2.1 Plant Materials and Growth Conditions

The thermotolerant eggplant line ‘E390’, intermediate heat-tolerant eggplant line ‘19009’ and thermosensitive eggplant line ‘HM’ were identified in a previous study (Liu et al. 2023). The eggplant line ‘19009’ was used for the virus-induced gene silencing (VIGS) assay, tobacco (Nicotiana benthamiana) was used for the functional verification assays and the A. thaliana Columbia (Col) ecotype and tomato (Solanum lycopersicum cv. Micro-Tom) were used to generate transgenic plants. Eggplant and tomato seedlings were grown under a 25/18°C day/night temperature with a 16/8 h light/dark photoperiod, a relative humidity of 60%–70% and a light intensity of 6000 lux. Tobacco seedlings were grown under a 25/18°C day/night temperature with a 16/8 h light/dark photoperiod, a relative humidity of 60%–70% and a light intensity of 3000 lux. A. thaliana seedlings were grown under a 22/18°C day/night temperature with a 16/8 h light/dark photoperiod, a relative humidity of 60%–70% and a light intensity of 3000 lux.

2.2 HT, Exogenous H₂O₂, and Exogenous Flavonoid Treatments

For high-temperature treatment (HTT), several plants of the eggplant lines ‘E390’, ‘19009’ and ‘HM’ with three or four true leaves were transferred into a growth chamber at different temperatures (25, 33, 38 or 42°C) with a 16/8 h light/dark photoperiod, a relative humidity of 60%–70% and a light intensity of 6000 lux for 72 h. The roots, stems and second true leaves (from the bottom to the top) of the eggplant line ‘19009’ not subjected to HTT were collected for tissue-specific analysis. The second true leaves of the eggplant line ‘19009’ were collected after treatment at different temperatures for 24 h, and the second true leaves of the eggplant lines ‘E390’ and ‘HM’ were collected at 0, 1, 6, 12, 24, 48 and 72 h post-42°C HTT for gene expression analysis.

For H₂O₂ treatment, water, 50 mM and 100 mM H₂O₂ solution was used to spray the leaves of the eggplant line ‘19009’ until three or four true leaves were dripping. The second true leaves of eggplant seedlings were collected at 24 h post-H₂O₂ treatment for gene expression analysis.

For flavonoid treatment, according to a previous study (Kang et al. 2022), rutin (0, 0.25, 0.75 and 2.5 g) was separately weighed and dissolved in 0.5 L of ethanol, and the rutin ethanol solution was sonicated in a water bath with ice for 30 min before application. The solution was used for spraying the leaves of the eggplant line ‘19009’ with three or four true leaves until the leaves were dripping. After spraying for 12 h, the treated plants were subjected to HTT (42/42°C day/night temperature, 16/8 h light/dark photoperiod) for 4 days. After HTT for 4 days, the heat injury index was measured. The eggplant heat injury index was rated as follows: 0, healthy; 1, first true leaf wilted; 2, second true leaf wilted; 3, third true leaf wilted; 4, all leaves wilted; 5, the stem wilted. The heat injury index was equal to Σ (incidence grade×the number of corresponding grade plants)/(highest grade×total plants) × 100. The treated plants were subsequently subjected to phenotypic observation and determination of the chlorophyll fluorescence and photosynthetic parameters, and the leaves were collected for determination of the REL and MDA content. Three biological replicates and three technical replicates of each biological replicate were included for all the samples. All the samples were placed in liquid nitrogen and then stored at −80°C for further experiments.

2.3 Bioinformatics Analysis

DNA and amino acid sequence alignment were performed using DNAMAN software. The phylogenetic evolutionary trees were generated using MEGA X via the neighbour-joining method and visualised and optimised using iTOL V5 (Letunic and Bork 2021). Gene collinearity analysis among the HSFA8 genes from A. thaliana, eggplant, tomato, pepper and potato were conducted with MCScanX tools (Wang et al. 2012) and visualised via multiple synteny plot tools in TBtools software (Chen et al. 2020). Information on the HSFA8 genes from A. thaliana, eggplant, tomato, pepper and potato was derived from the genome databases TAIR10 (https://www.arabidopsis.org/), S. melongena HQ-1315 (http://eggplant-hq.cn/Eggplant/home/index), ITAG4.0 (https://solgenomics.net/organism/Solanum_lycopersicum/genome/), Capsicum annuum zunla (http://peppersequence.genomics.cn/) and DM 1-3 516 R44-v6.1 (http://spuddb.uga.edu/dm_v6_1_download.shtml). The 1500 bp nucleotide sequences upstream of the start codon of SmHSFA8 were uploaded to the PlantCARE database (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) with default parameters to predict cis-elements (Lescot 2002).

2.4 DNA Affinity Purification Sequencing (DAP-Seq) Analysis

Genomic DNA was extracted from mature leaves of the eggplant line ‘E390’ using a MICH DNA Clean Beads (NGS-0201-100, Bluescape Hebei Biotech, Baoding, China) with two biological replicates. The construction of the library was carried out via the MICH TLX DNA-Seq Kit (NGS 0602-24, Bluescape Hebei Biotech, Baoding, China). DAP-seq experiments were performed according to a published protocol (Bartlett et al. 2017). Specifically, SmHSFA8 protein was prepared using the TNT SP6 Coupled Wheat Germ Extract System (L5030, Promega, Beijing, China), then immobilised on Magne HaloTag Beads (G7281, Promega, Beijing, China) and incubated with the constructed DNA library. Unbound DNA fragments were removed by washing the beads three times. The SmHSFA8-binding DNA fragments were then eluted and amplified with the indexed primers. Subsequent sequencing was performed with an Illumina NovaSeq instrument (NovaSeq. 6000, Illumina, San Diego, USA). The reads were mapped to the eggplant genome ‘HQ-1315’ using Bowtie2 (Langmead and Salzberg 2012). To identify SmHSFA8-binding sites, DAP-seq peaks were screened with MACS2 (fold enrichment > 2 and q < 0.05) (Zhang et al. 2008). Next, the locations of the DAP-seq peaks upstream or downstream of the transcription start site within the 2000 bp region were analysed via Homer software (Heinz et al. 2010). The SmHSFA8-targeted genes were identified via BEDtools (Quinlan and Hall 2010). Motif identification was performed using the Meme-ChIP suite 5.0.5 (Machanick and Bailey 2011).

2.5 RNA Extraction and Quantitative Real-Time Polymerase Chain Reaction (qRT‒PCR)

The total RNA of the roots, stems and second true leaves of eggplant and the leaves of A. thaliana and tomato were isolated and extracted using an Eastep Super Total RNA Extraction Kit (LS1040, Promega, Beijing, China). The primers for qRT‒PCR were designed using Primer 5.0 and listed in Table S1. All qRT‒PCR analyses were performed using ChamQ Universal SYBR qPCR Master Mix (Q711–02/03, Vazyme, Nanjing, China). The relative gene expression was calculated via the 2^−ΔΔCt method (Livak and Schmittgen 2001).

2.6 Subcellular Localisation Assay

The coding sequences (CDSs) of SmHSFA8 without the termination codon were cloned and inserted into the pGreenII-C18 vector. The recombinant plasmids and nuclear localisation signal (NLS-DsRed) were transformed into Agrobacterium tumefaciens GV3101 (psoup) strain (AC1002, Weidi Biotechnology, Shanghai, China), which was subsequently mixed at a 1:1 ratio and injected into the leaves of 5- to 6-week-old tobacco. After 72 h, the tobacco leaves were collected for fluorescence microscopy observation (BX53F, Carl Zeiss, Jena, Germany).

2.7 Transcriptional Activation Region Analysis

The CDSs of SmHSFA8 were truncated according to the conserved domain. The segmented sequences were cloned and inserted into the pGBKT7 (BD) vector. The recombinant plasmids were subsequently transformed into Y2HGold yeast strain (YC1002, Weidi Biotechnology, Shanghai, China). Positive transformants were selected by culturing on SD/-Trp and SD/-Trp-Ade-His media. The positive transformants were subsequently stained with X-α-Gal (SL0940, Coolaber, Beijing, China).

2.8 Transcriptional Activity Analysis

Transcriptional activation analysis assays were performed using a Dual-Luciferase Reporter Assay System (E1910, Promega, Beijing, China). The CDSs of SmHSFA8 were ligated into the pEAQ-PBD vector. The empty vector, reporter gene (GAL-LUC) and generated constructs were separately transformed into Agrobacterium tumefaciens GV3101 strain (AC1001, Weidi Biotechnology, Shanghai, China). The Agrobacterium tumefaciens strain harbouring the empty vector or generated construct and reporter at a 1:1 ratio was coinfiltrated into the leaves of 5- to 6-week-old tobacco. After 72 h, the tobacco leaves were collected, and the activities of REN and LUC were measured according to the protocol of the Dual-Luciferase Reporter Assay System.

2.9 VIGS Assay

To identify the functions of SmHSFA8 and SmF3H in the eggplant response to HTS, SmHSFA8 and SmF3H were knocked down through a VIGS assay. A 300 bp fragment of SmHSFA8 or SmF3H was ligated into the pTRV2 vector, respectively. The Agrobacterium tumefaciens GV3101 strain harbouring pTRV1 was mixed with the Agrobacterium tumefaciens GV3101 strain harbouring pTRV2-SmHSFA8 or pTRV2-SmF3H plasmid at a 1:1 ratio and then incubated in the dark at 28°C for 2 h. The mixed bacterial solution was injected into the cotyledons of 2-week-old eggplant seedlings, which were placed at 20°C in the dark for 48 h. Then, the eggplant seedlings were transferred to a standard culture chamber under a 25/18°C day/night temperature with a 16/8 h light/dark photoperiod, a relative humidity of 60%–70% and a light intensity of 6000 lux. After 2 weeks, the second true leaves of the transformed plants were collected for gene expression analysis. The SmHSFA8-silenced and SmF3H-silenced eggplant plants were subsequently subjected to HTT (42/42°C day/night temperature, 16/8 h light/dark photoperiod) for 4 days. The treated plants were first subjected to phenotypic observation, and then the leaves were collected for determination of the REL, MDA content, flavonoid content and gene expression.

2.10 Genetic Transformation of Tomato

To overexpress SmHSFA8 in tomato, the CDSs of SmHSFA8 without the termination codon were ligated into the BG-plant-GFP vector. The generated construct was introduced into Agrobacterium tumefaciens GV3101 strain, which was subsequently transformed into tomato plants via a previously described method (Fan et al. 2018). The positive transformants were identified according to the transcript level of SmHSFA8 and three independent positive transformed lines (T2) were used for subsequent experiments. The wild-type (WT) and OE-SmHSFA8 transgenic tomato plants (4-week-old) were subjected to HTT (42/42°C day/night temperature, 16/8 h light/dark photoperiod) for 4 days. After HTT for 4 days, the heat injury index was measured. The tomato heat injury index was rated as follows: 0, healthy; 1, single leaf wilted; 2, a branch of the leaves wilted; 3, multiple branches of the leaves wilted; 4, all leaves wilted; 5, the stem wilted. The tomato heat injury index was equal to Σ (incidence grade×the number of corresponding grade plants)/(highest grade×total plants) × 100. The treated plants were subjected to phenotypic observation and determination of the chlorophyll fluorescence and photosynthetic parameters, and the leaves were subsequently collected for determination of the REL, MDA content, H₂O₂ content, flavonoid content and gene expression.

2.11 Genetic Transformation of A. thaliana

To overexpress SmHSFA8 in A. thaliana, the CDSs of SmHSFA8 without the termination codon were ligated into the pBI121-GFP vector, whereas the 1500 bp nucleotide sequences upstream of the start codon of SmHSFA8 were ligated into the CP013 vector. The generated constructs were subsequently introduced into Agrobacterium tumefaciens GV3101 strain. Transgenic A. thaliana lines were generated via the floral-dip method (Clough and Bent 1998) and the transgenic A. thaliana plants were screened using 1/2 MS media supplemented with 50 mg/L kanamycin. The seeds of the T2 transgenic A. thaliana lines were harvested for subsequent experiments. WT and OE-SmHSFA8 transgenic A. thaliana plants were planted in the soil for PCR and gene expression analysis, and they were subsequently subjected to HTT (43/43°C day/night temperature, 16/8 h light/dark photoperiod) for 24 h and recovery treatment (22/22°C day/night temperature, 16/8 h light/dark photoperiod) for 5 d. The phenotypes of the plants were observed, and the flavonoid content and gene expression were determined after HTT for 24 h, whereas the seedling survival rate and fresh weight were determined after HTT and recovery treatment for 5 d. In addition, WT and OE-SmHSFA8 transgenic A. thaliana plants were seeded into MS media, and they were subsequently subjected to HTT (44°C) for 50 min and recovery treatment (22/22°C day/night temperature, 16/8 h light/dark photoperiod) for 5 or 7 days. The seedling survival rate was measured after HTT and recovery treatment for 5 days, whereas the number of lateral roots was counted after HTT and recovery treatment for 7 days. HT phenotypes of the plants cultivated in the soil were analysed using 27 seedlings (n = 27), whereas those in MS media were analysed using 50 seedlings (n = 50). Three and four repeated experiments were performed, respectively.

2.12 GUS Activity Staining Assay

A GUS activity staining assay was performed using CP013-ProSmHSFA8-GUS transgenic A. thaliana lines. The plants were stained using a GUS stain kit (SL7160, Coolaber, Beijing, China), and GUS expression was observed after decolorisation.

2.13 Determination of Physiological and Biochemical Indices

The REL of the leaves was determined using a conductivity metre (DDS-307, Shanghai instrument, Shanghai, China) and the MDA content was measured via a reaction with thiobarbituric acid (TBA) (Sun et al. 2012). The flavonoid content of the leaves was determined via a flavonoid content determination kit (HTC-1-W, Comin, Suzhou, China) and the H₂O₂ content was measured via a H₂O₂ content determination kit (H2O2–1-Y, Comin, Suzhou, China). Chlorophyll fluorescence and photosynthetic parameters were evaluated via a chlorophyll fluorescence imager (PlantExplorer, PhenoVation, Wageningen, Netherlands) (Liang et al. 2024). The plants were dark-adapted for 10–30 min. Then, the chlorophyll fluorescence of the entire plant was measured from the top. The fluorescence intensity was measured by applying a saturating pulse of 3000 μmol photons m⁻²s⁻¹ for 1 s, and the fluorescence intensities were stored at a 12-bit resolution. In addition, the leaf area of plants was determined via a ImageJ software.

2.14 Yeast One-Hybrid (Y1H) Assay

The CDSs of SmHSFA8 were recombined into the pGADT7 (AD) vector and the truncated segment sequences upstream of the start codons of SmHSP70, SmHSP21, SmHSP70B, SmF3H and SmEGY3 were separately recombined into the pAbAi vector. The Y1H assay was performed according to the manufacturer's instructions (PT4087-1, Clontech, Takara, China). The AD-SmHSFA8 vector was transformed into Y1HGold yeast strain (YC1001, Weidi Biotechnology, Shanghai, China) harbouring the pAbAi-recombinant plasmid. Then, they were spotted onto SD/-Leu and SD/-Leu + Aureobasidin A (AbA) (60231ES03, YEASEN, Shanghai, China) plates.

2.15 Electrophoretic Mobility Shift Assay (EMSA)

The CDSs of SmHSFA8 were cloned and inserted into the pMAL-c2X vector, and upon expression, SmHSFA8-MBP fusion protein was obtained. Biotin-labelled probes for SmHSP70, SmHSP21, SmHSP70B, SmF3H and SmEGY3 were synthesised by Sangon Biotech (Shanghai, China). The probe sequences were listed in Table S1. The protein and biotin-labelled probes were incubated as previously described (Wu et al. 2022). EMSAs were performed using a LightShift Chemiluminescent EMSA kit (20148, Thermo Scientific, Waltham, USA) according to the manufacturer's protocols.

2.16 Dual-Luciferase Reporter (DLR) Assay

The CDSs of SmHSFA8 and SmEGY3 were inserted into the pEAQ vector as effectors, whereas the promoter sequences of SmHSP70, SmHSP21, SmHSP70B, SmF3H, SmEGY3 and SmCSD1 were recombined into the pGreenII-0800 vector as reporters. The Agrobacterium tumefaciens GV3101 strain harbouring the pEAQ/pEAQ-SmHSFA8 plasmid mixed with the Agrobacterium tumefaciens GV3101 (psoup) strain harbouring the pGreenII-0800-ProSmHSP70/-ProSmHSP21/-ProSmHSP70B/-ProSmF3H/-ProSmEGY3 plasmid at a 20:1 ratio was infiltrated into the leaves of 5- to 6-week-old tobacco. While the Agrobacterium tumefaciens GV3101 strain harbouring the pEAQ/pEAQ-SmEGY3/pEAQ-SmHSFA8/pEAQ-SmHSFA8+pEAQ-SmEGY3 plasmid mixed with the Agrobacterium tumefaciens GV3101 (psoup) strain harbouring the pGreenII-0800-ProSmCSD1 plasmid at a 20:1 ratio was also infiltrated into the leaves of 5- to 6-week-old tobacco. After 72 h, the tobacco leaves were collected, and the activities of REN and LUC were measured according to the protocol of the Dual-Luciferase Reporter Assay System.

2.17 Yeast Two-Hybrid (Y2H) Assay

The CDSs of SmHSFA8 were recombined into the BD vector, whereas the CDSs of SmEGY3 were recombined into the AD vector. The recombinant plasmids were subsequently transformed into Y2HGold yeast strain. Positive transformants were selected by culturing on SD/-Trp-Leu and SD/-Trp-Leu-Ade-His media. Self-activation of SmHSFA8 was inhibited by 3-amino-1,2,4-triazole (3-AT) (SL0930, Coolaber, Beijing, China). The positive transformants were subsequently stained with X-α-Gal.

2.18 Bimolecular Fluorescence Complementation (BIFC) Assay

The CDSs of SmHSFA8 without the termination codon were recombined into the pSPYNE-35S (YNE) vector, and the CDSs of SmEGY3 without the termination codon were recombined into the pSPYCE-35S (YCE) vector. The recombinant plasmids and nuclear localisation signal (NLS-DsRed) were transformed into Agrobacterium tumefaciens GV3101 strain, which was subsequently mixed at a 1:1:1 ratio and injected into the leaves of 5- to 6-week-old tobacco. After 72 h, the tobacco leaves were collected for fluorescence microscopy observation (BX53F, Carl Zeiss, Jena, Germany).

2.19 Luciferase Complementary Imaging (LCI) Assay

The CDSs of SmHSFA8 without the termination codon were recombined into the 1300-nLUC (nLUC) vector, and the CDSs of SmEGY3 were recombined into the 1300-cLUC (cLUC) vector. The recombinant plasmids and auxiliary plasmid (P19) were transformed into Agrobacterium tumefaciens GV3101 strain, which was subsequently mixed at a 1:1:1 ratio and injected into the leaves of 5- to 6-week-old tobacco. After 72 h, D-luciferin potassium salt D (1 mM) (40902ES01, YEASEN, Shanghai, China) was injected into the tobacco leaves, and the tobacco leaves were collected for fluorescence observation via a fluorescence chemiluminescence gel imaging system (Universal Hood II, Bio-Rad ChemiDOCXRS + , Shanghai, China).

3.1 SmHSFA8 Activates SmEGY3 Expression and Enhances the Function of SmEGY3 by Interacting With It

In our previous studies, SmEGY3 played positive roles in the heat tolerance of eggplant by promoting SmCSD1-mediated H₂O₂ production and H₂O₂-mediated retrograde signalling pathway (Liu et al. 2024). As a bait protein, SmEGY3 was used to screen the interaction proteins in the cDNA library in eggplants after HTT. SmHSFA8, a heat shock transcription factor, was identified to interact with SmEGY3 (Figure 1A). The interaction result was confirmed by LCI and BiFC assays (Figure 1B,C). Moreover, we also found that SmHSFA8 could bind the promoter of SmEGY3 and activate its expression based on the results of Y1H, EMSA and DLR assays (Figure 1D–F). In the DLR assay, when SmHSFA8 protein was cotransferred with SmEGY3 protein, the expression of SmCSD1 was significantly enhanced (Figure 1G), suggesting that the interaction between SmHSFA8 and SmEGY3 enhanced the activation function of SmEGY3 on SmCSD1. These results suggested that SmHSFA8 might participate in regulating heat tolerance in eggplants. In eggplants, SmHSFA8 contained the typical DNA-binding domain, HR-A/B region, temperature-dependent repression domain, AHA domain and NLS motif (Figure S1). Phylogenetic analysis revealed that HSFA8 genes were relatively conserved in A. thaliana, potato, tomato, pepper and eggplant, and that they were clustered in the same branch as the HSFA1s genes, which functioned in regulating heat tolerance in plants (Yoshida et al. 2011) (Figure S2A). In addition, the homologues of HSFA8 gene were identified in specific Solanaceae species (potato, eggplant and tomato), but they were not found in Capsicum species (Figure S2B), suggesting that the homologues of HSFA8 gene possessed species specificity.

3.2 Expression Pattern and Transcriptional Activity Analyses of SmHSFA8 in Eggplants

To analyse the expression pattern and molecular characteristics of SmHSFA8 in eggplants, qRT‒PCR, GUS staining assay, subcellular localisation assay, transcriptional activation region analysis and transcriptional activity analysis were performed. SmHSFA8 was expressed in the roots, stems and leaves of eggplant (Figure S2C). After HTT, SmHSFA8 was upregulated in the thermotolerant (E390) and thermosensitive (HM) eggplant lines (Figure S2D). SmHSFA8 was induced by exogenous H₂O₂ treatment (Figure S2E). In addition, the transcript level of SmHSFA8 increased in a dose-dependent manner with temperature (Figure S2F). Several cis-elements, including LTRs, TGA elements, and CGTCA, TCCC and TGACG motifs, were found in the promoter of SmHSFA8 (Figure S3A). The GUS staining assay results indicated that the activity of GUS driven by the SmHSFA8 promoter was strongly induced in the stems and leaves of three independent transgenic A. thaliana plants (Figure S3B). SmHSFA8-Green fluorescent protein (GFP) fusion protein was localised in the cell nucleus (Figure S4A). The transcriptional activation region of SmHSFA8 was located at amino acid residues 193-374 (Figure S4B). The DLR assay results suggested that SmHSFA8 had transcriptional activation activity in planta, suggesting that SmHSFA8 functioned as an activator in response to HT (Figure S4C).

3.3 SmHSFA8 Positively Regulates the Heat Tolerance of Plants

To evaluate the function of SmHSFA8 under HTS, two representative SmHSFA8-silenced (by VIGS) eggplant plants were obtained, and the expression of SmHSFA8 was significantly lower in the SmHSFA8-silenced eggplant plants than in the control plants (pTRV2 empty vector-transformed plants) (Figure 2A). There was no significant difference in phenotypes between the control plants and SmHSFA8-silenced eggplant plants at room temperature (RT) (Figure 2B). The control and SmHSFA8-silenced eggplant plants were subsequently subjected to HTT for 4 days. The control plants presented robust flat leaves, whereas the SmHSFA8-silenced eggplant plants presented dwarfism with shrunken and drooped leaves (Figure 2B). In addition, the MDA content and REL significantly increased in the SmHSFA8-silenced eggplant plants compared with those in the control plants after HTT for 4 days (Figures 2C,D).

SmHSFA8 was overexpressed in tomato to further determine its function under HTS. Three OE-SmHSFA8 transgenic tomato lines were identified and were found to have more highly SmHSFA8 expression than the WT plants (Figure 2E). The leaf area of the OE-SmHSFA8 transgenic tomato plants was significantly greater than that of the WT plants at RT (Figure 2F and Figure S5). After HTT for 4 days, the leaves of the WT plants were shrunken and drooped, whereas the OE-SmHSFA8 transgenic tomato plants still grew normally (Figure 2F). The heat injury index of the OE-SmHSFA8 transgenic tomato plants was lower than that of the WT plants after HTT for 4 d (Figure 2G). There was no significant difference in the chlorophyll fluorescence, Fv/Fm, Fq'/Fm', rETR and NPQ between the control plants and SmHSFA8-silenced eggplant plants at RT (Figure 2H–L). After HTT for 4 days, the chlorophyll fluorescence, Fv/Fm, Fq'/Fm' and rETR of the OE-SmHSFA8 transgenic tomato plants were significantly greater than those of the WT plants (Figure 2H-K), whereas NPQ was significantly lower than that of the WT plants (Figure 2L). Compared with those of the WT plants, the REL and MDA content significantly decreased in the OE-SmHSFA8 transgenic tomato plants after HTT for 4 d (Figure 2M,N). Moreover, the expression of SlEGY3 and H₂O₂-mediated retrograde signalling genes (SlAPX2, SlZAT10 and SlWRKY33) and H₂O₂ content significantly increased in the OE-SmHSFA8 transgenic tomato plants at RT and HTT compared with the WT plants, and the expression of SlERF6 significantly increased in the OE-SmHSFA8 transgenic tomato plants at HTT (Figure S6). These results suggested that SmHSFA8 promoted H₂O₂ content and H₂O₂-mediated retrograde signalling pathway by directly activating SlEGY3 expression, thus enhancing the heat tolerance of tomato plants.

SmHSFA8 was also overexpressed in A. thaliana. Three OE-SmHSFA8 transgenic A. thaliana lines were identified (Figure S7A), and were found to have more highly SmHSFA8 expression than the WT plants (Figure S7B). No significant difference was found between the WT and OE-SmHSFA8 transgenic A. thaliana plants at RT. One day after HTT, the leaves of the WT plants were shrunken and drooped, but those of the OE-SmHSFA8 transgenic A. thaliana plants were still flattened. After recovery treatment for 5 days, the WT plants were small and even died, whereas the OE-SmHSFA8 transgenic A. thaliana plants exhibited normal growth (Figure S7C). The seedling survival rate and fresh weight were significantly greater in the OE-SmHSFA8 transgenic A. thaliana plants than in the WT plants after HTT for 1 d following recovery treatment for 5 days (Figure S7D,E). Moreover, the phenotypes of the WT and OE-SmHSFA8 transgenic A. thaliana plants in MS media under HTS were also analysed. After a brief period of HTT and recovery treatment for 7 days, the roots maintained normal growth with abundant lateral roots in the OE-SmHSFA8 transgenic A. thaliana plants but were suppressed in the WT plants (Figure S7F,H). A large proportion of the WT plants died, but the OE-SmHSFA8 transgenic A. thaliana plants maintained normal growth after a brief period of HTT and recovery treatment for 5 days (Figure S7G,I). These results demonstrated that SmHSFA8 played a positive role in regulating the heat tolerance of plants.

3.4 DAP-Seq Analysis of Candidate Target Genes and Downstream Pathways of SmHSFA8

A DAP-Seq assay was performed to elucidate the molecular mechanism by which SmHSFA8 regulates heat tolerance in eggplants. A total of 2272 SmHSFA8-binding peaks were identified based on two biological replicates (Figure 3A). Among these peaks, 12.59% were located within the promoter regions of the target genes (Figure 3B), which corresponded to 286 genes (Table S2). Six motifs with high confidence were identified using the Meme-ChIP suite 5.0.5 that might be the binding sites of SmHSFA8 (Figure 3C). In particular, the sequences of Motifs 3 and 6 were highly similar to those of HSEs, which are typical motifs of the binding sites of HSFs (Littlefield and Nelson 1999; Andrási, Pettkó-Szandtner, and Szabados 2021). HSFs usually participate in regulating the heat tolerance of plants by activating HSPs expression (Kotak et al. 2007; Hua 2009; Chen, Galli, and Gallavotti 2022). Several SmHSPs containing motifs similar to Motifs 3 and 6 were identified, including SmHSP70, SmHSP70B and SmHSP21. Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis revealed that the target genes of SmHSFA8 were involved in multiple metabolic processes, such as flavonoid biosynthesis (Figure 3D), and that SmF3H was involved in flavonoid biosynthesis. Moreover, SmHSP70, SmHSP70B, SmHSP21 and SmF3H were induced by HTS (Figure 3E–H). SmHSP70B, SmHSP21 and SmF3H were more highly expressed in the ‘E390’ eggplant line than in the ‘HM’ eggplant line after HTT (Figure 3F–H).

3.5 SmHSFA8 Regulates HSPs Expression in Plants

Y1H, EMSA and DLR assays were conducted to determine the effect of SmHSFA8 on SmHSPs. Y1H results revealed that SmHSFA8 bound the promoters of SmHSP70, SmHSP70B and SmHSP21 (Figure 4A). The SmHSFA8-MBP fusion protein was expressed and purified from Escherichia coli. SmHSFA8 could bind to the regulatory regions of SmHSP70, SmHSP70B and SmHSP21 by recognising the predicted binding motif (Figure 4B–D). DLR assay results revealed that SmHSFA8 was able to activate the expression of SmHSP70, SmHSP70B and SmHSP21 (Figure 4E). In addition, compared with the control and WT plants, the expression of HSP70, HSP70B and HSP21 significantly decreased in the SmHSFA8-silenced eggplant plants (Figure S8A–C) but significantly increased in the OE-SmHSFA8 transgenic A. thaliana and tomato plants at RT and HTT (Figure S8D–I).

3.6 SmHSFA8 Regulates Flavanone-3-Hydroxylase (F3H) Expression and Flavonoid Content in Plants

Y1H, EMSA and DLR assays were also conducted to determine the effect of SmHSFA8 on SmF3H. The results revealed that SmHSFA8 could bind the promoter of SmF3H and activate its expression (Figure 5A–C). Compared with the control and WT plants, the expression of F3H and flavonoid content significantly decreased in the SmHSFA8-silenced eggplant plants but increased in the OE-SmHSFA8 transgenic A. thaliana and tomato plants at RT and HTT (Figure 5D–I). These results demonstrated that SmHSFA8 regulated the heat tolerance of plants by promoting SmF3H-mediated flavonoid biosynthesis.

Three SmF3H-silenced lines generated via VIGS were produced from eggplants (Figure 6A). When exposed to HTT for 4 days, the control plants presented normal growth with relatively flat leaves; however, the SmF3H-silenced eggplant plants presented inhibited growth with shrunken, drooped, or even dried-up leaves (Figure 6B). Compared with those of the control plants, the MDA content and REL significantly increased in the SmF3H-silenced eggplant plants after HTT for 4 days (Figure 6C,D). Under RT condition and 4 days of HTT, the flavonoid content significantly decreased in the SmF3H-silenced eggplant plants compared with the control plants (Figure 6E). These results revealed that SmF3H affected the flavonoid biosynthesis and heat tolerance in eggplants.

3.7 Flavonoids are Involved in Regulating the Heat Tolerance of Eggplant

To investigate the effect of flavonoids on the heat tolerance of eggplant, the determination of flavonoid content and exogenous flavonoid treatment assays were performed. The flavonoid content in the ‘HM’ eggplant line significantly decreased, whereas it gradually increased in the ‘E390’ eggplant line under HTS (Figure S9). No significant difference in phenotypes between the 0.5, 1.5 and 5 mg/mL exogenous flavonoid-treated eggplant plants and the control plants was found at RT (Figure S10A). After exposure to HTT for 4 days, the leaves of the control plants presented wilting symptoms, whereas the leaves of the 5 mg/mL exogenous flavonoid-treated eggplant plants remained robust (Figure S10A). The chlorophyll fluorescence of the control plants was similar to that of the exogenous flavonoid-treated eggplant plants at RT. However, the chlorophyll fluorescence of the 5 mg/mL exogenous flavonoid-treated eggplant plants remained stable after HTT for 4 days compared with other eggplants plants (Figure S10B). The Fv/Fm, Fq'/Fm' and rETR of the exogenous flavonoid-treated eggplant plants were significantly greater than those of the control plants after HTT for 4 days (Figure S10C–E), while the NPQ of the 5 mg/mL exogenous flavonoid-treated eggplant plants was lower than that of the control plants after HTT for 4 days (Figure S10F). In addition, the heat injury index of the exogenous flavonoid-treated eggplant plants was lower than that of the control plants (Figure S10G) after HTT for 4 days, and the MDA content and REL of the exogenous flavonoid-treated eggplant plants were significantly lower than those of the control plants (Figure S10H and S10I).

Furthermore, the SmF3H-silenced eggplant plants were treated with 5 mg/mL exogenous flavonoids. Compared with those of SmF3H-silenced eggplant plants, the leaves of the control plants and the SmF3H-silenced eggplant plants treated with exogenous flavonoids remained robust after HTT for 4 d (Figure 7A). The chlorophyll fluorescence, Fv/Fm, Fq'/Fm' and rETR of the control plants and the SmF3H-silenced eggplant plants treated with exogenous flavonoids were significantly greater than those of the SmF3H-silenced eggplant plants after HTT for 4 days (Figure 7B–E), whereas thier NPQ was lower than that of the SmF3H-silenced eggplant plants (Figure 7F). The MDA content and REL of the control plants and the SmF3H-silenced eggplant plants treated with exogenous flavonoids were significantly lower than those of the SmF3H-silenced eggplant plants after HTT for 4 days (Figures 7G,H). In addition, the flavonoid content was significantly lower in the SmF3H-silenced eggplant plants than in the control plants and the SmF3H-silenced eggplant plants treated with exogenous flavonoids (Figure 7I). These results indicated that flavonoids are involved in regulating the heat tolerance of eggplant.