic tree and sequence alignment

The maximum likelihood (ML) model was employed for the phylogenetic analysis with a bootstrap value of 1000 replicates. The GPAT protein sequences of G. max, A. thaliana, P. vulgaris, V. vinifera, L. culinaris, and C. arietinum were aligned using ClustalW (Thompson et al. 1997). Using the aligned data, a phylogenetic tree was generated using the MEGA v11 program (Kumar et al. 2016). The evolutionary tree was created using the Interactive Tree of Life (iTOL) browser (Letunic and Bork 2011).

Comparative mapping between bean, soybean, and Arabidopsis (Synteny analysis)

To determine the orthologs of the GPAT genes from A. thaliana, G. max, and P. vulgaris, the Multiple Collinearity Scan Toolkit (MCScanX) was utilized (Wang et al. 2012). The protein sequences of orthologous genes were discovered using Phytozome v13. A synteny map of the GPAT genes in Arabidopsis, bean, and soybean was constructed using TbTools.

Gene ontology (GO) analysis, homology modeling, and protein–protein interactions of PvGPAT proteins

Previously found GPAT protein sequences were used to model the proteins. Simulations of the 3D structures of proteins were acquired using Phyre2 (Kelley et al. 2015). Simulations with a confidence interval > 90% were preferred. The Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) database (Szklarczyk et al. 2021) was used to identify the physical and functional links of the protein–protein interactions. The collected data were classified and linked with a confidence score (0.7) for every protein–protein interaction. The maximum number of interacting proteins was set to 5. Functional genomics approaches must be applied to annotate new sequencing data in plant biotechnology research. GO data (molecular functions, cellular components, and biological processes) of the PvGPAT proteins were accessible using the Blast2GO, Omics Box (Götz et al. 2008), STRING database, and DAVID: Functional Annotation Tool online browsers (Huang et al. 2007) tools to get functional information on the PvGPAT proteins.

Transcriptomic gene expression levels of PvGPAT genes

To examine the expression patterns of PvGPAT genes, Illumina RNA-seq data were collected from the NCBI database’s Sequence Read Archive (SRA) data bank (https://www.ncbi.nlm.nih.gov/sra). To locate relevant RNA-seq data, salt and drought stress accession numbers were obtained. The SRR8284480 (control) and SRR8284481 (14 days of drought) accession numbers were utilized for drought stress analysis. The SRR13434024 (control), SRR13434021 (3 h salt), SRR13434019 (6 h salt), SRR13434026 (12 h salt), and SRR14120315 (24 h salt) accession numbers were utilized for salt stress analysis. The reads per kilobase of the exon model of one million mapped reads (RPKM) technique was used to standardize the gene expression data. In addition, expression data corresponding to different developmental stages of bean plants (mature green pods, young pods, stems, roots, nodules, leaves, flowers, and flower buds) were obtained from the Phytozome v13 database in the form of FPKM. Heatmap analyses were created in the TbTools program using RPKM and FPKM data.

Applications for plant growth, enhanced agents, and stress

The Phaseolus vulgaris cv ‘’Serra’’ and ‘’Elkoca-05’’ used in this study were acquired from Erzurum Technical University, Science Faculty, Department of Molecular Biology and Genetics in Türkiye. Seeds were surface sterilized in a solution containing 1% (v/v) NaOCl for five minutes. Afterward, the seeds were planted in perlite-filled pots moistened with Hoagland's solution, a plant nutrient solution. The seedlings were kept in the pots until they had three leaves in a controlled environment growth chamber with a 250 mmol m−2 s−1 photosynthetic photon flux at 25 °C and 70% relative humidity. Hogland's solution was used to hydrate the seedlings as needed. Identical-sized plants were subsequently transferred to a hydroponic media supplemented with 0.2L of prepared 1/10 Hoagland's solution. Macronutrients, different concentrations of ions, and micronutrients are present in the Hoagland solution. The same environmental conditions were maintained for incubating common bean plants. Then, on the seventh day after the transfer, melatonin was administered as a foliar spray at concentrations ranging from 0 to 200 MEL 24 h before the stress treatment. For nine days under the same growth conditions, moderate salinity stress was applied using Hoagland solution supplemented with 150 mM NaCl. However, for 24 h, drought-stressed bean plants produced under identical conditions were preserved in Hoagland solution supplemented with 0 (control) or 20% PEG6000 (Aygören et al. 2023). The leaf tissues of the bean genotypes were taken into liquid nitrogen after these times and maintained at − 80 °C until the study was conducted. Three replicates of the bean genotypes used in the study were grown, and qRT-PCR was performed with three biological replicates.

Total RNA isolation, cDNA synthesis, and qRT-PCR analysis

Trizol® Reagent was used to isolate total RNA. The RNA quality was checked using a Multiskan Go spectrophotometer, and the RNA was photographed on a 1.5% agarose gel. A SensiFAST™ cDNA Synthesis Kit (Cat No: Bio-65053, UK) was used for synthesizing complementary DNA (cDNA) according to the manufacturer's procedure. Six PvGPAT genes were chosen for qRT-PCR analysis based on RNAseq data. qRT-PCR was carried out using the RotorGene® Q Real-Time PCR System with the ABT SYBER Green (Cat No: Q03-02-01, Ankara, Türkiye) mixture. The qRT-PCR mixture was 20 µL, of which the cDNA volume was 200 ng, both the forward and reverse primers were 0.4 µL, and the 2 SYBER Green Mix was 10 µL, with the reaction volume being 20 µL with double distilled water. The reaction was performed as follows: 95 °C for 10 min, followed by 40 cycles of 94 °C for 15 s, 60 °C for 30 s, and 72 °C for 30 s. The β-actin gene from P. vulgaris was used as a housekeeping gene. The qRT-PCR data were standardized using the 2−∆∆CT technique for relative quantification (Livak and Schmittgen 2001). Supplementary file 1 contains information on the PvGPAT primers used in this investigation. GraphPad Prism 9 (http://www.graphpad.com/) was used for statistical analyses based on one-way ANOVA and Dunnett's test at the 0.05 significance level.

Results

Characterization of GPAT genes in Phaseolus vulgaris

The P. vulgaris, Arabidopsis, and G. max GPAT genes were identified via a keyword search using the Phytozome v13 browser and the Pfam number (PF01553) from the Pfam database. Different alternative versions of the same gene were deleted due to this gene screening, and GPAT genes containing the PlsC (phosphate acyltransferase) and glycerol-3-phosphate acyltransferase domains were identified. There are 10 Arabidopsis genes, 28 soybean genes, and 13 bean genes. Sup. file 2 contains the protein sequences of the discovered GPAT genes. In addition to the PlsC domain, several bean genes include an N-terminal haloacid dehalogenase (HAD)-like hydrolase domain, an LPLAT superfamily member, and an N-terminal glycerol-3-phosphate acyltransferase domain (GPAT_N) (Fig. 3). For example, the PvGPAT1 gene contains the GPAT_N domain and the LPLAT superfamily. Additionally, PvGPAT4, 6, 7, 8, 9, 11, and 12 genes have the HAD domain. The PvGPAT2 and PvGPAT3 genes do not contain additional domains. Table 1 shows the 13 PvGPAT genes identified from the bean genome and their identification information. GPAT genes were found on the bean chromosomes PvChr2, PvChr3, PvChr6, PvChr7, PvChr8, and PvChr10 (Fig. 1). Most of the genes (PvGPAT7, PvGPAT8, PvGPAT9, PvGPAT10, and PvGPAT11) are on the PvChr8 chromosome. Table 1 also includes amino acid numbers, predicted isoelectric points, molecular weights (kDa), stability states, and gene instability indices. PvGPAT10 had the maximum number of amino acids (539), whereas PvGPAT2 and PvGPAT3 had the lowest number of amino acids (376). Its theoretical alkaline isoelectric point varies from 8.41 to 9.46. PvGPAT9 had the lowest value, while PvGPAT12 had the highest value. Seven bean GPAT proteins were unstable, whereas the other six were stable. The instability index varies between 31.84 and 51.38. These were the PvGPAT4 and PvGPAT1 values, respectively. When their molecular weights were compared, PvGPAT10 (60.40 kDa) was the highest, and PvGPAT2 was the lowest (43.26 kDa).

Table 1 Information about PvGPAT proteins
Full size table
Fig. 1
figure 1

Location of PvGPAT genes on bean chromosomes. The gene pairs colored blue and red are segmental duplicate genes. The gene pairs colored green are tandem duplicate genes

Full size image

PvGPAT4/PvGPAT10 and PvGPAT7/PvGPAT12 were segmentally duplicated genes, whereas PvGPAT8/PvGPAT9 were tandemly duplicated genes, according to gene duplication analyses. Table 2 displays the Ka and Ks values, Ka/Ks ratios, and duplication types of these duplicate genes. According to the Ka/Ks data, these beans were exposed to functional limitation and negative selection pressure throughout the evolutionary process. The process of GPAT gene duplication and divergence (MYA) was calculated using the synonymous mutation rate of one substitution per synonymous region per year as T = Ks/2 (= 6.56E9) (Büyük et al. 2016). The PvGPAT4/PvGPAT10 genes were isolated 11.04 million years ago, the PvGPAT7/PvGPAT12 genes 67.96 million years ago, and the PvGPAT8/PvGPAT9 genes 45.7 million years ago.

Table 2 Duplicate GPAT gene pairs in the bean genome
Full size table

The subcellular localization of bean GPAT genes is given in Sup. file 3. Using the WoLF PSORT database, it was determined that the intracellular localizations of PvGPAT genes were localized within in twelve different regions such as the nucleus, chloroplast, and mitochondria. Although PvGPAT3 was found in eight different subcellular compartments, PvGPAT1 and PvGPAT12 were found in only two different subcellular compartments. Only the PvGPAT1, PvGPAT3, and PvGPAT4 genes were found in the chloroplast organelle.

Phylogenetic and gene structure analyses of the GPAT genes

A phylogenetic tree was drawn from the protein sequences of the GPAT genes of Arabidopsis, bean, grape, chickpea, lentil, and soybean plants. The phylogenetic tree was constructed to determine the phylogenetic relationships between GPAT proteins. A phylogenetic tree of GPAT proteins was drawn using the maximum likelihood method with a bootstrap value of 1000 repeats and shaped in the ITOL program (Fig. 2). When the phylogenetic tree was examined, the GPAT proteins were divided into 3 groups: Group I: ATS1-like, Group II: GPAT9-like, and Group III: Other GPATs. Group I contains chloroplastic GPAT genes of different plants. These genes are orthologs of the ATS1 gene of Arabidopsis. The GPAT genes in Group II are orthologs of the Arabidopsis GPAT9 gene. Group III included genes other than these genes.

Fig. 2
figure 2

Phylogenetic tree of the GPAT genes. A phylogenetic tree was created using the Mega v11 program with amino acid sequences from soybeans, Arabidopsis, grapes, chickpeas, lentils, and common beans. The tree was branched into 3 groups with a total of 79 proteins, containing Group I: ATS1-like, Group II: GPAT9-like, and Group III: Other GPATs. The figures are bootstrap values calculated from 1000 replicates

Full size image

The phylogenetic tree, domain, and conserved motif analysis results created with PvGPAT proteins are presented in Fig. 3. Ten conserved motifs were found in the conserved motif analysis, with amino acid counts ranging from 32 to 50. PvGPAT4 and PvGPAT10 had the highest number of motifs (9). PvGPAT1 does not contain any motifs. PvGPAT2 and PvGPAT3 had the lowest number of motifs and had only 3 motifs. Motif 1 is found in all GPAT proteins except PvGPAT1. Sup. file 4 contains a list of the best probable match sequences for the motifs. Additionally, on the right side of Fig. 3, a visual representation of the gene structure analysis of the PvGPAT genes is presented. The PvGPAT genes contained multiple exon regions; some PvGPAT genes had almost the same exon–intron patterns. As a result of gene structure analysis, exon numbers were determined to be 2, 3, and 12, and intron numbers were defined as one, two, and eleven. PvGPAT1, PvGPAT2, and PvGPAT3 had the highest exon numbers. The other GPAT genes shared almost the same intron–exon sequence. Almost all PvGPAT genes contain the HAD and LPLAT superfamily domains. Unlike the others, PvGPAT2, PvGPAT3, and PvGPAT4 contain the acyltransferase domain, and PvGPAT1 contains the GPAT_N domain.

Fig. 3
figure 3

Phylogenetic relationship between PvGPAT proteins, the conserved motif diagram of PvGPAT proteins, additional domains contained in PvGPAT proteins, and the gene structure of PvGPAT genes. A description of each colored box is shown in the figure

Full size image

Gene ontology analysis

Gene Ontology (GO) analyses of bean GPAT genes were performed by Blast2GO® and Omics Box tools with default parameters (level 1). Additionally, gene ontology analyses were supported by the STRING database and DAVID: Functional Annotation Tool online browser (Sup. file 5). The percentages of participation of these genes in the analyses of molecular function, cellular components, and biological processes were determined. According to the findings obtained, the GPAT genes function in two different molecular functions; hydrolase activity (42.86%) and transferase activity (57.14%). GPAT proteins were found in two different cellular compartments: the chloroplast stroma (8.33%) and an integral component of the membrane (91.67%). It was also involved in six separate biological processes involving arbuscular mycorrhizal association (4.54%), cutin biosynthetic process (40.91%), phosphatidylglycerol biosynthetic process (4.55%), CDP-diacylglycerol biosynthetic process (4.55%), response to symbiotic fungus (4.54%) and dephosporylation (40.91%) (Fig. 4). In parallel with the GO analyses of GPAT genes previously performed in different plants, the GO results obtained from this study also showed that the GPAT genes in beans are functionally involved in glycerol-3-phosphate acyltransferase activity. In addition, these findings indicate that the GPAT gene is involved in the synthesis of various lipid products and their membrane-bound cellular components in biological processes (Safder et al. 2021).

Fig. 4
figure 4

The participation rates of the PvGPAT gene in biological processes, cellular components, and molecular functions are depicted as a pie chart

Full size image

Identification of cis-acting elements in the promoter region of PvGPAT genes

The 2000 bp upstream promoter region of the GPAT genes was removed and analyzed in the PlantCARE database to determine the transcriptional regulation of the bean GPAT genes. Thirty-eight different cis-acting elements associated with stress were defined in the promoters of the GPAT genes and these data were visualized with TbTools (Fig. 5 and Sup. file 6). According to this analysis, the MBS element associated with the drought stress response was found in PvGPAT1, 2, 8, 12, and 13 genes. ABRE, which is also effective in combating drought and salinity stress, was detected among the PvGPAT1, 3, 5, 6, 8, 10, 11, 12, and 13 genes. The DRE-core element was found only in the PvGPAT3 and PvGPAT5 genes.

Fig. 5
figure 5

Cis-acting elements in the promoter regions of PvGPAT genes. Using the TbTools tool, this study was created by choosing cis-acting elements associated with various stress conditions

Full size image

Three-dimensional structural modeling and protein–protein interactions of PvGPAT proteins

The GPAT proteins in the Protein Data Bank (PDB) were searched with BLASTP and three-dimensional homology modeling of these proteins was performed in a visually intelligible manner using Phyre2, a database used to predict the structure and function of GPAT proteins. Supp. File 7 shows the three-dimensional homology patterns of the proteins identified in this analysis. In the three-dimensional structure analysis of PvGPAT proteins, simulations with a modeling confidence level greater than 90% were preferred. When the three-dimensional structure of the PvGPAT proteins was examined, it was observed that they had many alpha-helix structures but fewer beta-sheet structures. Except for that of PvGPAT1, the three-dimensional structures of the other GPAT proteins were similar. According to three-dimensional homology modeling predictions of PvGPAT proteins, these proteins also have different interaction domains. However, they differ in their location and type of binding sites. However, functional research is required to better understand these proteins.

The Protein–protein interactions of PvGPATs were detected via the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) interface using PvGPAT protein sequences. The PvGPAT1 protein interacted most with other GPAT proteins. PvGPAT6 did not interact with other GPAT proteins. The PvGPAT5, PvGPAT8, and PvGPAT9 proteins interacted only with the PvGPAT1 protein. The V7AR36_PHAVU, V7BKF3_PHAVU, V7BFX5_PHAVU, V7BEX4_PHAVU, and V7CF69_PHAVU proteins were the 5 proteins that interacted most with PvGPAT proteins (Figs. 6, 7). The NCBI IDs of these five proteins and the molecular functions in which they played a part in are given in Table 3. Accordingly, PvGPAT proteins mostly interacted with proteins involved in glycerol degradation, diacylglycerol/triacylglycerol biosynthesis, and phosphatidylcholine acyl regulation pathways.

Fig. 6
figure 6

Protein–protein interactions (PPIs) of PvGPAT proteins with each other and with other proteins

Full size image
Fig. 7
figure 7

Metabolic pathway diagram of PvGPAT proteins and the 5 proteins with which they interact most

Full size image
Table 3 Information about the proteins that interacted most strongly with PvGPAT proteins
Full size table

Synteny analysis of GPAT genes from P. vulgaris and other plants

A synteny map of the GPAT genes of common bean, soybean, and Arabidopsis plants was constructed with the TbTools tool (Fig. 8). Seven syntenic links were discovered in the synteny study of A. thaliana and P. vulgaris, and 28 syntenic relationships were discovered in the synteny analysis of P. vulgaris and G. max. As a result of the analysis applied between P. vulgaris and A. thaliana, it was identified that AtATS1 gene with PvGPAT1 gene, AtGPAT2 and AtGPAT3 genes with PvGPAT5 gene, AtGPAT6 gene with PvGPAT7 and PvGPAT12 genes and PvGPAT13 gene with AtGPAT4 and AtGPAT8 genes were orthologous genes. As a result of the synteny analysis applied between P. vulgaris and G. max, it was identified that GmGPAT1 and GmGPAT17 genes with PvGPAT1 gene, GmGPAT16 gene with PvGPAT3 gene, GmGPAT5, 15, 21 and 25 genes with PvGPAT4 and PvGPAT10 genes, GmGPAT3, 8, 18 and 27 genes with PvGPAT5 gene, GmGPAT19 and GmGPAT28 genes with PvGPAT6 gene, GmGPAT2, 7, 13 and 26 genes with PvGPAT7 gene, GmGPAT20 and GmGPAT23 genes with PvGPAT8 gene, GmGPAT2, 13 and 26 genes with PvGPAT12 gene, and GmGPAT6 and GmGPAT12 genes with PvGPAT13 gene were orthologous genes.

Fig. 8
figure 8

Synteny analysis showing orthologous relationships between PvGPAT genes and AtGPAT and GmGPAT genes

Full size image

In silico gene expression analysis of PvGPAT genes in different bean tissues and under different stress conditions

A heatmap was drawn using 0, 3, 6, 12, and 24 h salt stress RNAseq data (Fig. 9). When this map was constructed, an increase in the expression level of the PvGPAT genes PvGPAT4, 9, and 12 was observed, while a decrease in the expression level of the PvGPAT11 gene was observed. According to the salt stress data, the genes that exhibited high expression levels, regardless of the increasing salt stress level, were PvGPAT1, 3, 7, and 13. The expression of the PvGPAT5 and PvGPAT10 genes increased during the 3rd hour of salt stress and decreased during the following hours. 0 and 14-day drought stress RNAseq data were used. Accordingly, drought stress caused a decrease in the expression of PvGPAT9, 10, and 11 genes and an increase in the expression of PvGPAT2, 7, 8, and 12 genes. Like those under with salt stress, the expression levels of PvGPAT1, 3 and 13 were always high.

Fig. 9
figure 9

Transcriptomic expression levels of GPAT genes in beans under salt (0–3-6–12-24 h) and drought (0–14 days) stress. This heatmap was created using RPKM values of salt and drought RNA-seq data from the NCBI SRA library

Full size image

A heatmap map was drawn with RNA-seq data generated from 8 different tissues of the bean (Fig. 10). The highest expression levels were detected for PvGPAT7 and PvGPAT13 in flower bud tissue. The lowest expression levels were detected for PvGPAT710 and PvGPAT12 in green mature pod tissue and for PvGPAT8 and PvGPAT9 in root tissue. PvGPAT1, 3, 7, and 13 were expressed at high levels in all tissues. The expression levels of PvGPAT genes change in different tissues.

Fig. 10
figure 10

In silico expression levels of GPAT genes in eight different bean tissues. This heatmap was created using FPKM values of tissue expression patterns in Phytozome v13

Full size image

Quantitative real-time (qRT-PCR) analyses

Primers suitable for the PvGPAT2, PvGPAT5, PvGPAT7, PvGPAT9, PvGPAT12, and PvGPAT13 genes were designed, and qRT‒PCR was performed. qRT‒PCR analyses were performed on two bean cultivars (Elkoca-05 and Serra) treated with 150 mM salt and 20% polyethylene glycol (PEG6000) to evaluate the changes in the expression levels of the PvGPAT genes. In addition, in the present study, the effect of melatonin application, which has been shown to have a curative effect on salt and drought stress, on the expression levels of PvGPAT genes was also evaluated. In a previous study, it was reported that the Elkoca-05 cultivar was relatively more resistant than the Serra cultivar was (Aygören et al. 2023). In the present study, two different treatment regimens were used to determine whether the response to stress conditions differed between the cultivars. The expression levels of six PvGPAT genes (Sup. file 1) (PvGPAT2, 5, 7, 9, 12, and 13) were examined in this study. In both cultivars, 0 mM salt + 0 mM MEL (control), 150 mM salt + 0 mM MEL (salt stress), 0 mM salt + 200 mM MEL (melatonin), or 150 mM salt + 200 mM MEL were applied (Fig. 11). As a result of salt stress application in the Elkoca-05 cultivar, a considerable increase in the expression levels of the PvGPAT5, PvGPAT7 and PvGPAT9 genes was observed in comparison to those in the control, while no considerable change was observed in the expression levels of the other genes. The application of melatonin had no effect on the expression levels of PvGPAT genes compared to those in the control group. The application of 150 mM salt + 200 mM MEL resulted in a considerable increase in the expression levels of PvGPAT5, 9, and 12 genes in comparison to those in the control, but there was no considerable change in the expression levels of the other GPAT genes. Salt stress application to the Serra cultivar resulted in a considerable increase in the expression levels of the PvGPAT2, 5, and 7 genes, with no considerable change in terms of the expression levels of the other genes. Melatonin application resulted in a considerable increase in the expression levels of the PvGPAT7 and PvGPAT12 genes in comparison to those in the control group but no considerable change in the expression levels of the other GPAT genes. As a result of 150 mM salt + 200 mM MEL application, a considerable increase in the expression levels of PvGPAT2, 5, 9, and 12 genes was observed in comparison to those in the control, while no considerable change was observed in terms of the expression levels of the other GPAT genes (Fig. 11). According to the qRT‒PCR results, the PvGPAT5 and PvGPAT7 genes responded similarly to the salt stress treatment, and the PvGPAT5, PvGPAT9, and PvGPAT12 genes responded similarly to the salt + melatonin treatment. However, the response of the PvGPAT7 and PvGPAT12 genes to melatonin application differed in the Serra cultivar.

Fig. 11
figure 11

Expression level plots of PvGPAT genes in two cultivars (Elkoca-05 and Serra) under salt stress and melatonin treatment

Full size image

For drought stress, 0% PEG + 0 mM MEL (control), 20% PEG + 0 mM MEL (drought stress), 0% PEG + 200 mM MEL (melatonin), and 20% PEG + 200 mM MEL were applied to both cultivars (Fig. 12). As a result of drought stress in the Elkoca-05 cultivar, no significant difference in the expression levels of PvGPAT genes was found compared to those in the control. A significant increase in the expression levels of the PvGPAT2 and PvGPAT12 genes was observed in the melatonin treatment group compared to the control group, but no considerable change in the expression levels of the other GPAT genes was observed. After the application of 20% PEG + 200 mM MEL, a considerable increase in the expression levels of PvGPAT2 and 5, 7, 9, and 13 genes was observed in comparison to those in the control group, while no considerable change was observed in terms of the expression level of the other GPAT genes (Fig. 12). As a result of drought stress in the Serra cultivar, a considerable increase in the expression levels of the PvGPAT2 and PvGPAT9 genes was observed in comparison to those in the control, while no considerable change was observed in terms of the expression levels of the other genes. When melatonin was applied, there was a considerable increase in the expression levels of the PvGPAT2 and PvGPAT5 genes in comparison to those in the control, but there was no considerable change in the expression levels of the other GPAT genes. As a result of 20% PEG + 200 mM MEL application, a considerable increase in the expression levels of the PvGPAT5 and PvGPAT9 genes was observed in comparison to those in the control, while no considerable change was observed in terms of the expression levels of the other GPAT genes. According to the qRT‒PCR results, the PvGPAT2 gene responded similarly to both melatonin and PvGPAT5; additionally, the PvGPAT9 gene responded similarly to salt + melatonin. However, the response of the PvGPAT2 and PvGPAT9 genes to drought application differed among the Serra cultivar. The responses of the PvGPAT2, PvGPAT7, and PvGPAT13 genes to drought + melatonin application differed among the Elkoca-05 cultivar. According to the qRT‒PCR results, a considerable increase in the expression of the PvGPAT5 and PvGPAT9 genes occurred in response to both salt stress + melatonin and drought stress + melatonin.

Fig. 12
figure 12

Expression level plots of PvGPAT genes in two cultivars (Elkoca-05 and Serra) under drought stress and melatonin treatment

Full size image

Discussion

With the use of comprehensive bioinformatics tools, the identification and characterization of different gene families in plants can be performed. In this study, we analyzed the gene structure, evolutionary relationships, selection pressure, and expression profiles of these genes in response to different abiotic stresses, mainly to determine the roles of the GPAT genes in P. vulgaris. This research will create a detailed knowledge base for future research on GPAT genes. GPAT genes (containing the PlsC domain), which play important roles in various physiological, biochemical, and metabolic pathways, have been characterized in many plants, including rice (26) (Safder et al. 2021) and maize (15) (Xu et al. 2019). In this study, 13 PvGPAT genes containing the PlsC domain were characterized. In addition to the PlsC domain, the PvGPAT genes contain the GPAT_N and HAD domains and the LPLAT superfamily. Research in Brassica napus revealed that the BnGPAT19, 20, 21, and 22 genes include extra hydrolase and acyltransferase domains (Wang et al. 2020). This difference may indicate that GPAT genes have other activities in addition to acyltransferase activity. As in PvGPAT1, an additional GPAT_N domain is found in AtATS1 of Arabidopsis, GmGPAT1 of G. max, and ZmGPAT12 of maize (Waschburger et al. 2018; Zhu et al. 2019). Individuals from the same family's amino acid sequences possess comparable motifs and, as a result, perform similar tasks (Bailey et al. 2006).

The phylogenetic tree constructed with beans and other plants was divided into three groups. Although GPAT genes are found in different locations in cells, they are found mainly in the chloroplast, endoplasmic reticulum, and mitochondria. Genes localized in a similar location exhibited the same branching pattern. Phylogenetic analyses of GPAT proteins in maize have shown that GPAT genes can be divided into three groups: I, II, and III (Zhu et al. 2019). Similarly, Cui et al. (2019) examined the GPAT genes of cotton cultivars phylogenetically and discovered that these GPAT genes were grouped into three groups. Waschburger et al. (2018) examined GPAT genes in various plants and investigated how these genes are related to Arabidopsis GPAT genes. The subcellular components in which these genes were located were also determined. The phylogenetic tree revealed that GPAT proteins with the same cellular localization and similar intron‒exon distributions had similar branching patterns. PvGPAT2 and PvGPAT3 had high affinities for AtGPAT9 and were located in the endoplasmic reticulum, while PvGPAT1 had a strong relationship with AtATS1 and was located in the chloroplast. When the phylogenetic tree was examined, genes located in the chloroplast were grouped into the ATS1-like group. In these investigations, soluble GPAT genes were collected in a group, and all the GPAT genes in this group were found in the chloroplast (Waschburger et al. 2018; Cui et al. 2019; Zhu et al. 2019). The exon‒intron architectures of orthologous and paralogous genes are conserved across plant species, making them valuable tools for estimating the distance between species (Wang et al. 2020).

Duplicate gene pairs were screened to better understand the evolutionary process of the GPAT genes. Duplicate gene pairs may have suffered gain-of-function, loss-of-function, or neofunctionalization throughout their evolution (Lynch and Conery 2000). The nonsynonymous substitution ratio (Ka), synonymous substitution ratio (Ks), and Ka/Ks ratio are widely used to analyze the direction of evolution and the strength of selection in coding sequences (Li et al. 2009). These algorithms use diverse substitution or mutation models based on various assumptions that consider different sequence features (Muse 1996). This enables us to make several predictions about evolutionary distance. Ka/Ks > 1 suggests positive selection-accelerated evolution; Ka/Ks = 1 indicates neutral selection; and Ka/Ks 1 indicates functional limitation with purifying (negative) selection (Ilhan et al. 2018; Kasapoglu et al. 2020; Kızılkaya et al. 2020). Duplicate gene pairs in beans have been subjected to purifying selection throughout the evolutionary process. Research on GPAT genes from four distinct cotton species revealed that the Ka/Ks ratio of certain duplicated GPAT genes was less than one and that the Ka/Ks ratio of other duplicated genes was greater than one. This implies that the GPAT genes were subjected to positive and negative selection pressure throughout the evolutionary process (Cui et al. 2019).

When the bean GPAT proteins were aligned, 4 conserved motifs important for acyl transferase activity were identified (Sup. file 8). The same situation was observed for Arabidopsis thaliana GPATs in a previous study. According to a genome-wide analysis of Arabidopsis thaliana GPATs, four conserved amino acid motifs (Blocks I-IV) are required for acyltransferase activity (Lewin et al. 1999). Motif 3 contains Block I and Block II, while Motif 1 contains Block III and Block IV. Block I (VCNHRTLLDP), Block II (VRLTR), Block III (VCPEGT), and Block IV (IVPVA) are four conserved amino acid sequences that are critical for acyltransferase activity (Maas-Fernández et al. 2010). Numerous amino acid residues, including histidine (H) and aspartic acid (D) residues in Block I, glycine (G) residues in Block III, and proline (P) residues in Block IV, have been found to play a role in catalysis. In Arabidopsis, the conserved regions include glutamate and serine in Block III and phenylalanine and arginine in Block II, which are important for binding to the glycerol-3-phosphate substrate (Gonzales-Baro et al. 2007). These findings showed that the sequences of the four blocks (I-II-III-IV) of the GPAT genes in all the organisms were conserved throughout the evolutionary process. When the intron‒exon models of the GPAT genes were examined, it was determined that they exhibited a similar organization to that of other species. For example, Waschburger et al. (2018) found that GPAT genes are highly similar to AtATS1 and AtGPAT9 according to intron‒exon analysis and that they contain 12 exons and 11 introns, as we detected in beans. Accordingly, the intron‒exon organization of GPAT genes is conserved among different species.

The presence of cis-elements associated with abiotic stresses such as drought and salinity in the promoter region of bean GPAT genes indicates that GPAT genes also play a role in stress metabolism. The distribution of different cis-acting elements in promoter regions is one reason for differences in the regulation and function of genes (Hernandez-Garcia and Finer, 2014). Promoter regions allow specialized cis regulators to regulate gene expression by inhibiting or increasing the transcription of some genes in response to abiotic and biotic stress. Numerous cis-acting elements, including drought-sensitive elements (DREs), MYB elements linked to drought and salinity, and abscisic acid-sensitive elements, which are sensitive to high salt, dehydration, and cold temperatures (ABREs), have been found (Yamaguchi-Shinozaki and Shinozaki 2005). As a result of the promoter region analysis performed by Xu et al. (2019) in maize GPAT genes, many cis-elements, such as LTRs, TC-rich regions, and MBSs, were found to be associated with stress and defense. Cui et al. (2019) identified many stress-sensitive cis-acting elements via promoter region analysis of cotton GPAT genes. Considering the findings of this study and those of previous research, it can be concluded that GPAT genes play a role not only in plant growth but also in stress adaptation.

While the amino acid sequence determines protein structure, the three-dimensional structure of a protein determines its biochemical functionality. Proteins with similar amino acid sequences generally have similar biochemical functions (Alberts et al. 2002). Although GPAT proteins are structurally very similar to each other, they also have different interaction domains. In addition to having similar biochemical activities, it also has additional activities. Protein–protein interactions (PPIs) govern various biological processes, including intercellular interactions and metabolic and developmental control (Braun, 2012). Noncovalent contacts between some residue side chains underlie protein folding, protein aggregation, and protein–protein interactions (PPIs) (Ofran and Rost 2003). PvGPAT proteins mostly interact with proteins involved in glycerol degradation, diacylglycerol/triacylglycerol biosynthesis, and phosphatidylcholine acyl regulation pathways. Glycerol kinase is the first enzyme implicated directly in glycerol breakdown; it phosphorylates glycerol to glycerol-3-phosphate, which can then enter the triglyceride synthesis pathway (Ditlecadet et al. 2011; Chen et al. 2023). The first step in the de novo production of triacylglycerol is catalyzed by GPAT (Cao et al. 2006). These contacts result in various interactions and relationships between proteins. For example, in Arabidopsis, the AtGPAT9 protein interacts homomeratically and heterometrically with the AtLPAAT2 and AtLPCAT2 proteins, which are involved in glycerolipid metabolism and participate in glycerolipid biosynthesis together (Shockey et al. 2016). Furthermore, plants that are stressed change the makeup of their membranes, and lipids such as phosphatidic acid have been linked to stress signaling (Lavell and Benning 2019). With the data obtained in this study, information about the most effective genes involved in glycerolipid biosynthesis is presented for future functional research.

Most of the PvGPAT genes were more highly expressed in the flower tissue of the bean. Like the AtGPAT genes, the PvGPAT genes exhibited variable expression patterns in different tissues. Previous research has shown that AtGPAT1 is required for pollen fertility (Zheng et al. 2003), while AtGPAT4, AtGPAT5, and AtGPAT8 are required for cutin or suberin synthesis (Beisson et al. 2012). Furthermore, AtGPAT9 is involved in the manufacturing of polar lipids and TAG in growing leaves, as well as the creation of lipid droplets in developing Arabidopsis pollen grains (Singer et al. 2016). The expression of AtGPAT6 was greater in sepals, petals, and stamens than in other tissues, indicating that AtGPAT6 may play a significant role in Arabidopsis flower formation (Cui et al. 2019). GPAT genes are associated with plant flower development and stress responses (Shen et al. 2023). These findings show that GPAT genes are associated with plant flower development (Shen et al. 2023).

The PvGPAT gene exhibited different expression profiles under salt and drought stresses. The PvGPAT gene was generally upregulated in response to salt stress, creating a stress response. The basis of an effective salt stress response is reprogramming the expression of many genes. The expression levels of some genes increase, while the expression levels of others decrease (Mareri et al. 2022). Under drought stress, most of the PvGPAT genes were downregulated. In response to drought stress, the expression levels of many genes and gene families, as well as those of various hormones, are up- or downregulated (Mareri et al. 2022). In silico expression analysis of GPAT genes under salt, cold, and UV stress in maize (Xu et al. 2019) and under cold and salt stress in cotton (Cui et al. 2019) revealed different expression patterns. In these investigations, GPAT genes were generally upregulated in response to stress. In response to environmental stresses, genes involved in lipid biosynthesis were unregulated, thus increasing lipid accumulation in cell membranes. Thus, plants develop resistance to various stress factors (Nam et al. 2022). When melatonin and stress were combined, an increase in the expression level of most PvGPAT genes was observed, suggesting that melatonin has a healing effect on stress. Melatonin increases the expression of genes suppressed by salt stress and alleviates the suppressive effects of stress on gene expression (Wei et al. 2015). A study has proven that melatonin helps plants heal against drought and salt stress (Wei et al. 2015).

In the present study, as a result of the various treatments applied, the expression levels of several genes in the two bean cultivars exhibited the same changes, while they exhibited different changes in the expression of other genes. Taken together, these findings show that drought, salt, and melatonin application create different responses at the gene expression level among different individuals of the same species. Genetic diversity in plants enables adaptation to different environmental conditions so that some cultivars can survive. However, cultivars that lack genetic diversity may become sensitive to different stresses. Genetic diversity refers to the differences between individuals of the same species, such as quality traits and tolerance to various stresses. One of the reasons for this difference stems from the differences in the expression levels of several genes (Salgotra and Chauhan 2023). The GPAT gene family is involved in both plant development and stress metabolism. This study emphasizes the importance of GPAT genes in plants and will guide future research on GPAT genes.

Conclusion

Characterization of the GPAT genes in the bean genome was carried out using bioinformatics tools in our study. Thirteen GPAT genes were discovered in the bean genome, and these genes were located across the six chromosomes of the bean. The response of the PvGPAT gene to salt, drought, and melatonin application in two bean cultivars was examined at the expression level via qRT‒PCR. According to the qRT‒PCR results, stress and melatonin application caused a significant increase in the expression level of the PvGPAT5 and PvGPAT9 genes in both cultivars. The expression levels of the other genes differed between the two cultivars. Although the cultivars are different individuals of the same species, they have some genetic differences, which reveals the differences in their responses to environmental stimuli. The results obtained from this study showed that GPAT genes are also involved in stress responses in addition to glycerolipid biosynthesis. Additionally, this study is a preliminary study for future functional and genomic research on GPAT genes.