S-HsfA2 negatively regulates extreme heat tolerance and growth in Arabidopsis.

(A) Schematic diagrams of HsfA2-III and its corresponding product S-HsfA2 (top). Mutations in the LxLxLx motif (underlined) generated the dominant version of S-HsfA2 (S-HsfA2L-A). The predicted AlphaFold 3D structure of S-HsfA2 (AFDB accession: AF-J7G1E7-F1) is shown (bottom). (B) Thermotolerance of the WT control, S-HsfA2-overexpressing (S-HsfA2-OE) lines, and S-HsfA2L-A-overexpressing (S-HsfA2L-A-OE) lines in both medium (top) and soil (bottom). A representative image is shown. All green cotyledons were considered surviving, and all yellow cotyledons were considered dead. The survival rates are presented as the means ± SDs of three independent experiments. Different letters indicate significant differences according to one-way ANOVA, P < 0.05. (C) RNAi-mediated S-HsfA2 knockdown (targeting the retained intron sequences; see 1A; S-HsfA2-KD) (left) and thermotolerance analyses of S-HsfA2-KD lines (right). Three S-HsfA2-KD lines and the WT control at 12 DAS were treated at 42°C for 1 h and then subjected to RT-PCR splicing analysis and RT-qPCR analysis. YELLOW-LEAF-SPECIFIC GENE 8 (YLS8) served as a loading control. The data are presented as the means ± SDs of two independent experiments. The significance of differences between the experimental values was assessed by Student’s t test (*P <0.05). (D) S-HsfA2-OE resulted in reduced root length (left) and a dwarf phenotype (right) in Arabidopsis seedlings. Mutations (L to A) in the LxLxLx motif (S-HsfA2L-A-OE) partially rescued these growth defects. The data are presented as the means ± SDs of at least two independent experiments. Different letters indicate significant differences according to one-way ANOVA, P < 0.05. DAS, days after sowing.

Comparative transcriptome analysis between S-HsfA2-OE and S-HsfA2L-A-OE revealed differentially expressed heat-responsive genes.

(A) Experimental scheme for use in mRNA-Seq analysis. (B) The number of HRGs (42°C/22°C) for S-HsfA2-OE and S-HsfA2L-A-OE. (C) Venn diagram showing 848 HRGs shared between S-HsfA2-OE and S-HsfA2L-A-OE, in which differentially expressed HRGs are shown on the left. (D) Bubble diagram showing the top 15 pathways associated with the 848 shared HRGs according to the GO enrichment analyses. The X-axis represents the enrichment factor (the ratio of the number of genes enriched in the GO pathway or GO term to the number of annotated genes), and the Y-axis represents the name of the pathway. The bubble size represents the number of HRGs involved. The bubble color indicates the enrichment degree (p value) of the pathway. (E) Heatmap showing the expression profiles (S-HsfA2-OE was set as 1) of selected representative differentially expressed HRGs involved in heat tolerance. DAS, days after sowing.

S-HsfA2 specifically binds the HRE identified by ChIP-seq in 35S:S-HsfA2-GFP Arabidopsis plants.

(A) Representative images of the subcellular localization of S-HsfA2-GFP in the root cells of two independent transgenic Arabidopsis lines (#4 and #6) treated without (normal) or with heat stress (42°C for 1 h). GFP was used as a negative control. Scale bar, 50 μm. Compared with the WT control, the two lines presented a heat stress-sensitive phenotype (right). (B) ChIP-Seq analysis using an anti-GFP antibody in two 35S:S-HsfA2-GFP lines revealed a centrosymmetric 7-bp motif and 80 putative target genes of S-HsfA2. (C) S-HsfA2 binding to HREs was verified by EMSA and Y1H assays. The mutated HRE (mHRE) was used as a negative control. Unlabelled HRE is used as a competitor. For the EMSAs, the specific binding signal is marked with an arrow, whereas the asterisk indicates a nonspecific signal. EMSAs were repeated two times, and typical images are shown. (D) The production of His6-DBD and His6-tDBD was confirmed by immunoblotting with an anti-His6 antibody (left). Mock proteins (empty vector control) were used as a negative control. Coomassie Brilliant Blue (CBB)-stained proteins are shown as loading controls. The binding of His6-DBD and His6-tDBD to HSE or mutated HSE (mHSE) was verified by EMSA (right). The HRE and HSE sequences are shown, and the corresponding mutations are shown in lowercase. DAS, days after sowing.

HRE confers heat induction in Arabidopsis.

(A) Qualitative GUS histochemical staining of GUS reporter transgenic Arabidopsis seedlings. At least twenty seedlings were analysed, and typical images are shown. Scale bar, 0.5 mm or 1 mm for HRE-35Sm:GUS (#2). Heat-induced GUS activity analyses of two HRE-35Sm:GUS transgenic Arabidopsis lines at the GUS mRNA level via RT-qPCR (B) and at the GUS protein level via Western blotting with anti-GUS and anti-Tubulin antibodies (C). For the GUS mRNA level, the control (22°C) was set as 1. The GUS protein level is expressed as the relative band intensity of GUS to tubulin (the control (22°C) was set as 1). (D) Heat time course analysis of the GUS mRNA level in HRE-35Sm:GUS transgenic Arabidopsis plants. In (B), (C), and (D), the data are presented as the means ± SDs of at least two independent experiments. Different letters indicate significant differences according to one-way ANOVA, P < 0.05. (E) Western blot analysis verified the successful isolation of INPUT nuclear protein (NP) without cytoplasmic protein (CP) contamination via an anti-histone H3 (nuclear marker) antibody and an anti-glucose pyrophosphorylase (UGPase) (cytoplasmic marker) antibody (top). EMSAs revealed that HRE probes bound to NPs from heat shock-treated (37°C, 42°C) or untreated (22°C) Arabidopsis plants (bottom). EMSAs were repeated two times, and typical images are shown. The bound complex is indicated by an arrow. (F) Abundances of GUS protein in crossed GUS reporter plants harboring 35S:S-HsfA2-Flag or 35S:S-HsfA2L-A-Flag effector (+effector) or an uncrossed reporter control (-effector) were determined by immunoblotting assays using anti-GUS and anti-Tubulin antibodies. The relative GUS expression level in the −effector was set as 1. The data represent the means from two independent experiments.

S-HsfA4c is similar to S-HsfA2 and represents a new HSF.

(A) Schematic diagrams of HsfA4c splice variants and their RT-PCR splicing analyses in two-week-old Arabidopsis plants under heat stress. The YLS8 gene served as a loading control. gDNA, genomic DNA control. The asterisk indicates the in-frame stop codon. S-HsfA4c binding to the HRE was verified by Y1H (B) and EMSA with bacterially expressed and purified His6-S-HsfA4c (C). The above experiments were performed two times, each yielding similar results. (D) Thermotolerance of the 35S:GFP vector control and 35S:S-HsfA4c-GFP overexpression (S-HsfA4c-OE) lines. A representative image is shown (left). The survival rates of the seedlings are presented as the means ± SDs of two independent experiments. Different letters indicate significant differences according to one-way ANOVA, P < 0.05. (E) Thermotolerance of the WT control and two T-DNA HsfA4c insert-knockdown mutants (hsfa4c-1 and hsfa4c-2). Abundances of HsfA4c-I and HsfA4c-II in hsfa4c-1 and hsfa4c-2 confirmed by RT-qPCR (top), a representative image (middle), and survival rate analysis (bottom). The data are presented as the means ± SDs of three independent experiments. Different letters indicate significant differences according to one-way ANOVA, P < 0.05. (F) Representative images of the subcellular localization of S-HsfA4c-GFP in Arabidopsis root cells treated without (normal) or with heat stress (42°C for 1 h). DAPI, 4,6-diamidino-2-phenylindole (DAPI; nuclei staining). Scale bar, 40 μm. DAS, days after sowing.

Heat-responsive HSP17.6B is regulated by S-HsfA2 and confers tolerance in Arabidopsis.

(A) Heat-induced GUS expression assays of Arabidopsis plants harboring a series of HSP17.6Bp-driven GUS reporter transgenes via GUS histochemical staining (top) and heat-induced GUS mRNA level analysis via RT-qPCR (bottom). For each GUS reporter transgene, mixed Arabidopsis plants (10 DAS) from three independent transgenic lines were used. The data are presented as the means ± SDs of two independent experiments (the WT control was set as 1). Different letters indicate significant differences according to one-way ANOVA, P < 0.05. (B) Schematic diagrams of the HSP17.6B gene (top) showing HRE-HRE-like and ChIP-qPCR products. ChIP-qPCR experiments were performed with an anti-Flag antibody and mouse IgG (mock control) in 35S:S-HsfA2-Flag (S-HsfA2-OE) Arabidopsis plants. (C) Relative expression levels of HSP17.6B (the WT control was set as 1) in S-HsfA2-OE and S-HsfA2-KD plants were analysed via RT-qPCR under heat (45°C for 3 h) stress. (D) Relative expression levels of HSP17.6B (the WT control was set as 1) in the T-DNA mutant (hsp17.6b-1), three lines of 35S:HSP17.6B-Flag/hsp17.6b-1 (HSP17.6B-KI), and two lines of 35S:HSP17.6B-Flag (HSP17.6B-OE) were determined via RT-qPCR under heat (37°C for 2 h) stress. Schematic diagrams of 17.6b-1 are shown (top). (E) The survival rates of the WT control, hsp17.6b-1, and HSP17.6B-OE and HSP17.6B-KI plants were analysed after heat stress (a representative image is shown on the left). (F) HSP17.6B-OE plants (15 DAS) presented defects in growth, as indicated by decreased chlorophyll content and fresh weight (left). The dwarf but heat-resistant phenotype of the HSP17.6B-OE plants grown on soil (right). In (C) to (F), the data are presented as the means ± SDs of at least two independent experiments. The significance of differences between the experimental values was assessed by Student’s t test (*P <0.05). DAS, days after sowing.

S-HsfA2 serves as a negative binding regulator of HsfA2 by interacting with the DBD of HsfA2.

(A) Five HSFs that interact with S-HsfA2 were identified via Y2H analysis of 19 Arabidopsis HSFs. (B) GST pull-down verified that S-HsfA2 interacted with HsfA2 in vitro. (C) Further Y2H and BiFC assays confirmed that S-HsfA2 interacted with the DBD of HsfA2. (D) Schematic diagram of the streptavidin-agarose bead pull-down assay used to verify the prebinding of GST-S-HsfA2 with His6-HsfA2 to decrease the biotin-labelled HSE bait-binding capacity of His6-HsfA2. The data are the means of two independent experiments (the control (0 h) was set as 1). (E) Dual-luciferase (LUC) assays in tobacco leaves. A representative bioluminescence image of the sHSP17.6Bp:LUC reporter coexpressing the indicated effector groups. The effector-driven reporter expression activity was expressed as the relative ratio of LUC activity to Renilla luciferase (REN) activity (the GUS effector control was set as 1). The data are presented as the means ± SDs of three independent experiments. Different letters indicate significant differences according to one-way ANOVA, P < 0.05.

A proposed model for S-HsfA2 balancing Arabidopsis heat tolerance and growth.

Under normal conditions, HsfA2 and its target HSP17.6B gene are not expressed. However, these two genes are induced by heat stress. Furthermore, HsfA2 undergoes alternative splicing under severe heat stress to generate the splicing transcript HsfA2-III, which is translated to S-HsfA2. The overexpression of HSP17.6B not only confers thermotolerance but also retards seedling growth, reflecting the hyperactivation of the HSR mediated by HSP17.6B overexpression. S-HsfA2 balances the gain (extreme heat tolerance, EHT) and loss (growth retardation, GR) of HsfA2-regulated HSP17.6B overexpression to ensure proper growth in two ways: ① antagonistic repression of HSP17.6B overexpression through a noncanonical HSR and ② a negative binding regulator of HSFs that inactivates the HSP17.6B promoter-binding capacity of HsfA2.