The roles of Arabidopsis HSFA2, HSFA4a, and HSFA7a in the heat shock response and cytosolic protein response

Previously, we found that Arabidopsis plants transformed with a construct containing the promoter of Oshsp17.3 from rice fused to the β-glucuronidase gene (GUS), Oshsp17.3Pro::GUS (Oshsp17.3p), showed a GUS signal after heat shock (HS) or azetidine-2-carboxylic acid (AZC) treatment. HS and AZC trigger the heat shock response (HSR) and cytosolic protein response (CPR), respectively, in the cytosol by modulating specific heat shock factor (HSF) activity. Here we further identified that AtHSFA2 (At2g26150), AtHSFA7a (At3g51910), AtHSFB2a (At5g62020), and AtHSFB2b (At4g11660) are HS- and AZC-inducible; AtHSFA4a (At4g18880) is AZC-inducible; and AtHSFA5 (At4g13980) is less AZC- and HS-inducible. To investigate the roles of these 6 AtHSFs in the HSR or CPR, we crossed two independent Oshsp17.3p transgenic Arabidopsis plants with the AtHSF-knockout mutants athsfa2 (SALK_008978), athsfa4a (GABI_181H12), athsfa5 (SALK_004385), athsfa7a (SALK_080138), athsfb2a (SALK_137766), and athsfb2b (SALK_047291), respectively. As compared with the wild type, loss-of-function mutation of AtHSFA2, AtHSFA4a, and AtHSFA7a decreased HS and AZC responsiveness, so these 3 AtHSFs are essential for the HSR and CPR. In addition, loss-of-function results indicated that AthsfB2b is involved in regulating the HSR in Arabidopsis. Furthermore, analysis of the relative GUS activity of two double knockout mutants, athsfA2/athsfA4a and athsfA2/athsfA7a, revealed that AtHSFA2, AtHSFA4a, and AtHSFA7a function differentially in the HSR and CPR. Transcription profiling in athsf mutants revealed positive or negative transcriptional regulation among the 6 AtHSFs in Arabidopsis plants under HS and AZC conditions. Tunicamycin treatment demonstrated that these 6 AtHSFs are not involved in the unfolded protein response.


Background
Protein homeostasis is crucial for maintaining normal cellular function. Plants, being sessile organisms, cannot escape from their growing environments. Extremes in environmental factors can result in stressful conditions that inevitably damage proteins directly or cause cells to synthesize misfolded proteins, which can lead to perturbed cell function and stress-induced cell death. Plants have evolved an extensive network of chaperone systems to restore protein folding or to remove irreversibly unfolded proteins (Mehdy 1994;Shinozaki and Yamaguchi-Shinozaki 1996;Bukau et al. 2006;Cramer et al. 2011;Redondo-Gómez 2013).
Accumulation of unfolded proteins within cells, eliciting compartment-specific chaperones and pathways, is termed the unfolded protein response (UPR). The UPR initiates the dissociation of the endoplasmic reticulum (ER) chaperone, immunoglobulin binding protein, and ER master sensors, such as inositol-requiring 1 and protein kinase R-like ER kinase, to activate downstream effectors to restore protein homeostasis in the lumen of the ER. A cytosolic process, the cytoplasmic protein response (CPR), increases the synthesis of molecular chaperones such as heat shock proteins (HSPs). In contrast to the better-understood UPR of the ER, the regulatory molecules in the CPR are not well elucidated.
The heat-shock response (HSR), predominantly a response to maintain protein-folding homeostasis in the cytosol, causes transcriptional activation of HSPs under thermal stress (Aparicio et al. 2005;Jungkunz et al. 2011). The expression of HSP genes is mainly under the control of heat shock transcription factors (HSFs) (Schöffl et al. 1998;Nover et al. 2001). The number of HSFs is characteristically higher in plants than in other organisms. For example, Arabidopsis and rice have 21 and 25 HSFs, respectively, but Drosophila, C. elegans and yeast have only one HSF (Nover et al. 2001;Guo et al. 2008;Scharf et al. 2012). The multiplicity of members of the HSF family in plants may contribute to their fitness to face varied environmental challenges such as extreme temperatures, drought, and salinity (Busch et al. 2005).
Plant HSFs are classified into three classes (A, B, and C) on the basis of structural characteristics and phylogenetic comparison. Class A HSFs contain a DNA binding domain, an oligomerization domain, nuclear localization domains, and transcriptional activation domains. Classes B and C lack a transcriptional activation domain (Nover et al. 2001). Recent studies of tomato HSFA1a mutants and an Arabidopsis HSFA1a/1b/1d/1e quadruple mutant revealed that members of HSFA1 genes can function as master regulators for the HSR and play important roles in cross-regulation for abiotic stress responses (Mishra et al. 2002;Liu et al. 2011). Increasing evidence shows functional diversification among different HSF members.
In addition to heat shock (HS), a proline analog, azetidine-2-carboxylic acid (AZC), can induce accumulation of abnormal-misfolded proteins in the cytosol to trigger the CPR by modulating HSFA2 activity (Yeh et al. 2007;Sugio et al. 2009;Nishizawa-Yokoi et al. 2011). In the current study, we fused the promoter of AZC-inducible small HSP (sHSP), Oshsp17.3, with the β-glucuronidase gene (GUS) (Oshsp17.3Pro::GUS) and transformed into Arabidopsis AtHSF mutants, and detected GUS activity in response to AZC and HS (Guan et al. 2010). Our results allowed us to characterize the roles of Arabidopsis HSFs in the HSR and CPR.

Plant materials
The Arabidopsis thaliana ecotype Col-0 was used in this study as the wild type (WT). Seeds were surfacesterilized in commercial bleach that contained 5% (v/v) sodium hypochlorite and 0.1% (v/v) Triton X-100 solution for 10 min, rinsed in sterilized water, and stratified at 4 °C for 2 days in the dark. Seeds were germinated on growth agar plates [1/2 Murashige and Skoog medium (MS; Duchefa), 1% sucrose (w/v), 0.8% agar (w/v)].

RNA isolation and RT-PCR
Total RNA was extracted from 10-day-old Arabidopsis seedlings as described (Guan et al. 2010). The firststrand cDNA was synthesized with 1 μg total RNA by using the SuperScript III First-Stand Synthesis System (Invitrogen). PCR amplification corresponding to different AtHSFs shown in Fig. 3 were 30 s at 94 °C, 30 s at 52 °C, and 30 s at 72 °C, then 5 min at 72 °C. Primers used for analysis of gene expression were designed by use of NCBI Primer-BLAST (https ://www.ncbi.nlm.nih.gov/ tools /prime r-blast /) and are in Table 1. DNA from 15 μl of each PCR reaction was fractionated by electrophoresis on 1.2% (w/v) agarose gel with 0.01% (w/v) ethidium bromide in 1× Tris-Acetate EDTA buffer. The gel was Table 1 Oligonucleotides used in RT-PCR

Gene
Primer name Sequence

Stress treatment of transgenic Arabidopsis mutants
For HS treatment, 10-day-old F3-generation Arabidopsis seedlings were incubated in shaking buffer [1% sucrose (w/v), 5 mM potassium phosphate buffer, pH 6.8] at 39 °C for 1 h, then 23 °C for 20 h of recovery. For AZC treatment, 10-day-old F3-generation Arabidopsis seedlings were incubated in shaking buffer with or without 5 mM AZC at 23 °C for 4 h, rinsed in sterilized water, then incubated in shaking buffer at 23 °C for 15 h of recovery. For tunicamycin (Tm) treatment, 10-day-old F3-generation Arabidopsis seedlings were incubated in shaking buffer with or without 5 μg/ml Tm at 23 °C for 4 h, rinsed in sterilized water, then incubated in shaking buffer at 23 °C for 15 h of recovery. All samples were frozen by liquid nitrogen and stored at − 80 °C.

GUS staining
GUS staining was described previously (Guan et al. 2010). In brief, 10-day-old seedlings were treated and incubated in the fixation solution (0.3% formaldehyde, 0.1% Triton X-100, 0.1% β-mercaptoethanol, 100 mM sodium phosphate buffer, pH 7.0) for 60 min. Then the fixation solution was replaced with washing solution (100 mM sodium phosphate buffer, 1 mM EDTA, pH 7.0) twice for 15 min. Washed seedlings were vacuum-infiltrated for 5 min in GUS staining buffer (1 mM X-Gluc, 0.5 mM ferricyanide, 0.5 mM ferrocyanide, 0.1% Triton X-100, 10 mM EDTA, 100 mM sodium phosphate buffer, pH 7.0), then incubated at 37 °C for 24 h. The staining reaction was stopped by adding distilled water, the color of chlorophyll was removed with 70% ethanol (v/v) several times, and seedlings were soaked in 95% ethanol (v/v) for 1 h. Plants were photographed to record deposition of the GUS.

Statistical analysis
Data are shown as mean ± SE from three independent experiments. Statistical differences were analyzed by Student t test or Duncan multiple range test. P < 0.05 was considered statistically significant.

AtHSFA2, AtHSFA4a, and AtHSFA7a are not responsive to Tm
AZC typically induces the UPR and CPR. The data in Fig. 2 indicated that AtHSFA2, AtHSFA4a, and AtHS-FA7a are essential for the HSR and AZC response in Arabidopsis. Studies have shown AtHSFA2 as a crucial regulatory component of the CPR (Sugio et al. 2009). To understand whether these AtHSFs are involved in the UPR, we examined the effect of Tm treatment (UPR induction) in the AtHSF mutants tested. Tm did not activate the expression of the 6 AtHSF genes (Fig. 4a). On GUS analysis, no Tm responsiveness was detected in the mutant plants tested (Fig. 4b, c). These results confirmed that AtHSFA2, AtHSFA4a, and AtHSFA7a function in the CPR.

Discussion
To adapt to biotic and abiotic stresses, plants have evolved a complex set of molecular responses, which often exhibit features sharing substantial overlap pathways and components. HSF/HSP responses are recognized as central chaperone components against unfolded protein accumulation, a signal for triggering HSR, UPR, or CPR based on distinct subcellular localization (Aparicio et al. 2005;Swindell et al. 2007;Yeh et al. 2007). Many reports have shown that HSFs are important for resistance to heat and other environmental stresses (Mishra et al. 2002;Charng et al. 2007;Banti et al. 2010;Liu et al. 2011). Using an HS-and AZC-sensitive promoter-GUS fusion system (Guan et al. 2010) together AtTubulin level was an internal control. Two biological repeats were performed, and similar results were obtained. b, c Relative GUS activity of seedlings treated with Tm. Data are mean ± SE GUS activity relative to that of non-stress control condition (Ctrl) from three independent experiments with knockout plants, we aimed to identify the contribution of AtHSFA2, AtHSFA4a, AtHSFA5, AtHSFA7a, AtHSFB2a, and AtHSFB2b to the responses induced by HS, AZC, and Tm. Plant HSFs are regulated by HS and AZC, including up-and downregulation. We found the expression of AtHSFA2, AtHSFA4a, AtHSFA7a, AtHSFB2a, and AtHS-FB2b induced > twofold with 1-h HS treatment and then reduced after prolonged heat incubation (Fig. 1). As well, AZC upregulated AtHSFA2, AtHSFA4a, AtHSFA7a, AtHSFB2a, and AtHSFB2b expression > 2.9-fold during treatment. However, Tm did not affect the expression of the 6 AtHSFs (Fig. 4a). Despite a slight difference in plant material and treatment time, the results are similar to published microarray data (Busch et al. 2005;Schramm et al. 2008;Sugio et al. 2009), finding that AtHSFA2, AtHSFA4a, AtHSFA7a, AtHSFB2a, and AtHSFB2b are important for stress response networks.
Studies have shown that AtHSFA2 and AtHSFA7a knockout mutants lose acquired thermotolerance, and AtHSFA2 mutants also show reduced tolerance to AZC (Charng et al. 2007;Siddique et al. 2008;Sugio et al. 2009). In this study, loss-of-function mutation of AtHSFA2 significantly repressed relative GUS activity under HS and AZC treatment (Fig. 2b-e). By contrast, null mutation of AtHSFA4a and AtHSFA7a only slightly repressed relative GUS activity under HS and AZC stress. These results agree with others showing that AtHSFA2 is closely related to the regulation of HSR as well as CPR (Busch et al. 2005;Nishizawa et al. 2006;Ogawa et al. 2007;Sugio et al. 2009;Jung et al. 2010), whereas AtHS-FA4a and AtHSFA7a have a lesser effect on HSR and CPR. Furthermore, as compared with AtHSFA2 knockout alone, double knockout with AtHSFA2 and AtHS-FA4a or AtHSFA7a showed more significant repression of HS-induced GUS activity (Fig. 2b-e). Thus, AtHSFA2, AtHSFA4a, and AtHSFA7a may be linked to activation of different target genes/pathways in the HSR. However, AtHSFA2 appears to be a functionally redundant factor to AtHSFA4a and AtHSFA7a for AZC-induced CPR because the GUS activity of AtHSFA2-knockout plants was similar to that with double knockout of AtHSFA2 and AtHSFA4a or AtHSFA7a under AZC treatment (Fig. 2b-e). Ikeda et al. (2011) reported that AtHsfB1 and AtHs-fB2b, sharing functional redundancy in repressive activities, were able to suppress the accumulation of AtHSFA2 and AtHSFA7a transcripts and were indispensable for acquired thermotolerance. As compared with AtHSFA2 knockout, AtHsfB2b knockout slightly repressed GUS activity in response to HS treatment (Fig. 2b, c). We also revealed no significant change in HS-induced AtHSFA2 and AtHSFA7a transcript levels with AtHsfB2b knockout (Fig. 3a). These results suggest that AtHSFB2b may mediate the HSR but not CPR. Of note, AtHSFB2a is highly AZC-and HS-inducible, but we did not find a significant reduction in GUS activity with AtHsfB2a knockout during AZC treatment. However, we cannot absolutely exclude the role of AtHsfB2a in AZC-induced CPR because of its high expression under AZC and HS treatment.
In conclusion, we confirmed and characterized the roles of AtHSFA2, AtHSFA4a, AtHSFA5, AtHSFA7a, AtHSFB2a, and AtHSFB2b in the HSR and CPR. For simplifying our result, we propose a working model to show the roles of following AtHSFs in CPR and HSR (Fig. 5). AtHSFA2, AtHSFA4a, and AtHSFA7a function independently in the HSR, but AtHSFA2 may function redundantly with AtHSFA4a and AtHSFA7a in the CPR. AtHSFB2b has some role in mediating the HSR, and AtHSFA5 and AtHSFB2a cannot mediate the HSR and CPR. These 6 AtHSFs are not involved in the UPR.
Authors' contributions C-AL, S-JW, and C-HY conceived the concept and designed the experiment. K-FL and M-YT have equally contributed towards this manuscript. K-FL and M-YT performed the experiments and analyzed the data. C-HY wrote the manuscript. All authors read and approved the final manuscript.