Genome-Wide Identification and Characterization of RdHSP Genes Related to High Temperature in Rhododendron delavayi

Abstract

Heat shock proteins (HSPs) are molecular chaperones that play essential roles in plant development and in response to various environmental stresses. Understanding R. delavayi HSP genes is of great importance since R. delavayi is severely affected by heat stress. In the present study, a total of 76 RdHSP genes were identified in the R. delavayi genome, which were divided into five subfamilies based on molecular weight and domain composition. Analyses of the chromosome distribution, gene structure, and conserved motif of the RdHSP family genes were conducted using bioinformatics analysis methods. Gene duplication analysis showed that 15 and 8 RdHSP genes were obtained and retained from the WGD/segmental duplication and tandem duplication, respectively. Cis-element analysis revealed the importance of RdHSP genes in plant adaptations to the environment. Moreover, the expression patterns of RdHSP family genes were investigated in R. delavayi treated with high temperature based on our RNA-seq data, which were further verified by qRT-PCR. Further analysis revealed that nine candidate genes, including six RdHSP20 subfamily genes (RdHSP20.4, RdHSP20.8, RdHSP20.6, RdHSP20.3, RdHSP20.10, and RdHSP20.15) and three RdHSP70 subfamily genes (RdHSP70.15, RdHSP70.21, and RdHSP70.16), might be involved in enhancing the heat stress tolerance. The subcellular localization of two candidate RdHSP genes (RdHSP20.8 and RdHSP20.6) showed that two candidate RdHSPs were expressed and function in the chloroplast and nucleus, respectively. These results provide a basis for the functional characterization of HSP genes and investigations on the molecular mechanisms of heat stress response in R. delavayi.

1. Introduction

During growth and development, the sessile nature of plants makes them vulnerable to various kinds of biotic (herbivore and pathogen attacks) and abiotic stresses (drought, flooding, salinity, low and high temperature) [1,2]. The different stresses can result in cell disturbances and secondary stresses, including membrane damage, protein denaturation, oxidative stress, and osmotic stress at a cellular level [3,4]. To withstand complex environments, plants have evolved complex mechanisms to respond and adapt to these stresses [5]. Among them, heat stress has become one of the main abiotic stresses threatening higher plant productivity and survival throughout the world [6]. The HSP family is one such gene family that is involved in tolerance against heat stress [7,8,9]. However, the molecular mechanisms of the HSP family genes involved in regulating heat stress response are still unclear, and need to be explored.

Heat shock proteins (HSPs) are molecular chaperones induced by heat stress, which are present in all living organisms, from bacteria to protists, plants, fungi, and animals [10,11,12,13]. In plants, the HSP family genes play essential roles in plant growth and development as well as in response to various environmental stresses (heavy metals, salinity, drought, low- and high-temperature stress, and biotic stress) [14,15,16,17]. The HSP gene family has been identified in many plants, such as Arabidopsis, soybean (Glycine max), foxtail millet (Setaria italica), tobacco (Nicotiana tabacum), and rice (Oryza sativa) [18,19,20,21,22]. The HSP gene family has been divided into five subfamilies based on their molecular weight and domain composition, including the HSP90 subfamily (heat shock protein 90), HSP100 subfamily (the Casein lytic proteinase/heat shock protein 100), HSP20 subfamily (small heat shock proteins), HSP60 (chaperonin proteins), and HSP70 (70-kDa-heat shock proteins) [18].

The roles of HSP genes in various stress responses have been reported in many plants [23]. The Arabidopsis HSP21, a plastidial small heat shock protein that interacts with the pTAC5 protein, is essential for chloroplast development in Arabidopsis under heat stress [24]. The HSP60 protein has been reported to be involved in the folding and aggregation of many proteins that are transported to organelles [25]. The Arabidopsis hsp60 mutants displayed severe defects in chloroplast and embryo development and caused cell death [26]. The overexpression of the SlHSP70-1 gene in transgenic tomato caused internode elongation in tomato [27]. In addition, HSP family genes also play important roles in response to various environmental stresses [28]. The overexpression of the rice OsMSR3 gene in Arabidopsis can enhance tolerance to cadmium and copper stresses [29]. Two HSP genes (CaHSP26 and CaHSP22.5) from sweet pepper (Capsicum annuum) can protect the PSII system by increasing oxidation resistance and photochemical activity after chilling stress [30]. The overexpression of the NtHSP gene in tobacco alleviates NaCl stress, and MeHSP90.9 increases cassava resistance to drought stress by regulating abscisic acid (ABA) and hy0drogen peroxide (H₂O₂) [31]. The maize ZmHSP16.9 gene, encoding a cytosolic class I sHSP, can enhance the heat tolerance in transgenic tobacco [32]. The HSP100/ClpB protein, located in the chloroplast, plays an active role in regulating the heat resistance of tomato (Solanum lycopersicum) [33]. In addition, HSP90 genes in barley play roles in regulating hypersensitive responses to stripe rust [34].

R. delavayi is an evergreen shrub with large scarlet flowers that make it highly attractive as an ornamental species [35]. However, heat stress causes considerable damages to R.delavayi, including leaf scorching, sunburns on leaves and stems, and leaf senescence and abscission [36]. Although the functions of HSP genes in some model plants have been well studied, the role of RdHSP genes in R. delavayi under heat stress remains largely unknown. This study conducted a comprehensive overview of the RdHSP gene family in R. delavayi, including a phylogenetic analysis, motif/gene structure analysis, chromosome location, gene duplication events, and subcellular localization. We also investigated the expression patterns of the RdHSP genes using our RNA-seq data and qRT-PCR to select candidate genes involved in heat stress. Our findings will be useful resources for future studies to unravel the functions of the RdHSP genes and will contribute to our understanding of the evolutionary history of the HSP genes in different species.

2. Materials and Methods

2.1. Identification of HSP Family Genes in R. delavayi

The whole genome sequence of six Rhododendron species, including Rhododendron delavayi, Rhododendron williamsianum, Rhododendron simsii, Rhododendron ovatum, Rhododendron henanense, and Rhododendron irroratum, was downloaded from the Ericaceae Genome Resource (TEGR, http://www.tegr.com.cn). The hidden Markov model (HMM) profiles of the HSP20 (PF00011), HSP60 (PF00118), HSP70 (PF00012), HSP90 (PF00183), and HSP100 (PF02861) domains were used as queries to search the genomes of six Rhododendron species. According to the physical localization, the redundant and short sequences (length < 100 aa) were removed. All candidate genes were further screened for the presence of conserved domains using the Pfam and Simple Modular Architecture Research Tool (SMART) databases [37]. Finally, candidate genes with the conserved core domain were used as RdHSP family genes.

2.2. Chromosomal Localization, Structure, Cis-Regulatory Elements Analysis

The locations of RdHSP genes were achieved from the R. delavayi genome annotation information in the TEGR database. The MapInspect software (http://mapinspect.software.informer.com/) was used to draw the physical location of RdHSP genes on chromosomes.

The Gene Structure Display Server (GSDS, version: 2.0) software was used to analyze the intron-exon structures of RdHSP genes. The Multiple EM for Motif Elicitation (MEME) website (http://meme-suite.org/tools/meme, accessed on 17 April 2024) was used to identify the conserved motifs of RdHSP family proteins [38].

The upstream 2000 bp sequence of the RdHSP genes was manually cut and submitted to the online PlantCARE software (v. 1.0) to predict the cis-regulatory elements. The identified promoter cis-regulatory elements were manually filtered and plotted using GSDS software.

The subcellular localization prediction of the RdHSP family protein was performed using the online WoLF PSORT software (https://wolfpsort.hgc.jp/).

2.3. Phylogenetic Analysis

An unrooted phylogenetic tree of the HSP family proteins from six Rhododendron species was constructed using the MEGA-X software with the default parameters. The Model Generators were used to estimate the amino acid substitution models for HSPs’ evolution while using the default parameters.

2.4. Synteny Analysis

Synteny analyses among six Rhododendron species were conducted locally based on the method described [39]. The extracted protein sequences were then aligned pairwise to identify conserved homologous gene pairs using Blastp (v2.7.1) with an E-value threshold of 1E-5. According to the results of homologous gene pairs, MCScanX software was used to identify the syntenic regions [39]. The duplication events of HSPs in six Rhododendron species were analyzed using MCScanX software [39].

2.5. Expression Patterns of RdHSP Genes in R. delavayi under Heat Stress

The expression patterns of RdHSPs in R. delavayi leaves under heat stress treatment were investigated using our RNA-seq data. The fragments per kilobase of transcript per million mapped reads (FPKM) was applied to estimate the gene expression. The Cluster 3.0 software was used to perform a hierarchical clustering analysis based on the log₂ (FPKM), which was visualized using Java TreeView (version:1.2.0).

2.6. Real-Time Polymerase Chain Reaction (qRT-PCR) Validation of RNAseq Data

To verify the expression patterns of the 12 candidate RsHSP genes in heat-treated R. delavayi, RT-qPCR was conducted on the Bio-Rad CFX96 Real-time PCR system using SYBR PrimeScript RT-PCR Kit II (Takara, Dalian, China) and gene specific primers (Table S1). The reference gene Rd18s rRNA was used as a reference for normalization in the qRT-PCR analysis [36]. The 2^−ΔΔCT method was applied to calculate the relative expression changes of selected RdHSP genes. Relative expression values were obtained from three biological repeats, and measurements were made for three technical repeats.

2.7. GO Enrichment Analysis

GOATOOLS software (version:1.2.3) was applied to conduct GO annotations for RdHSP genes. The functional enrichment analysis of RdHSP genes was performed using Fisher’s exact test. Moreover, Bonferroni correction was applied to minimize false positives, and functions were considered to be significantly enriched when their Bonferroni-corrected P-values (Padjust) were <0.05. The RdHSP genes involved in pathways were analyzed using the Kyoto Encyclopedia of Genes and Genomes pathway-mapping database (KEGG; https://www.kegg.jp/kegg/, accessed on 17 April 2024).

2.8. Subcellular Localization Analysis of Two Candidate RdHSP Genes

Total RNA extraction from the young leaves of R. delavayi was performed, which was reverse transcribed into cDNA using a Plant RNA Kit (TaKaRa Biotechnology Co., Ltd., Dalian, China) and the RT Reagent Kit (TaKaRa Biotechnology Co., Ltd., Dalian, China), respectively. The diluted cDNA (100 ng/μL⁻¹) was used as templates for cloning the full-length encoding sequence (CDS) of two candidate RdHSPs. The CDS without the stop codon of two RdHSP20 genes was ligated into a transient expression pCambia 2300 harboring the GFP expression cassette derived by the 35S promoter. Then, the constructed plasmid was transferred to Agrobacterium GV3101 through the conventional freezing–thawing method. Next, Agrobacterium GV3101 with empty pCambia 2300-GFP or pCambia 2300-RdHSP20s-GFP was injected into 5-week-old N. benthamiana leaves. After the infiltrated N. benthamiana was cultured in the dark for 24 h and then in low light for 12 h, the GFP signals were detected by an Echo Revolve fluorescence microscope (Revolve FL).

2.9. Protein Interaction Prediction

To explore the differential proteins interacting with the RdHSPs involved in regulating R. delavayi heat stress response, the RdHSPs associated with heat stress tolerance were identified based on phylogenetic analysis and expression levels, which were submitted to the STRING v10 database (https://string-db.org). Moreover, the construction of the PPI networks was performed based on the Arabidopsis active interaction sources, including experiments, co-expression, and databases. The cytoscape software (v. 3.8.1) was used to visualize the PPI networks. The strong interconnected regions were identified using the ClusterONE software with the parameter sets (minimum density = 0.01, minimum size = 2, and edge weights = combined score). The edges and nodes in the network represent interactions and proteins, respectively. The interactions between the RdHSPs involved in heat stress response and potential differential proteins in R. delavayi were screened by STRING.

2.10. Statistical Analysis

The experimental design of the article consists of three biological replicates and three technical replicates. Data were statistically analyzed using IBM^® SPSS^® Statistics 20 (IBM, Armonk, NY, USA). An analysis of variance (ANOVA) and mean separation were performed using a t-test or one-way ANOVA with the least significant difference (LSD) at p < 0.05.

3. Results

3.1. Identification and Chromosomal Distribution of the RdHSP Family Genes

In the present study, a total of 76 RdHSP genes were identified in the R. delavayi genome (Table S2). The RdHSP proteins ranged in size from 133 aa (RdHSP20.12) to 2558 aa (RdHSP90.7), with an average length of about 632 aa (Table S2). The calculated molecular weight of the predicted RdHSP proteins ranged from 15.01 kDa (RdHSP20.12) to 286.53 kDa (RdHSP90.7), and the isoelectric points of these proteins ranged from 4.7 (RdHSP70.8) to 9.18 (RdHSP20.13) (Table S2). The predicted ranges of the instability index, aliphatic index, and GRAVY of the RdHSP family proteins were 25.32–57.45, 67.78–113.06, and −0.799–0.096, respectively (Table S2). Based on the genome annotation, the identified 76 RdHSP genes were unevenly distributed on 13 R. delavayi chromosomes. As shown in Figure 1, chromosome 13 contained the largest number of RdHSP genes (13), while only two RdHSP genes were distributed on chromosome 2. The percentage of RdHSP genes per chromosome varied from 0.06% on chromosome 2 to 0.47% on chromosome 13. Interestingly, three tandem duplicated gene pairs, such as RdHSP70.7 and RdHSP70.8, RdHSP70.15 and RdHSP70.16, as well as RdHSP70.28 and RdHSP70.29, were present on chromosomes 03, 07, and 13, respectively.

3.2. Phylogenetic Analysis of the HSP Gene Family among Six Rhododendron Species

The unrooted phylogenetic trees of the HSP subfamily genes from six Rhododendron species were constructed using the MEGA 10.0 software with the default parameters setting neighbor-joining method (Figure 2). The gene information for HSP genes from six Rhododendron species was collected and stored in Table S3. Our result showed that the RdHSP family genes of R. delavayi were divided into five subfamilies based on domain composition: HSP20, HSP60, HSP70, HSP90, and HSP100 (Figure S1 and Figure 3a). Moreover, each RdHSP subfamily was divided into different groups based on the predicted subcellular localization and phylogenetic tree topology (Figure 2 and Table S2). The RdHSP100 subfamily was classified into two groups: group I (cytoplasm-localized, one gene) and group II (chloroplast-localized, three genes) (Figure 2a and Table S2). The RdHSP90 subfamily was divided into six groups, including group I (cytoplasm-localized, two genes), group II (mitochondrion-localized, two genes), group III (chloroplast-localized, one gene), group IV (endoplasmic reticulum-localized, one gene), group V (mitochondrion-localized, one gene), and group VI (cytoplasm-localized, one gene) (Figure 2b and Table S2). The RdHSP20 subfamily was classified into seven groups: group I (cytoplasm-localized, three genes), group II (chloroplast-localized, one gene), group III (cytoplasm-localized, one gene), group IV (cytoplasm-localized, one gene), group V (plasma membrane-localized, one gene), group VI (cytoplasm-localized, two genes), and group VII (cytoplasm-localized, six genes) (Figure 2c and Table S2). The RdHSP60 subfamily was classified into 10 groups: group I (cytoplasm-localized, 3 genes), group II (chloroplast-localized, 2 genes), group III (cytoplasm-localized, 5 genes), group IV (cytoplasm-localized, 1 gene), group V (endoplasmic reticulum-localized, 1 gene), group VI (cytoplasm-localized, 1 gene), group VII (cytoplasm-localized, 1 gene), group VIII (cytoplasm-localized, 2 genes), group IX (chloroplast-localized, 1 gene), and group X (nucleus-localized, 2 genes) (Figure 2d and Table S2). The RdHSP70 subfamily was classified into 5 groups: group I (chloroplast-localized, 21 genes), group II (mitochondrion-localized, 2 genes), group III (chloroplast-localized, 3 genes), group IV (mitochondrion-localized, 2 genes), and group V (endoplasmic reticulum-localized, 2 genes) (Figure 2e and Table S2).

3.3. Conserved Domain, Protein Motif, and Gene Structure Analyses of the RdHSP Family Genes

To better analyze the conserved domain, motif composition, and genetic structural diversity of RdHSP genes, the distribution of the conserved domain in the RdHSP family proteins was explored (Figure 3a). Our result showed that the RdHSP20, RdHSP60, RdHSP70, RdHSP90, and RdHSP100 subfamily proteins contained the conserved HSP20/α-crystallin domain (PF00011), the GroEL/chaperonin-like (PF00118), CCT/dnaK/HSP70 domain (PF00012), HSP90 (PF00183), and Clp (PF02861), respectively. Furthermore, the motif composition of RdHSP family proteins were investigated using the MEME suite. A total of 10 putative conserved motifs were identified (motifs 1–10, Figure 3b). Motif 1, motif 2, and motif 3 were distributed across all members of the RdHSP20 subfamily. Motif 7 and motif 5 were present in most members of the HSP20 subfamily. All members of the RdHSP60 subfamily contained motif 1, and 94.7% of RdHSP60 subfamily proteins contained motif 2, motif 3, and motif 4. The RdHSP70 subfamily proteins contained motif 1, motif 2, motif 5, motif 8, and motif 9, and the RdHSP90 subfamily proteins contained motif 1, motif 2, and motif 3. The motifs 1–5 and motifs 7–10 were distributed across the RdHSP100 subfamily proteins. As expected, most members in the RdHSP60, RdHSP90, and RdHSP100 subfamilies had similar motif compositions, suggesting functional similarities among the HSP proteins within the same subfamily.

The intron–exon distribution of RdHSP family genes was explored and visualized using the Gene Structure Display Server 2.0 (GSDS, Figure 3c). A total of 76 RdHSP genes possessed exons varying from 1 to 43. Twelve RdHSP genes lacked intron or had only one exon, including RdHSP20.1, RdHSP20.12, RdHSP20.14, RdHSP20.15, RdHSP20.2, RdHSP20.3, RdHSP20.4, RdHSP20.7, RdHSP20.8, RdHSP70.2, RdHSP70.21, and RdHSP90.6. The majority (62 of 76) of the RdHSP genes have 1 to 20 introns, and the RdHSP90.7 gene contained 43 exons and 42 introns, which was the greatest number of exons in the total RdHSP genes. As expected, gene structure analysis revealed that most genes in the same subfamily had similar intron/exon distribution, including the numbers and length of exons. For example, most genes of the RdHSP20 subfamily had one exon and lacked introns. The members of the RdHSP60 subfamily had more than five introns. Thus, our results indicate that RdHSP genes in the same subfamily had similar gene structures, which further verify the phylogenetic relationship of these RdHSP genes (Figure 2).

3.4. Cis-Element Analysis in the RdHSP Gene Promoters in R. delavayi

Cis-elements, commonly distributed in 5′ upstream regions of genes, are used as the binding sites of the transcription factors involved in transcriptional regulation. In the present study, the 2000 bp upstream region of the transcription start site (TSS) of RdHSP genes was applied to the analysis of cis-regulatory elements using the PlantCARE online database (Figure 4). There was a total of 19 functionally annotated cis-elements in the promoter of most RdHSP genes, which were classified into the following four categories: light-responsive elements (Sp1, chs-CMA1a, G-Box, TCCC-motif, Gap-box, TCT-motif, AAC-motif, GT1-motif, and GATA-motif), stress-responsive elements (GC-motif, LTR, TC-rich, MBSI, and MBS), hormone-responsive elements (CGTCA-motif, GACG-motif, TATC-box, GARE-motif, P-box, ABRE, O2-site, and TGA-element), and cis-acting elements related to plant development (AREHD-Zip1, CAT-box, NON-box, and GCN4_motif).

3.5. Syntenic Analysis of the RdHSP Gene Family in R. delavayi

To further investigate the potential evolutionary mechanisms of RdHSP family genes, we conducted the collinearity analysis of RdHSP genes in the R. delavayi genome by the all-vs.-all local BLASTP and MCScan methods (Figure 5). A total of 8 segmental duplication gene pairs with 15 RdHSP genes were detected in the R. delavayi genome, which only accounted for 19.73% of RdHSP family genes (Figure 5a and Table S4). Moreover, all RdHSP genes with syntenic regions were distributed on all chromosomes. Subsequently, the ratio of non-synonymous to synonymous substitution (Ka/Ks) was estimated using the pairwise model by maximum likelihood (PAML v8.0). Ka/Ks < 1 and Ka/Ks > 1 represent purifying and positive selection, respectively (Hurst, 2002). In this study, the Ka/Ks ratios of five RdHSP gene pairs were less than one, indicating that purifying selection played an important role in the expansion of these genes, which maintained the similar functions of the RdHSP gene family in R. delavayi. The Ka/Ks ratios of three RdHSP gene pairs were greater than one, demonstrating that these genes are under positive selection (Table S4). Moreover, Ks was usually used to estimate the evolutionary dates of genome or gene duplication events. The WGD/segmental duplicated events in R. delavayi occurred from 15.02 (Ks = 0.4507) to 90.73 mya (Ks = 2.8429).

In addition, to further understand the phylogenetic mechanisms of the RdHSP family, comparative syntenic maps were conducted among six Rhododendron species (R. delavayi, R. williamsianum, R. simsii, R. ovatum, R. henanense, and R. irroratum) (Figure 5b–e). In the present study, there were 57 orthologous HSP gene pairs between R. delavayi and R. henanense (Table S5), 42 orthologous HSP gene pairs between R. delavayi and R. irroratum (Table S6), 48 orthologous HSP gene pairs between R. delavayi and R. ovatum (Table S7), 46 orthologous HSP gene pairs between R. delavayi and R. simsii (Table S8), and 35 orthologous HSP gene pairs between R. delavayi and R. williamsianum (Table S9). The number of orthologous events of RdHSP-RgHSP was much greater than that of RdHSP-RwHSP.

The Ka, Ks, and Ka/Ks values of the orthologous syntenic gene pairs between peach and strawberry were also computed to analyze the evolutionary selection in the RdHSP gene family, and we counted the nonsynonymous (Ka) and synonymous substitutions (Ks) among the orthologous gene pairs, as well as the Ka/Ks ratios among the six Rhododendron species (Figure 6). The results showed that the Ka/Ks ratios of the orthologous gene pairs were less than one, suggesting that the orthologous HSP genes of the six Rhododendron species were subjected to purifying selection during evolution, and similar mechanisms affect the horizontal evolution of HSP family genes across species.

3.6. GO and KEGG Enrichment Analysis of RdHSP Genes in R. delavayi

Gene ontology helps in the functional analysis of genes by determining their similarity with other genes of known function. In this study, GO enrichment analysis of the RdHSP genes was conducted to explore the biological functions of the RdHSP family genes in R. delavayi (Figure 7a and Table S10). The top 20 enrichment score GO terms, including 16 molecular functions and 4 biological processes, were enriched in the RdHSP genes relative to the complete GO database. In the biological process category, RdHSP genes were mainly enriched in response to heat, protein folding, response to hydrogen peroxide, and response to ethanol. In the molecular function category, the RdHSP genes were enriched in unfolded protein binding, ATP binding, adenyl nucleotide binding, adenyl ribonucleotide binding, purine ribonucleoside, triphosphate binding, purine nucleoside binding, purine ribonucleoside binding, purine ribonucleotide binding, nucleoside binding, purine nucleotide binding, ribonucleoside binding, ribonucleotide binding, carbohydrate derivative binding, nucleotide binding, nucleoside phosphate binding, and small molecule binding. The GO term enrichment suggested multiple roles of RdHSPs in the cell such as protein processing, growth- and development-related processes, stress responses, and metabolism. Moreover, KEGG enrichment analysis was conducted with RdHSP family genes. The twenty-one RdHSP genes were significantly enriched in five KEGG pathways, including protein processing in endoplasmic reticulum, lipid and atherosclerosis, endocytosis, spliceosome, RNA degradation, and plant–pathogen interaction (Figure 7b and Table S11), suggesting that these RdHSP genes mostly functioned in the regulation of gene expression and immune responses to stress.

3.7. Expression Patterns of the RdHSP Family Genes in R. delavayi under High Temperature

To evaluate the expression levels of RdHSP genes in R. delavayi under high-temperature treatments, the expression patterns of RdHSP genes in the leaves of R. delavayi treated with a high temperature of 38 °C were evaluated using our RNA-seq data, which were divided into four types (Figure 8). Twenty-one RdHSP genes were up-regulated at both 3 d and 6 d under high-temperature treatment, while 10 RdHSP genes showed decreased expression at both time points (Figure 8), which may be the long-term response genes in R. delavayi under high-temperature treatment. The expression levels of 9 RdHSPs (RdHSP20.4, RdHSP20.8, RdHSP20.6, RdHSP20.3, RdHSP20.10, RdHSP70.15, RdHSP20.15, RdHSP70.21, and RdHSP70.16) in R. delavayi treated with high temperature were more than 10 times those of the control. Moreover, the expression levels of 28 RdHSP genes were increased to a maximum at 3 d, and decreased sharply at 6 d of high-temperature treatment (Figure 8), suggesting that these RdHSP genes may be short-term response genes in R. delavayi under high-temperature treatment. Both long- and short-term response genes in R. delavayi might play important roles during the high-temperature treatment. In addition, there is no significant difference in the expression patterns of 14 RdHSP genes between the high-temperature treatment group and the control group.

3.8. Validation of RNA-Seq-Based Gene Expression

To further confirm the reliability of the RNA-seq results, qRT-PCR was conducted on 15 RdHSPs involved in response to heat stress selected at random with high expression levels. Comparative analysis of expression levels was conducted in the leaves of R. delavayi treated with high temperature, and the expression trends in the qRT-PCR results were consistent with the RNA-Seq data (Figure 9).

3.9. Protein Interaction Prediction

To further investigate potential protein–protein interactions among RdHSP proteins, we constructed an interaction network of RdHSP family proteins using the STRING database (Figure 10). The results showed that RdHSP60.1 and RdHSP100.20 had the highest number of interactions with other RdHSP proteins, respectively. For example, RdHSP100.2 interacted with seven other RdHSPs such as RdHSP20.11, RdHSP20.1, RdHSP90.1, RdHSP20.13, RdHSP20.6, RdHSP70.1, and RdHSP70.21, respectively. Diverse interactions were also detected in different RdHSP subfamilies, including interacting proteins from the same RdHSP subfamily (e.g., RdHSP60.1 and RdHSP60.10, RdHSP60.1 and RdHSP60.5, RdHSP60.1 and RdHSP60.13, RdHSP60.1 and RdHSP60.15, RdHSP60.1 and RdHSP60.7, RdHSP60.1 and RdHSP60.14, RdHSP60.1 and RdHSP60.2). These findings indicated that the RdHSP family genes may cooperate for regulating various physiological functions and in R. delavayi.

3.10. Subcellular Location of RdHSP Proteins

To verify the prediction results of the subcellular localization of RdHSP proteins, full-length coding sequences of two RdHSP genes (RdHSP20.6 and RdHSP20.8) without the stop codon were inserted into the pCAMBIA2300-GFP vector. These two genes were fused with the GFP protein driven by the 35 S promoter, which was transiently expressed in N. benthamiana leaves. The subcellular localization results showed that the GFP protein with no gene inserted is expressed in various organelles in N. benthamiana, but the GFPs fused with the RdHSP20.6 and RdHSP20.8 proteins are expressed in the chloroplast, indicating that RdHSP20.6 and RdHSP20.8 genes are expressed and function in the chloroplast (Figure 11).

4. Discussion

The HSP family genes play important roles in regulating plant growth and development as well as in response to various environmental stresses [40]. However, a systematic identification of the RdHSP family genes in R. delavayi has not been reported. In this study, we identified 76 RdHSP genes, including 15 RdHSP20 genes, 19 RdHSP60 genes, 30 RdHSP70 genes, 8 RdHSP90 genes, and 4 RdHSP100 genes. The number of HSP family genes varies among different plant species. For example, the number of the RdHSP family genes in R. delavayi is slightly higher than that in Arabidopsis. This result suggested that no species-specific expansion or gene loss occurred in R. delavayi. However, when compared to R. delavayi, wheat has a significantly higher number of HSP genes [41], suggesting that polyploid organisms may exhibit an elevated number of HSP genes due to gene duplication within each chromosome set. In addition, the number of members in different HSP subfamilies varies greatly among different species (Table S12). The number of genes in the HSP70 subfamily is significantly greater than that in the HSP60 subfamily, suggesting that HSP70 subfamily genes undergo significant expansion.

Gene duplication events such as WGD/segmental duplication, tandem duplication, and dispersed duplication were used as the main important expansion mechanisms for producing new genes and generating genetic novelty in plants [42]. Contrary to expectations, our results showed that 15 (17.44%) and 8 (9.30%) of the RdHSP family genes in R. delavayi were obtained from WGD/segmental duplication and tandem duplication, indicating that these two duplication events partially contributed to the expansion of the RdHSP family genes in R. delavayis. Different duplication events contributed differently to the expansion of the HSP gene family in different plants. For example, segmental duplication and tandem duplication events contributed to the expansion of HSP genes in cassava and Triticum aestivum, respectively [43,44]. In addition, the specific duplication models play different roles in the expansion of the different gene family [45]. For example, segmental duplication and tandem played important roles in expanding the WRKY and AP2/ERF family genes in plant species [46], and transposed duplications contributed to the expansion of the MADS-box and NBS-LRR gene families [45]. Selective pressures played important roles in the evolution of duplicated genes [47], and most duplicated RdHSP genes have undergone purification selection. In addition, synteny analysis showed that syntenic HSP gene pairs among six Rhododendron were detected in this study, and the number of orthologous events of RdHSP-RgHSP was much greater than that of RdHSP-RwHSP, suggesting that R. henanense and R. delavayi have a closer genetic relationship than R. williamsianum. In addition, the Ka/Ks ratio of segmentally duplicated gene pairs and the orthologous gene pairs was less than one, suggesting that purifying selection played important roles in the expansion of the HSP gene family, and similar mechanisms affect the horizontal evolution of HSP family genes across species. The Ka/Ks ratio within species was less than the ratio between species, implying that there are significant differences in the horizontal evolution rate of the HSP family genes among different Rhododendron species.

Cis-acting regulatory elements (CAREs) in promoters act as key molecular switches for the transcriptional regulation of complex and dynamic gene networks. In this study, analyses of the cis-acting elements in the RdHSP gene promoters were identified and divided into four types: light-, stress-, hormone-, and cis-elements related to plant development. These types of cis-acting regulatory elements were also identified in cassava, dove, and barley [48]. This result further demonstrated that RdHSP family genes played important roles in regulating various physiological processes and in response to complex environmental stresses. In addition, among stress-responsive cis-elements, the cis-element HSE was missing on all promoters of RdHSP family genes, and further revealed the reason why R. delavayi exhibits high sensitivity to heat stress.

Numerous studies have revealed that HSP family genes play essential roles in response to high-temperature stress [49]. Our RNA-Seq results revealed that 83.72% (72) of RdHSP family members in R. delavayi exhibited the differential expression profiles under high-temperature treatments, indicating that the RdHSP family genes played important roles in regulating high-temperature stress response. Furthermore, 21 RdHSP genes were up-regulated at both 3 and 6 d after high- temperature treatment, and the expression level of 28 RdHSP genes was increased to a maximum at 3 d, and decreased slightly at 6 d of high-temperature treatment, suggesting that these RdHSPs might be associated with the long- and short-term regulation of high-temperature treatment. In addition, the expression levels of nine RdHSP genes in R. delavayi treated with high temperature were more than 10 times higher than those of the control, implying that these RdHSPs played key roles in regulating the high-temperature treatment. Functional analysis revealed that the orthologous genes of nine RdHSP genes in Arabidopsis were involved in response to heat stress [50]. For example, the orthologous genes of RdHSP20.6, RdHSP20.3, RdHSP20.10, RdHSP70.15, and RdHSP70.21 6 in Arabidopsis were HSP21, HSP17.6B, HSP17.4B, BIP2, and HSP70-46, respectively. Previous studies have revealed that HSP70-4, HSP17.4B, and HSP21 played important roles in enhancing the heat tolerance, which further confirmed that these genes were involved in enhancing heat tolerance [51,52,53].

5. Conclusions

In the present study, the systematic analysis of the RdHSP gene family in R. delavayi was first performed, and a total of 76 RdHSP genes were identified and unevenly distributed on 13 chromosomes. The conserved motifs, gene structure, and evolutionary relationships of the RdHSP family genes were also established and analyzed. The exploration of the cis-acting elements in the RdHSP gene promoters demonstrated that the RdHSP genes played important roles in regulating plant development and in response to various environmental stresses. Expression analysis showed that 83.72% (72) of the RdHSP family genes in R. delavayi exhibited the differential expression profiles under high-temperature treatments, and the expression levels of 9 RdHSP genes (RdHSP20.4, RdHSP20.8, RdHSP20.6, RdHSP20.3, RdHSP20.10, RdHSP70.15, RdHSP20.15, RdHSP70.21, and RdHSP70.16) in high-temperature treated R. delavayi were more than 10 times those of the control, suggesting that these genes may be involved in positively regulating heat stress response. The above results could provide a basis for the functional characterization of HSP genes, and also provide candidate genes for the future improvement of heat tolerance in R. delavayi.