• Open Access

A novel membrane-bound E3 ubiquitin ligase enhances the thermal resistance in plants

Authors


Summary

High temperature stress disturbs cellular homoeostasis and results in a severe retardation in crop growth and development. Thus, it is important to reveal the mechanism of plants coping with heat stress. In this study, a novel gene that we identified from Brassica napus, referred to as BnTR1, was found to play a key role in heat stress response in planta. BnTR1 is a membrane-bound RINGv (C4HC3) protein that displays E3 ligase activity in vitro. We demonstrated that modest expression of BnTR1 is sufficient to minimize adverse environmental influence and confers thermal resistance on development without any detrimental effects in B. napus and Oryza sativa. Our investigation into the action mechanism indicates that BnTR1 is likely to be involved in mediating Ca2+ dynamics by regulating the activity of calcium channels, which further alters the transcripts of heat shock factors and heat shock proteins contributing to plant thermotolerance. Hence, our study identified BnTR1 as a novel key factor underlying a conserved mechanism conferring thermal resistance in plants.

Introduction

Recently, increased temperatures caused by global warming have begun to show negative effects on agricultural productivity around the world (Battisti and Naylor, 2009). Hence, developing thermally resistant crops has become an eminent challenge for researchers in plant biology and agriculture (Atkinson and Porter, 1996). The most effective strategy is to alter critical genes involved in thermal resistance so that the capability of the plant to heat stress can be boosted. Recent investigations indicate that plants have developed various mechanisms, for example, scavenging of reactive oxygen species (ROS) and production of antioxidants, maintenance of membrane stability, accumulation and adjustment of compatible solutes and heat shock protein (HSP)-mediated chaperone signalling and transcriptional activation, to cope with heat stress (Wahid et al., 2007). Accordingly, many genes have been identified to confer heat resistance when transferred into plants; for example, up-regulated ascorbate peroxidase (APX), glutaredoxin (GRX), heat shock factors (HSFs), HSPs and down-regulated FAD7 (omega-3 fatty acid desaturase) are all contributed to heat shock resistance in plants (Murakami et al., 2000; Panchuk et al., 2002; Shi et al., 2001; Sung et al., 2003; Wu et al., 2012).

However, most of the previous studies have mainly focused on downstream events of heat stress response (HSR) mechanism including HSFs and HSPs (Kotak et al., 2007). Based on the primary roles of HSP90, heat shock cognate 70 (HSC70) and some unspecified heat-labile proteins in the cytoplasm (Shi et al., 1998; Sung and Guy, 2003; Voellmy and Boellmann, 2007), the general model for heat shock sensing in plants established the link between heat stress and the induction of HSPs. However, the upstream components involved in heat stress signalling initiation are still unclear. Recently, an alternative complementing heat-sensing mechanism suggests that the primary temperature sensor of the cell is located in the plasma membrane and that Ca2+-permeable channels act as the earliest temperature-sensing component of the plant HSR (Saidi et al., 2009).

High temperatures induce changes in membrane fluidity and activate some unknown Ca2+ channels, resulting in an influx of Ca2+ (Saidi et al., 2009). However, the reason for the rise in intracellular Ca2+ ([Ca2+]cyt) after heat stress is not well understood. In animal cells, the [Ca2+]cyt levels increase by the inflow of extracellular Ca2+ through ion channels in the plasma membrane. Recent works on transient receptor potential (TRP), encoding cation channels, imply a linkage between proteins located in the plasma membrane and thermotolerance in Drosophila (Peier et al., 2002; Voets et al., 2005). In plants, HSRs are supposed to be initiated from the plasma membrane (Saidi et al., 2009). Recently, specific cyclic nucleotide-gated ion channels (CNGCs) in the plasma membrane are found to be involved in thermosensing, HSPs expression and Arabidopsis thermotolerance (Finka et al., 2012; Gao et al., 2012). All of these indicate that proteins in the plasma membrane play critical roles in HSRs. However, their functions and regulatory mechanisms at the molecular level remain largely elusive.

In this study, we identified that a novel yet-conserved gene in Brassica napus, referred to as Brassica napus thermal resistance gene 1 (BnTR1), is induced upon heat stress. BnTR1 belongs to a new type of membrane-bound RINGv (Really Interesting New Gene Variant) E3 ligase, and its homologues are conserved and widely spread in dicots and monocots. The effect of the gene in various plants suggests that BnTR1, when expressed at a modest level, is sufficient to minimize adverse environmental influence and confers strong resistance to high temperature in multiple plant species without affecting their normal development. Hence, BnTR1 is likely to be a critical gene involved in a conserved thermal resistance mechanism in plants.

Results

Identification of BnTR1 as a temperature-induced gene

To identify genes that are responsive to increased temperatures in Brassica napus, the comparison of transcriptional variations between 15-day-old leaves growing under heat stress (35 °C) with those under normal condition (22 °C) was made using the mRNA differential display technique. Twenty-six up-regulated and nine down-regulated transcripts were identified. An unannotated gene, whose transcript was up-regulated by the increased temperature, was cloned by rapid amplification of the 5′ cDNA end (5′ RACE) and named as BnTR1.

The open reading frame of the gene is made up of 861 base pairs, which encodes a polypeptide of 286 amino acid residues. Sequence analysis indicated that BnTR1 (accession no: GU204975) is a highly conserved protein, which shares 88%, 62%, 57% and 54% identity, respectively, to its homologues in Arabidopsis thaliana (At3 g47550), Ricinus communis (RCOM_1472160), Oryza sativa (Os08 g0436200) and Zea mays (LOC100281338) (Figure S1). In addition, phylogenetic tree analysis based on a neighbour-joining (NJ) bootstrap (Figure 1a) indicates that BnTR1 belongs to a novel gene family conserved in plants.

Figure 1.

BnTR1 belongs to a novel C4HC3 RING finger gene family. (a–c) BnTR1 and its homologues in higher plants. (a) Phylogenetic tree of BnTR1 protein in Brassica napus with its homologues in plants. The evolutionary history was inferred using the neighbour-joining method. Phylogenetic analysis was conducted in MEGA4. (b) Scheme of putative domains in BnTR1. RINGv, really interesting new gene variant domain; TMD, transmembrane domain. (c) Amino acid sequence of BnTR1. BnTR1 contains one RINGv (as underline indicated) and two putative transmembrane domains (TMD, as grey indicated). (d) E3 ubiquitin ligase activity of BnTR1. GST-BnTR1 fusion proteins were assayed for E3 activity in the presence of E1 (UBA1), E2 (UBCh5a) and ubiquitin (Ub). GST itself was used as a negative control. Protein bands with ubiquitin attached were detected by a ubiquitin antibody. kD, kilodalton.

BnTR1 is a functional E3 ligase

A NCBI Conserved Domain Search revealed that the BnTR1 protein contains a RING-variant domain (Really Interesting New Gene variant, SMART NO. 00744), which is a C4HC3 zinc-finger-like motif that is found in a number of cellular and viral proteins (Figure S2). Some of these proteins were shown to have the ubiquitin E3 ligase activity (Coscoy et al., 2001; Hoer et al., 2007; Nakamura et al., 2006). Therefore, we speculate that BnTR1 has E3 ligase activity as well. To ascertain this notion, we examined the potential ligase activity of BnTR1 using an in vitro ubiquitination assay. BnTR1 was expressed in Escherichia coli as GST-fused protein and then purified. The recombinant BnTR1-GST or GST control was incubated with UBA1, a human ubiquitin-activating enzyme (E1), and UBCh5a, a human ubiquitin-conjugating enzyme (E2). Ubiquitination activity was observed in the presence of purified BnTR1-GST, but not of GST (Figure 1d), indicating that BnTR1 possesses E3 ligase activity in vitro.

Subcellular localization of BnTR1

Sequence analysis revealed that there are two putative transmembrane domains (TMDs) located at the C-terminus of BnTR1, from residue 176 to 197 and 216 to 236 (Figure 1b,c), suggesting that BnTR1 may be a membrane protein. Hence, we examined the expression pattern and subcellular localization of BnTR1 in B. napus by immunofluorescent microscopy. We found that BnTR1 was expressed in all parts of the plant, including roots (Figure 2a), stems (Figure 2b) and leaves (Figure 2c). Consistent with this, at transcript level, BnTR1 was also found in different plant tissues (Figure 3a). At cellular level, BnTR1 was predominantly distributed at the cell periphery. Immunohistochemical analysis indicated that BnTR1 is a membrane protein (Figure 2d), while the control did not show any signal (Figure 2e). Consistent with the above findings, we found an eGFP-fused recombinant BnTR1 (BnTR1-eGFP), when transiently expressed in Arabidopsis protoplasts by PEG-mediated method, was densely localized at plasma membrane (Figure 2f and h), while the control eGFP was dispersed throughout the whole cell (Figure 2g and i).

Figure 2.

Subcellular localization of BnTR1. (a–c)Immunofluorescence analysis of BnTR1 expressing in various tissues of Brassica napus: root (a), stem (b) and leaf (c). Subcellular localization of BnTR1 was detected by indirect immunofluorescence using anti-BnTR1 antibody followed by FITC-conjugated secondary antibody and visualization under fluorescence microscope, and nuclei were stained with DAPI (blue). (d–e) The immunohistochemical staining of BnTR1 in the stem of B. napus. (d) Incubated with rabbit anti-BnTR1. (e) Incubated without rabbit anti-BnTR1. (f–i) Subcellular localization of BnTR1-eGFP fusion in Arabidopsis protoplasts. Cells were analysed by fluorescence microscopy and photographed after 16 h of incubation at 22 °C following PEG-mediated transient expression. (f) BnTR1-eGFP fusion and (g) eGFP as control under blue light. (h) BnTR1-eGFP fusion and (i) eGFP under white light. The scale bar, 200 μm. Red arrows indicate location of BnTR1.

Figure 3.

The expression of BnTR1. (a) The expression of BnTR1 in different tissues. Roots, stems and leaves of 4-week-old Brassica napus, and flowers just before anthesis were harvested for RNA-extracted and run q-PCR. The data are means of four separate runs, and SDs are indicated. (b–d) The expression of BnTR1 is sensitive to heat stress. (b) Schematic representation of the BnTR1 promoter. The right arrow indicated the initial transcriptional start sites. HSE, heat stress-responsive element; the luciferase gene substituted the sequence of BnTR1 after its transcription start point. (c) Level of BnTR1 transcripts induced by heat stress. Total RNAs were isolated from 4-week-old leaves of B. napus treated at the indicated temperatures and duration and then analysed by RT-PCR. The transcripts of GADPH were as a control. HS, heat stress; N, normal condition. (d) Regulation of BnTR1 promoter (pBnTR1) activity by heat stress in Arabidopsis protoplasts. The protoplasts were treated with 30 °C (filled squares) as heat stress and normal temperature 22 °C (empty triangles) as control for a continuous period. The expression of pBnTR1-LUC was indicated and normalized to GUS expression. The data represented the mean value of three individual experiments (SD).

Induction of BnTR1 by heat stress

The finding that BnTR1 expression was up-regulated in B. napus under increased temperatures suggests that it is a heat-responsive gene. Consistent with this notion, a sequence analysis of the promoter (accession no: GU204976) in silico of BnTR1 revealed a heat stress-responsive element (HSE) located at the upstream of the transcriptional start point (−242 to −251 nucleotides, AAACAATTTC) (Figure 3b). Hence, we examined whether BnTR1 was responsive to heat stress and found that exposure of B. napus seedlings to continuous heat treatment at 35 °C resulted in a time-dependent increase in the expression levels of the gene (Figure 3c).

To confirm that heat-induced increase in BnTR1 transcript levels was caused by stress-stimulated transcription activity, but not by an enhanced mRNA stability, we placed a luciferase gene under the control of the BnTR1 promoter and transfected the construct into Arabidopsis protoplasts for transient analysis. A 2- to 3-fold increase in luciferase activity was observed by shifting the protoplasts from normal temperatures (22 °C) to higher temperatures (30 °C) for 6 h (Figure 3d). Collectively, these findings suggested that BnTR1 is a heat-responsive gene.

BnTR1 enhances the resistance of plants to heat stress

The fact that BnTR1 expression is responsive to heat stress prompted us to analyse its function in heat stress resistance. Accordingly, we created homozygous transgenic lines overexpressing BnTR1 in B. napus and O. sativa, respectively, and chosen the single copy of the transgene per haploid for further analysis (Figure S3). Immunoblot analysis with a BnTR1-specific antibody revealed a 2- to 3-fold increase at protein levels in the transgenic lines of B. napus compared with the non-transgenics (CK) (Figure 4a lower panel), which correlated with a modest increase in transcription levels (Figure 4a upper panel and Figure S4). Comparable expression levels of BnTR1 were also found in BnTR1 transgenic O. sativa lines (Figure 5g).

Figure 4.

BnTR1 enhanced the thermal resistance in Brassica napus. (a) BnTR1 expressing levels were monitored in nontransgenics (CK) and BnTR1-overexpressing lines by Northern blot (upper panel) and immunoblotting (lower panel). Equal amounts of RNA (15 μg) or protein (20 μg) were load in each lane. Rubisco is shown as a loading control. (b) Comparison of survival rate after 35 °C treatment for 6 and 9 days, respectively. Values are means ± standard deviation. Student's t-test significant at **P < 0.01 between BnTR1-overexpressing plants and CK, n = 20 ± 5. 43-3, 44-21, and 3-17 are independent BnTR1-overexpressing lines of B. napus, respectively. The data represented the mean value of three individual experiments (SD). (c) Phenotypic observation of up-regulated BnTR1 lines compared with CK seedlings under the treatment of heat stress. The seedlings were grown at 22 °C for 14 days and then maintained at 35 °C for a period as indicated.

Figure 5.

Increased thermal resistance in BnTR1 transgenic Oryza sativa. Phenotypic observation of BnTR1 transgenic O. sativa lines compared with CK under the treatment of heat stress after the seedlings were grown at 26 °C for 40 days. (a) Before treatment. (b) 48 °C for 30 min. (c) Treated plants were grown at 26 °C for 3 days after heat shock. (d) Treated plants were grown at 26 °C for 4 days after heat shock. (e) Treated plants were grown at 26 °C for 45 days. (f) Comparison of fringes during the stage of grain filling. (g) BnTR1 expressing levels were monitored in CK and BnTR1 transgenic lines by Northern blot. CK, non-transgenics; Independent BnTR1 transgenic O. sativa lines used are 49-2, 17-1 and 164-2, respectively.

To characterize the function of BnTR1 in thermal resistance in plants, the seedlings of B. napus were grown under normal conditions at 22 °C for 2 weeks followed by heat stress at 35 °C for 6 days. BnTR1 transgenic plants exhibited a normal appearance, while the leaves of CK seedlings began to show yellow spots and gradually shrink. The transgenics with up-regulated BnTR1 could survive at least 9 days, while CK almost died at the 6th day (Figure 4c). After heat stress at 35 °C for 6 days, the survival ratio was 78%, 82% and 73% in the 3-17, 43-3 and 44-21 BnTR1 transgenic lines, respectively, compared with the untreated controls. However, the survival ratio in CK was reduced to 36%. After treated for 9 days, the survival ratio in CK was only 10%, while the three transgenic lines examined, 3-17, 43-3 and 44-21, kept survival ratio at 62%, 71% and 55%, respectively (Figure 4b). In the absence of thermal stress, the transgenic lines did not display any visible growth or developmental difference to the parental lines (Figure 4c).

In O. sativa, BnTR1 also elicited a similar thermal resistance function. Upon treatment at 48 °C for 30 min, the 40-day-old seedlings of BnTR1 transgenic O. sativa appeared to be unaffected. In contrast, the seedlings of CK shrank and became fragile and subsequently displayed great changes in morphology with lodging of plants and poor recovery of leaves (Figure 5a–d). Forty-five days later, CK were significantly shorter than BnTR1 transgenic lines (Figure 5e). Further analysis demonstrated that the agronomical traits of BnTR1 transgenics were unaffected and fringes appeared normal, while those of CK plants were dramatically disturbed and fringes were short and sloppy (Figure 5f).

Under continuous heat stress (30 °C–39 °C each day, Figure S5) during the stage of flowering, BnTR1 transgenic O. sativa plants displayed higher pollen viability than CK according to the susceptibility of pollen grain to I2-KI (Figure 6a–d). Upon exposing plants to this stress for 4 days, the pollen viability of CK was reduced to 30%, while that of the six independent transgenic lines retained at 65%–90%. Furthermore, the higher pollen viability in the transgenic lines persisted on the 5th and 6th days compared with CK (Figure 6e).

Figure 6.

BnTR1 enhanced the pollen viability to increase the yield in Oryza sativa under heat stress. (a–d) Pollen activities of BnTR1 transgenic and nontransgenic O. sativa were observed under heat stress for 4 days using microscope. The darkly stained pollens were fertile, and the lightly stained pollens were sterile. (a) Non-transgenics (CK). (b–d) BnTR1 transgenic O. sativa line 49-2, 17-1 and 164-2, respectively. (e) Scheme of pollen activity of BnTR1 transgenic O. sativa lines and CK. Pollen activities were determined at 4th, 5th and 6th day under heat stress, respectively. n = 500 ± 20 (SD). (f, g) Comparison of panicle structure between CK (f) and BnTR1 transgenic line 49-2 (g). (h, i) Comparison of the grains between CK (h) and BnTR1 transgenic line 49-2 (i). (j, k) Schematic comparisons of seed set rate (j) and grains yield per plant (k) between CK and BnTR1 transgenic lines after heat stress. n = 60 ± 5 (SD); CK, non-transgenics; 49-2, 17-1, 164-2, 117-9, 71-4 and 216-3 are independent BnTR1 transgenic lines, respectively. The data represented the mean value of three individual experiments (SD). The single and double asterisks represent significant difference determined by the Student's t-test at P < 0.05 and P < 0.01, respectively.

The lower pollen activity is expected to result in a lower seed set, which is an important factor affecting yields at high temperatures (Zinn et al., 2010). To analyse the effect of increased BnTR1 on the yields under heat stress, plants from the previous experiment were kept being challenged with heat stress in the developmental stages of flowering and grain filling for 2 weeks. The number of plump seed of CK displayed significantly reduced than the transgenic lines (Figure 6f–i). In parallel with this, the seed set rate of CK was reduced to 20%. In contrast, seed rate of BnTR1 transgenics was reduced to 40%–62% (Figure 6j). The grain weights of individual plant in BnTR1 transgenics were 6.0–9.4 g compared with 4.0 g in CK (Figure 6k).

Taken together, the results presented above demonstrate that BnTR1 enhances the thermal resistance of plants including B. napus and O. sativa.

BnTR1 minimizes the environmental impacts on the crops

Interestingly, BnTR1 transgenic O. sativa plants also exhibited better agronomical traits than CK when cultivated in the field condition. When cultivated in Chengdu, China, the T2 BnTR1 transgenic O. sativa produced 11.8-12.3 panicles and 104.5-117.1 primary branches of panicle per plant compared with 9.1 and 81.2 in CK, respectively (Figure 7a–d). The total grains per plant were 963–1026 in different independent BnTR1 transgenic lines compared with 872 in CK (Figure 7e), resulting in the fact that the yield of BnTR1 transgenics was 8%–17% more than that of CK (Figure 7f). In addition, we also analysed the effect of photoperiod and season on BnTR1 transgenic lines planted in Hainan, China, in winter. The data indicated that the grains per plant and primary branches per panicle of BnTR1 transgenics were significantly higher than those in CK, leading to the yield in BnTR1 transgenics from 19 to 44% higher than that in CK (Figure 7c–f).

Figure 7.

BnTR1 minimizes the influences environmental impacts during the growth of plants. (a, b) Comparisons of panicles (a) and primary branch of panicle (b) between BnTR1 transgenic Oryza sativa and CK. (c–f) Schematic comparisons of panicle (c), primary branch (d), total grain (e) and yield (f) per plant. Values in c–f are means ± SD (n = 25 plants). The single and double asterisks represent significant difference determined by the Student's t-test at P < 0.05 and P < 0.01, respectively. CK, nontransgenics; 49-2, 17-1 and 164-2 are independent BnTR1 transgenic O. sativa, respectively.

These results corrected well with that BnTR1-overexpressing plants displayed predominant agronomic traits compared with those of CK, practically in adverse conditions.

BnTR1 induces the expression of heat stress-related genes by regulating calcium dynamics

Heat shock proteins and HSFs are two major classes of factors involved in stress responses in planta, many of which are up-regulated in response to heat stress (Lee et al., 1995; Lindquist, 1992). In this study, we established a link between BnTR1, and these two types of heat-induced factors in B. napus. Quantitative PCR analysis revealed that the level of HSFA1a was ninefolds, and those of HSPs (HSP101, HSP83.1, HSP70, HSP18.2, HSP17.6) were about 4–12 folds higher in up-regulated BnTR1 transgenics than in CK under normal growth conditions (Figure 8). Upon heat stress at 35 °C, the expression level of HSFA1a was dramatically reduced compared with that under nonstressed condition in the BnTR1 transgenics, resulting in the reduction of HSPs, which were only 0.3–4 folds higher in transgenics than that in CK (Figure 8). It suggested that BnTR1 regulated the expression of HSFA1a in normal and heat-stressed condition, respectively. Hence, BnTR1 is likely to act as a master regulator for the expression of heat stress-responsive genes.

Figure 8.

Expressional analysis of genes related to heat stress in the overexpression of BnTR1 lines and CK of Brassica napus. RNA samples were purified from 4-week-old seedlings from up-regulated BnTR1 transgenic lines (grey column) and CK (empty column) of B. napus under the treatment of heat stress at 35 °C. The transcripts of BnTR1 and heat-related genes HSFA1a, HSFA1b, HSP83.1, HSP101, HSP70, HSP17.6 and HSP18.2 were examined by quantitative PCR, and the results were normalized to GAPDH abundance. The amount of CK at 0 h was set as 1. The data represent trends observed in at least three individual experiments using three (two to six) independent BnTR1 transgenic lines (43-3, 44-21 and 3-17) of B. napus leaves.

Previous studies have indicated the crucial role of calcium in regulation of the synthesis of HSPs under heat stress (Gao et al., 2012). Thus, we tested the possibility that BnTR1 regulated the activity of calcium channels. Whole-cell patch-clamp experiments in root cell protoplasts recorded that the current of calcium channels in expressed BnTR1 Arabidopsis lines 101-6, 10-1 and 1-7 ranged from −175 to −320 pA at −150 mV, which was 7–12 folds stronger than that in WT (−25 pA) (Figure 9a). When treated with 1 mm NaCl, the current of calcium channels in transgenic lines decreased from −60 to −30 pA at −150 mV, while the current of calcium channels in WT suddenly increased up to −260 pA (Figure 9b). The degree of channel activity depended on the concentration of internal calcium. Using isotope labelling, we found that the intensity of cellular 45Ca2+ concentration was 2–3 folds higher in BnTR1 transgenic B. napus than that in CK plants treated with 0.1 mm CaCl2. Along with the concentration of CaCl2 increasing from 0.3 to 3 mm, the intensity of 45Ca2+ in CK was equal to that in transgenics. After treated with 50 mm or 100 mm CaCl2, however, the 45Ca2+ was 20% to 35% lower in BnTR1 transgenics than that in CK (Figure 9c), indicating a more sensitive modulation of Ca2+ channel in transgenic plants.

Figure 9.

Up-regulated BnTR1 alters the calcium influx by converting of calcium channel activity. (a, b) Changes in the calcium channel activity in root cells of WT and overexpression of BnTR1 plant. Steady-state currents of Arabidopsis root protoplasts were recorded against the clamped voltages from −150 to 50 mV by whole-cell patch clamp. (a) BnTR1 greatly increased Ca2+ channel currents under normal conditions. The number of protoplasts studied in at least three independent experiments (n = 11 for WT, n = 15 for 1-7, n = 9 for 10-1 and n = 10 for 101-6); (b), BnTR1 restrained Ca2+ channel currents treated with 1 mm NaCl. The number of protoplasts studied in at least three independent experiments (n = 15 for WT, n = 20 for 1-7, n = 12 for 10-1 and n = 15 for 101-6). WT, wild type; 1-7, 10-1 and 101-6, independent BnTR1 transgenic A. thaliana. (c) The content of calcium in Brassica napus leaves cells by isotope labelling. Values are means ± standard deviation. The single and double asterisks represent significant difference determined by the Student's t-test at P < 0.05 and P < 0.01, respectively, n = 20 ± 5. CK, nontransgenics. 43-3, 44-21 and 3-17 are independent BnTR1-overexpressing lines of B. napus, respectively.

Discussion

Plant growth and development are severely impacted by heat stress. Our data suggest that BnTR1 is involved in the response to heat stress. Up-regulated BnTR1 altered the expression of heat stress-related genes and promoted the resistance of B. napus and O. sativa to heat stress (Figures 4 and 5). These results indicate that the function of BnTR1 linked to thermal resistance is conserved. The BnTR1 promoter was responsive to increased temperature (Figure 3d), providing further support for its role in HSR.

BnTR1 is a heat-induced gene and encodes a RING finger protein with functional E3 ligase activity

The phylogenetic tree indicates that the TR1 gene family exists only in plants (Figure 1a). The absence of BnTR1 homologues in other organisms, such as animals and bacteria, suggests that BnTR1 function is unique to plants, which is consistent with the view from an evolutionary perspective that the response of plants to heat stress differs from those observed in other organisms (Waters, 2003). The existence of TR1 from lower to higher plants indicates that it may play a conserved role during plants evolution.

The members of BnTR1 gene family possess two unique features: two putative TMDs and a RING finger domain that displays E3 ligase activity in vitro (Figure 1b–d). However, the RING finger domain in BnTR1 is characterized by a C4HC3 motif that is distinct from the C3HC4, C3HC2 and C3H2C3 motifs found in many well-known E3 ligases. This RING motif classifies BnTR1 as a member of a subgroup of MARCH E3 ligases (Figure S2), which are found in human lymphocytes and are involved in regulating membrane receptors (Lehner et al., 2005; Nathan and Lehner, 2009). To our knowledge, BnTR1 represents the first member of this subgroup E3 ligases related to plant stress tolerance. Ubiquitination plays an important role in the perception and signal transduction of various external environmental signals. A large number of E3 ligases mediate cellular responses to abiotic stresses, such as cold, salt and drought (Dong et al., 2006; Lee et al., 2001; Qin et al., 2008; Zhang et al., 2007) in plants. Recently, it is predicted that the ubiquitin–proteasome system also plays an important role in heat stress (Zhang et al., 2012). However, there are few documents on how E3 ligases participated in heat signalling pathway.

Under heat stress, BnTR1 displayed a slow response profile (Figure 3c). The elevation of BnTR1 transcript starts at about 3 h after the onset of heat shock, while the HSR is transient in nature, usually peaking 1–2 h after onset, providing protection from acute episodes of thermal stress (Gurley, 2001). But it is not clear whether BnTR1 is involved in the initial induction of HSR, because the background expression may be sufficient at the early stage. Furthermore, in silico analysis of the promoter identified one HSE recognition site (Figure 3b), which is known as the binding site for transcriptional activators in heat stress signalling. Promoter expression analysis of BnTR1 provided further support for its role in heat stress (Figure 3d). Thus, our results suggest that BnTR1 may play a positive role in heat stress signalling.

The cell surface localization distinguishes BnTR1 from other known factors involved in thermal resistance in plants, which commonly reside in the cytoplasm or nucleus and are believed to be involved in downstream events of heat response (Trofimova et al., 1999). As a result, high levels of expression of these factors are often needed to produce thermal-resistant effect in plants (Lohmann et al., 2004; Nover et al., 1996; Prändl et al., 1998; Schöffl et al., 1998). The membrane localization of BnTR1 indicates that it might be involved in upstream regulation of signalling transduction in HSR.

BnTR1 induces the expressions of multiple heat-responsive genes and enhances thermotolerance in various species

Plant developed a series of responsive pathways against high temperature, which mainly included HSFs and HSPs. HSFs regulate the expression of heat-shock-regulated genes including HSPs as transcription factors (Ikeda et al., 2011).

In our results, the transcription of HSFA1a and HSFA1b, key regulators of heat shock response (Busch et al., 2005; Lohmann et al., 2004; Mishra et al., 2002), was not affected by heat shock in CK under heat stress, in consistent with what has been reported in the previous analysis (Busch et al., 2005; Guan et al., 2013; Lohmann et al., 2004; Nishizawa et al., 2006; Prändl et al., 1998; Yoshida et al., 2011), while it was increased accompanying with up-regulated BnTR1, resulting in the elevations of the main HSPs (Figure 8). Evidence from Arabidopsis suggests that overexpressed transcriptional regulators such as HSFA1a and HSFA1b cause plants to constitutively express at least some HSPs, leading to enhanced basal thermotolerance (Prändl et al., 1998; Zhu et al., 2009). However, deficiency of hsf1/3 has clear negative effects on the expression of some HSP genes (Lohmann et al., 2004). Previous studies were consistent with our results. Moreover, there is no difference in the expressions of HSF4 and HSF7 between up-regulated BnTR1 plants and CK under heat stress, indicating little effect of HSF4 and HSF7 on thermotolerance, according to the functions of HSF4 and HSF7 as transcriptional repressors or attenuators in HSR (Czarnecka-Verner et al., 2000).

Our phenotypic analysis showed that the thermotolerance of up-regulated BnTR1 plants was higher than that of controls (Figures 4 and 5). However, it is unclear how BnTR1 regulates the expression of HSFA1a under normal and stressed condition. Nevertheless, these results demonstrate that BnTR1 mediates HSR.

BnTR1 regulates the [Ca2+]cyt dynamics

Internal calcium plays a fundamental role in plant growth and development under normal as well as stress conditions (Mahajan et al., 2008). Previous studies have established a link between heat response and the calcium-mediated signalling network (Gong et al., 1998). Ca2+ can induce the expression of the calmodulin (CaM) genes to activate the CaM-binding protein kinases (CBKs) or heat-activated MAP kinases (HAMKs), which further phosphorylate the HSFs and promote the expression of HSPs, in response to heat stress in plants. Recent works have shown that heat shock could activate specific CNGC-based thermosensors in the plasma membrane, resulting in an influx of Ca2+ (Saidi et al., 2011), which suggests that there is heat shock-responsive Ca2+ channel for cytosolic Ca2+ induction, which might be an early event during heat stress signals. Heat stress results in a rapid increase in the level of intracellular Ca2+ concentration. However, Ca2+ cytotoxicity may lead to cell injury even death when cytoplasmic calcium concentrations exceed physiological levels (Szydlowska and Tymianski, 2010). To reduce this toxicity, the activity of Ca2+ channel was controlled by some unclear mechanisms. Our studies showed that the increased level of BnTR1 enhanced and weakened the activity of Ca2+ channel under normal and abnormal conditions, respectively (Figure 9a,b), suggestive the regulation of the Ca2+ channel by BnTR1 in a direct or indirect manner. Along with an elevation in the free extracellular calcium concentration, the increase in 45Ca2+ of transgenic lines was less than that of CK (Figure 9c), in consistent with the change in Ca2+ channel activity. Hence, it is possible that BnTR1 controls the expression of heat stress-responsive genes by regulating the concentration of cytosolic Ca2+ in planta. The increased [Ca2+]cyt then initiates the stress signalling pathways for stress tolerance (Mahajan et al., 2008; Rudd and Franklin-Tong, 2001; Sanders et al., 1999). On the other hand, BnTR1 will limit the excessive expression of heat-responsive genes by reducing the concentration of Ca2+ under abnormal condition (Figures 8 and 9). In this way, a modest increase in the expression of BnTR1 is expected to regulate the expression of many heat-responsive factors and confer strong thermal resistance in plants.

Our studies with O. sativa demonstrate that a constitutive BnTR1 expression is able to increase heat tolerance throughout a plant's life without any noticeable negative effect on its development (Figure 6). In contrasts with many other thermal-resistant genes tested previously, for example, DREB1A, AtHSFA3 and BhHSF1, they often produce defective growth phenotypes (Hong et al., 2009; Ito et al., 2006; Kasuga et al., 1999; Wu et al., 2009; Yoshida et al., 2008; Zhu et al., 2009). Most importantly, the expression of BnTR1 increases the yield of crops under field conditions (Figure 7), indicating that BnTR1 is able to minimize adverse environmental impacts on the crops. Overall, BnTR1 is a novel yet-conserved gene that seems to play a key role in the thermal resistance of plants. Future studies will reveal the targets and mechanism of BnTR1 in enhancing resistance to heat stress in plants.

Experimental procedures

Plant materials and growth conditions

Brassica napus was grown under sterile conditions on MS (Murashige and Skoog, 1962) nutrient agar medium containing 2% (w/v) sucrose or on soil in a growth chamber (16-h light/8-h dark) at 22 °C and 70% relative humidity after vernalization. O. sativa was grown at 26 °C.

Amplification of BnTR1 cDNA and promoter fragments

The 3′ cDNA of BnTR1 was obtained by the mRNA differential display technique. 5′ RACE was employed to amplify the missing 5′ segment.

TAIL-PCR was performed to amplify the unknown 5′ flanking region of BnTR1 as described by Liu et al. (1995). The regulatory region of BnTR1 was obtained and named as ProBnTR1. The sequences of BnTR1 and ProBnTR1 had been submitted to NCBI with the accession numbers, GU204975 (BnTR1) and GU204976 (ProBnTR1).

In silico analysis

Phylogenetic analyses were conducted in MEGA, version 4 (Tamura et al., 2007). The evolutionary history was inferred using the NJ method. The sum of branch lengths of the TR1 tree is 2.18. Prediction of the TMD was obtained by www.ch.embnet.org/software/TMPRED_form.html. The BnTR1 promoter was analysed by http://bioinformatics.psb.ugent.be/webtools/plantcare/html.

Plasmid construction and plant transformation

BnTR1 was amplified and subsequently cloned into the BamHI-EcoRI sites of pGEX-6P-1 vector (Amersham). For transient expression in onion (Allium cepa) epidermal cells, eGFP was inserted into SmaI-SacI sites of pBI221 (Jefferson, 1987) yielding pBI221-p35S::eGFP. BnTR1 was inserted into BamHI-SmaI sites of pBI221-p35S::eGFP. For transient expression in protoplasts, ProBnTR1 and ProAt3g47550 were amplified and cloned into HindIII-BamHI sites of pBI221 to replace the CaMV 35S promoter.

For BnTR1 stable expression in transgenic plants, BnTR1 was cloned into the BamHI-Sac I sites of the pBI121 (Jefferson, 1987) in sense orientation (to generate BnTR1 transgenic B. napus and A. thaliana) or pCAMBIA1300 (CAMBIA, Australia) (to generate BnTR1 transgenic O. sativa) vector to replace the glucuronidase gene (GUS) yielding p35S:: BnTR1.

For AtTR1 stable expression in transgenic Arabidopsis, AtTR1 was cloned into the BamHI-SacI sites of the pBI121 in sense or antisense orientation. All of the above primers used are listed in Table S1.

Transformations of planta were performed as described by Clough and Bent (1998), Hiei et al. (1994), and Zhang et al. (2001). The BnTR1 transgenic lines used in this study were T2 homozygous O. sativa and T3 homozygous B. napus and A. thaliana, respectively.

Analysis of transient expression in Arabidopsis protoplasts was performed as described by Yang et al. (2006). The protoplasts were incubated at 30 °C for heat stress and at 22 °C for control.

Traits measurement and analysis

See Appendix S1.

Estimating the copy number of the transgenic plants

See Appendix S1.

Assay of E3 ubiquitin ligase activity

In vitro E3 ligase assay of BnTR1 was performed as described by Wang et al. (2001). Human E1 (UBA1, GI: 23510338), yeast ubiquitin (GI: 209599) and human E2 (UBCh5b, GI: 4507773) were purchased from Sigma Company (St Louis, MO).

Antibody preparation

BnTR1 antibody was prepared from the synthetic peptide corresponding to the nonconserved region from the 136th to 149th residues (DDWEDGVHLDSSDP) of BnTR1 (Takara Bio, Dalian, China). Synthetic peptide was conjugated to keyhole limper haemocyanin by the way of a C-terminally added cysteine and injected under the skin of rabbits. IgC fraction of sera was purified with a column filled with protein A and then affinity-purified with Sepharose-immobilized synthetic peptide after 60 days.

Subcellular localization

Immunohistochemistry and immunofluorescence of BnTR1 in B. napus were performed as described by Maryani et al. (2003). Transient expression of pBI221-p35S::eGFP or pBI221-p35S::BnTR1::eGFP in Arabidopsis protoplasts was determined as described by Yoo et al. (2007).

Assays of thermal resistance of plants

Fourteen-day-old Brassica seedlings were transferred into a climate chamber at 35 °C with 16-h light (60 μmol/m2/s)/8-h dark and 70% relative humidity. Seedlings of 40-day-old O. sativa were transferred to a growth chamber at 48 °C for 30 min. Plants of O. sativa during anthesis were transferred into a climate chamber under continuous heat stress (Figure S5).

Analysis of pollen viability and seed set

Analysis of pollen viability was performed as described by Wang et al. (2006). Pollen grains stained uniformly were considered viable. Panicles were harvested at physiological maturity. Numbers of filled and unfilled grains and grain weight per panicle were recorded. Seed set was estimated as the ratio of number of filled grains to total number of the reproductive sites (florets) and expressed as a percentage.

Analysis of gene expression

Total RNAs were extracted from the leaves of 4-week-old B. napus that grew under 22 °C or the treatment at 35 °C using an RNAeasy kit (Qiagen, Valencia, CA). Equal amounts of total RNA were used for RT-PCR analysis with SuperScriptTMIII Reverse Transcriptase (Invitrogen, Carlsbad, CA). Expression levels were determined using the iCycler 5 (Bio-Rad, Hercules, CA). Primers used in B. napus are listed in Table S2.

Cytosolic Ca2+ measurement

The leaves of 7-week-old B. napus were selected for extract of protoplasts. The separated protoplasts were added into WI solutions (Yoo et al., 2007) contained different calcium densities (0.1 mm, 0.3 mm, 1 mm, 3 mm, 50 mm and 100 mm) and 10 nm 45Ca as tracer. After shook at 54 rpm and 25 °C for 12 h in the dark, the mixture was centrifugated at 100 g for 3 min. Then, the protoplasts were washed three times using WI solutions adding 100 mm EDTA. The radioactivity was measured using a liquid scintillation counter (Perkin-Elmer, Boston, MA).

Patch clamp and data acquisition

Arabidopsis mature epidermal root cell protoplasts were isolated as described by Zhao et al. (2007). The whole-cell voltage-clamp currents of Arabidopsis root cells were recorded with an EPC-9 amplifier (Heka Instrument, Lambrecht, Germany) as described by Miao et al. (2006) and Wang et al. (2004). Data were analysed using PULSEFIT 8.7 (Heka, Lambrecht, Germany), IGOR 3.0 (Wavemetrics, Lake Oswego, OR) and ORIGIN 7.0 software (OriginLab Corp., Northampton, MA).

Acknowledgements

We thank Dr. Yu Jiang (University of Pittsburg, USA) for useful suggestions and comments in the preparation of the manuscript. We also thank Dr. Xinrong Ma (Chinese Academy of Sciences, Chengdu Branch) for technical assistance of the transformation using gene bombardment. This research was supported by the grants from the National Natural Science Foundation of China (31171586 and 30971557 to Y.Y.; 30971816 to X-F.L.; 30871322 to J-M.W.; 10975103 to N.L).

Author contributions

Z-B.L., J-M.W., J-B.X. and Y-F.Y. assembled the constructs for transformation, undertook the transformation experiments and performed all of the molecular biological analysis. Z-B.L., F-X.Y., L.Y. and J-B.X. performed the analysis of transgenic B. napus lines. M.G., X-J.W., F-J.X. and X.Z. performed the analysis of transgenic O. sativa lines and the agronomical traits. L. D. performed the patch-clamp analysis. J-B.X. performed transient expression analysis. Z-B.L. and N.L. performed the cytosolic calcium content experiments. Z-B.L., J-M.W., F-X.Y., X-F.L. and Y.Y. analysed the data. Y.Y. designed the study and wrote the manuscript.

Ancillary