Functional analysis of tomato CHIP ubiquitin E3 ligase in heat tolerance

Plants have evolved genetic and physiological mechanisms to mitigate the adverse effects of high temperature. CARBOXYL TERMINUS OF THE HSC70-INTERACTING PROTEINS (CHIP) is a conserved chaperone-dependent ubiquitin E3 ligase that targets misfolded proteins. Here, we report functional analysis of the SlCHIP gene from tomato (Solanum lycopersicum) in heat tolerance. SlCHIP encodes a CHIP protein with three tandem tetracopeptide repeat (TPR) motifs and a C-terminal U box domain. Phylogenetic analysis of CHIP homologs from animals, spore-bearing and seed plants revealed a tree topology similar to the evolutionary tree of the organisms. Expression of SlCHIP was induced under high temperature and was also responsive to plant stress hormones. Silencing of SlCHIP in tomato reduced heat tolerance based on increased heat stress symptoms, reduced photosynthetic activity, elevated electrolyte leakage and accumulation of insoluble protein aggregates. The accumulated protein aggregates in SlCHIP-silenced plants were still highly ubiquitinated, suggesting involvement of other E3 ligases in ubiquitination. SlCHIP restored the heat tolerance of Arabidopsis chip mutant to the wild type levels. These results indicate that tomato SlCHIP plays a critical role in heat stress responses most likely by targeting degradation of misfolded proteins that are generated during heat stress.


Results
Identification and sequence analysis of tomato SlCHIP. In order to identify the CHIP homologs in tomato, we conducted BLAST searches against sequenced tomato genome using Arabidopsis CHIP protein sequence as query. To verify a tomato protein as a CHIP homolog, we relied not only on the sequence similarity to Arabidopsis CHIP but also on the presence of both the TPR motifs and U-box domain. Using these criteria, we identified a single tomato gene (Solyc06g083150) encoding a CHIP homolog, which was named SlCHIP. Several additional tomato genes encode proteins with high sequence similarity to some parts of Arabidopsis and tomato CHIP proteins but are not considered to be CHIP homologs because they contain only TPR motifs or U box domain. One such gene is Solyc09g082540 that encodes a protein containing three TPR motifs at the N-terminus but lacks the U-box domain. This TPR gene, named SlTPR28, was included in the study as a control.
Using gene-specific primers, we PCR-amplified the coding sequences of both SlCHIP and SlTPR28 using cDNA from tomato cultivar Zheza809 as templates (Fig. 1A). Based on the coding sequence and genome annotation, tomato SlCHIP gene contains eight exons and seven introns, which are identical to Arabidopsis AtCHIP gene. SlCHIP and SlTPR28 encode two proteins of 276 and 318 amino acids, respectively. The similarity between the deduced SlCHIP and SlTPR28 proteins is 23.9%. SlCHIP shares 68.8% amino acid identity with AtCHIP protein. The deduced SlCHIP protein contains three TPR motifs and one U-box domain based on SMART analysis (http://smart .embl-heide lberg .de/). However, SlTPR28 protein contains only three TPR motifs but no U-box domain (Fig. 1B,C).

Phylogenetic analysis of SlCHIP homologs from different organisms. CHIP ubiquitin E3 ligases
have been identified in all eukaryotic organisms. In contrast to the CHIP proteins from animals, which have been well characterized, there have been only a few studies on plant CHIP proteins, exclusively in Arabidopsis. Therefore, we searched the sequenced genomes of representative plants along the evolutionary tree for genes encoding CHIP homologs, again using the criteria of high sequence similarity and presence of both the TPR motifs and the U-box domain. Using these criteria, we identified 14 CHIP-encoding genes from 10 spore-bearing and seed plants. Like animals, most plants contain a single gene encoding the chaperone-dependent U box E3 ubiquitin ligase. However, some plants including the spreading earthmoss (Physcomitrella patens), maize (Zea mays), purple false brome (Brachypodium distachyon) and soybean (Glycine max) contain two genes encoding CHIP proteins. Maize, soybean and purple false brome are known to be polyploid plants that have gone genome duplications during their evolutionary history, which could account for the presence of more than one CHIP genes in their genomes.
To analyze the evolutionary relationship of the conserved protein family, we performed phylogenetic analysis of CHIP homologs from different organisms including animals, spore-bearing and seeds plants. As shown in Fig. 2, there are three major clades in the phylogenetic tree. All the CHIP proteins from the animals clustered in one clade, while those from spore-bearing and seed plants clustered in two separate clades (Fig. 2). These results indicate that the topology of phylogenetic tress for CHIP homologs from animals, spore-bearing and seed plants is in the agreement with the evolutionary tree of the organisms. Furthermore, CHIP proteins from monocot and dicot plants also generally clustered separately in the major clade of seed plants (Fig. 2) www.nature.com/scientificreports/ CHIP homologs from maize and purple false brome have a relative distant relationship (Fig. 2). As expected, one CHIP homolog from each of the two monocot plants clustered with those from other monocot plants but the other homolog was actually grouped closer with CHIP homologs from dicot plants (Fig. 2). The evolutionary significance of the sequence variation of the CHIP homologs from maize and purple false brome is unclear but could reflect potential functional diversification of the two CHIP homologs in these plants.

Differential induction of SlCHIP by heat stress and stress hormones. Arabidopsis mutant plants
for CHIP gene AtCHIP are compromised in tolerance to a spectrum of abiotic stresses including high temperature but are normal in resistance to the hemibiotrophic bacterial pathogen Pseudomonas syringe and the necrotrophic fungal pathogen Botrytis cinerea 29 . As a major environmental stress during the whole plant life cycle, heat stress is a major limiting factor for tomato growth and production. In order to explore whether SlCHIP is involved in tomato response to heat stress, we used RT-qPCR to examine expression of the gene after 0-, 3-, 6-and 9-h treatment at 45 °C. For comparison, we included the tomato SlTPR28 gene in the expression analysis as control. As shown in Fig. 3A, both genes were induced by approximately twofold after 3-h of exposure to the high temperature. Increase in time of exposure to the high temperature caused further increase in the transcript levels of SlCHIP, which were more than 7-times higher after 6-h of heat stress (Fig. 3A). After peaking at 6 h of heat treatment, the transcript levels of SlCHIP rapidly declined and reached close to the basal levels after 9 h at the high temperature (Fig. 3A). By contrast, no further induction in the transcript levels of SlTPR28 was detected after the initial increase (Fig. 3A). These results indicated that SlCHIP is highly responsive to heat stress. In plants, different hormones play important roles in diverse biological processes including stress responses. To further examine the involvement of SlCHIP and SlTPR28 in stress responses in tomato, we analyzed their response to JA, SA and ABA, three plant hormones with roles in plant responses to different stresses. SA, JA and ABA were applied to tomato cultivar Zheza809 plants and leaf samples were collected at 0-, 1-, 3-and 6-h after hormone treatments. For comparison, we also analyzed the expression of AtCHIP in Arabidopsis Col-0 plants. As shown in Fig. 3B, both SlCHIP and SlTPR28 were down-regulated in their transcript levels by application of exogenous JA. On the other hand, AtCHIP appeared to be significantly induced by JA treatment (Supplemental Fig. S1). Both SlCHIP and SlTPR28 were induced by SA but with different kinetics (Fig. 3C). AtCHIP expression was not substantially altered during the first 3 h after SA treatment and displayed significant decline after 6 h of SA treatment (Supplemental Fig. S1). Therefore, there appeared to be significant difference between Arabidopsis and tomato in SA-regulated expression of the homologous CHIP genes. On the other hand, there was a strikingly similarity in the ABA-regulated expression of SlCHIP in tomato and AtCHIP in Araidopsis. Both CHIP genes were strongly induced during the first hour after ABA treatment but rapidly declined to the basal levels by the third hour of ABA treatment ( Fig. 3D and Supplemental Fig. S1). By contrast, no induction of SlTPR28 was observed in tomato plants after ABA treatment (Fig. 3D). These results demonstrated that SlCHIP was responsive to plant stress hormones and its strong induction by SA and ABA suggested potential roles in tomato responses to both biotic and abiotic stresses.   Table S1). As controls, pTRV2 VIGS constructs for SlTPR28 (pTRV2-SlTPR28) and tomato PDS gene for phytoene desaturase (pTRV2-PDS), as well as the pTRV2 empty vector (mock) were also coinfiltrated 37 . The bleaching phenotype due to the silencing of the tomato in the pTRV2-SlPDS infiltrated plants was used to verify the efficiency of the VIGS procedure. Silencing efficiency was also assessed by testing target gene expression level in the terminal leaflets of the fifth leaves using RT-qPCR with gene-specific primers. The transcript levels of SlCHIP and SlTPR28 in the tomato plants infiltrated with their respective silencing vectors were reduced by more than 80%, when compared to those in pTRV2-infiltrated control plants (Fig. 4A). By contrast, there was no significant cross-silencing of SlCHIP in SlTPR28-silenced plants or SlTPR28 in SlCHIP-silenced plants as their transcript levels were both similar to those in pTRV2-infiltrated plants (Fig. 4A). These results indicate that TRV-mediated silencing of SlCHIP and SlTPR28 was effective and specific ( Fig. 4A).
To determine the effect on tomato heat tolerance, we placed the silenced and control plants in a growth chamber at 45 °C. After 9 h at the high temperature, control plants were very normal with no leaf wilting symptoms, while SlTPR28-silenced plants displayed only minor but visible wilty phenotype (Fig. 4B). Importantly, shoots of all SlCHIP-silenced plants displayed substantially wilting symptoms after 9-h at the high temperature. More than 80% of leaflets of the fourth leaves in SlCHIP-silenced plants were visibly curled (Fig. 4B). Furthermore, approximately 23% of leaflets of the fourth leaves in SlCHIP1-silenced plants showed significant water-soaking symptoms and tissue collapse after 6-h at 45 °C (Fig. 4C). These results indicated that SlCHIP played a critical role in tomato heat tolerance.
To provide more quantitative assessment of the impact of SlCHIP silencing on tomato heat tolerance, we compared heat-induced electrolyte leakage in leaves of control and silenced tomato plants. As shown in Fig. 5A, there was a slight but statistically insignificant increase in electrolyte leakage in control or SlTPR28-silenced tomato leaves after exposure to 45 °C for 9 h (Fig. 5A), consistent with little development of heat stress symptom in these plants (Fig. 4B). By contrast, there was approximately a twofold increase in electrolyte leakage after heat stress in SlCHIP-silenced tomato leaves (Fig. 5A), in agreement with the substantial development of watersoaking and wilting symptoms (Fig. 4). Heat stress causes harmful effects on a variety of biological processes in plants including photosynthesis. Therefore, we also compared SlCHIP-and SlTPR28-silenced tomato plants with control plants for the effect of heat stress on CO 2 assimilation rate of the terminal leaflets of the fourth leaves immediately after heat treatment. As shown in Fig. 5B, after 9-h exposure at 45 °C, no statistically significant reduction of CO 2 assimilation rate was observed in leaves of control or SlTPR28-silenced tomato plants. On the other hand, silencing of SlCHIP caused more than 55% reduction in CO 2 assimilation rate after 9-h exposure at 45 °C (Fig. 5B). These results further indicated that silencing of SlCHIP, but not SlTPR28, substantially compromised tomato heat tolerance.
Silencing of SlCHIP caused increased accumulation of insoluble proteins during heat stress. Heat stress usually leads to denatured or damaged proteins and these proteins accumulate as insoluble, detergent-resistant protein aggregates in plant cells 28 . To ascertain whether the compromised heat tolerance of SlCHIP-silenced plants was associated with increased accumulation of insoluble, detergent-resistant protein aggregates in plant cells, we isolated and quantified total and insoluble proteins from leaves of control, SlCHIPor SlTPR28-silenced plants after 0-, 3-, 6-, and 9-h of heat treatment at 45 °C. As shown in Fig. 6, there was little increase, from 3.7 to 4.2%, in the insoluble proteins as percentages of total proteins in control plants after 9 h at the high temperature. Insoluble proteins as percentages of total proteins also increased marginally (from 3.71 to 4.7%) in SlTPR28-silenced plants (Fig. 6). By contrast, in SlCHIP-silenced plants, insoluble proteins as percentages of total proteins rose from 3.72 to 6.4%, which represented a 73% increase after 9 h at the high temperature (Fig. 6). Therefore, the increase in insoluble protein aggregates in SlCHIP-silenced plants were 5-6 times higher than that in control or SlTPR28-silenced plants after 9 h of heat stress. Taken together, these results indicated a critical role of SlCHIP in protection against proteotoxicity by targeting insoluble protein aggregates in tomato leaf cells during heat stress.
Increased ubiquitination of insoluble protein aggregate in SlCHIP-silenced plants. E3 ubiquitin ligase CHIP has been shown to prohibit plastid-destined precursor protein accumulation in the cytosol by regulating precursor degradation through 26S proteasomes 35 . Arabidopsis CHIP mitigates proteotoxicity additively with NBR1-mediated autophagy under heat stress 29 . In order to determine the role of SlCHIP in the ubiquitination of heat-induced insoluble protein aggregates, we compared the extent of ubiquitination of insoluble protein aggregates among control, SlCHIP-and SlTPR28-silenced plants after 0 and 9 h under heat stress. Soluble and insoluble proteins were isolated, fractionated by SDS-PAGE and assessed for the levels of ubiquitination by proteins blotting using an anti-ubiquitin monoclonal antibody. As shown in Fig. 7 and Supplementary Fig. S2, levels of ubiquitinated protein were all very low in the soluble proteins of all samples with or without heat stress. In the insoluble fractions, the levels of ubiquitination in mock-, SlCHIP-and SlTPR28-silenced plants were also similarly low without heat stress ( Fig. 7 and Supplementary Fig. S3). After 9 h of heat treatment, the levels of ubiquitinated proteins all increased in control, SlCHIP-and SlTPR28-silenced plants (Fig. 7) www.nature.com/scientificreports/ higher levels of ubiquitinated proteins in SlCHIP-silenced plants after 9 h of heat treatment (Fig. 7). In fact, the levels of ubiquitinated proteins were also elevated even after 1 (Fig. 8A). However, only 16-20% of leaves of the atchip mutant plants expressing SlCHIP displayed yellowish and curled symptoms, similar to the 14-18% yellowish and curled leaves in wild type Col-0 plants after exposure for 9 h at 45 °C (Fig. 8A). We also collected leaf samples at 0-, 3-, 6-and 9-h of high temperature for analysis of insoluble protein aggregates. As shown in Fig. 8B, increase in insoluble proteins as percentages of total proteins in atchip mutants was nearly 10 times of that in Col-0 wild-type plants. By contrast, insoluble proteins as percentages of total proteins in the transgenic atchip mutant plants expressing SlCHIP increased only slightly from 3.27 to 3.78% (Fig. 8B). These results demonstrated that SlCHIP was able to restore the heat tolerance of the atchip mutant plants close to the wild-type levels.

Discussion
CHIP proteins are present in a large number of eukaryotic organisms including animals and plants 29,36-39 . As a U-box E3 ubiquitin ligase, CHIP collaborates with molecular chaperones for recruitment of misfolded client protein substrates and targets their ubiquitination and degradation by the proteasome system 29,40,41 . As a central component in cellular protein quality control, CHIP has been extensively analyzed in animal systems for its role in a myriad of physiological and pathological processes 37,[42][43][44][45][46] . Plants as sessile organisms are constantly exposed to a variety of unfavorable environmental conditions including extreme temperatures and mechanisms for maintaining the integrity of the proteome of plant cells are essential for plant growth, development and survival 47,48 . CHIP proteins are also present in plants but so far there have been only a very few studies on plant CHIP proteins only in Arabidopsis [49][50][51][52][53][54] . In the present study, we have searched the genomes of a substantial number of spore-bearing and seed plants and identified genes encoding CHIP proteins that contain both TPR motifs at the N-terminus and a C-terminal U-box domain. Like many animal organisms, a majority of plants contain a single CHIP gene, further underscoring the evolutionarily conserved role of the chaperone-dependent E3 ubiquitin ligase in different organisms. In some plants including maize and soybean, there are two genes encoding CHIP proteins most likely due to their polyploid nature with genome duplication during the evolutionary history 55,56 . Interestingly, phylogenetic analysis revealed that while the two CHIP proteins from soybean are highly similar and are both clustered with CHIP proteins from other dicot plants, the CHIP proteins from maize and Brachypodium distachyon have diverged significantly in protein sequences and as a result, were placed in different subclades in the phylogenetic tree with one of the two CHIP proteins from each species clustered with those from dicot plants. Whether the structural difference of the CHIP proteins in these plants is associated with functional divergence is unclear but can be addressed through analysis of their expression patterns, biochemical and molecular properties and, most importantly, the phenotypes of their mutants. As a central E3 ubiquitin ligase in protein quality control, CHIP acts together with chaperones in targeting degradation of misfolded proteins, which are generated at high levels in plant cells under heat stress. To determine the role of tomato CHIP in response to high temperature, we have analyzed the response of tomato SlCHIP www.nature.com/scientificreports/ gene expression to high temperature and discovered it to be heat-inducible (Fig. 3A). Unlike genes encoding HSP proteins with rapid heat induction, increase in the transcript levels of SlCHIP was relatively slow but steady during the first 6 h of heat treatment (Fig. 3A). This heat-induced increase in the expression of SlCHIP during the early hours of heat stress was followed by a rapid decline after 6 h at the high temperature (Fig. 3A). The early induction of SlCHIP during heat stress would lead to elevated levels of SlCHIP proteins necessary for the rapid increase in misfolded proteins under the high temperature. The transient nature of the heat-induction of SlCHIP expression could be attributed to potential deleterious effect of prolonged or excessive production of the ubiquitin E3 ligase, which could lead to non-specific ubiquitination and degradation of cellular proteins. Consistent with this possibility, overexpression of Arabidopsis AtCHIP gene reduces plant tolerance to extreme temperatures 54 . In addition to heat stress, SlCHIP was responsive to plant stress hormones JA, SA and ABA. While the responses of CHIP to JA and SA appeared to be complex and varied between tomato and Arabidopsis, ABA-induced expression of CHIP was rapid and strong in both Arabidopsis and tomato. ABA as an important plant stress hormone plays a critical role in plant responses to heat stress 57,58 . ABA-induced CHIP gene expression preceded heat-induced CHIP expression and it will be of great interest to determine whether ABA signaling plays a role in heat-induced expression of the CHIP genes during plant response to high temperature. To directly address the role of SlCHIP in plant heat tolerance, we have taken two molecular approaches. First, we have successfully demonstrated that expression of SlCHIP in the Arabidopsis atchip mutant could restore the www.nature.com/scientificreports/ heat tolerance of the heat-sensitive mutant (Fig. 8), supporting the conserved role of the CHIP proteins from two different plant species in plant heat tolerance. In addition, we employed VIGS to suppress the expression of SlCHIP and comprehensively analyzed the impact on tomato heat tolerance. VIGS is rapid, simple and particularly suitable for characterization of phenotypes that might be lethal in stable lines 59 . Potential drawbacks of VIGS include cross-silencing of sequence-related off-target genes. However, there is only a single CHIP gene with no close homologs in the tomato genome. In addition, we have included a gene encoding a protein containing TPR motifs as negative control in the study and found no significant cross-silencing among the two related genes based on RT-qPCR analysis of the transcript levels for the genes and the specific nature of the effects of their silencing on the heat tolerance and associated physiological and biochemical parameters. Virus-induced silencing of SlCHIP caused increased water-soaking and wilting symptoms in tomato plants after incubation at 45 °C, which were much less in SlTPR28-silenced plants and barely observed in control plants infiltrated with the pTRV2 empty vector. These observations provided direct and compelling evidence for a critical role of SlCHIP in tomato heat tolerance. Increased development of heat stress symptoms in the SlCHIP-silenced tomato plants was associated with elevated electrolyte leakage (Fig. 5A), which is a hallmark of stress-induced injury of plant cells associated with reduced integrity of cell membrane, increased production of ROS and programmed cell death 60 . Furthermore, we observed reduced capacity of photosynthesis in the leaves of SlCHIP-silenced but not in SlTPR28-silenced or control tomato plants after exposure to high temperature (Fig. 5B). Chloroplasts are one of the most sensitive organelles in plants to heat stress and apparently CHIP has a critical role in protection of chloroplasts from heat stress even though the heat-regulated E3 ubiquitin ligase is mostly present in the cytosol. Previously, it has been shown that in Arabidopsis, CHIP and HSC70 mediate plastid-destined precursors degradation by the proteasome system 35 . These proteins are synthesized in the cytosol as unfolded precursors and can accumulate in the cytosol under conditions when their import into plastids is hindered 35 . These protein precursors, if not promptly degraded, can accumulate in the cytosol to form nonspecific aggregates and cause severe cellular damage. Under high temperatures, denatured proteins increase and the capacity of cellular protein quality control will be under a great pressure. Under such conditions, maintaining plastid-destined protein precursors at unfolded conformation may become increasingly insufficient, leading to their misfolding and formation of protein aggregates. Reduced levels of CHIP proteins as in SlCHIP-silenced tomato plants would compromise the ability to degrade these misfolded proteins or protein aggregates, leading to toxic effects on cellular structures including the photosynthetic organelles of plant cells.
As a chaperone-dependent E3 ubiquitin ligase, CHIP targets ubiquitination of misfolded proteins for degradation by the 26S proteasome system. In the absence of CHIP, these misfolded proteins will form protein aggregates, which could be targeted for degradation by other degradative pathways such as selective autophagy 29 . Indeed, increased heat sensitivity in SlCHIP-silenced tomato plants was associated with elevated levels of protein aggregates (Fig. 6). Interestingly, protein blotting using an anti-ubiquitin monoclonal antibody showed that the protein aggregates accumulated in SlCHIP-silenced tomato plants were still highly ubiquitinated (Fig. 7). One possible explanation for the counterintuitive observation is that SlCHIP specifically or preferentially recognize those misfolded but still soluble proteins through their associated chaperone proteins and targets then for ubiquitination and degradation by the proteasome system. Those soluble misfolded proteins, if not promptly degraded, will then interact nonspecifically to form insoluble protein aggregates. Apparently there are additional ubiquitin E3 ligases that recognize these misfolded proteins during or after their aggregation for ubiquitination and ultimately degradation by selective autophagy or other pathways. Therefore, it will be of importance to identify and analyze the unknown E3 ligases and associated protein degradation pathways that function coordinately in the recognition, ubiquitination and processing of stress-induced abnormal proteins that are generated under stress conditions. In addition, a large number of studies in animals have established that CHIP E3 ligases have important roles not only in protein quality control but also in signaling through targeting of specific regulatory proteins [37][38][39]43,45 . It is highly conceivable that CHIP may have a similar signaling role in plant stress responses. A comprehensive analysis of the roles and mode of action of CHIP proteins will provide important new insights into the molecular link between protein quality control networks and plant stress responses.

Methods
Plant materials and growth conditions. The tomato cultivar Zheza809 was used in this study. Arabidopsis T-DNA insertion mutant atchip-1 (Salk_048371) in Col-0 accession has been previously described 29 . Tomato and Arabidopsis plants were grown in a growth room at 22-24 °C on a photoperiod of 12-h light (600 µmol m −2 s −1 ) and 12-h dark. Heat treatment was performed by placing plants in a growth chamber at 45 °C with 200 µmol m −2 s −1 light for 9 h to test heat tolerance. Virus-induced Silencing of SlCHIP and SlTRP28. SlCHIP and SlTPR28 fragments of 286 and 364 bp, respectively, were amplified and cloned into vector pTRV2 to generate pTRV2-SlCHIP and pTRV2-SlTPR28, respectively. Plasmids pTRV2, pTRV2-SlCHIP and pTRV2-SlTPR282 were transformed into Agrobacterium tumefaciens strain GV3101 competent cells by electroporation and transformants were selected on YEB plates containing kanamycin (50 ng/ml), rifampicin (50 ng/ml) and gentamycin (25 ng/ml). Plates were kept at 28 °C in an incubator for 48 h and positive colonies were confirmed by colony PCR. Agrobacteria carrying pTRV2, pTRV2-SlCHIP or pTRV2-SlTPR28 were used to perform VIGS according to the protocol as described 62 . Infiltration was performed on 30 tomato plants for each gene and the experiments were repeated three times. RT-qPCR was performed to determine the silencing efficiency for each gene using RNA isolated from the terminal leaflets of the fifth leaves. Only those tomato plants with more than 80% reduction in the transcript levels for the silenced gene were used in the subsequent assays for heat tolerance.

Cloning of SlCHIP and
Assays of heat tolerance. Selected 6-week old tomato and 4-week old Arabidopsis plants were placed in a growth chamber at 45 °C with 200 µmol m −2 s −1 light for 9 h to test heat tolerance. Three replicates were performed with 12 plants for each replicate for every type of plants. Tomato leaf samples were collected with 0-, 3-, 6-and 9-h high temperature. Heat stress symptoms including wilting and water soaking were observed at indicated times by taking imaging of treated plants or counting of symptomatic leaves.
Determination of electrolyte leakage caused by high temperature was performed after heat stress as previous described 63 . Briefly, seven leaf disks from the leaflets of the fourth and fifth leaves were sampled, rinsed with deionized water and put into tubes with 20 ml deionized water for 20 h at 24 °C. The conductivity was measured before and after tubes autoclaved using a potable MW802 pH/EC/TDS meter (Milwaukee Instruments, Inc., Rocky Mount, NC, USA). The CO 2 assimilation rates of the terminal leaflets of the fourth leaves were determined in the silenced and pTRV2 plants with an infrared gas analyzer-based potable photosynthesis system (LI-6400; Li-COR, Lincoln, NE, USA).

Generation of SlCHIP transgenic Arabidopsis lines. In order to generate transgenic SlCHIP-express-
ing Arabidopsis lines, the full length coding sequence for SlCHIP gene was amplified from plasmid pMD18-SlCHIP and inserted into the plant transformation vector pFGC5941-3ΧHA with the 1.5 Kb Arabidopsis CHIP gene native promoter from Arabidopsis Col-0 plant. The resulting plasmid was confirmed by colony PCR and transformed into atchip-1 mutant plants (Salk_048371). Transformants were screened by selecting for resistance to Basta. Transgenic plants expressing SlCHIP were determined by western blot using an anti-HA monoclonal antibody. Homozygous T2 transformants were selected and used in the study.
Soluble and insoluble proteins extraction and western blotting. Tomato and Arabidopsis leaf samples were collected at 0-, 3-, 6-and 9-h after heat treatment. Leaf samples were ground in liquid nitrogen and homogenized in a detergent containing extraction buffer as described previously 29 . Total proteins were first passed through filtration to remove debris and soluble and detergent-resistant insoluble proteins were separated by low speed centrifugation as previously described 28 . The protein quantity was measured using BCA (Bicinchoninic acid) Protein Assay Kit (SK1070, Coolaber, Beijing, China) based on the manufacturer's instruction. Ubiquitinated proteins were detected by western blotting with an anti-ubiquitin monoclonal antibody (Sigma, USA). The antigen-antibody complexes were detected by enhanced chemiluminescence using luminal as previously described 28  www.nature.com/scientificreports/ Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.