Genome-wide characterization of the MBF1 gene family and its expression pattern in different tissues and stresses in Zanthoxylum armatum

Background Multiprotein bridging factor 1 (MBF1) is a crucial transcriptional coactivator in animals, plants, and some microorganisms, that plays a necessary role in growth development and stress tolerance. Zanthoxylum armatum is an important perennial plant for the condiments and pharmaceutical industries, whereas the potential information in the genes related to stress resistance remains poorly understood in Z. armatum. Results Herein, six representative species were selected for use in a genome-wide investigation of the MBF1 family, including Arabidopsis thaliana, Oryza sativa, Populus trichocarpa, Citrus sinensis, Ginkgo biloba, and Z. armatum. The results showed that the MBF1 genes could be divided into two groups: Group I contained the MBF1a and MBF1b subfamilies, and group II was independent of the MBF1c subfamily.. Most species have at least two different MBF1 genes, and MBF1c is usually an essential member. The three ZaMBF1 genes were respectively located on ZaChr26, ZaChr32, and ZaChr4 of Zanthoxylum chromosomes. The collinearity were occurred between three ZaMBF1 genes, and ZaMBF1c showed the collinearity between Z. armatum and both P. trichocarpa and C. sinensis. Moreover, many cis-elements associated with abiotic stress and phytohormone pathways were detected in the promoter regions of MBF1 of six representative species. The ERF binding sites were the most abundant targets in the sequences of the ZaMBF1 family, and some transcription factor sites related to floral differentiation were also identified in ZaMBF1c, such as MADS, LFY, Dof, and AP2. ZaMBF1a was observed to be very highly expressed in 25 different samples except in the seeds, and ZaMBF1c may be associated with the male and female floral initiation processes. In addition, expression in all the ZaMBF1 genes could be significantly induced by water-logging, cold stress, ethephon, methyl jasmonate, and salicylic acid treatments, especially in ZaMBF1c. Conclusion The present study carried out a comprehensive bioinformatic investigation related to the MBF1 family in six representative species, and the responsiveness of ZaMBF1 genes to various abiotic stresses and phytohormone inductions was also revealed. This work not only lays a solid foundation to uncover the biological roles of the ZaMBF1 family in Z. armatum, but also provides some broad references for conducting the MBF1 research in other plants. Supplementary Information The online version contains supplementary material available at 10.1186/s12864-022-08863-4.


Background
Flooding, drought, salinity, metal ions, extreme temperature and other adverse conditions severely inhibit plant growth and development, often resulting in changes in plant physiology, biochemistry, morphology, senescence and even death [1]. Therefore, carrying out studies on plant stress and improving plant resistance can improve the economic values of the plant. The vegetative growth, reproduction, and development of plants are significantly impacted by a vast array of biotic and abiotic stresses [2]. During long-term evolution, plants have developed a comprehensive regulatory network to facilitate the rapid detection of environmental changes and simultaneously conduct adaptive responses [3]. Environmental signals can be sensed by internal stress-induced factors in plants, and then a broad range of transcription factors (TFs) will be activated to mediate the plant appropriate responses, including WAKY, Trihelix, NAC, MYB, ERF, C2H2, bZIP, bHLH, LBD, MBF1, and other families [1,[4][5][6]. In previous studies, approximately 7% of the plant genomecould be encoded TFs, including approximately 4.7% of Arabidopsis thaliana, 6.4% of poplar, and 7.7% of Zanthoxylum armatum [2,7,8]. These TFs, as signal molecules, interact with specific cis-acting elements in the promoter region of downstream stress response genes to regulate the signal cascade network in plants [8][9][10].
Multiprotein bridging factor 1 (MBF1), as one of the crucial transcriptional coactivators, is widely distributed in animals, plants, and some microorganisms, and plays a necessary role in gene expression and regulation during stress responses [11,12]. There are three members in the MBF1 family, including MBF1a, MBF1b, and MBF1c [13], each with a flexible N-terminal region and highly conserved helix-turn-helix (HTH) domain in its C-terminal [11]. The function of the N-terminus was to bind with the activating proteins, and the C-terminus was contributed to the protein folding and dimer formation of functional MBF1. In previous studies, MBF1, as a transcriptional coactivator, has been shown to bridge the regulatory factor and TATA-box binding protein. It continually shifts its the locations in the cytoplasm and nucleus to conduct its functions [14]. The ubiquitous expression of MBF1s was investigated in various tissues of different species, such as roots, stems, leaves, flowers, fruits, seeds and anther organs [15], but some specific expression abundances were found among the three different MBF1 genes during plant development [13,16]. The overexpression of AtMBF1c can increase the number of seeds in Arabidopsis and soybeans [17,18], and the flowering time was also earlier in Arabidopsis [17]. MBF1c rapidly accumulates and interacts with trehalosephosphate synthases to respond to heat stress [19]. The overexpression of MBF1 genes also increases antioxidant enzyme activity under drought stress and NaCl treatment [20,21]. A subsequent study found that an increase in the Ca 2+ signal could enhance tolerance to salt stress in MBF1c overexpressing plants [22]. MBF1 genes appear to be the negative regulator for cold stress response in hot pepper [15], but have no effect on seedling in Arabidopsis and R. raetam [15]. Although some previous reports described the MBF1 genes in other species, there is a paucity of information at the whole genome level. Therefore, more functions of MBF1 need to be continuously uncovered in plants.
Zanthoxylum (Rutaceae) is one of the most important economic and medicinal tree species in southwestern China and some regions of Asia, such as Japan, South Korea, India, Pakistan, Nepal and so on [23,24]. Due to the special numbing taste of its fruits, it is a very famous and essential ingredient in Sichuan hot-pot in China, and is often used to flavour some delicious dishes in many countries [25,26]. Zanthoxylum armatum DC., commonly named green prickly ash, is not only enriched in special numbing substances, but also has a special fragrance, and the colour of its fruit is green, all of which contribute to its increasing popularity in recent public diets [27]. As one of the crucial spices to these cuisines and regimens, this agricultural product creates commercial value in billions of dollars in China, as shown on the website of huajiao.cn [24,28]. Meanwhile, Z. armatum is also used for economic development as a vital medicinal plant in several countries [29][30][31]. For instance, Z. armatum is planted as the second most important medicinal species in Pakistan [32]. The fruit pulp of Zanthoxylum could be extended to treat skin diseases, toothache, fever, and stomach ache [24,26,30]. Plant-based medications are a high proportion of prescription used in regions of health care scarcity [33]. Fortunately, Z. armatum exhibits a high tolerance to drought, heat, and nutrient poor soils, so it is also selected to colonize barren mountains as a pioneer tree. However, Z. armatum is susceptible to flooding, cold, and some disease, which seriously restrict its large-scale application and development, especially in tropical and subtropical rainy regions. Therefore, exploring stress resistance is a more effective strategy to breed superior varieties and expand the cultivation regions of Z. armatum, thereby increasing the availability of its products and economic value. In a previous study, 38 WRKY transcription factors were identified in response to drought in Zanthoxylum bungeanum Maxim (red prickly ash) [25]. However, the potential genetic information of the genes related to stress resistance remains poorly understood in Z. armatum (green prickly ash).
Recently, the genomic dataset of Z. armatum was released with important resources [34], providing a wealth of information to explore the gene families associated with resistance functions. Thus, in the present study, the MBF1 family was isolated and identified in Z. armatum. To comprehensively obtain the detailed characteristics of the MBF1 family, we also selected another five species, including Arabidopsis thaliana (Herbage), Oryza sativa (Poaceae), Populus trichocarpa (Xylophyta), Citrus (Homology), and Ginkgo biloba (Gymnosperm). A series of analyses were performed in the present study to identify features, such as phylogenetic relationships, gene structures, motif compositions, chromosomal distributions, gene duplications, cis-acting elements in promoters, protein tertiary structures, and target gene sites. The gene expression profiles of the ZaMBF1 family were also carried out in 26 different tissues by using publicly available and proprietary RNA-seq datasets. Furthermore, we also investigated the expression patterns of ZaMBF1 genes in response to various abiotic stresses and phytohormone inductions by qRT-PCR analysis, including water-logging, cold stress, ethephon, methyl jasmonic acid (MeJA), and salicylic acid (SA) treatment. Thus, the results described in the present study represent valuable information for future functional studies associated with the MBF1 family.

Isolation and identification of MBF1 in six species
To comprehensively analyse the information of the MBF1 family in the present study, six representative species were selected to isolate the MBF1 genes, including A. thaliana as the herbage model plant, O. sativa as the Poaceae model plant, P. trichocarpa as the woody model plant, C. sinensis as the homology plant, G. biloba as the gymnosperm plant, and Z. armatum. Fortunately, most of these species had three MBF1 genes, and there were five members in C. sinensis (Table 1) To further clarify the characteristics of MBF1 family in Z. armatum, we analysed the protein molecular weight (MW), isoelectric point (pI), grand average of hydropathicity (GRAVY), and Pfam annotation. It was clearly found that all twenty MBF1 proteins exhibited overall hydrophilicity (from -0.94 to -0.33). The same Pfam IDs suggested that these proteins contained the same conserved domain (Additional file 3: Table S3). In general, the ZaMBF1 family has shown typical molecular characteristics compared with other different species, which will be helpful for further study to uncover its specific biological functions.

Evolutionary and phylogenetic analysis of the MBF1 family
In the present study, a total of 202 amino acid sequences from 53 species were collected to construct a phylogenetic analysis and illustrate the evolutionary relationships of MBF1 family proteins (Fig. 1a). The results showed that all of the MBF1s were divided into two groups based on their subfamilies. Group I contained the MBF1a and MBF1b subfamilies, and group II independently comprised the MBF1c subfamily. MBF1a and MBF1b were detected to have a closer relationship, which suggested that they might conduct more similar functions. However, MBF1c was displayed to be a relatively independent group, which indicated that it might play some specific roles related to stress response and the development process in plants. Moreover, in Anacardium occidentale, a member of the Anacardiaceae family, three MBF1 proteins were adjacent to the ZaMBF1 protein. Notably, both of ZaMBF1a and ZaMBF1b were located in the MBF1a subfamily. Combined with their similar sequence information (Table 1), we suggest that ZaMBF1a and ZaMBF1b might play similar roles or even redundant functions.
To explore the origin and evolution of MBF1 genes, we investigated 53 species from different kinds of plants.
Notably, the number of MBF1 genes ranged from 2 to 8, and most plants contained 3 or 4 genes (Fig. 1b). More MBF1 genes were detected in annual or perennial herbs (Additional file 1: Table S1), such as Brassica rapa FPsc (6), Caulanthus amplexicaulis (7), Eruca vesicaria (8), Gossypium darwinii (7), Gossypium hirsutum (8), Kalanchoe laxiflora (6), Lotus japonicus (7), and Triticum aestivum (7). At the same time, these species mainly exhibit the duplication of the MBF1a and MBF1b subfamilies. These results could explain one of the reasons why herbaceous plants are often much more resistant to abiotic stress. Table 1 The information summary of MBF1 sequences and annotation Note: Gene ID is the ID in their genome database, gDNA is the base length of DNA sequence, CDS is the base length of conding sequence, GC % is the content of G and C base in Coding sequence, AA is the number of amino acid, MW is the molecular weight (kDa), GRAVY is the grand average of hydropathicity, pI is the theoretical isoelectric point

Gene structure and subcellular location analysis
To comprehensively explore the information of the MBF1 family, the six species in this study were selected for gene structure analysis (Fig. 2). Similarly, the MBF1 genes were clustered in two major branches. The MBF1a and MBF1b subfamilies belonged to the same class, and all of them had 2-3 introns and 3-4 exons (Fig. 2a). The exon lengths of MBF1a and MBF1b were similar in each plant, but longer introns were detected in O. sativa and G. biloba, especially GbMBF1b. Both ZaMBF1a and ZaMBF1b were showed four exons and three introns. Furthermore, most of the MBF1c genes only had one exon, but the CDS of ZaMBF1c was separated by a short intron. Additionally, the species with larger genomes (G. biloba and Z. armatum) do not exhibit UTR regions in their MBF1 members.
To further analyse the biological functions, the motif and conserved domains were identified in the twenty MBF1 proteins of various plants. Ten conserved motifs were identified in the MBF1 protein sequence by MEME software (Fig. 2b), and three significant motifs were detected and displayed in Additional file 4: Table S4. Motif 2 was discovered to be related to the MBF1 domain, which is involved in the aMBF1 superfamily ( Fig. 2c, Additional file 5: Table S5). Motif 5 was only detected in the MBF1c subfamily in some species. Similar motifs and domains were detected in ZaMBF1a and ZaMBF1b, which indicated that they might endow some similar biological functions.
Moreover, subcellular localization predictions were made for all MBF1 genes of the six species (Fig. 3a). The results showed that most of them were located in the cytoplasm, and the rest were located in the nucleus, except CsMBF1b.4. All the ZaMBF1 were mainly located in the cytoplasm, especially ZaMBF1b. ZaMBF1a and ZaMBF1c could also be related to peroxisome, which might imply an important role as resistance factors. Moreover, the signal peptide prediction is additional evidence for the localization of genes. The results showed that no signal peptide sequence was detected in any of the three ZaMBF1 genes ( Fig. 3b-d), which indicated that they were not secreted proteins or transmembrane proteins, so they could only be freed in the cytoplasm or localized in the nucleus. The consistent result in a previous study was that the subcellular localization of MBF1 was dependent on plant development stage, interaction with other transcription factors, and environmental conditions, and it could shuttle back and forth between the cytoplasm and the nucleus [14].

Analysis of chromosomal locations and collinearity
The chromosomal location analysis showed that the MBF1 genes were closer together on chromosomes when their relationships were closer (Additional file 6: Figure S1). For example, the four CsMBF1b were located in nearby sites on CsChr5. OsMBF1a and OsMBF1b, GbMBF1c.1 and GbMBF1c.2 also displayed the similar located patterns. A very important finding was that three ZaMBF1 genes were located on three different chromosomes. Although ZaMBF1a and ZaMBF1b showed the close gene structure and phylogenetic relationships, they were still observed on ZaChr26 and ZaChr32, respectively. This result again suggested that a gene duplication event might have occurred in ZaMBF1a and ZaMBF1b.
To confirm our results, we carried out collinearity analysis by the MCScanX method using the genome data of Z. armatum. The results showed that obvious collinearity occurred in the genomic profile of Z. armatum, and the three ZaMBF1 genes also showed the gene replication events among each other (Fig. 4). The higher sequence The phylogenetic and gene structure analyses of MBF1 genes from six representative species. a The genomic structure of various MBF1 genes. The blue box was UTR regions, green box was exons, and organe line was introns. The phylogenetic tree was also conducted using the amino acid sequences of MBF1 proteins in each species. b The motif distribution of each MBF1 member. c The conserved protein domain distribution of each MBF1 protein, aMBF1 was archaeal ribosome-binding protein aMBF1 domain Fig. 3 The subcellular localization and signal peptide prediction of MBF1 genes. a the subcellular localization of MBF1 genes from six representative species, the data is the certainty of each MBF1 located in the various sites. b-d the signal peptide prediction of ZaMBF1 family Fig. 4 The synteny analysis of the MBF1 genes between Z. armatum and other five species. The red lines highlighted the syntenic MBF1 pairs between Z. armatum and the species similarity between repetitive gene pairs was due to the ZaMBF1 family involvement in the similar biological regulation processes, and the gene replications were likely to promote the evolution of the ZaMBF1 genes in Z. armatum.
To further analyse the evolutionary relationship of MBF1 genes in the plants, the comparative syntenic maps were first constructed between Z. armatum and other species. The results showed that a rich synteny was detected between Z. armatum and eitherA. thaliana or O. sativa, but it was not present in MBF1 family genes ( Fig. 5a-b). However, the paralogous MBF1 genes were investigated in the woody plants in this study (Fig. 5c-d).
MBF1c acted as one pair of syntenic genes observed in Z. armatum and P. trichocarpa or C. sinensis. Surprisingly, C. sinensis, as a proximal relationship species, was also found to have a syntenic gene with A. thaliana (Fig. 5d). These results suggested that MBF1c may play a vital role in the evolutionary process of the MBF1 gene family, and some considerable differentiation of MBF1 genes might have been carried out during the evolution of perennial woody plants. Meanwhile, few syntenic relationships were detected between Z. armatum and G. biloba (Fig. 5e), which might indicate significant differences between angiosperms and gymnosperms during the long history of evolution.

Cis-elements and tertiary structure analysis of MBF1
To further determine the molecular functions and expression pattern of the MBF1 family, the promoter regions of 2000 bp upstream from the transcription starting site were isolated from the genomic data of six species. A series of cis-acting elements were identified as being associated with abiotic stress (Additional file 7: Table S6). Furthermore, most promoters exhibited the responses to drought, light, anaerobic, and low-temperature conditions (Fig. 6, Additional file 8: Table S7). The eight cis-elements detected in ZaMBF1a were mainly related to drought and light responsiveness. An exclusive cis-element associated with circadian control was obtained in ZaMBF1a among the MBF1 promoters of six species. Meanwhile, 12 cis-elements were investigated from ZaMBF1b, which were mainly involved in responses to light, anaerobic and low-temperature induction. The cis-elements of ZaMBF1c were likewise focused on light, anaerobic induction and low-temperature responsiveness. In addition, some phytohormone pathways related to stress were also detected in these promoter regions (Fig. 6, Additional file 8: Table S7). All of the species were detected to have at least one cis-element involved in abscisic acid (ABA) responsiveness in the present study, except Z. armatum. We also could not observe any cis-elements related to gibberellin (GA) responsiveness in Z. armatum and O. sativa. However, there were more cis-elements associated with MeJA and salicylic acid (SA) responsiveness in Z. armatum. These results indicated that the pathways of MeJA and salicylic acid might play vital roles in biotic and abiotic stress resistance in Z. armatum. Notably, almost every cis-element was detected in G. biloba, which might explain its strong vitality in the Mesozoic era.
Moreover, the tertiary structures of the ZaMBF1 amino acid sequence were investigated by the SWISS-MODEL server (Fig. 7). Both the MBF1 domain and helix-turnhelix domain were detected in each protein. A similar 3D structure was observed in ZaMBF1a and ZaMBF1b (Additional file 9: Table S8), which suggested that these genes may perform shared functions. However, ZaMBF1c displayed a relatively different tertiary structure, especially in the helix-turn-helix domain.

TF binding site analysis of the ZaMBF1 family
According to the results from the Plant Transcriptional Regulatory Map, a total of 382 TF binding interactions were observed in the ZaMBF1 family, and their length ranged from 8 to 28 bp (Additional file 10: Table S9). Most of the target genes were associated with the responses to plant stress, such as the WRKY, Trihelix, NAC, MYB, LBD, and C2H2 (Fig. 8). The ERFs were the most frequent target sites in the sequences of the ZaMBF1 family, which indicated that a strong interaction might be carried out between ERFs and ZaMBF1 proteins. Meanwhile, some ARR sites were also detected in ZaMBF1a, which are important members of the cytokinin signal transduction pathway. It is worth noting that some TF sites involved in floral differentiation were identified in ZaMBF1c (Fig. 8c), including one site related to MIKC_MADS, two sites related to LFY, one site related to Dof, and one site related to AP2. Meanwhile, a site related to MIKC_MADS was also detected in ZaMBF1b. These results suggested that the ZaMBF1 family not only plays the vital roles in the response to stress, but also performs some important functions in the reproductive growth process. To explore the role of the MBF1 family in the growth and development of Z. armatum, the expression abundance of MBF1 genes was investigated in the transcriptome dataset. Based on the expression profiles from the 12 various organs in Z. armatum (Fig. 9a, Additional file 11: Table S10), ZaMBF1a exhibited the opposite expression patterns as ZaMBF1c. ZaMBF1a was highly expressed in all of the samples except for seeds, and high abundances were observed in roots, stems, and mature leaves. This result suggested that ZaMBF1a plays a crucial role in the growth of Z. armatum. However, ZaMBF1c had a relatively high expression in seeds but a relatively lower expression in fruits, husks and other samples. This result indicated that ZaMBF1c might be associated with the seed development process. ZaMBF1b often showed moderate expression compared with ZaMBF1a and ZaMBF1c, which implied that its role in the growth and development of various organs might be redundant.
Meanwhile, similar expression patterns were still observed during the male and female floral differentiation processes (Fig. 9b, Additional file 12: Table S11). ZaMBF1c was significantly up-regulated in the initiation stage of floral sex differentiation, but down-regulated in the subsequent floral differentiation and development process. This result also confirmed that the ZaMBF1c conducts some important functions in the reproductive growth process. However, ZaMBF1a exhibited a complementary expression pattern in the male and female flower samples, but ZaMBF1c showed an opposite expression patterns. These results suggested that both ZaMBF1a and ZaMBF1c play critical roles in vegetative and reproductive development. Additionally, ZaMBF1b was not detected an obvious expression pattern, which indicated that it might be a redundant gene that serves as a supplementary functional member in the growth and development of Z. armatum.
To further understand the role of the ZaMBF1 family in response to stress, we performed the expression patterns of ZaMBF1 genes after infecting the leaves with rust disease (Fig. 9c). Both ZaMBF1b and ZaMBF1c were observed to be significantly up-regulated in the 7 days after infection (INCZ7), which suggested that this time was a vital point for ZaMBF1 genes to respond to leaf rust disease. However, we could not find any information about the ZaMBF1a gene in the transcriptome dataset after infection. This result indicated that the expression of ZaMBF1a may not be significantly induced by leaf rust disease. All of these results will be helpful for future research on gene function and molecular mechanisms.

Expression profiles of the ZaMBF1 family in response to different treatments
To deeply uncover the roles of the ZaMBF1 family in response to stress, various treatments were carried out to detect the expression of three ZaMBF1 genes by qRT-PCR (Fig. 10). Totally, each treatment could significantly induce or inhibit the expression of ZaMBF1 genes in Z. armatum.
Similar expression patterns were observed in ZaMBF1a and ZaMBF1b after water-logging treatment (Fig. 10a-c). Both of them were up-regulated at 6 and 12 h, but gradually down-regulated to the normal level (similar to 0 h) in the subsequent days. However, ZaMBF1c was found to have a very significant up-regulation after water-logging treatment (Fig. 10c), the expression of which was approximately eightfold higher after two days treatment Fig. 8 The transcription factor binding sites of ZaMBF1 members in Z. armatum. a The predicted information of the target genes interacted with ZaMBF1a. b The predicted information of the target genes interacted with ZaMBF1b. c The predicted information of the target genes interacted with ZaMBF1c. The pie charts were the percentage of the various transcription factors in each ZaMBF1 member, respectively compared to the baseline at 0 h. Even though the seedlings were treated 3 days with water-logging, a very high relative expression of ZaMBF1c was still detected. These results suggested that the ZaMBF1 family could play an important role in the response to water-logging stress, especially the ZaMBF1c gene.
Additionally, there was no significant pattern in the expression of ZaMBF1a and ZaMBF1b after cold treatment ( Fig. 10d-f ). However, ZaMBF1c was very significantly down-regulated in the cold environment, and was significantly inhibited after 6 h of treatment (Fig. 10f ). These results implied that ZaMBF1a and ZaMBF1b might not be directly related to cold stress, while ZaMBF1c showed a negative response to cold stress in Z. armatum.
Due to the crucial target sites of ERFs in the sequences of the ZaMBF1 family, ethephon (ETH) was selected to treat the seedlings and investigate the expression patterns of the three ZaMBF1 genes (Fig. 10g-i). The confirmatory results showed that significant down-regulation was observed at 6 h after ETH treatment among the three MBF1 genes, especially the ZaMBF1a genes, which had more ERF target sites (Fig. 8). However, the expression were gradually up-regulated to the normal level by 9 h after ETH treatment. All of these results suggested that negative regulation might also be used to respond to ETH signalling in the ZaMBF1 family, and that ZaMBF1a might be a vital member due to its enriched ERF elements.
The effects of MeJA treatment showed that ZaMBF1a and ZaMBF1b had comparable similar expression patterns (Fig. 10 j-l), and they were gradually up-regulated within the 12 h posttreatment and returned to normal levels at 24 h posttreatment. However, ZaMBF1c was found to be very significantly up-regulated at 3 h posttreatment (Fig. 10 l), and very high expression was also detected again at 24 h after treatment. These results indicated that the ZaMBF1 genes also play an important role in the response to MeJA stress, especially the ZaMBF1c gene.
To confirm the information of the ZaMBF1 family response to phytohormones involved in plant stress resistance, SA treatments were also conducted to detect the expression of the three ZaMBF1 genes (Fig. 10 m-o). ZaMBF1a was significantly induced within the 9 h of treatment, but significantly downregulated at 24 h after treatment. Similar results were observed for ZaMBF1b. However, significant up-regulation was observed in the expression of ZaMBF1c after SA treatment (Fig. 10 o), and the expression level of 3 h after treatment was approximately 12-folds higher than that of the control. To sum up, the three ZaMBF1 genes could serve their functions in response to plant stress via phytohormone signalling pathways. Both ZaMBF1a and ZaMBF1b performed their functions within 12 h after treatment, whereas the ZaMBF1c could continue to work via their powerful functions in the plant stress resistance.

Discussion
Multiprotein Bridging Factor 1 (MBF1) is an ancient super-family across plant life, that has a typical helixturn-helix structure [14]. The first study of MBF1 was reported from silkworm as a non-DNA binding transcription co-factor [35]. In the present study, we selected MBF1 from six representative species to uncover its characteristics. Compared with A. thaliana, all other species have at least two different MBF1 genes, and MBF1c is an essential member. Interestingly, the CDS length of most MBF1c was slightly longer than that of the MBF1a and MBF1b genes in each species. MBF1a and MBF1b were more closely relationship, whereas MBF1c displayed relatively independence according to the phylogenetic analysis of 202 amino acid sequences from 53 species. The consistent results were that MBF1 family could be divided into two groups, and the structure of Group I was four exons and three introns. The structure of group II contained zero, one, or two introns [14]. Thus, MBF1a and MBF1b could be deemed to belong to the Group I subfamily, and MBF1c could be deemed to belong to the group II subfamily in the current study. This might be the reason why the MBF1a and MBF1b subfamilies were alternately clustered into a large group in the present The 0 h was selected as the control sample to estimate the relative expression in each treatment, and the asterisks were the significant differences, * indicated the p < 0.05; ** indicated the p < 0.01 study (Fig. 1a). In addition, the gene structure with few or no introns might have contributed to rapid transcription and translation to enhance the ability to respond to external stress [36,37]. This might explain why G. biloba, a famous living plant fossil, survived the ice age due to an increased number of MBF1c members. Furthermore, the short sequence of MBF1c might be caused by the limited genome assembly in C. sinensis. It is worth noting that the gene structures of ZaMBF1a and ZaMBF1b appeared to be symmetrical and obtained more similar motifs and domains (Fig. 2), and collinearity was also observed between ZaMBF1a and ZaMBF1b (Fig. 4), all of which suggested that a gene duplication event occurred in ZaMBF1a and ZaMBF1b. Gene replication events, which usually depended on tandem and segmented replications, could promote the expansion of gene family members and enhance the potential regulation abilities to adapt to the various environmental changes [38][39][40]. Previous studies suggested that tandem repeats might provide some similar genetic functions and expression patterns [37,41]. Thus, the detailed functions and the essential members of these ZaMBF1 genes could be revealed in further studies.
The expansion of gene family members and the mechanism of genomic evolution were mainly due to gene duplication events, including tandem and segmented replications [37]. Duplication events related to the resistance gene families have been reported in many plant species [1, 37,41], but there have been no findings related to the MBF1 family. In this study, three ZaMBF1 genes were located on different chromosomes without any tandem duplication, whereas the obvious collinearity was present among them. In a previous study, an independent whole genome duplication (WGD) event was followed by extreme amplification of transposable elements in Z. armatum, which promoted the genome size expansion of Z. armatum [34]. Thus, gene replication or WGD events might have contributed to the evolution of ZaMBF1 genes in Z. armatum [42,43]. Additionally, the analysis of collinear regions between species suggested that ZaMBF1c exhibited collinearity only with the MBF1c genes in P. trichocarpa and C. sinensis. These results indicated that the orthologous genes may distinctly evolve in similar species types, which was consistent with previous studies on the collinearity of WRKY genes in pine/apple/ rice/corn and NAC genes in A. thaliana/apple/grape/ rice [1,44]. Meanwhile, Zanthoxylum is relatively recent genus and diverged from Citrus approximately 36.5-37.7 million years ago [45]. Z. armatum and C. sinensis, both of them belonging to the Rutaceae family, were estimated to have diverged approximately 32.9 million years ago, and Z. armatum experienced an independent wholegenome duplication event approximately 26.6 MYA after its split with Citrus [34]. Thus, these reasons might result in the syntenic gene pairs of MBF1c obtained in Z. armatum vs. C. sinensis and C. sinensis vs. A. thaliana, but not between Z. armatum and A. thaliana. Furthermore, the syntenic region of MBF1c was also present between A. thaliana and C. sinensis, but was not detected among ZaMBF1c with O. sativa and G. biloba, implying that MBF1c developed after the divergence of dicots and monocots under the clade of angiosperms [1].
The transcription factor binds to cis-elements located in the gene promoter sequences to play their core roles regulating the expression of the genes and controlling the growth, development, and responses to the environment in plants [46,47]. The structure of the promoter could suggest the possible functions and regulatory mechanism of the genes [3]. Furthermore, the copy number of the cis-elements could also determine the transcription level and cause the specific phenotype [48]. Our present study investigated cis-elements associated with MBF1 in six different species. Fortunately, a total of 255 cis-elements were detected in six representative species, including a response to drought inducibility, light, anaerobic conditions, and so on. The highest number of cis-elements was observed in O. sativa and A. thaliana, implying that Poaceae and herbage have multiple stress resistance pathways. Additionally, most species have cis-elements involved in phytohormone pathways, such as the responsiveness to ABA, GA, MeJA, and SA, which suggested that MBF1 could interact with plant hormones to defend against plant stress [15,19,20]. Notably, almost every cis-element can be detected in G. biloba, which might explain why this species could have survived the long evolution of the Mesozoic era to become a living plant fossil. Additionally, a previous study found that MBF1s could regulate the expression of ABA REPRESSOR1 (ABR1) under MV-induced (Methylviologen) stress [15,19,49]. The ectopic expression of HbMBF1a enhanced ABA insensitivity in A. thaliana [50]. However, we could not detect any cis-elements related to ABA and GA responsiveness in the promoter sequences of Z. armatum, suggesting that the ZaMBF1 genes are not involved in the plant stress response process via ABA and GA pathways. Thus, detailed information on the ZaMBF1 genes needs to be explored in further studies by exogenous induction, expression detection, and functional identification in Z. armatum.
Due to the necessity of flooding and cold resistance during the cultivation of Z. armatum and the detection of cis-elements associated with anaerobic induction and low-temperature responsiveness in ZaMBF1 promoter regions, these abiotic stresses were selected to treat the seedlings and carry out the expression pattern analysis of ZaMBF1 genes. Meanwhile, more MeJA and SA cis-elements were detected in the promoter regions and the crucial target sites of ERFs in the sequences of the ZaMBF1 family in Z. armatum, so these factors were also chosen to investigate the changes in abundance of the ZaMBF1 family in the present study. In a previous study, the ethylene response factor repressed MBF1 expression by binding to its promoter region [51]. The consistent result was that the expression of ZaMBF1 genes was very significantly down-regulated at 6 h but gradually up-regulated to the normal level in the subsequent times after ETH treatment. Thus, negative regulation might occur to respond to ETH signal in ZaMBF1 family, especially for ZaMBF1a, due to the high frequency of ERF binding sites in its sequence. Similar to previous studies, the 6 h point after various treatments was a crucial point in the ZaMBF1 genes response to external induction in the current investigation [20]. ZaMBF1c was observed to be more sensitive to these treatment and exhibited significant resistance, even after 3 days. In fact, MBF1c is a vital member that responds to the environmental changes in various species [19,22,49] and can interact with TPS gene family to defend against environmental stress [52]. Moreover, the function of MBF1 genes has been of great interest in many plants [18,19,22,53,54], but they appeared to have different roles in the responsiveness to cold stress [14]. The ectopic expression of AtMBF1 did not influence the response to cold stress in Arabidopsis seedlings [17], and a similar result was also observed in R. raetam [55]. The transcript abundance of BocMBF1 could be significantly activated in Brassica oleracea [56]. However, a significant down-regulation was detected as the cold stress response in the overexpression of CaMBF1 in hot pepper [15]. The consistent result was that ZaMBF1b and ZaMBF1c were very significantly down-regulated after cold stress in the current study, which is probably because both possessed more cis-elements associated with low-temperature responsiveness in their promoter regions. These results indicated that negative regulation may exist in response to low-temperature in Z. armatum [15]. Recently, water stress has seriously affected plant growth with global warming, which is one of the fatal factors in the worldwild decline in agricultural and forestry productivity [57]. Plants activate the antioxidant defence system and synthesize many stress response genes involved in signal transduction networks to avoid cell damage and improve their tolerance [58][59][60]. In previous studies, the overexpression of MBF1 resulted in the accumulation of fewer reactive oxygen species and increased antioxidant enzyme activity [14]. The ability to contend against oxidative stress is severely impaired in mbf1abc triple mutant plants of Arabidopsis, and the roots and cotyledons of this triple mutant undergo cell death after treatment with H 2 O 2 [61]. Fortunately, the consistent results in the present study were that all of the ZaMBF1 genes showed the powerful responsiveness to the water logging treatment, especially the ZaMBF1c genes. However, the regulatory functions of ZaMBF1 family are poorly understood. Thus, we suggest that these genes could be further researched to reveal their functions and molecular mechanisms related to anaerobic and cold tolerances, which would benefit the breeding of some high-quality varieties for perennial rainy areas in further studies.
Moreover, MBF1 plays a significant role in growth development and stress tolerance and shows ubiquitous patterns of expression in various organs [14]. In previous studies, the MBF1 genes were detected in roots, stems, leaves, flowers, fruits, seeds and other tissues of various species [13,15,16]. However, the expression levels of three MBF1 genes showed significant diversity in different tissues [14]. In the present study, the consistent results were that ZaMBF1a is predominantly expressed in all tissues except seeds, and seeds also showed a relatively high expression compared with that in other organs. This implied that ZaMBF1a might be an essential member associated with plant development. Meanwhile, ZaMBF1c was very significantly up-regulated in seeds but not in fruits and husks, suggesting that ZaMBF1c mainly participates in reproductive organ development. The confirmed results were that ZaMBF1c was significantly up-regulated in the initiation stage of floral sex differentiation, but extremely down-regulated in the subsequent floral differentiation and development process. Interestingly, we not only detected the target genes associated with the stress response in the sequence of ZaMBF1c, but also identified some target genes related to floral differentiation in the present study (Fig. 8). Thus, we suggest that ZaMBF1c could be triggered to improve plant stress tolerance and simultaneously induce the transition from vegetative growth to reproductive growth by activating LFY, Dof, and other MADS-box genes, when the plants suffer environmental stress. This regulatory mechanism could contribute to preserving their genotypes and rescuing the plants from disadvantageous conditions.
To sum up, our study carried out a genome-wide investigation of the MBF1 family in six representative species. A series of bioinformatics analyses were performed in the present study, and the expression patterns of ZaMBF1 genes in response to various external abiotic stresses and phytohormone inductions were investigated by qRT-PCR analysis. ZaMBF1a might be an essential member for plant development, and ZaMBF1b may serve redundant functions for ZaMBF1a in Z. armatum. Additionally, ZaMBF1c played an important role during the reproductive organ development process, and it also demonstrated strong responsiveness to environmental changes. However, our study identified only the MBF1 genes to provide comprehensive bioinformatics results based on genome-wide and transcriptome profiles. Future studies should focus on generating transgenic Zanthoxylum lines to reveal the molecular functions of ZaMBF1 genes [22]. The proteins that interact with ZaMBF1 could be explored by yeast two hybridization, bimolecular fluorescence complementarity and co-immunoprecipitation assay technique [52]. Promoter identification could also be conducted to uncover the genetic information of ZaMBF1c related to floral meristem differentiation in Z. armatum. Moreover, the content of various antioxidant enzymes and morphological characteristics could also be investigated in the transgenic Zanthoxylum lines [21], or even RNA-seq and metabolomics are an effective approach to isolate vital genes interacting with ZaMBF1 [50]. All the above studies will help to reveal the molecular functions and regulatory mechanisms of ZaMBF1 genes, which will contribute to breeding high-quality varieties of Z. armatum with meaningful stress tolerance.

Conclusion
In this study, we identified MBF1 genes in six representative species and characterized their gene structures, phylogenetic relationships, motif compositions, chromosomal distributions, cis-acting elements, gene duplications, and collinearity analysis. All the species have at least two different MBF1 genes, and MBF1c is usually an essential member in most species. The gene replications occurred among three ZaMBF1 genes, and ZaMBF1c showed collinearity between Z. armatum and both P. trichocarpa and C. sinensis. A total of 255 cis-elements involved in abiotic stress and phytohormone pathways were identified in the promoter regions of six species, which may reflect the response pathways of the MBF1 family in each species. Moreover, the ERF binding sites were the most frequently targeted in the sequences of the ZaMBF1 family, and ZaMBF1c may play a role promoting the transition from vegetative to reproductive growth in Z. armatum. Furthermore, ZaMBF1a is an essential member in various organs in Z. armatum, and ZaMBF1c might be associated with the male and female floral initiation processes. In addition, all of the ZaMBF1 genes could be significantly induced by water-logging, cold stress, ethephon, methyl jasmonate, and salicylic acid treatments, especially ZaMBF1c. Our findings will boost functional genomics research on stress resistance genes in Zanthoxylum. Further studies could focus on confirming the genetic roles of these genes in breeding highquality varieties with stress resistance in Z. armatum.

Evolutionary relationships of MBF1 proteins
A total of 202 MBF1 proteins from 53 species were downloaded from the Phytozome database [69]. All the protein information of various species is shown in Additional file 1: Table S1, and they all have MBF1 and HTH conserved domains. The neighbour-joining phylogenetic analysis of all the MBF1 proteins was carried out using MEGA6.0 software with 1000 bootstrap replicates [70].

Chromosomal locations, gene duplication and collinearity
In the present study, the chromosomal positions of each MBF1 gene in six species were collected from their genome data. The location of each MBF1 gene was displayed by TBtools software [74]. MCScanX software was applied to analyse MBF1 gene duplication and collinearity events [75]. The threshold was set as follows: the E-value was 1e-10, and the alignment number was 5.

Cis-regulatory elements and protein 3D structure analysis
The sequence of 2000 bp upstream from the transcription starting site of each MBF1 gene was isolated by Seqkit software (https:// github. com/ shenw ei356/ seqkit), which was used to conduct cis-regulatory element analysis as promoter region. The cis-regulatory elements were investigated by the Plant-CARE database (http:// bioin forma tics. psb. ugent. be/ webto ols/ plant care/ html/) [76]. All cis-regulatory elements associated with abiotic stress and phytohormones were identified and displayed by TBtools software [74]. Moreover, the tertiary structures of the ZaMBF1 protein were investigated by the SWISS-MODEL server (https:// swiss model. expasy. org/) [77].

Transcription factor binding site analysis of the ZaMBF1 family
To further clarify the transcription factor (TF) binding sites in the ZaMBF1 family, the target genes were scanned by the Plant Transcriptional Regulatory Map website (http:// plant regmap. gao-lab. org/) with the Binding site prediction tool. Due to the lack of a genomic dataset in this software, the homologous species, C. sinensis, was selected to perform this project.

Transcriptomic data analysis of the ZaMBF1 expression patterns
To explore the expression patterns of ZaMBF1 in Z. armatum, the available transcriptomic data were collected from the NCBI database [34] and our studies [78]. A total of 26 different tissues were collected to analyse the ZaMBF1 expression profiles in Z. armatum. The young flower (YF), young leaf (YL), husks (HK), seeds (SD), terminal bud (TB), petiole (PT), and prickles (PK) were obtained from public transcriptome data (BioProject: PRJNA721257), and the root (RT), stem (SM), mature leaf (ML), fruit (FT) and leaf bud (LB) were obtained from our previous study (GEO number: GSE142491). The FPKM was used to estimate the significance among three ZaMBF1 genes in each sample by one-way ANOVA test, and the level of significance was set to p < 0.05. Meanwhile, eight samples were used to uncover the expression patterns of ZaMBF1 during male and female floral differentiation and development processes, including male stage 1 (M1), male stage 2 (M2), male stage 3 (M3), male stage 4 (M4), female stage 1 (F1), female stage 2 (F2), female stage 3 (F3), and female stage 4 (F4) (GEO number: GSE195749), and the significance of gene expression was calculated by DEseq2 software [79], all of the up-or down-regulation are illustrated by the first comparison component. Additionally, the six transcriptome profiles were also analysed after infecting the plant with leaf rust disease from our recent studies, including INCZ0, INCZ1, INCZ3, INCZ7, INCZ14, INCZ21. All of the transcriptome data were collected from three biological replicates. The expression of ZaMBF1 were displayed as heatmap by TBtools software [74].

Exogenous treatments and quantitative real-time PCR (qRT-PCR) analysis
One-year-old seedlings of Z. armatum were planted in a plastic pot as the experimental material in this study, which was located in Sichuan Agricultural University (30.70°N, 103.87°E), Chengdu City, Sichuan Province, China. The seedlings were treated with the different abiotic stress and phytohormone inductions, including water-logging, cold stress, ethephon, methyl jasmonic acid (MeJA), and salicylic acid (SA). Each treatment had at least three biological repeats. Because the seedlings died within approximately one week under flood conditions, similarly grown seedlings of Z. armatum (50-60 cm high with 35-45 leaf nodes) were selected for enough irrigation with water to submerge the soil for 0, 6 h, 12 h, 1 d, 2d, 3d, and 6 d as the water-logging treatment in this study. Meanwhile, the 5-6 °C is the lowest temperature for the survival of Z. armatum in the wild, so similarly grown seedlings were grown at 6 °C for 0, 0.5, 1, 2, 3, 6, 9, 12, and 24 h as the cold stress treatment. For phytohormone analysis, similarly grown seedlings of Z. armatum were treated with 2.5% v/v ethephon (40% ingredient), 100 M MeJA or 100 M SA (Sigma, Santa Clara, CA, USA) solution for 0, 3, 6, 9, 12, and 24 h as described in previous studies [1,42]. At selected time points of various treatments, similarly sized leaves were collected and flash frozen in liquid nitrogen, and deposited in an ultra-low temperature freezer (Thermo, Waltham, MA, USA).
The total RNA of each sample was extracted using a FastPure Plant Total RNA Isolation Kit (Vazyme, Nanjing, China) according to the manufacturer's protocol, and cDNA was synthesized using a PrimeScript ® II First