Background

Transcription factors (TFs) play essential regulatory roles in many crucial biological processes in plants, which are classified into different families according to their conserved DNA-binding domains [1,2,3]. The programmed and regulated interactions between TFs and downstream genes influence metabolic processes and phenotypic traits [2, 4,5,6]. Currently, about 100 TF families have been identified in Arabidopsis, including MYB, the basic leucine zipper (bZIP), the basic/helix-loop-helix (bHLH), and WRKY [4, 7, 8]. Among them, bZIP transcription factor family is one of the largest and most diverse families in plant [8]. In addition, the bZIP TFs were named after a highly conserved bZIP domain, which contained two structural characteristics on a contiguous helix [8, 9]. Firstly, a basic region of roughly 18 amino acid residues was followed by an invariant N-X7-R/K-X9 motif for nuclear localization and sequence-specific DNA binding. Secondly, the leucine zipper domain was composed of multiple amphiphilic-helices, with one leucine or other hydrophobic amino acid appearing per six amino acids. These amino acids, such as valine, methionine, L-Isoleucine, and L-Phenylalanine, frequently interact via hydrophobic surfaces to create a-helical homodimers or heterodimers [10]. To bind DNA, two subunits engage via the hydrophobic sides of their helices, forming a superimposed coiled-coil shape. In plant, bZIPs preferentially bind to palindromic or pseudo-palindromic ACGT core Cis-regulatory Elements such as G-box (CACGTG), C-box (GACGTC), A-box (TACGTA), and ABRE (CCACGTGG) [11].

Nowadays, bZIP transcription factor family has been identified in model plants such as Arabidopsis, rice, and maize, and many horticulture plants [9, 12,13,14,15,16,17,18]. In Arabidopsis, there are 78 bZIP members, which can be divided into 13 groups based on their functional structure (designated A-M) [8, 9]. bZIPs have been linked to responses to a number of abiotic/biotic stimuli, including high salinity, drought, heat, cold, pathogen infection, as well as hormone signaling, including abscisic acid (ABA), ethylene, and light signaling [8, 13]. Specifically, subgroup B of the bZIP TFs, namely bZIP17 and bZIP28, exhibit a pivotal function in abiotic stress responses and exhibit a close association with the ABA pathway [19, 20]. Overexpressing bZIP17 or bZIP28 in Arabidopsis leads to enhanced tolerance to diverse environmental stresses, whereas their knock-out plants exhibit highly sensitive phenotypes [21,22,23,24]. Besides, AtbZIP17 could collaborate with AtbZIP28 to regulate development-related genes, particularly those associated with stress maintenance and root elongation [23, 25, 26]. The survival rate and germination rate of TabZIP28-overexpressed Arabidopsis were higher than those of wild type after heat stress [27]. In rice, overexpression of OsbZIP60 (a homologous gene of AtbZIP17) could enhance heat and drought resistance [28]. A recent study shown that tobacco plants' tolerance to salt, osmotic, drought, and heat stress was increased by overexpressing the EcbZIP17 gene from finger millet [29]. Whilst subgroup B of bZIP assumes pivotal roles in the intricate tapestry of plant development and stress responses, the molecular function of it under heat stress in wax gourd is unknown.

Wax gourd (Benincasa hispida (Thunb.) Cogn, 2n = 2x = 24) is an important vegetable crop of Cucurbitaceae family, which has been grown as an annual herb in China for about 1,500 years and has an annual planting area of more than 200,000 hm2 [30, 31]. The fruits of wax gourd contain abundant nutrients and metabolites, and can be used as both food and medicine [32, 33]. As a fruit vegetable, wax gourd is sensitive to environmental influences, including sodium chloride, heavy metals, low and high temperatures, and drought stress. Undoubtedly, heat stress is one of the major constraints to crop production and food security worldwide, with a yield loss of 6–7% expected for every 1 °C increase in seasonal mean weather coupled with intense heat stress [34]. High temperature resulted in significant quality and production losses during wax gourd growth, particularly in open field farming [35]. Hence, the heat resistance plays a pivotal role in the breeding of wax gourd. a draft genome sequence of wax gourd was published, serving as a fundamental resource to expedite genetic advancements in this crop [36]. However, up to now, no comprehensive analysis of the bZIP gene family in wax gourd has been reported. Consequently, it holds immense importance to systematically investigate the bZIP family in wax gourd under heat stress conditions.

In the present study, we identified the bZIP gene family based on the wax gourd genome data, and analyzed the phylogeny to determine the relationships between these genes. The gene structure, promoter sequences, and physicochemical properties of the proteins were also analyzed. Subsequently, the subgroup B gene, BhbZIP58, was selected for further expression and functional study. Our results provide systematic information on the bZIP gene family in wax gourd, which will facilitate the study of function and regulatory mechanisms of bZIP family members.

Results

Identification of bZIP TFs in wax gourd

A total of 61 candidate BhbZIP genes were identified from wax gourd genome, the nomenclature and related basic information of which were listed in Table S1. Since the bZIP domain of Bhi09M001946 was severely deleted, so it was not classified in bZIP family. To ensure the accuracy of the results, we also corrected and confirmed each bZIP using the IGV-sRNA tool. The core conserved domain of BhbZIP proteins was discovered to be 50 amino acids long on average, with the bZIP domain comprised of a basic region and a leucine zipper (Fig. 1). According to the position of genes on the chromosome, the bZIP genes of wax gourd were designated from BhbZIP1 to BhbZIP59. Besides, two genes, BhiUN613M6 and BhiUN916M5, located on loose scaffolds were named as BhbZIP60 and BhbZIP61, respectively. Chromosome mapping showed that 59 members of bZIP family were unevenly distributed on 12 chromosomes (Fig. 1). The chromosome with the most BhbZIP genes (12 genes) was chromosome 4, while chromosomes 2 and 12 had only one member. According to a report, two or more genes from the same family that were present in a 200 kb region of a chromosome are considered to be a gene cluster [37]. Three small gene clusters were found based on the BhbZIP location information: BhbZIP2-BhbZIP3, BhbZIP28-BhbZIP29, and BhbZIP47-BhbZIP48. Among the members, BhbZIP13 had the minimum number of amino acids (136 aa), and BhbZIP58 had the maximum number of amino acids (763 aa). As shown in Table S1, molecular weight of BhbZIP proteins ranged from 15.71 to 82.68 KDa and the isoelectric point (PI) ranged from 4.45 to 10.56. The average hydrophilicity (GRAVY) of all members of bZIP family was negative, indicating that they were hydrophilic proteins, which was consistent with the potential functions of TFs. Secondary structure prediction results revealed that most BhbZIP proteins were primarily consisted of α-helix and random coil, with the fraction of extended structure and β-turn being the least (Table S2). It was consistent with the ability of the bZIP domain to construct a continuous a-helical structure [9]. The results of subcellular localization prediction showed that all transcription factors located in the nucleus (Table S2).

Fig. 1
figure 1

Chromosome distribution of the wax gourd bZIP genes. Chr1-12 represent chromosome numbers 1–12 and blue letters represent bZIP genes. The gene names in red represented gene clusters

Full size image

Phylogenetic and conserved domain analysis of BhbZIP genes

To categorize and investigate the evolutionary relationship of the BhbZIP genes, a phylogenetic tree was created using the bZIP proteins of wax gourd and Arabidopsis. Phylogenetic analysis showed that the BhbZIP genes were divided into 12 subgroups (A, B, C, D, E, F, G, H, I, J, K, S) based on the relationship with AtbZIPs (Fig. 2). We found that the short sequences of wax gourd and Arabidopsis were concentrated in subgroup S and the long sequences were concentrated in subgroup B. The number of wax gourd and Arabidopsis bZIPs in each subgroup was similar approximately. Among the subgroup, the subgroup M was deleted and only one bZIP gene was identified in subgroup B, J, H, and K, while the largest group was the subgroup S with 16 BhbZIP genes. Furthermore, we noticed that several subgroups of wax gourd contained the unique domain, with group G, D, and C having MFMR and MFMR_assoc domains, DOG1 domain, and bZIP_C domain, respectively.

Fig. 2
figure 2

Phylogenetic tree of wax gourd and Arabidopsis bZIP family. Red circles represent wax gourd, blue squares represent Arabidopsis. The dendrogram was drawn by IQ-TREE with the Maximum Likelihood method. Different groups were marked with different colors. The various sizes of circles on the branch represent bootstrap values. The conserved protein domains were represented by the outermost color blocks

Full size image

Exon/intron structures and conserved motifs of BhbZIP genes

To understand the conserved structure and evolution of wax gourd bZIP family, we discovered 20 conserved motifs using MEME analysis combined with a phylogenetic tree (Fig. S2). Motif1 was a basic region of the bZIP structural domain, while the leucine zip was not very conserved and was distributed in motif4, motif6 and motif7. The number of leucine repeats in the ZIP domain varied considerably, from 3 times (Group D) to 8 times (Group C and group S). Motif2 was the DOG1 domain in the Pfam databases. As shown in Fig. 3B, all of the bZIP members had motif1. In contrast, motif2, motif5, and motif3 were present only in Group D. Similarly, motif10 and motif11 were only found in Group A. Many motifs existed in a specific subgroup, which might be related to specific biological functions.

Fig. 3
figure 3

Gene structures and protein motifs of bZIP gene family in wax gourd. A Phylogenetic tree of wax gourd bZIP gene family. Different colors represented different subgroups, and the majuscule was the name of the subgroup. B Protein motifs in bZIP members. The colorful boxes delineated different motifs, and the black box (motif1) represented bZIP domain. C Gene structures of wax gourd bZIP gene family. Exons, 5’ UTR, and 3’ UTR were respectively displayed using red, dark blue, and light blue rectangles. Black lines denoted introns. The clustering was performed according to the results of phylogenetic analysis

Full size image

To ensure the accuracy of the bZIP gene structure, we corrected and complemented them using transcripts. Most BhbZIP genes in the same subgroups had a similar exon–intron structure (Fig. 3C). For instance, the genes of subgroup S contained only one or two exons, while those of subgroups D and G had a high number of exons. Besides, all the members of Group G had two five-prime untranslated regions (UTR).

Intra- and inter-species collinearity analysis of BhbZIP genes

The members of bZIP family in wax gourd were generated through four replication events, with no tandem duplication existed (Table S4). Based on the analysis of gene replication events, the number of BhbZIP genes generated by Dispersed, WGD, Transposed, and Proximal was 30 (51%), 14 (24%), 11 (18%), and 4 (7%), respectively. On the chromosomes of wax gourd, a total of 10 collinear gene pairs of bZIP genes were found (Fig. 4), and listed in Table S5. Nine of them were from subgroup S and only one was from subgroup F, and all genes from collinearity were generated by WGD replication events. Especially, there were two collinear pairs consisting of three genes that had collinearity relationships with each other. Two small clusters of bZIP genes, BhbZIP2-BhbZIP3 and BhbZIP28-BhbZIP29, formed two groups of proximal replicates with sequence similarity bigger than 70%, and all belonged to Group I. Furthermore, we calculated the Ka/Ks ratios of 12 sets of BhbZIP genes to explore the evolutionary selection pressure. The result showed that all the Ka/Ks values of collinear gene pairs of BhbZIP were lower than 1, which indicated that these genes were subjected to purified selection (Table S5).

Fig. 4
figure 4

Collinearity analysis of bZIP family in wax gourd. Chromosomes 1–12 were represented by grey rectangles. The different color lines and heatmaps along the rectangles respectively represented GC content and gene density on the chromosomes. The different colour lines represented collinear gene pairs of BhbZIPs between the wax gourd chromosomes

Full size image

To explore the collinearity relationship between wax gourd and other species, three dicotyledons (Arabidopsis, grape, and cucumber) and one monocotyledon (rice) were chosen (Fig. S3). A total of 49 collinear pairs were found between wax gourd and Arabidopsis. Among them, some BhbZIP genes had collinearity relationships with two to three genes of Arabidopsis, suggesting that these genes were crucial in the evolution of bZIP family in wax gourd. The number of collinearity gene pairs of wax gourd vs cucumber, and wax gourd vs grape, were 35 and 60, respectively (Table S6). However, only 11 collinearity pairs were found between rice and wax gourd. According to Venn diagram (Fig. S3), four of the BhbZIP pairs were collinear with the four species. Additionally, a total of 11 collinear gene pairs were identified between wax gourd and cucumber/grape/Arabidopsis species, while no found in the rice genome.

Cis-regulatory elements analysis of BhbZIP promoters

To better understand how bZIPs participated in the regulation of biotic and abiotic stresses, the 2000 bp upstream promoter sequences of the BhbZIP genes were used to search the Cis-regulatory Elements using PlantCARE online software. Cis-regulatory Elements were classified into 5 categories according to their functions (Fig. 5). The BhbZIP promoter region contained lots of hormone response elements, including abscisic acid (AAGAA-motif, ABRE, ABRE4, F-box), methyl jasmonate (CGTCA-motif), gibberellin (GARE-motif, P-box, TATC-box), salicylic acid (TCA, TCA-element, as-1), ethylene (ERE), jasmonic acid (CTAG-motif), auxin (AuxRE, AuxRR-core, TGA-element), which was consistent with the function of bZIP transcription factor in response to abiotic stress. They also contained many stress response elements such as heat stress related elements (STRE, CCAAT-box, AT-rich), cold stress response elements (LTR), anaerobic response elements (ARE), damage and defence response elements (WRE3, WUN-motif or W box). In this study, we also found a large number of functional elements related to light response, such as Box 4, G-box, GT1-motif, I-box, and so on, which was consistent with previous reports of bZIP regulation of photosynthesis [9]. A few BhbZIP genes also contained regulatory elements related to growth and development, such as endosperm expression-related elements (GCN4), meristems expression-related elements (CAT-box), seed expression-related elements (RY-element), differentiation-related elements (HD-Zip1), zein metabolism regulatory related elements (O2-site), circadian regulatory elements (circadian). Interestingly, almost all BhbZIP genes contained cis-elements of MYB and MYC. Conducting a thorough analysis of the promoter region of the BhbZIP gene family contributes significantly to the profound exploration of their intricate biological functionalities.

Fig. 5
figure 5

Analysis of Cis-regulatory Elements in BhbZIP promoters. The different colors and numbers of grids indicated the numbers of different Cis-regulatory Elements. The numbers at the bottom of the graph represented the total number of Cis-regulatory Elements in BhbZIP promoters. The histograms with different colors on the right represented the sum of the Cis-regulatory Elements in each category

Full size image

protein–protein interaction network of BhbZIP proteins

Based on studies of AtbZIP and gene homology, we could speculate about the potential functions of most BhbZIP proteins. An interaction network was constructed by using the STRING database based on the search for protein families (Fig. 6A) and multiple sequences (Fig. 6B). The result showed that 12 members were involved in the ABA activation signal pathway (NOG243340), 7 of which belonged to subgroup A, and 5 belonged to subgroup S (Table S3). In the growth and development processes, BhbZIP55 (subgroup A), BhbZIP40 (subgroup C), and BhbZIP21 (subgroup D) involved in seed germination (NOG257560), seed maturation (NOG10040), and flower development regulation (NOG259341), respectively.

Fig. 6
figure 6

Interaction network of BhbZIP proteins. A Search for the orthologs in Arabidopsis matching BhbZIPs. B The BhbZIP protein–protein interaction network. This network was predicted by the online software STRING. The BhbZIP proteins were shown in the red font below the Arabidopsis orthologs. Purple lines indicated experimentally validated protein interactions; green lines indicated predicted protein interactions based on gene neighborhood; blue lines indicated predicted protein interactions based on gene co-occurrence; and black lines indicated protein coexpression

Full size image

Some bZIP genes were associated with other members. BhbZIP1 was linked with several genes including BhbZIP14, BhbZIP26, BhbZIP55, BhbZIP6, BhbZIP39, BhbZIP53, and BhbZIP7 (Fig. 8B). BhbZIP40 was in the core region and interacts with several genes, such as BhbZIP13, BhbZIP17, BhbZIP52, BhbZIP42, and BhbZIP9. There was an interaction between AtbZIP17, a heat stress-related transcription factor, and AtbZIP60, and their homologous protein in wax gourd, BhbZIP58, was linked to BhbZIP60 by a purple line, suggesting that they might interact.

Expression patterns of BhbZIP58 in different tissues and abiotic stress

We randomly selected 12 BhbZIPs from the A-S subgroups and conducted expression analysis under heat stress. Remarkably, our findings revealed that BhbZIP58 exhibited significant induction to heat stress rapidly (Fig. S4). In Arabidopsis, the pivotal role of subgroup B within the bZIP gene family in abiotic stress has been well-documented [20, 21, 25, 26]. However, there is no relevant literature or reports regarding the involvement of subgroup B of the bZIP gene family in abiotic stress in wax gourd. So, we chose the subgroup B member of wax gourd (BhbZIP58) to study its expression patterns in various tissues obtained from plants at the reproductive stage under normal growing circumstances by the real-time quantitative PCR (qRT-PCR). As shown in Fig. 7A, BhbZIP58 was expressed in all organs, with the highest expression in root and the lowest expression in fruit.

Fig. 7
figure 7

Expression level of BhbZIP58 in different tissues (A), under heat stress (B), cold stress (C), salt stress (D), ABA treatment (E), and drought stress (F). Data were means ± SEM of three biological replicates, and the transcription level of BhbZIP58 at 0-24 h and different tissues were statistically evaluated by Duncan's test. Different letters meaned significant differences at 0.05 levels

Full size image

To explore the expression of BhbZIP58 in different stresses, firstly, we wondered if BhbZIP58 was influenced by the circadian clock. We sampled seedlings under normal growth conditions and performed qPCR assay on BhbZIP58 with the circadian-regulated gene LHY (Bhi12G002045) as control. Results showed that there was no significant change in the expression of BhbZIP58 among the 24 h (Fig. S5), indicating thar BhbZIP58’s expression with insusceptible by circadian rhythms. Next, we performed five treatments including high temperature, low temperature, salt stress, drought, and ABA to detect BhbZIP58 expression. When exposed to high temperature stress, BhbZIP58 was up-regulated and had two expression peaks, with the highest peak at 1 h and the second peak at 24 h (Fig. 7B). In the process of cold stress, salt stress and drought stress, the expression of BhbZIP58 had a similar trend, which decreased at first and increased at 3 h or 6 h. For ABA treatment, the expression was almost unchanged at the beginning, but rose sharply at 12 h and reached to the peak at 24 h (Fig. 7E).

Subcellular localization and transcriptional activity analysis of BhbZIP58

First of all, we used the primers ranging from the start codon to the stop codon to clone BhbZIP58 and found that it could not be cloned. Subsequently, we compared its protein sequence with the homologues of other cucurbits and found that the gene had a redundant sequence, possibly due to an error in the genome annotation file (Fig. S6). Consequently, we designed two primers to clone BhbZIP58 based on the corrected sequence (Fig. S7).

The fluorescent protein (GFP) tag was fused to the C terminus of BhbZIP58 under the control of the super promoter to determine the subcellular localization. Agrobacterium tumefaciens-mediated transient expression was used to transfer the 35 s:BhbZIP58-GFP recombinant construct and the control into tobacco leaves. The result showed that BhbZIP58-GFP fusion protein was found only in the nucleus (Fig. 8A). To investigate whether BhbZIP58 acted as a transcriptional activator, transcriptional activation assays were performed. As shown in Fig. 8B, BhbZIP58 protein exhibited self-activation, as the yeast AH109 cells transformed with the pGBKT7-BhbZIP58 and pGBT9 (PC) grew well on the SD/-Trp/-His medium and turned blue on the selective medium containing X-α-Gal.

Fig. 8
figure 8

Subcellular localization and transcriptional activation assay of BhbZIP58 protein. A Subcellular localization of the 35S::BhbZIP58-GFP fusion protein in tobacco leaves. Free GFP served as a control. A DAPI staining assay was conducted to confirm the nuclear localization. White bars = 25 μm. B Growth of yeast cells (strain AH109) transformed with pGBT9 vector (a positive control), pGBKT7 vector (a negative control), or pGBKT7-BhbZIP58 vector on SD/-Trp or SD/-Trp/-His with or without X-α-gal. 10 × , 100 × , and 1000 × represent yeasts that were diluted with the original concentration for 10-, 100-, and 1000- fold, respectively

Full size image

Functional analysis of the BhbZIP58 gene under heat stress

To better understand the biological role of BhbZIP58 under heat stress, we transferred it into Arabidopsis for heterologous overexpression. We generated a 35S:BhbZIP58 construct and transformed it into Arabidopsis. The expression of BhbZIP58 in the transgenic lines was significantly higher than that in the wild type (Fig. S8). All transgenic lines displayed similar phenotypes under heat stress. The overexpression strain and the wild type grew identically under normal conditions (Fig. 9A). After heat stress treatment, most wild-type plants were visibly stressed, with thin plants, yellowing leaves and even death, but overexpressed plants remained healthy and only 10% died. (Fig. 9B, C). These results indicated that BhbZIP58 could improve the heat tolerance.

Fig. 9
figure 9

Functional analysis of BhbZIP58 under heat stress. A-B Overexpression of the BhbZIP58 gene in Arabidopsis under different heat stress. C Survival statistics of overexpression lines and WT under heat stress. Asterisks indicate significant differences between the transgenic lines and WT according to Student’s t-test (n = 3, **P < 0.01)

Full size image

Discussion

The bZIP genes play crucial roles in regulating plant growth, development, and response to biotic and abiotic stresses including high temperature, drought, salt, cold, and ABA treatment [2, 8, 9]. So far, bZIP gene family has been reported in several plant species, such as Arabidopsis, rice, maize, tomato, and cucumber [8, 12,13,14, 17]. However, no study on the bZIP gene family of wax gourd has been reported. In this study, we systematically studied the bZIP gene family in wax gourd. A total of 61 bZIP genes were identified from wax gourd genome, which distributed unevenly on the 12 chromosomes (Fig. 1). The number of bZIPs in wax gourd was similar to that of cucumber, watermelon, pomegranate and poplar, which might be caused by ancient polyploid events [14, 16, 38, 39]. For the accuracy of the follow-up results, we corrected and confirmed each BhbZIP gene using the transcriptome data of our research group. The basic biochemical properties (MW, PI, GRAVY, secondary structure, subcellular localization) of the BhbZIP protein sequences were predicted (Tables S1, and S2). Then we analyzed the BhbZIP proteins by multiple sequence alignment and found that the BhbZIP proteins had two basic domains: a highly conserved DNA-binding basic region and a diverse leucine zipper, which was consistent with that of other plants, such as Arabidopsis, rice, maize [12, 14, 15, 40].

The classification and functional research of bZIP family in Arabidopsis had been extensively studied, which were divided int 13 subgroups based on their function and structure [9]. To further understand the evolutionary relationships between Arabidopsis, a phylogenetic tree of BhbZIP and AtbZIP genes was constructed. Then the 61 BhbZIP genes were divided into 12 subgroups according to the classification criteria and genetic relationship of Arabidopsis, which was different from cucumber and watermelon [14, 38] (Fig. 2). When compared with Arabidopsis, the subgroup M of wax gourd was deleted and some subgroups had relatively smaller numbers of members, which implied that BhbZIP might lose during the evolution process of wax gourd. The short sequences of AtbZIP and BhbZIP were concentrated in subgroup S and the long sequences were concentrated in subgroup B. These findings showed that the conserved regions of these genes had a similar evolutionary relationship between wax gourd and Arabidopsis. Subsequently, we predicted the motifs of BhbZIP protein sequences on the MEME website and obtained the gene structure of BhbZIPs from wax gourd reference genome. The result showed that the same subgroup had similar motif distribution patterns and gene structures, which was similar to that of other plants (Fig. 3) [14, 16, 38]. The foundational bZIP core domain, known as motif1, exhibited its presence across all BhbZIP proteins, while certain distinctive motifs were exclusive to specific subgroups. This intriguing observation suggested the BhbZIP might assume an indispensable role, considering the diverse functions attributed to bZIP.

Previous studies showed that during the evolution process, gene families typically underwent tandem duplication or large-scale segmental duplication events [41]. Since the bZIP gene family was relatively large, we analyzed the tandem duplication events and segmental duplication events of BhbZIPs. Interestingly, we found no tandem duplication events existed in bZIP genes of wax gourd. Meanwhile, the results of DupGen_finder implied that whole-genome duplication (WGD) and dispersed duplication (DSD) were the main driving forces of bZIP gene family expansion and evolution in wax gourd. Nine of the collinear gene pairs were discovered from the subgroup S and only one from the subgroup F, implying that the reason for the large number of members in the subgroup S might be due to chromosomal segmental duplication. Moreover, the values of non-synonymous (Ka) and synonymous (Ks) and the ratio of non-synonymous to synonymous substitutions can be used to evaluate whether a protein-coding gene was under selective pressure [42]. The calculated Ka/Ks ratio of all BhbZIP gene pairs was less than 1, which indicated that these genes might have experienced strong purifying selection pressure with limited function divergence during evolution. Canon's study showed that collinearity analysis implied that sequences from two species from the same ancestral sequence had similar functions [43]. To investigate the evolution of bZIP family of wax gourd, we examined the collinearity analysis of bZIP genes from other species, such as cucumber, Arabidopsis, grape, and rice (Fig. S3). Our results showed that more BhbZIP collinear gene pairs existed in Arabidopsis and grape than cucumber, which might be due to the loss of bZIP genes during the evolution of Cucurbitaceae. Some BhbZIP genes exhibited multiple collinearity pairs with other species, suggesting their potential significance in evolution, potentially as ancestral genes. While only a few collinear gene pairs were found in wax gourd vs rice, probably due to the dramatic divergence of bZIP in dicotyledonous and monocotyledonous plants, which was similar to the result of potato [18]. According to Venn diagram (Fig. S3), four of the BhbZIP gene pairs were collinear with the four species, suggesting that these genes might have existed before these species diverged and have remained collinear ever since. Of note, a total of 11 collinear gene pairs were identified between wax gourd and cucumber/grape/Arabidopsis species, while no found in the rice genome, this suggested that these gene pairs might have emerged after dicotyledones and monocots differentiation.

Plant bZIP proteins exhibit a relaxed DNA-binding specificity for DNA sequence motifs containing an ACGT core [11]. We predicted the Cis-regulatory Elements of the BhbZIPs promoter and identified many ACGT-containing promoter elements, such as G-box and ABRE. Some cis-elements were associated with hormone regulation, including auxin, gibberellin, abscisic acid, and methyl jasmonate responsive elements, and some were induced by environmental stress, including ARE, STRE, LTR, and WUN-motif, which was consistent with previous reports [9, 42]. MYB and MYC elements that participate in broad range of stresses like drought, salinity and low-temperature existed in almost all the BhbZIPs [44]. However, only a small number of regulatory elements related to growth and development existed in BhbZIPs. A protein interaction network indicated the potential interactions and co-expression of different BhbZIPs, which corresponded to the formation of dimers between bZIP proteins [9]. Interestingly, the heat shock elements (HSE), which often existed in the promoters of numerous genes induced by heat stress, were didn’t found in BhbZIPs promoters. There are two members, bZIP17 and bZIP28, in the subgroup B of the Arabidopsis bZIP family, while the corresponding subgroup B of wax gourd was only BhbZIP58. Phylogenetic analysis showed that BhbZIP58 had high homology with AtbZIP17. Therefore, we hypothesized that the BhbZIP58 gene in wax gourd also has similar function to the Arabidopsis. In this study, we analyzed the response of BhbZIP58 to 5 kinds of stresses (Fig. 7B-F). The results showed that BhbZIP58 responded to all the 5 stresses, especially to heat stress. In addition, bZIP17/28 in Arabidopsis were involved in the ABA pathway [25]. Interestingly, BhbZIP58 expression was up-regulated after ABA treatment, indicating that BhbZIP58 was likewise engaged in the ABA signaling pathway. In addition, BhbZIP58 had transcriptional activation activity and was localized in the nucleus, which suggest that BhbZIP58 acts as a transcriptional activator to regulate gene expression in response to stress (Fig. 8A,B). The function of BhbZIP58 was preliminarily determined by overexpression in Arabidopsis under heat stress, indicating it could improve the ability of wax gourd to resist heat stress (Fig. 9). This result was similar to that of AtbZIP17, EcbZIP17 and OsbZIP60 in improving the heat tolerance of plants [21, 28, 29]. However, the regulatory mechanism of BhbZIP58 in response to heat stress requires further experimental verification.

Conclusions

In the present study, 61 BhbZIP genes were identified and divided into 12 subgroups by the bioinformatics method. We corrected and confirmed the CDS and gene structure of each BhbZIP gene using transcripts to ensure the accuracy of the results. It was worth noting that whole-genome duplication (WGD) and dispersed duplication (DSD) were the main driving forces of bZIP gene family expansion in wax gourd. BhbZIP gene might have undergone intense purifying selection pressure during the evolutionary process. In addition, BhbZIP58 could rapidly respond to high temperature and its expression level was the highest in the root. BhbZIP58 located in the nucleus and had transcriptional activation activity, and that its overexpression improves heat stress tolerance in Arabidopsis. This study established the foundation for further research into the regulatory mechanism of BhbZIP58 under heat stress in wax gourd, and promoting the molecular breeding of wax gourds.

Materials and methods

Identification of BhbZIP genes

We download the genome annotation and the genome sequences file of wax gourd from the Cucurbit Genomics Database (CuGenDB) (http://cucurbitgenomics.org/v2/) [45]. The Hidden Markov Model (HMM) configuration files of PF00170, PF07716, and PF03131 were downloaded from Pfam (https://pfam.xfam.org/) [46]. The raw bZIP proteins were found in the wax gourd genome using the hmmsearch function of the HMMER software (version 3.3), and multiple sequence alignment was carried out using clustalw (version 2.1) [47, 48]. Next, the hmmbuild function was used to build a wax gourd-specific bZIP gene family HMM model, and the E-value threshold was set to e−5. Using the updated HMM file, the members of bZIP gene family were once more identified from the wax gourd genome [49]. Candidate proteins that lacked the bZIP conserved domain or were severely deleted were excluded from further analysis using InterPro (version 91.0) and the National Center for Biotechnology Information (NCBI) Batch CD-Search Tools (https://www.NCBI.nlm.nih.gov/structure/bwrpsb/bwrpsb.cgi). To ensure gene accuracy, we also used the IGV-sRNA software (https://gitee.com/CJchen/IGV-sRNA) to correct and confirm the genes. Finally, we used ClustalX (version 2.1) and WebLogo (http://weblogo.threeplusone.com/create.cgi) for multiple sequence alignment and visualization, respectively [47, 50].

Analysis of physicochemical properties of proteins

Isoelectric point (PIs), molecular weight (MWs), mean hydrophilicity, and charge were all examined using the ExPASy-ProtParam website (https://web.expasy.org/protparam/) [51]. Wax gourd genome annotation file was used to map the chromosome positions of family members through TBtools software (version 1.104) [52]. The online software Plant-mPLoc (http://www.csbio.sjtu.edu.cn/bioinf/plant-multi/) was utilized to predict the subcellular localization of BhbZIPs by uploading protein sequences [53]. Additionally, the secondary structure of BhbZIPs was predicted through the online software SOPMA (https://npsa-prabi.ibcp.fr/cgi-bin/npsa_automat.pl?page=/NPSA/npsa_sopma.html) [54].

Conserved domain and phylogenetic tree analysis

All bZIP proteins sequence in Arabidopsis could be accessed from the TARE website (https://www.arabidopsis.org/). Pfam website (http://pfam.xfam.org/) was used to identify the position of conserved domain of BhbZIP and AtbZIP protein [46]. The following three steps should be taken to create a phylogenetic tree between AtbZIP and BhbZIP. First, the sequence of all bZIP protein will be aligned using mafft software [55]. Second, the gapthreshold value was set to 0.3 while trimming the multi-sequence alignment file using Trimal software [56]. The final step was to use IQ-TREE software to find the optimal model and build a maximum likelihood tree, and set the alert value to 1000 [57]. Then ITOL website (https://itol.embl.de/, version 6.6) was utilized to beautify the phylogenetic tree [58].

Analysis of protein conserved motif and Gene structure

MEME (http://meme-suite.org/) website was used to identify the conserved motif of BhbZIP proteins [59]. The maximum motif parameter of the protein was set to 20, the rest of the parameters remain unchanged [9]. Gene structure information of BhbZIP was obtained from genome annotation files. The IGV-sRNA was then used to correct and confirm the gene structure of each BhbZIP including UTR regions, exons and introns by bam files of the transcriptome data. Then a phylogenetic tree was created specifically for only BhbZIPs using the method described above. Finally, visualization of the phylogenetic tree, conserved motifs, and gene structure were performed by ggplot2 (version 3.4.0) and ggtree (versionv3.6.1) [60, 61].

Collinearity and gene duplication analysis of the BhbZIPs

The genome sequence and annotation data of rice, Arabidopsis, cucumber and grape were obtained from the Ensembl Plants website (http://plants.ensembl.org/index.html). BLASTP (version 2.13.0) and MCsanx software were then used to examine the inter-species collinearity relationship between wax gourd and rice, cucumber, grape, and Arabidopsis, and the intra-species collinearity relationship of wax gourd [62]. MCscanx and DupGen_finder attempts to identify the different modes of duplicated gene pairs in wax gourd by using grape as an outgroup [62, 63]. The synonymous and non-synonymous substitution (Ka/Ks) of the collinear gene pairs were determined using the ParaAT and KaKs_Calculator software [64, 65]. GC content and gene density of wax gourd genome were extracted by bedtools, and R program circlize was used for visualization [66].

Cis-regulatory elements analysis and prediction of the interaction network

The 2000-bp DNA sequence upstream of the start codon was extracted from each wax gourd bZIP gene using the Tbtools software (version 1.104) [52]. Then, we analyzed Cis-regulatory Elements using the online tool PlantCRAE (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) and categorize Cis-regulatory Elements. We kept only the promoters on the positive chain and eliminated a very small number of promoters, which were visualized with ggplot2 [61]. STRING software (https://string-db.org/) was used to analyze the interaction of BhbZIP proteins (orthologs in Arabidopsis), and the confidence parameter was set to 0.4 [67].

Plant material and stress treatments

The wax gourd seeds were provided by Vegetable Research Institute, Guangdong Academy of Agricultural Sciences [30]. The seeds were germinated and then sown in pots which were placed in an artificial climate chamber at 28 °C and 14/10 h for light /dark cycles. The same-growing and one-month-old seedlings were chosen and placed in the incubators at 42℃ and 8℃ for heat and cold stress treatment, respectively [68, 69]. We used 10% PEG6000 to simulate drought treatment and 9 g/L NaCl to simulate salt stress, both of which were treated with irrigated roots [70]. For ABA treatment, wax gourd leaves were sprayed with 0.2 g/L ABA [69]. At 0 h, 1 h, 3 h, 6 h, 12 h, and 24 h, the wax gourd leaves were collected, accordingly. The various wax gourd tissues were obtained from plants at the reproductive stage under normal growing circumstances. Three biological replicates were performed for all treatments, and samples were submerged in liquid nitrogen and then moved to a refrigerator at -80℃.

Total RNA extraction and RT-qPCR

Total RNAs were extracted from different samples using a TransZol Up Plus RNA Kit (TransGen Biotech, Beijing, China) by the specification. Then the RNAs were reverse transcribed into the first-strand cDNA by a FastKing gDNA Dispelling RT SuperMix Kit (Tiangen Biotech, Beijing, China). BhbZIP58 was chosen for research using Real-time quantitative PCR (qPCR) to conduct the expression profile under abiotic stresses. Three replicates were conducted on a CFX96 Touch™ Real-Time PCR detection system (BIO-RAD, USA). Actin gene (Bhi10M001911) of wax gourd was used as a reference for normalization, whose primers were listed in Table S7 [71]. PCR reactions were performed using iTransStart® Green qPCR SuperMix (TransGen Biotech, Beijing, China). The relative mRNA expression level for each gene was calculated by the 2−ΔΔCT method [72]. Gene-specific DNA primers for qPCR were presented in Table S7. To find the expression profiles with significant differences, the mean Ct values were statistically examined using Duncan’s test. P ≤ 0.05 was regarded as statistically significant.

Subcellular localization analysis in tobacco

BhbZIP58 gene was cloned in two separate segments (F1-R1, F2-R2) and then the total sequence was cloned using F1 and R2 primers (Table S7). The full-length coding sequence (CDS), without the stop codon of BhbZIP58 was cloned and inserted into the pSuper-1300 GFP vector to fuse with the GFP tag. Then using the freeze–thaw method, the recombinant plasmid produced and empty vector were respectively introduced into Agrobacterium tumefaciens strain GV1301. The tobacco was grown in a greenhouse under the 16-h light/8-h dark cycle at 25℃ for five weeks. Agrobacterium was then incubated overnight and the bacteria were collected by centrifugal separation, diluted to an OD600 of 1 with a resuspension and injected into tobacco leaves. The infected plants were cultured for two days as normal after being kept in the dark for 24 h. Near the injection site of the leaves, the confocal laser scanning microscope (Zeiss LSM710, Jena, Germany) was utilized to observe the GFP fluorescence. Primers used for vector construction were listed in Table S7.

Transcriptional activation activity analysis of BhbZIP58

pEASY®-Basic Seamless Cloning and Assembly Kit (TransGen Biotech, China) and NdeI and BamHI restriction endonucleases (New England Biolabs, USA) were used to insert the CDS of BhbZIP58 into the pGBKT7 vector. On a medium devoid of tryptophan (Trp) and histidine (His), pGBT9 was applied as a positive control (PC) that could activate the HIS3, ADE2, and MEL1 reporter genes as well as α-galactosidase activity in the transformed yeast cells. pGBKT7 served as a negative control (NC). The yeast strain AH109 was transformed with the NC, PC, and pGBKT7-BhbZIP58 plasmids, which were then cultivated on synthetic-defined (SD)/-Trp or SD/-Trp/-His medium with or without X-α-Gal.

Heterologous overexpression in Arabidopsis

The CDS of BhbZIP58 was inserted into the pBI121 vector and controlled by the 35S promoter. A.tumefaciens strain GV3101 was transformed with the recombinant construct. Arabidopsis Columbia ecotype (Col-0) was used as the WT, which provided by Vegetable Research Institute, Guangdong Academy of Agricultural Sciences. The Arabidopsis plants were grown in a growth room at 22℃ with a 16-h light/8-h dark cycle. Agrobacterium-mediated transformation of Arabidopsis was performed using the floral dip method [73]. Transgenic plants were identified by kanamycin resistance and further confirmed by qPCR. Arabidopsis T3-generation plants were used for functional identification. For the seedling survival assays, we planted 20 seeds of WT and three overexpression lines. The control group was grown at 22 °C for 12 days, while the treatment group underwent the following heat stress conditions: incubation at 22 °C for 5 days, followed by exposure to 37 °C for 90 min, then returning to 22 °C for 90 min, subsequently subjecting the seedlings to 44 °C for 90 min, and finally returning at 22 °C for 7 days [74]. Three biological replicates were performed for all treatments. The survival rate of the 20 Arabidopsis seedlings was then used to represent the transgenic lines.