Genome-wide analysis of the Brachypodium distachyon (L.) P. Beauv. Hsp90 gene family reveals molecular evolution and expression profiling under drought and salt stresses

The structure, evolution, and function of heat shock proteins 90 (Hsp90s) have been investigated in great detail in fungi and animals. However, studies on the Hsp90 genes in plants are generally limited. Brachypodium distachyon (L.) P. Beauv., as a model plant for cereal crops, has become a potential biofuel grass. During its long evolution, the Hsp90 gene family in Brachypodium has developed some strategies to cope with adverse environments. How the Hsp90 gene family in Brachypodium evolved in different plant lineages and what its role is in plant responses to drought and salt stresses remains to be elucidated. We used a set of different bioinformatics tools to identify 94 Hsp90 genes from 10 species representing four plant lineages and classified into three subgroups. Eight BdHsp90 genes were detected from B. distachyon. The number of exon-intron structures differed in each subgroup, and the motif analysis revealed that these genes were relatively conservative in each group. The fragments duplication and tandem duplication, which are the prime powers for functional diversity, generally occurred during the duplication of the whole plant genome. Transcriptional analysis of the BdHsp90 genes under salt and drought stress conditions indicated that the expression of these genes was delayed or increased at different stress time points; The expression was more affected in that of Bradi3g39630, Bradi4g06370, and Bradi1g30130. Our findings suggest the involvement of BdHsp90s in plant abiotic stress response, and further consolidate our views on the stress response mechanism of Hsp90 in general.


Introduction
Plants are often exposed to various abiotic stresses, including drought, salinity, light intensity, heat shock, chilling, and chemical pollutants. These stresses act simultaneously on plants, causing cell injury and inducing many responses, such as osmotic and oxidative stress responses [1]. Various land plants have developed specific responses or tolerance mechanisms to cope with adverse environments. Among them, many stress proteins, such as heat shock protein (Hsp) chaperones, are induced to guard cells against these harmful stimuli [2]. PLOS  The Hsps are classified into five categories based on their molecular weight, Hsp100 (e.g., 100 kDa), Hsp90, Hsp70, Hsp60, and small Hsp (sHsp). The 90-kDa heat shock protein (Hsp90) is an ATP-dependent molecular chaperone, with a highly conserved sequence from bacteria to higher eukaryotes and homologs in different organisms [3]. The core structure of Hsp90 contains the N-terminal ATP binding domain, the middle domain, and the C-terminal dimerization domain [3]. There is a notable feature of Hsp90s, which is a long, highly charged linker domain between the N-terminal and the middle domain in eukaryotic organisms [4].
Hsp90s are involved in regulating and maintaining the conformation of a variety of proteins, as well as in protecting normal cells from stress stimuli [5]. In fungi and in animals, Hsp90s play extensive roles in stress signal transduction, such as the folding of steroid hormone receptors, protein kinases, and transcription factors, as well as the activation of a substrate to initiate stress signal transduction [6][7][8]. Recent studies on Hsp90s in plants have mostly focused on evolutionary analysis and physiological functions [9][10][11]. A number of Hsp90 genes have been identified in many plants, and strongly up-regulated by temperature, drought, salinity, and heavy metal stresses [12][13][14]. A recent proteomics and phosphoproteomics analysis in plants revealed that a number of Hsp90 proteins are both present under drought and salinity stress and are probably involved in signal transduction during the response to stress [15][16][17].
Although the Hsp90 genes from plants have been known for more than a decade, our understandings of the stress response mechanisms of plant Hsp90s, their roles as molecular chaperones, and their molecular interactions with other clients and co-chaperones is limited [18]. The Hsp90 system in plants differs from that in animals in that it has an additional subcellular compartment, the plastid; Rapid molecular responses are other character subjected to sudden environmental changes [19]. It is necessary to better understand the proteins that are involved in signal transduction and other stress processes and how they activated, processed, and trafficked within plant cells [20]. The study, therefore, focused on the Hsp90 protein family in Brachypodium distachyon (L.) P.Beauv. to provide a comprehensive sequence-based understanding of the different family members in plants and to highlight the stress responses under drought and salt conditions. The similarities and differences between Hsp90 proteins of different plant origin were explored. The findings of this study provide a basis for functional analysis of Hsp90 in plants.

Sequence retrieval and identification
Hsp90 gene families were identified from 10 species representing four plant lineage from unicellular green algae to multicellular plants. The first search was performed using "Hsp90" as a keyword in the Phytozome v12.1 (https://phytozome.jgi.doe.gov/pz/portal.html) database. Seven Arabidopsis thaliana Hsp90 genes were first searched and then used as a query in the BLAST against phytozome v12.1. E value of candidate sequences was below 1E-10, and redundant parts were excluded manually. Sequences were collected from the following four major plant lineages: the unicellular green algae Chlamydomonas reinhardtii; the moss Physcomitrella patens; the monocotyledonous angiosperms B. distachyon, Oryza sativa (rice), Triticum aestivum (wheat) and Zea mays (maize); and the dicotyledonous angiosperms A. thaliana (thale cress), Glycine max (soybean), Medicago truncatula (legume), and Gossypium raimondii (cotton). Candidate sequences were further confirmed by Pfam (http://pfam.xfam.org/) [20] and checked by SMART (http://smart.embl-heidelberg.de/) [21,22]. Finally, we obtained all protein and corresponding coding sequences (CDS) and genomic sequences of the Hsp90 genes.

Gene chromosomal location and phylogenetic analysis
Locations of 94 Hsp90 genes were mapped by the MapInspect program and further modified manually. Phylogenetic trees were executed by Bayesian inference using the Markov Chain Monte Carlo (MCMC) method [23]. Initially, multiple sequence alignments were performed for full proteins based on the MUSCLE program (http://www.ebi.ac.uk/Tools/msa/muscle/) [24,25]. Bayesian inference phylogenyetic construction was performed by MrBayes v 3.2 using General Time Reversible (GTR) model with Γ distributed rates (gamma-distributed rate variation) [26,27]. The set conditions of bayesian analysis are mcmc ngen = 8×10 5 and samplefreq = 100. As the average standard deviation was below 0.01, the program was terminated. After discarding the burn-in samples, the remaining data were used to generate a Bayesian tree, which was shown by using FigTree v1.4.2.

Dating the duplication events
Tandem duplications were involved in multiple members of this family within the same or neighboring intergenic regions, and clustered these genes together with a maximum of 10 extra genes [32]. Segmental duplications of each Hsp90 gene were queried in the Plant Genome Duplication Database (PGDD, http://chibba,agtec.uga.edu/duplication/). To calculate the occurrence of segmentally duplicated genes, the Ks value was searched in PGDD under the following conditions: Ks > 0, Ks 1, and anchor number was set to !3 between the same species. Based on synonymous substitutions per year (λ), the number of segmentally duplicated genes was 6.5 × 10 −9 for B. distachyon [33], 6.5 × 10 −9 for rice [34], 6.1 × 10 −9 for soybean [33], 1.5 × 10 −8 for A. thaliana [35], and 1.5 × 10 −8 for cotton [36]. The approximate age (T) of duplication events of Hsp90 gene pairs was then calculated using the equation T = Ks/2λ [37].

Plant materials and stress treatments
The uniform seeds from the diploid inbred line of Bd21 were surface sterilized using 75% alcohol and 15% sodium hypochlorite, and then rinsed three times in sterile water. Subsequncely, the seeds were submerged in water (26˚C) in complete darkness for 3 days. On the fourth day, seedlings were transferred to plastic pots containing full-strength Hoagland solution. The conditions were set at 16/8h (light/dark) photocycle, 28/26˚C (day/night) and 70% relative humidity. Seedlings with three leaves were set in the following conditions: salinity stress (200 mM NaCl) and moderate drought stress [20% (w/v)]polyethylene glycol 6000 (PEG 6000). Leaf and root samples of control seedlings were harvested at 0h. The corresponding samples of treated seedlings were harvested at 12, 24, and 48h. Each sample was collected from 20 plants, with three replicates. All samples were immediately stored at -80˚C until used.

Total mRNA extraction and qRT-PCR analysis
Total RNA was isolated from frozen samples using TRIzol Reagent (Invitrogen) according to the manufacturer's instructions. Genomic DNA removal and cDNA synthesis were operated by using a PrimeScript1RT reagent Kit with gDNA Eraser (TaKaRa). Gene-specific primers of each Hsp90 gene in B. distachyon were designed using the on-line tool Primer3Plus (www. bioinformatics.nl/cgi-bin/primer3plus/primer3plus.cgi) [38]. The primers were examined by blasting primer sequences in the NCBI database (http://www.ncbi.nlm.nih.gov/tools/primerblast /index.cgi?LINK_LOC=BlastHome), and all primers were specifically in accordance with the respective sequence of its targeted gene. The primer sequences for the qRT-PCR assays are listed in S1 Table. Transcription levels of each Hsp90 gene in B. distachyon were quantified by a CFX96 Real-Time PCR Detection System (Bio-Rad) using the intercalating dye SYBR-green and following the 2(-Delta Delta C(T)) method [39]. The B. distachyon constitutively expressed Ubiquitin gene was used as a reference for normalization [40]. qRT-PCR was performed in a 20 μL volume reaction system containing 10μL 2× SYBR1Premix ExTaq™ (TaKaRa), 2μL 10-fold diluted cDNA, 0.15μL of each gene-specific primer and 7.7μL ddH 2 O. The PCR conditions were as follows: 95˚C for 3 min, 40 cycles at 95˚C for 20s, 61˚C for 15s and 72˚C for 10s. Triplicates for each PCR and three biological replicates were performed for each gene. The results for the qRT-PCR assays are listed in S1 Table. The qRT-PCR efficiency was determined by five serial ten-fold dilutions of cDNA. Statistical analyses were conducted using independent Student's t tests with SPSS statistics software (version 17.0). The hierarchical clustering results were performed using Cluster 3.0 and TreeView softwares.

Identification of Hsp90 genes in B. distachyon and in the other nine representative plant species
To clarify the origin and evolution of the Hsp90 genes in plants, we identified 10 species representing four major plant lineages. A total of 94 Hsp90 genes were obtained (S2 Table). Each gene contained conserved HATPase_c and HSP90 domain (S3 Table), and PF02518 and PF00183 families by Pfam (S3 Table). Among all Hsp90 genes, three were identified in the unicellular green algae, 11 in the moss, 42 in the monocotyledonous angiosperms and 38 in the dicotyledonous angiosperms (S1 Fig). Furthermore, the Hsp90 genes identified in the 10 plant species encode proteins ranging from 278 to 1054 amino acids (aa) in length, but the majority are approximately 695 aa. The predicted isoelectric points range between 4.79 and 6.28, and Molecular weight range between 40.6 and 123.6 kDa. The detailed information of these identified Hsp90 genes were listed in S2 Table. Phylogenetic relationships of the Hsp90 gene family The Bayesian inference, a standard approach for the estimation of branch support, was executed by MrBayes as posterior probabilities within the time run [41]. To deeply explore the evolutionary relationships of Hsp90 genes within various plant species, the full amino acid sequences of 94 identified proteins were to execute multiple sequence alignment using the MUSCLE program. An unrooted phylogenetic tree was constructed using the Markov Chain Monte Carlo (MCMC) method based on Bayesian inference (Fig 1). Finally, the resulting tree topology classified these plant genes into three subgroups. Interestingly, the genes in C. reinhardtii were lined the first in each group. Moreover, the distribution of Hsp90 genes from different individual species had slight differences and unique patterns in each group (Fig 1). Of the 94 plant Hsp90 genes, 53 is in Group I, 14 in Group II and 27 in Group III. From an evolutionary perspective, the resulting phylogram (S2 Fig) is in agreement with the timeline of plant evolution: unicellular green alga occurred first and was followed by the moss, monocotyledonous angiosperms, and dicotyledonous angiosperms. Within A. thaliana, four genes were in Group I, one in Group II, and three in Group III. Almost the same distribution pattern was observed for the Hsp90 gene family members from A. thaliana (Fig 1). Overall, the distribution characters of the Hsp90 gene family were similar among the 10 plant species.

The analysis of structural characters of the Hsp90 gene family
To investigate the possible molecular mechanisms underlying the expansion of the Hsp90 gene family, the exon-intron structure in all chosen plant lineages was investigated using the online GSDS. The relative length of introns and the corresponding exon sequences within individual Hsp90 gene paralogs are listed on the neighbor-joining phylogenetic tree (Fig 2). Our results revealed that in the same group, the gene members had similar exon-intron structures. Although the number of introns in individual Hsp90 genes ranged from 1 to 21, there were no significant differences in the number of introns between individual genes within the same group, except for those in C. reinhardtii (Fig 2). Group I, the largest family comprising 56% of the genes in the subfamily, had 1-3 introns. Group II, the smallest family, which contained 16% of the genes in the subfamily, had 14-16 introns. The number of introns in Group III, which included 28% of the genes, ranged between 17 and 21. In contrast, the number of introns in genes of the unicellular green alga (Cr) did not follow these patterns, and were 7, 11, and 8 introns in Group I, Group II, and Group III, respectively (Fig 2).
To explore the possible function for the structural evolution of the Hsp90 gene family, the conserved protein motifs were identified in these genes using MEME. Ten conserved motifs were detected in the 94 Hsp90 genes (Fig 3).The structure of individual Hsp90 gene paralogs is shown in Fig 3. The number of motifs contained in the individual Hsp90 gene was primarily 7-10. Similarly with the patterns of introns observed in each gene, the number of conserved  (Fig 4). However, the gene numbers on each chromosome were directly proportional to the length of the corresponding chromosome, suggesting that the Hsp90 genes in multiple plant species have no obvious chromosomal preferences. Furthermore, the genes were clustered in certain chromosomal regions or dispersed individually in other locations, thus corroborating the mapping results reported in other plants [42,43].
There are the three dominant evolutionary events in gene duplication, including segmental duplication, tandem duplication, and transposition (retroposition and replicative transposition) [44], among which segmental duplication and tandem duplication commonly occur in plants based on polyploidy and unequal crossing-over, respectively [40]. To gain some insights into the expansion pattern of the Hsp90 genes, we identified tandem duplicates and segmental duplication clusters based on the gene locus and searched the PGDD [45,46] to locate  Table), suggesting that segmental duplication events played an important role in the evolution of the Hsp90 superfamily in those species. The details were as follows: one pair of genes were identified as segmental duplications (Bd3g38897 and Bd4g32941) in B. distachyon, two pairs of segmentally duplicated genes were identified in rice (Os08g38086 and Os09g29840, Os08g39140 and Os09g30412), and nine pairs of segmental duplications were found in soybean. The precise information of gene duplication events in these species are listed in Table 1.

Subcellular localization and evolution relationships of the Hsp90 gene family
Subcellular localization is a key characteristic of protein functional research. It can take part in the cell activity and function efficiently in correct subcellular location [47]. In our study, based on three bioinformatics tools, including TargetP1.1 [48], WoLF PSORT and Predotar v. 1.03 [49], subcellular localization was conducted in 94 Hsp90 proteins (Fig 5 and S5 Table). Interestingly, based on the above analysis, the results showed that these Hsp90s in Group I are largely distributed in the nuleus (20%) and cytoplasm (80%). Moreover, all of their ends contained the conserved MEEVD sequence. The Hsp90s in Group II were mostly districted in the ER (78%), and all of their ends contained KDEL sequence, which is a specific retention sequence in the endoplasmic reticulum [49]. The Hsp90s in GroupIII were mostly districted to the chloroplasts (40%) and mitochondria (45%), and all of their ends contained extra sequences differed from Group I and Group II. The results were in accordance with previous results [19].

Expression profiles of Hsp90 genes in root and leaf tissues under osmotic stress
We tested the response mechanism in two different stress conditions by firstly constructing eight B. distachyon genes primers using the on-line tool Primer3Plus [37]. Unfortunately, seven of eight genes were constructed of their primers and standard curve (S1 Table, S3 Fig).
The expression profiles of the seven Hsp90 genes in the root and leaf tissues of B. distachyon were analyzed using real time PCR (Fig 6, Fig 7). Under normal conditions, the expression level of Hsp90 genes in the root is no higher than that in the leaf, except for Bradi1g30130 (3.5-fold) and Bradi3g3889700 (3-fold) (Fig 6). In addition, it is worth noting that the expression level of Bradi3g3889700 in the root is lower than the expression level of the other six genes, while the expression level of Bradi1g30130 was higher than that of the other six genes (Fig 6). These results indicate that the transcriptional profiles of the Hsp90 genes in the B. distachyon have specific tissue expression.
Drought and salt stresses were two common types of abiotic stresses. The common response that they induced was that of osmotic stress, the same as heat stress. In our study, the expression pattern of four Hsp90 genes in the leaf tissue showed first down-regulation and then up-regulation under salt stresses, when compared with the control condition, while Bra-di1g39590, Bradi5g02307 and Bradi3g39620generally showed down-regulation in most stress time points (Fig 6). Moreover, under drought stress, the expression pattern of three Hsp90 genes, including Bradi5g02307, Bradi3g39620, and Bradi3g39590 totally showed significant down-regulation, while the other four genes showed significant up-regulation in comparison to the control condition, especially in 24h or 48h points (Fig 6).
In root tissue, compared with control condition, Bradi3g3889700 showed almost no expression changes, and Bradi1g30130 showed significant up-regulation under salt stress, while the other five Hsp90 genes showed a down-regulation pattern (Fig 6). Furthermore, under drought stress, Bradi3g39620, Bradi3g39630 and Bradi3g3889700 showed no expression changes, compared with the control. The other four genes showed obvious up-regulation pattern, especially in Bradi5g02307, Bradi1g30130 and Bradi4g06370, which might be involved in stress responses. From the clustering analysis, and in contrast to the express pattern of other genes, Bra-di1g30130 showed significant up-regulation pattern under two stresses, especially in the 48h stress point (Fig 7), suggesting that it might participate in stress tolerance. Bradi3g3889700, as Evolution and expression profiling in Brachypodium distachyon (L.) Hsp90 gene family under abiotic stresses Evolution and expression profiling in Brachypodium distachyon (L.) Hsp90 gene family under abiotic stresses the leaf-specific stress response gene, showed no change in root tissue under the two stresses, which is consistent with its subcellular localization (chloroplast).

Discussion
The Hsp90 genes family, is large and has been identified in almost studied eukaryotic species. To date, the necessary roles of the Hsp90 genes in higher plants have been investigated in response to biotic and abiotic stresses [10,50]. In the current study, we explored the phylogenetic relationships and characters of the 94 Hsp90 genes from 10 species using a variety of bioinformatics tools. To study the drought response profiles of the plant Hsp90 genes, the eight Hsp90 genes in B. distachyon were all checked for the expression analysis under drought and salt stresses using the real-time PCR method combined with the multiple sequence alignment. The results showed that the evolution of plant Hsp90 genes is relatively conserved from unicellular green algae to dicotyledonous, while the expression profile of the Hsp90 genes in different tissues of B.distachyon is divergent in response to drought and salt stresses.

Evolution profile and duplication events of the Hsp90 gene family
In this study, 94 Hsp90 genes were selected from 10 species representing the four major plant lineages. From these results, we can see that Hsp90 gene members in 10 plants species are different, while the number of the Hsp90 genes in higher plants is more than that of lower plants.
Using MCMC method based on Bayesian inference, phylogenetic trees were supported by high bootstrap values, and showed that these Hsp90 genes belong to three major subgroups. These findings are in accordance with those of previous studies [9,11]. Interestingly, three Cr occurred first in each group, suggesting that the characters of the Hsp90 gene family were conserved throughout the evolutionary history. The development of the intron was an important process in genomic evolution, and an adaptive measure for speciation evolution [51]. Interestingly, our results showed that genes in each subgroup possess similar exon-intron structures with similar numbers of introns, while the genes in different subgroups manifested slightly different exon-intron structures. In addition, the pattern also can be found in the individual species. Using subcellular location analysis, Hsp90 genes were almost all distributed in nuleus/ cytoplasm in Group I. For Group II, the Hsp90 genes were almost all located in the ER. Compared with above two groups, Hsp90 genes in Group III were almost all located in the chloroplasts/mitochondria. Based on the exon-intron structure and subcellular location, we speculate that our evolution classification in plant Hsp90s is creditable. Gene duplication events play critical roles in the gene evolution, and have developed into a major mechanism for the establishment of new functions [51]. In this study, 23.4% (22 of 94) Hsp90 genes were considered to be derived from segmental duplication events, and these segmental duplication gene pairs occurred in the same subfamilies, suggesting that they can show new evolutionary functions after duplication. The estimated time of segmental duplication events ranged from 8.1 to 61.8 Mya, demonstrating that these Hsp90 duplicated genes may experience the split in whole genome duplication for each species. For B. distachyon, there was only one pair segmental genes found (BdGF14e/ BdGF14b), and it was estimated to be originated at about 61.8 Mya. As Table 1 was seen [52][53][54][55], the divergence time of Poaceae was approximately 50-70 Mya. The whole genome duplication of cotton, soybean and rice was approximately 13-20, 5-13(59) and 53-94Mya, respectively, suggesting that the duplication of B. distachyon Hsp90 gene family occurred before the divergence of the genome from cotton, and approached to point of genome its divergence from rice and soybean. The expression profiling and possible response mechanism of the B. distachyon Hsp90 genes in response to osmotic stresses It is clear that the plant Hsp90 gene family has divergent functions, and that they participate in a wide range of biological processes, especially occurring in response to abiotic stresses [56]. In addition, previous investigations suggest that plant Hsp90 genes have different response mechanisms in response to different stresses. Most of the Hsp90 genes perform their functions through participating in biological signaling pathways, such as the abscisic acid (ABA) signaling pathway and the ER stress signaling pathway [56,57].
The B.distachyon cytoplasm Hsp90 genes in response to osmotic stresses through the ABA signaling pathway The hormone ABA is a very important phytohormone that play important roles not only in seed germination and development, but also in response to a wide range of abiotic stresses, such as exogenous ABA, salinity, drought, temperature stresses [58]. The response mechanism of plants against many abiotic environmental stresses is through the ABA-dependent signaling pathway. [59][60][61].
Based on previous experiment results, Jacob P. et al, (2017) [62] provides a possible stress response mechanism of Hsp90 mediated with ABA signaling. In normal conditions, HsfA1s are bound by Hsp90/70 and their co-chaperones ABI5 (ABA insensitive five) and DREB2A (dehydration responsive element binding protein 2A) were linked through the E3 ligase DRIP1/2, and degraded by 26 proteasomes [63,64]. Upon stress application, the high number of misfolded and denatured proteins triggers the recruitment of Hsp90/70 to its client and releases the HSFA1s [62]. Especially, heat and drought stresses will lead to phosphorylated E3 ligase degradation due to SnRK2 activation, and DREB2A and ABI5 accumulation [64][65][66], which can then enter the nucleus, cooperatively or separately bind their target DNA, and activate the expression of the target genes [62]. In our study, based on cellular location and evolution classification, four Hsp90 genes in the B. distachyon, including Bradi3g39590, Bra-di3g39620, Bradi3g39630, and Bradi5g02307, were distributed in the cytoplasm (Fig 1, Fig 4). Of the four genes, the expression pattern of Bradi3g39620 and Bradi5g02307 generally showed an up-regulated trend in response to osmotic stress, while the other two genes showed downregulated trends, which might be involved in plant development under stress conditions. Moreover, the evolution distribution of Bradi3g39620 was similar with that of At5g56100 (AtHsp90.2) and relatively homologous (72%), and the evolution distribution of Bradi5g02307 was similar to that of At5g52640 (AtHsp90.1), suggesting that the up-regulation of the two cytoplasm Hsp90 genes may participate in the osmotic stress response of ABA signaling through freeing of the relative transcript factors, which activate downstream gene expression in stress response and tolerance. Moreover, OsHSP90-2 and OsHSP90-4 were also found upregulated to be drought, cold, heat and salt stresses [67], which is consistent with our results. However, in a report that constitutively reduced cytosolic Hsp90 by using siRNA method, the expression of genes generally responsible for stress responses such as ABA stimulus, drought stress, and jasmonic acid response were enhanced [68]. These results showed that some cytoplasm Hsp90 genes might be the stress response "monitor", by positively and passively fining and regulating to take part in cell protection and control against adverse environments.
The ER B.distachyon Hsp90 genes response to osmotic stresses through the ubiquitin proteasome system (UPS) Abiotic stresses usually cause protein function disorder in the ER, including protein aggregation, misfolding, and denatureation, and activate the unfolded protein response (UPR) [69].
Under normal conditions, ER organelle is the origin of initial protein synthesis. Hsp90 is required to play important roles in correct protein folding. When ER was subjected to diverse environments, such as drought, salt, and heat, the UPR was triggered. In higher eukaryotes, IRE-1(a approximately 100kDa type I transmembrane protein) contains Kinase and endoribonuclease domains, and is generally considered to be the most important signaling pathway from the ER [69]. Hsp90/IRE1 interaction might potentially cause the IRE1 response and affect the death of the cell during ER stress [69]. In our study, based on cellular location and evolution classification, Bradi1g30130 in the B. distachyon was distributed in the ER (Figs 1 and 4). The evolutionary distribution of Bradi1g30130 was similar to that of At4g24190 (AtHsp90.7) and homologous (73%). The expression pattern of Bradi1g30130 generally showed a significant up-regulated trend in response to drought and salt stresses, especially at 48h time points in the leaf tissue. AtHsp90.7 was previously shown to be more sensitive than AtHsp90.2 in response to drought and salt stresses [56]. Excessive Hsp90 in ER may reduce the shift and targeting of plasma membrane or vacuole membrane ion transporters to cope with salt stress by relieving cytosolic Na + [70]. Additionally, excessive Hsp90, especially in the chloroplast or ER, might trigger a general homeostasis itself, either by changing the organelle import/export system or the native protein homeostasis to enhance normal salt and osmotic stress tolerances [56]. These results suggest that the Hsp90 level in plant is critical for homeostasis of stress response and/or tolerance proteins.

Conclusions
Based on the evolution analysis and molecular structure from 10 species representing four major plant lineages, the Hsp90 gene family was found to be relatively conservative, and showed function diversity in the cellular location, gene duplication and evolution distribution between each group. Under drought and salt stresses, BdHsp90 gene family in root and leaf tissue showed different expression pattern in each group, especially in Bradi1g30130 and Bra-di3g3889700, suggesting that they might be involved in different stress responses. Given the above mentioned findings and those published in the literatures, we suggest that Hsp90 gene family plays roles in cytoplasm ABA signaling, ER stress protection and plant development under stress. Despite the detailed mechanism of BdHsp90 involvement in stress not being clearly understood, their function characterization will provide new insights into stressresponsive pathways.