Apart from energy generation, mitochondria perform a signalling function determining the life and death of a cell under stress exposure. In the present study we have explored patterns of heat-induced synthesis of Hsp101, Hsp70, Hsp17.6 (class I), Hsp17.6 (class II) and Hsp60, and the development of induced thermotolerance in Arabidopsis thaliana cell culture under conditions of mitochondrial dysfunction. It was shown that treatment by mitochondrial inhibitors and uncouplers at the time of mild heat shock downregulates HSP synthesis, which is important for induced thermotolerance in plants. The exposure to elevated temperature induced an increase in cell oxygen consumption and hyperpolarization of the inner mitochondrial membrane. Taken together, these facts suggest that mitochondrial functions are essential for heat-induced HSP synthesis and development of induced thermotolerance in A. thaliana cell culture, suggesting that mitochondrial–nuclear cross-talk is activated under stress conditions. Treatment of Arabidopsis cell culture at 50°C initiates a programmed cell death determined by the time course of viability decrease, DNA fragmentation and cytochrome c release from mitochondria. As treatment at 37°C protected Arabidopsis cells from heat-induced cell death, it may be suggested that Hsp101, Hsp70 and small heat-shock proteins, the synthesis of which is induced under these conditions, are playing an anti-apoptotic role in the plant cell. On the other hand, drastic heat shock upregulated mitochondrial Hsp60 synthesis and induced its release from mitochondria to the cytosol, indicating a pro-apoptotic role of plant Hsp60.
The mitochondrion is a semi-autonomous organelle and has its own DNA (mtDNA), but its genetic capacity is limited. The plant mitochondrial proteome might contain as many as 2000–3000 different gene products (Millar et al., 2005), but only 33 proteins, three rRNAs and 20 tRNAs are encoded by the Arabidopsis thaliana mitochondrial genome (Unseld et al., 1997). Therefore the majority of mitochondrial proteins are encoded by the nuclear DNA, synthesized on cytosolic ribosomes and subsequently transported into the mitochondria. This genetic division of labour requires a precise co-ordination of gene expression between two genomes to allow the growth and division of functional mitochondria.
Results recently obtained shed light on the mechanisms controlling the expression of nuclear genes, the products of which participate in oxidative phosphorylation, mtDNA replication, transcription, translation and mitochondrial biogenesis. Two classes of nuclear regulatory proteins have been defined in mammalian cells: DNA-binding transcription factors (NRF-1, NRF-2, SP-1 and others) and transcriptional co-activators (PGC-1α, PGC-1α and PRC) (Scarpulla, 2006). In the yeast Saccharomyces cerevisiae, for example, when their mitochondrial functions are damaged by the loss of mtDNA, a pathway of communication from mitochondria to the nucleus, mediated by the transcription factors Rtg1 and Rtg3 and a cytoplasmic protein Rtg2, is activated. As a result, the expression of genes encoding Cit2 (peroxisomal citrate synthase), Dld3 (cytosolic d-lactate dehydrogenase), Mrp13 (mitochondrial ribosomal protein), as well as the COX5A and COX6 genes (subunits of the cytochrome c oxidase complex) is upregulated (Liu and Butow, 2006).
Interestingly, expression of CIT2 and DLD3 is also upregulated during mild heat shock in yeast (Gasch et al., 2000; Sakaki et al., 2003), suggesting that mitochondrial–nuclear cross-talk is activated under stress conditions. A few examples of mitochondrial regulation related to HSP expression have been reported in plants. In maize mitochondrial mutants, the expression of some HSP genes was shown to be elevated (Kuzmin et al., 2004). The line of A. thaliana expressing a transgene encoding a maize mitochondrial sHSP, ZmHSP22, demonstrated altered expression of other nuclear HSP genes (Rhoads et al., 2005). Microarray analysis of rotenone-treated Arabidopsis cells indicated the upregulation of gene sets involved in mitochondrial chaperone activity, such as HSP60, HSP10 and others (Lister et al., 2004). Short heat treatment enhanced A. thaliana anoxia tolerance, suggesting that heat-shock proteins (HSPs) may play a role in survival in limited-oxygen conditions. Otherwise, anoxic conditions induce the expression of HSP genes (Loreti et al., 2005).
Although these findings support the conclusion that there is a pathway of communication between mitochondria and the nucleus, the precise mechanisms for how the two genomes interact, integrate and affect each other’s genetic activity are poorly understood. Studies in different cell types have indirectly suggested different mechanisms of signalling from mitochondria to nucleus, including O2 tension (Bailey-Serres and Chang, 2005; Poyton and Dagsgaard, 2000); reactive oxygen species (ROS) (Rhoads et al., 2006); and altered Ca2+ homeostasis (Biswas et al., 2005).
Results obtained in the past decade revealed that oxidative phosphorylation is not the only function of mitochondria. Mitochondria produce the biosynthetic precursors (Liu and Butow, 2006; Møller, 2001) and actively participate in regulation of programmed cell death (PCD) in mammalian and plant cells (Vianello et al., 2007). Our recently obtained data (Rikhvanov et al., 2005) suggest that mitochondria are also involved in the execution of adaptive response. We have shown that mild heat shock induces an increase in the electrochemical potential across the inner mitochondrial membrane (mtΔΨ) of yeast cells (Rikhvanov et al., 2005). A similar event has been shown to occur in animal cells (Balogh et al., 2005). The addition of mitochondrial inhibitors and uncouplers during mild heat shock suppressed the mtΔΨ and inhibited heat-induced synthesis of Hsp104 and elevation of thermotolerance to subsequent drastic heat shock (Rikhvanov et al., 2005).
We believe that the general process of mitochondrial signalling is common to simple eukaryotes such as yeast and to humans. However, the mode of expression of nuclear-encoded proteins may be quite different among particular proteins and organisms, so the questions remain: do mitochondria regulate the expression of other HSPs, except of yeast Hsp104 and, if this is true, does the same rule apply to other organisms? To answer these questions, in the present work we studied the effect of mitochondrial inhibitors and uncouplers on the development of induced thermotolerance and synthesis of various HSPs using cell culture of A. thaliana.
Exploring heat-shock conditions required for induction of HSP synthesis and cell death in A. thaliana cell culture
First, we determined the heat-treatment regime required for induction of HSPs and development of induced thermotolerance. Also, we wanted to know at which temperature the cells begin to die and monitor the dependence between HSP synthesis and progression of cell death. To do that, cell culture grown at 26°C was separately treated at 26, 37, 39, 43, 46 or 50°C at the intervals indicated in Figure 1 legend, and was incubated additionally at 26°C for 120 min. The synthesis of Hsp101, Hsp17.6 (class I) and Hsp17.6 (class II) was undetectable at 26°C, but was dramatically elevated after heat shock at 37°C (Figure 1b). Further rise in temperature suppressed their synthesis. In contrast, the level of the constitutively expressed mitochondrial heat-shock protein Hsp60 was unchanged after heat shock at 37°C, but gradually increased as the temperature rose (Figure 1b). Anti-Hsp101 antibodies specifically recognize Hsp101 (product of AT1 G74310 gene locus), while other antibodies have wide substrate specificity and detect homologous polypeptides. There are three Hsp60s putatively localized to the mitochondria (Hill and Hemmingsen, 2001), and multiple Hsp17.6 (Vierling, 1991) proteins. It appears that these related HSPs behave in a similar way in the case of heat shock of different severity and duration.
Results of microarray experiments generated from genevestigator microarray data sets (Zimmermann et al., 2004; http://https://www.genevestigator.ethz.ch) confirmed that, in Arabidopsis cell cultures treated at 38°C, the level of HSP101 and HSP17.6C-C1 gene transcripts is strongly elevated, but the level of HSP60 is unchanged (Figure S1).
Apoptotic cell death is not immediate (Swidzinski et al., 2002; Wu and Wallner, 1983; Figure 6a). So to assess the effect on cell viability after heat shock of the treatments indicated above, we incubated culture cells for 48 h at 26°C and then stained them with TTC (2,3,5-triphenyltetrazolium chloride). As seen in Figure 1c, the heat shock at 37 and 39°C hardly affected the cell’s ability to reduce TTC, but exposure at 43, 46 and 50°C diminished it, correlating with elevation of the Hsp60 level and inhibition of Hsp101 and sHSP synthesis (Figure 1b). Thus culture cells did not undertake futile efforts to resist heat damage by synthesizing stress proteins important for thermotolerance when imposed heat shock was beyond a certain threshold, suggesting that a lack of Hsp101 and sHSP induction at high temperatures is the cause, rather than a result, of cell death. This observation is in line with the hypothesis that HSP induction and cell death are mutually exclusive responses to stress in mammal cells (Samali et al., 1999a).
Mild heat shock induces the cell’s ability to survive subsequent, more severe heat-shock exposure. Mild heat shock at 37°C significantly increased the cell’s capacity to survive drastic heat shock at 50°C, but pre-treatment at 39°C resulted in an insignificant effect on induced thermotolerance (Figure S2a). Note that there was marked induction of Hsp101, Hsp17.6 (class I) and Hsp17.6 (class II) synthesis at 37°C (Figure 1b), but only subtle induction of Hsp101, and no increase in the level of Hsp17.6 (class I and II) at 39°C.
The reliability of results obtained by the TTC-reduction method was supported by regrowth assay on agar media (Figure S2b). Arabidopsis cells grown at 26°C and directly transferred to 50°C failed to form callus; alternatively, cells that received preliminary treatment at 37°C did not lose this ability. Therefore mild heat shock at 37°C had no deleterious effect on cells, effectively induced synthesis of Hsp101, Hsp17.6 (class I) and Hsp17.6 (class II) proteins, and induced thermotolerance to subsequent drastic heat-shock exposure. Taking into account the above results, in subsequent experiments we used pre-treatment at 37°C as the inducer of adaptive response in Arabidopsis cell culture.
Effect of oxidative phosphorylation inhibitors on heat-shock response of A. thaliana cell culture
Further, we inquired whether mitochondrial dysfunction caused by application of mitochondrial inhibitors and uncouplers results in the inhibition of heat-shock response in A. thaliana. To assess this, we added inhibitors to cell cultures of A. thaliana and determined changes in protein levels after incubation at 26 or 37°C, followed by recovery (Figure 2a). First, all chemicals were added at concentrations that suppressed the heat-shock response in yeast experiments (Rikhvanov et al., 2005). Immunoblotting showed (Figure 2b) that treatment with 150 μm NaN3, 20 μm carbonyl cyanide m-chlorophenylhydrazone (CCCP), 10 μm antimycin A and 1 mm 2,4-dinitrophenol (DNP) at the ordinary temperature of incubation (26°C) induced Hsp70 synthesis, but there was no increase in the level of Hsp101, Hsp17.6 (class I) and Hsp17.6 (class II). However, the presence of these agents during treatment at 37°C strongly inhibited the induction of their synthesis. No treatment had any appreciable effect on the Hsp60 level (Figure 2b).
To measure the effect of mitochondrial inhibitors and uncouplers at the level of transcript accumulation, we monitored changes in transcript levels of HSP genes in a similar kind of experiment using semi-quantitative multiplex RT–PCR analysis. As a control, we used the EIF4A1 gene, as its expression was shown to be unchanged in Arabidopsis cell culture during treatment at 38°C (Figure S1). The HSP101, HSP17.6C and HSP60 gene-expression profiles in cells treated at 26 and 37°C in the presence or absence of 150 μm NaN3 and 20 μm CCCP (Figure 2d) resembled, in general, those observed in immunoblotting experiments (Figure 2b). Mild heat shock (37°C) increased the level of HSP101 and HSP17.6C-C1 transcripts, but this increase was diminished in the presence of azide and CCCP. Unlike the immunoblotting experiment, treatment with azide and CCCP at 26°C moderately increased the level of HSP17.6C-C1 transcript (Figure 2d). Therefore, on the one hand, treatment with mitochondrial inhibitors and uncouplers at ordinary incubation temperature can activate expression of HSP70 and HSP17.6C-C1; on the other hand, these agents suppress HSP expression induced by mild heat shock. We obtained qualitatively the same results when HSP expression was determined directly after a mild heat shock, without subsequent 2-h recovery periods (Figure S3).
The level of Hsp60 synthesis probably increased in accordance with elevation of heat-shock temperature, achieving a maximum at 50°C (Figure 1b). The presence of mitochondrial inhibitors and uncouplers during heat exposure at 50°C (10 min) and subsequent incubation at 26°C (120 min) did not decrease the heat-induced level of Hsp60 (Figure 3). Moreover, the combined effect of heat shock and chemicals at high concentrations led to an additional increase in Hsp60 level (Figure 3c). This fact suggests that disturbance of mitochondrial function by mitochondrial inhibitors and uncouplers specifically suppresses mild heat shock-induced synthesis of Hsp101, Hsp70, Hsp17.6 (class I) and Hsp17.6 (class II), but does not inhibit elevation of Hsp60 synthesis induced by hard heat shock.
To reveal whether the inhibition of HSP synthesis leads to inhibition of induced thermotolerance, we studied the effect of treatments with mitochondrial inhibitors and uncouplers on the ability of Arabidopsis cells to reduce TTC. Cells were treated with inhibitors at 26 or 37°C, washed three times to remove inhibitors, resuspended in the fresh medium, allowed to recover at 26°C for 120 min, then exposed to drastic heat shock at 50°C for 10 min. Survival was assessed after 48 h incubation at 26°C. As seen in Figure 4b, the addition of 150 μm azide, 20 μm CCCP, 10 μm antimycin A and 1 mm DNP during treatment at 37°C severely suppressed induced thermotolerance. However, the presence of agents in these concentrations at 26°C also significantly suppressed the ability of Arabidopsis cells to withstand drastic heat shock at 50°C (Figure 4b, insert). This result is in contrast with previous data obtained using yeast cells, when the same concentrations had no negative effect on yeast tolerance to severe heat shock after treatment at 26°C (Rikhvanov et al., 2005). Apparently such a contradiction may be explained by our failure to remove completely chemicals from the Arabidopsis cultures after preliminary treatment at 26 or 37°C. Indeed, the residual amounts of inhibitors in the cells could enhance the lethal effect of drastic heat-shock treatment (Figure S4). So, it was clear that we could not draw any conclusions about the effect of these agents (at given concentrations) on induced thermotolerance, as they also suppressed basic thermotolerance. To circumvent this problem, we have undertaken similar experiments using much lower concentrations of mitochondrial inhibitors and uncouplers. Treatments with azide, CCCP, antimycin A and DNP at 26°C at concentrations of 50, 4, 2 and 200 μm, respectively, had no negative effects on the ability of Arabidopsis cells to reduce TTC after a damaging heat shock at 50°C (Figure 4c, insert). Moreover, azide and DNP increased their thermotolerance to some extent. Nevertheless, these low concentrations of chemicals effectively suppressed the development of induced thermotolerance after treatment at 37°C in a similar way to high concentrations (Figures 4c and S5). Furthermore, treatment with inhibitors at 37°C (without subsequent heat shock at 50°C) had no negative effect on vitality, indicating that the suppression of induced thermotolerance is not because of generalized irreversible cell damage caused by the these agents (Figure S6). Therefore the mitochondrial inhibitors and uncouplers used at low concentrations specifically suppressed the induced thermotolerance of A. thaliana.
The additional immunoblotting experiments revealed that, in general, low concentrations of mitochondrial inhibitors and uncouplers suppressed the heat-induced synthesis of Hsp101, Hsp17.6 (class I) and Hsp17.6 (class II) in a similar manner to high concentrations (Figure 2c). However, RT–PCR experiments showed that low inhibitor concentrations suppressed the heat-induced expression of HSP101, but did not change AtHSP17.6C-C1 expression (Figure 2e). Taken together, these results suggest that disturbance of mitochondrial functions by mitochondrial inhibitors and uncouplers at the time of mild heat shock suppressed HSP expression in Arabidopsis cell culture.
To perform their functions, most HSPs require ATP (Rikhvanov et al., 2007). Mitochondrial inhibitors and uncouplers undoubtedly would decrease the ATP level in Arabidopsis cells, and this event, in principle, could impair the function of HSPs. Moreover, azide is a potent inhibitor of anti-oxidant enzymes. Therefore, whether or not inhibitors suppress induced thermotolerance by impairing HSP (anti-oxidant enzyme) functions or inhibiting expression of HSPs must be verified. To differentiate between these possibilities, cell culture was treated at 26 and 37°C (120 min) without any additions, to allow HSP expression. Azide (50 μm) was added during the recovery period at 26°C (120 min). After that, the cells were washed, resuspended in fresh medium and subjected to 50°C (10 min), as indicated above. As can be seen in Figure S7, azide had a minor effect on cells’ ability to induce thermotolerance (cf. Figure 4c). Therefore the ability of azide to suppress induced thermotolerance appears to be determined by inhibition of heat-induced HSP expression. The suppression of expression of HSPs, important for development of a defence program, evidently leads to inhibition of induced thermotolerance to severe heat shock.
Mild heat shock increases the cellular respiration rate and induces hyperpolarization of inner mitochondrial membrane in A. thaliana cell culture
The results shown above suggest that mitochondrial functions are required for development of a defence program in Arabidopsis cell culture during mild heat shock. Therefore it was interesting to compare the oxygen consumption rates of cell culture at elevated temperature to determine changes in respiratory activity when HSP synthesis is induced. To exclude the possibility that change in respiratory rate is due to the Q10 effect (the proportional change in the rate of respiration with a 10°C change in temperature), the Arabidopsis cells incubated at 26°C were treated at 37°C (120 min) and allowed to recover (26°C, 30 min), then the respiration rates were measured in a polarographic chamber at 26°C. The oxygen consumption rate of Arabidopsis plants was shown to increase during exposure to mild, elevated temperatures that correlates with the enhanced expression of some genes encoding mitochondrial proteins (Rizhsky et al., 2004). In line with these data, a twofold increase in oxygen consumption of A. thaliana cells was observed after incubation at 37°C for 120 min (Figure 5a). The presence of 50 μm azide (complex IV inhibitor) and 2 μm antimycin A (complex III inhibitor) during incubation at 26 or 37°C significantly suppressed respiration, indicating that both inhibitors were quite effective in inhibiting mitochondrial functions, even at low concentrations. The combined effect of 2 μm antimycin A and 2 mm benzohydroxamic acid, an alternative oxidase inhibitor, produced almost complete inhibition of oxygen consumption at both temperatures tested. The latter fact indicates that elevation of oxygen consumption induced by heat shock was mediated mainly by activation of respiratory activity, and excludes the involvement of non-respiratory pathways in enhanced oxygen consumption (Møller et al., 1988). The addition of uncouplers 4 μm CCCP and 200 μm DNP did not significantly change the respiration rate (Figure 5a).
The effects of inhibitors of the electron-transport chain and of uncouplers on mitochondrial function are well documented, and they are known to dissipate the mtΔΨ of inner mitochondrial membranes in fundamentally different ways. DNP and CCCP enable free movement of protons across the inner mitochondrial membrane, resulting in collapse of the H+ gradient. Azide and antimycin A impair H+ extrusion through the inner membrane because of inhibition of electron flow, which consequently leads to the same effect, depolarization of the inner mitochondrial membrane. Mild heat shock was shown to elicit the hyperpolarization of inner mitochondrial membrane in yeast (Rikhvanov et al., 2005) and animal cells (Balogh et al., 2005). To assess the change in mtΔΨ during heat shock in Arabidopsis cell culture, we used tetramethylrhodamine methyl ester (TMRM) as a fluorescent probe. This indicator dye was accumulated by mitochondria and exhibited a red shift in both their absorption and fluorescence emission spectra in proportion to mtΔΨ (Scaduto and Grotyohann, 1999). TMRM added to cells at 26°C gave a specific staining of mitochondria, reflecting noticeable mtΔΨ in the cells under these conditions. Mild heat shock at 37°C induced a strong increase in the accumulation of TMRM, indicating an increase in electrochemical polarization of the inner mitochondrial membrane (Figure 5b). The treatment by inhibitors effectively suppressed TMRM staining in control and in heat-shocked cells (data not shown). Therefore the induction of HSP synthesis by mild heat shock is accompanied by activation of respiratory mitochondrial activity and hyperpolarization of inner mitochondrial membrane. Agents that are capable of depolarizing the inner mitochondrial membrane also inhibited the HSP expression induced by mild heat shock.
Heat-induced execution of PCD in A. thaliana cell culture
Swidzinski et al. (2002) have shown that treatment of Arabidopsis cell cultures at 55°C for 10 min was sufficient to induce PCD, as characterized by timing of cell death and DNA degradation. As shown in Figure 6a, cells collected immediately after heat shock at 50°C (10 min) were characterized by only modestly decreased capacity to reduce TTC. However, subsequent incubation at 26°C progressively diminished cell viability. Apoptotic cell death is not immediate, and such a time course of viability decrease indicates the execution of PCD in Arabidopsis cell culture (Swidzinski et al., 2002; Vacca et al., 2004). Indeed, DNA fragmentation in the control samples and in the cells collected at time point zero after heat shock at 50°C was not detectable, but it became quite visible when the majority of cells had lost their viability (Figure 6b).
Cytochrome c released from mitochondria acts as an important molecule in the early stage of PCD in animal cells (Garrido et al., 2006) and plant cells (Vianello et al., 2007). To address the question of whether drastic heat shock initiates PCD in A. thaliana cell culture via cytochrome c release, the effect of heat shock on cytochrome c distribution in mitochondrial and cytosolic protein fractions was determined. As expected, at normal temperature cytochrome c and the mitochondrial protein isocitrate dehydrogenase were found to be present in mitochondrial fractions, but not in cytosolic fractions (Figure 6c). Drastic heat shock at 50°C induced a decrease in the level of cytochrome c in mitochondrial fractions and its appearance in cytosolic fractions. At the same time, isocitrate dehydrogenase localized in mitochondrial matrix has not been detected in the cytosolic extracts. Therefore drastic heat shock at 50°C appears to result in the execution of PCD in Arabidopsis cell culture.
Role of HSPs in heat shock-induced PCD in A. thaliana cell culture
Hsp60 is suggested to play an anti-apoptotic as well as a pro-apoptotic role in animal cells (Didelot et al., 2006). In particular, Samali et al. (1999b) have demonstrated that mammalian Hsp60 is released from mitochondria in cytosol during the induction of apoptosis with staurosporine, and that release of mitochondrial Hsp60 may accelerate caspase activation. In order to determine the subcellular distribution of Hsp60 in Arabidopsis cells, cytosolic and mitochondrial fractions were analysed. In control cells (26°C) as well as in cells subjected to mild heat shock (37°C), Hsp60 was found in the mitochondrial fraction. Drastic heat shock (50°C) resulted in a decrease in Hsp60 level in the mitochondrial fraction, and induced its appearance in the cytosolic fraction. The discrepancy between the level of Hsp60 at 50°C in Figures 1 and 7 is probably because of the loss of some amount of protein during fractionation. Mild heat shock significantly prevented the release of this protein from mitochondria to the cytosol at 50°C (Figure 7). Release of cytochrome c during treatment at 50°C was also inhibited by mild heat shock, but the degree of inhibition was significantly less than in the case of Hsp60. Such a result is in line with data obtained on animal cells (Li et al., 2000; Mosser et al., 2000), and suggests that mild heat shock may inhibit the execution of PCD in Arabidopsis cells upstream and downstream of cytochrome c release. Hsp101 and Hsp17.6 (class I), which are localized in cytosol in A. thaliana (Agarwal et al., 2001; Vierling, 1991), were found in the mitochondrial fraction of cells treated at 37°C, and subjected to drastic heat shock at 50°C. During drastic heat shock, these proteins are known to form oligomeric complexes with heat-aggregated proteins (Lee et al., 2005). So it is quite likely that these complexes co-sediment with mitochondria. However, the presence of Hsp101 (but not Hsp17.6) in the mitochondrial fraction of cells treated only at 37°C is intriguing. It is unlikely that the majority of cellular proteins become aggregated at this temperature. It is worth noting that cytosolic hexokinase was shown to be associated with the mitochondria, and plays a role in the control of PCD in plants (Kim et al., 2006). In contrast, Hsp70 was evenly distributed between mitochondrial and cytosolic fractions, reflecting wide substrate specificity of SPA-820AP antibodies that detect the cytosolic and organellar forms of Hsp70.
The heat-shock response is a highly conserved phenomenon, characterized by rapid induction of HSPs that function primarily as molecular chaperones to ensure the correct function of many cellular proteins under conditions of elevated temperature. The expression of heat shock-regulated genes was controlled by the binding of heat-shock transcription factors (Hsfs) to the highly conserved heat-shock element (HSE) in the promoters of target genes (Miller and Mittler, 2006). It appears that the transcription factor DREB2A is also involved in the initiation of this process (Sakuma et al., 2006). While the regulation of HSP expression at DNA level is more-or-less defined, the physiological and biochemical nature of the signals that activate transcriptional factors during mild heat shock remains conjectural.
As we have shown earlier (Rikhvanov et al., 2005), mild heat shock elicits hyperpolarization of the inner mitochondrial membrane in yeast. The uncouplers or mitochondrial inhibitors that are capable of dissipating the potential on the inner mitochondrial membrane under particular experimental conditions prevent the synthesis of Hsp104 induced by mild heat shock, and thus inhibit the development of induced thermotolerance. In the current investigation, we have shown that our conclusions, obtained by using simple unicellular eukaryotes, are valid for higher plants.
Data obtained in the current work suggest a linkage between the functioning of plant mitochondria and HSP expression. Mild heat shock elicits hyperpolarization of the inner mitochondrial membrane in Arabidopsis cell culture (Figure 5b). Agents that are capable of preventing an increase in mtΔΨ also inhibited HSP expression (Figure 2) and development of thermotolerance (Figure 4) induced by a mild heat shock. The inhibitors did not affect, or slightly increased, basic thermotolerance after treatment at normal growth temperature (Figure 4c), had no negative effect on the viability of cell culture at 37°C (Figure S6), and specifically inhibited induced thermotolerance (Figure S7). Therefore the effect of mitochondrial inhibitors and uncouplers on heat-shock response in A. thaliana cell culture may be explained by its ability to suppress the rise in mtΔΨ. This supposition is in line with our earlier results showing that antimycin A suppressed the rise of mtΔΨ, inhibited Hsp104 synthesis and the development of induced thermotolerance in respiring yeast cells subjected to a mild heat shock, but had no influence in cells that did not respire (Rikhvanov et al., 2005). Thus mitochondrial dysfunctions during mild heat shock not only result in downregulation of the HSP104 gene of S. cerevisiae (Rikhvanov et al., 2005), but also suppress the expression of the HSP101 gene, the HSP104 homologue in A. thaliana, HSP17.6-C1 (class I and II) genes, and probably heat-induced HSP70 genes.
DNA microarray experiments have shown that heat shock and mitochondrial inhibitors induce different but partly overlapping sets of genes. Thus, in yeast cells, several HSP genes are induced by antimycin A (Epstein et al., 2001; Sakaki et al., 2003). Arabidopsis plants were found to show a significant common response to antimycin A, aluminium, cadmium, hydrogen peroxide and virus infection (Yu et al., 2001). These results are corroborated by immunoblotting and PCR analysis showing that mitochondrial inhibitors are capable of inducing expression of nuclear genes, including HSPs, at normal temperatures (Barrett et al., 2004; Howarth, 1990; Massie et al., 2003; Maxwell et al., 2002; Sweetlove et al., 2002). In some cases, treatment with these inhibitors increased cell tolerance to a lethal heat exposure (Howarth, 1990; Massie et al., 2003). We did not observe an elevation of Hsp101, Hsp17.6 (class I) and Hsp17.6 (class II) content in immunoblotting experiments when cells were treated with inhibitors at normal growth temperature, but there was an apparent increase in Hsp70 level (Figure 2b). Moreover, the RT–PCR analysis showed slight, but distinct, elevation of the HSP17.6C-C1 transcript level after azide and CCCP treatment at 26°C (Figure 2d,e). At the same time, all inhibitors suppressed the development of induced thermotolerance and expression of heat-regulated genes (Figure 2). It appears that the functional state of mitochondria determines the expression of nuclear heat-regulated genes and, depending on environmental conditions, the mitochondrial dysfunction could activate or suppress their expression.
Taken together, our data support the hypothesis (Maxwell et al., 2002; Yu et al., 2001) that the plant mitochondrion plays a crucial role in conveying intracellular stress signals to the nucleus leading to altered expression of stress genes. Thus perturbation of mitochondrial functions as result of inhibitor application (Sakaki et al., 2003), or mutations in genes coding components of the electron transport chain (Kuzmin et al., 2004) or expressing a transgene encoding a maize mitochondrial sHSP, ZmHSP22 in A. thaliana (Rhoads et al., 2005), were shown to affect the expression of nuclear-encoded HSP genes. Probably such regulation is not limited to heat-shock conditions only, and is valid for other stress situations. For instance, mutation in a gene encoding a protein with high similarity to the 18-kDa Fe–S subunit of complex I (NADH dehydrogenase) leads to reduced cold induction of stress-responsive genes such as RD29A, KIN1, COR15A and COR47 in Arabidopsis plants (Lee et al., 2002).
Considering the effect of mitochondrial inhibitors and uncouplers on the heat-shock response, one can point out that additions of mitochondrial inhibitors and uncouplers would inevitably result in the inhibition of ATP production in the cell. ATP is required for transcription and protein synthesis, therefore the ability of chemicals to suppress HSP expression at 37°C may be determined by ATP shortage. The fact that treatment by mitochondrial inhibitors and uncouplers was able to induce expression of nuclear genes (HSPs among them) at ordinary temperatures (Barrett et al., 2004; Epstein et al., 2001; Howarth, 1990; Massie et al., 2003; Maxwell et al., 2002; Sakaki et al., 2003; Sweetlove et al., 2002; Yu et al., 2001) (Figure 2) argued strongly against this. Moreover, mitochondrial inhibitors and uncouplers did not affect the elevation of Hsp60 level after heat exposure at 50°C (Figure 3). This suggests that involvement of ATP shortage in inhibition of Hsp101, Hsp17.6 (class I) and Hsp17.6 (class II) synthesis by these agents, observed during mild heat shock (Figure 2a), is unlikely as, if it were so, a similar effect should be observed for Hsp60.
There are a number of reports showing that treatment with salicylic acid (SA), an important component of the plant signalling pathway, protects plants from drastic heat shock (Clarke et al., 2004; Dat et al., 1998, 2000; Larkindale and Knight, 2002; Larkindale et al., 2005). One group of authors has shown that SA at normal temperature induced HSP17.6C-C1 expression, but had no effect on HSP101 expression in A. thaliana (Clarke et al., 2004). Others have shown that sodium salicylate inhibited heat shock-induced gene expression in yeast (Giardina and Lis, 1995) and in Drosophila (Winegarden et al., 1996). It is worth noting that, although SA was able to induce Hsp70/Hsc70 synthesis in tomato cells and leaf discs, the induction of Hsp17 and HsfA2 by heat shock was significantly repressed in the presence of SA (Cronje, 2002). On the other hand, SA was known to disrupt normal mitochondrial function, acting as both an uncoupler and an inhibitor of mitochondrial electron transport (Maxwell et al., 2002; Norman et al., 2004). Therefore SA may be a natural compound that affects the mitochondrial regulation of nuclear gene expression during biotic and abiotic stresses.
Another important aspect of the current investigation is the possible participation of HSPs in inhibition of PCD in plant cells. Heat shock at 50°C appears to result in the execution of PCD in Arabidopsis cell culture (Figure 6). As pre-treatment at 37°C suppressed this process (Figures 4 and 7), one may suggest that mild heat shock-induced synthesis of Hsp101, Hsp70 and Hsp17.6 (class I and II) (Figures 1 and 2) protects plant cells from PCD. It is well known that mild heat shock inhibits the execution of PCD in animal cells induced by diverse apoptotic stimuli (Bettaieb and Averill-Bates, 2005; Jiang et al., 2005). While the anti-apoptotic role of HSPs in animals is well defined (Didelot et al., 2006), to date there is no information about a similar role for HSPs in plants. The progression of heat-induced PCD was accompanied by cytochrome c release from mitochondria. Among a variety of events occurring during heat-induced PCD, the release of cytochrome c has been shown to occur in both mammalian (Garrido et al., 2006) and plant mitochondria (Vacca et al., 2006). In the latter case, such a release was prevented by addition of anti-oxidant enzymes, suggesting participation of ROS in this process. Mild heat shock inhibited cell death (Figure 4), but suppression of cytochrome c release was rather modest (Figure 7). It appears that HSPs, induced by mild heat shock, may inhibit execution of PCD upstream and downstream of cytochrome c release. Such a supposition is in line with the result obtained by Li et al. (2000) on mammalian cells, showing that mild heat shock or overexpression of HSP70 prevented cell death induced by a lethal heat shock, but not cytochrome c release. However, other authors have shown that mild heat shock or overexpression of HSP70 could inhibit cytochrome c release in mammalian cells (Mosser et al., 2000). So we suggest that the release of cytochrome c from mitochondria per se does not necessarily lead to cell death, both in mammalian (Garrido et al., 2006) and in plant cells.
It appears that the progression of heat-induced PCD was accompanied by induction of Hsp60 synthesis (Figure 1b). The induction of Hsp60 synthesis was not correlated with the development of a defence program because the increase of Hsp60 content was accompanied by an increase in PCD rate (Figures 1c and 6). HSP60 gene expression was shown to be upregulated in the course of PCD induced by heat shock at 55°C in Arabidopsis cell culture (Swidzinski et al., 2002). In the latter case, however, upregulation of HSP60 gene expression was modest, which may be explained by differences in experimental protocols. Swidzinski et al. (2002) collected samples after 10 min exposure at 55°C, but we used cells that were heat-shocked for 10 min at 50°C and additionally incubated for 120 min at 26°C. PCD is a form of cellular death requiring active gene expression, therefore cells were apparently in a more advanced phase of cell suicide in our case. Alternatively, a heat-induced increase in the level of Hsp60 synthesis may be the course of HSP60 transcript stabilization or preferred translation. Hsp60 is known to play an anti-apoptotic as well as a pro-apoptotic role in animal cells (Didelot et al., 2006). In particular, Samali et al. (1999b) have demonstrated that human Hsp60 is released from mitochondria in cytosol after induction of apoptosis with staurosporine and, supposedly, play a role in triggering PCD through caspase cascade activation. Similar data have been obtained in our experiments. Induction of PCD in A. thaliana cell culture is accompanied by the release of Hsp60 from mitochondria to the cytosol. This event is significantly prevented by mild heat shock (Figure 7), and accompanied by increased thermotolerance. Based on these results, we suggest that upregulation of mitochondrial Hsp60 protein synthesis and its release from mitochondria may be an early feature of heat-induced PCD in plants. Of course, such proposals are rather speculative and require detailed experimental confirmation.
Cell culture growth conditions
Cell culture of A. thaliana (ecotype Columbia) was obtained in March 2005 in our laboratory, as described by Mathur and Koncz (1998). Briefly, 14-day-old sterile seedlings were chopped with scissors into a 250-mL Erlenmeyer flask with 60 ml medium. The flask was placed on a shaker (80 r.p.m.) and incubated in darkness at 26°C. Medium composition was optimized after several experiments, and finally consisted of mineral salts according to Murashige and Skoog (1962), 3% sucrose, 0.5 mg l−1 thiamin–HCl and 0.1 mg l−1 2,4-D. Cells were subcultured every 14 days by sixfold dilution of the cell culture with fresh medium. For solid media, agar (7 g l−1) was added. For performing experiments, 8-day-old cultures in a logarithmic phase of growth were used (Figure S8).
Treatments of Arabidopsis cell culture
To test the effect of various regimes of heat-shock exposure on survival and HSP synthesis, culture cells were treated separately at 26°C (120 min), 37°C (120 min), 39°C (120 min), 43°C (60 min), 46°C (60 min) or 50°C (10 min), and allowed to recover at 26°C. For protein extraction and determination of survival, samples of cells were collected after 2 and 48 h incubation, respectively, at 26°C. To study the effect of mild pre-treatment heat shock on tolerance to drastic temperature exposure, cells were treated at 37 or 39°C for 120 min, allowed to recover at 26°C for 120 min, then exposed to 50°C for 10 min. The effect of NaN3, DNP, antimycin A and CCCP on HSP synthesis and induced thermotolerance was studied by the addition of inhibitors, at the concentrations indicated, in cell culture for 120 min incubation at 26 or 37°C. After three successive cycles of washing, the culture was resuspended in fresh medium and allowed to recover at 26°C for 120 min. Then one part of the culture was used for protein and RNA extraction; the other part was exposed at 50°C for 10 min. Survival was determined after 48 h incubation at 26°C. To study the execution of heat-induced PCD, cells were treated at 50°C for 10 min, and samples for survival determination and DNA-fragmentation analysis were collected at different intervals after heat exposure.
Protein extraction, SDS electrophoresis and immunoblotting
The samples collected were washed with K–Na phosphate buffer pH 7.0 and stored at –70°C until protein extraction. Cells were resuspended in protein-isolation buffer (0.1 m Tris–HCl pH 7.4–7.6, 3 mm SDS, 1 mm 2-mercaptoethanol), frozen in liquid nitrogen and ground with quartz powder. The insoluble cell components were discarded by centrifugation (15 000 g, 20 min), and total protein was precipitated by cold acetone and dissolved in the sample loading buffer (0.625 m Tris–HCl pH 6.8, 0.08 m SDS, 0.1 m 2-mercaptoethanol, 10% glycerol, 0.001% bromophenol blue). Protein content was determined according to Lowry. Proteins were separated by SDS–PAGE, transferred onto nitrocellulose membrane, and probed with antibodies to Hsp101 and Hsp17.6 class I and II (gift of E. Vierling and M. Escobar, University of Arizona, USA), Hsp70 (SPA-820AP, StressGen, http://www.assaydesigns.com) and Hsp60 (SPA-807, StressGen).
Cell survival was determined by using TTC reduction and regrowth (culture growth for 10 days after stress) assays, as described by Wu and Wallner (1983), adapted as follows: approximately 0.1 g fresh cells were suspended in 1.5 ml K–Na phosphate buffer (pH 7.0) containing 0.05% TTC. The incubation with TTC was carried out in anaerobic conditions at 26°C for 3 h in the dark. Formazan, the product of the TTC reduction, was extracted with 95% ethanol at 60°C for 15 min. Formazan concentration was quantified by absorbance at 485 nm and calculated per gram FW. For regrowth assay, cells following treatment were transferred on Petri plates with solidified MS medium. Plates were photographed after 10 days’ incubation at 26°C in the dark.
Reverse transcription (RT)–PCR analysis
Total RNA was isolated from Arabidopsis cell culture using the SV Total RNA Isolation System (Promega, http://www.promega.com) according to the manufacturer’s protocol. Equal amounts of total RNA were used for each experiment to synthesize the first cDNA strand using the oligodT18 primer and REVERTA kit (AmpliSense, Allergen, http://www.interlabservice.ru). RT–PCR assays were performed using cDNAs as templates and primers designed to amplify parts of the coding sequences of HSP101, EIF4A1, HSP60 and HSP17.6C-C1 genes according to the manufacturer’s instructions (AmpliSense). The structure and sizes of PCR primers were based on the corresponding cDNA sequences available from the http://www.arabidopsis.org database for A. thaliana (Table 1). For amplification of HSP101 (AT1 G74310), HSP17.6C-C1 (AT1 G53540) and EIF4A1 (AT3 G13920), specific primers were used. Primer pairs used for amplification (HSP60) were deduced from the cDNA sequences of two HSP60 transcripts, AT2 G33210 and AT3 G23990 (Table 1). A PCR of the EIF4A1 gene, shown to be unchanged under heat stress, was used for normalization between samples.
The absence of genomic DNA in the cDNA samples was verified by PCR amplification without a reverse-transcription step. Semi-quantification of target transcripts was carried out using hot-start PCR, selecting a number of cycles in the linear range of amplification. Equal amounts of cDNA, obtained in the reactions of first-strand synthesis with 100 ng of total RNA and 5 ρm of each primer were used in PCR reactions. Twenty-five amplification cycles were performed in accordance with the results of preliminary experiments where amplification was carried out in dynamics for multiplex PCR and for each gene alone (Figure S9). The following optimized PCR conditions were used: denaturation at 95°C for 40 sec; annealing at 58°C for 1 min; extension at 72°C for 1.5 min with pre-heating at 95°C for 3 min; and final extension at 72°C for 7 min using a Mastercycler gradient thermocycler (Eppendorf, http://www.eppendorf.com). Equal volumes of RT–PCR products were separated elecrophoretically on 1.5% agarose gel. The gels were stained using ethydium bromide followed by visualization of stained bands with a UV-transilluminator (Bio-Rad, http://www.bio-rad.com).
Assay of oxygen consumption rate
The respiration rates of cell cultures were determined at 26°C using a Clark-type oxygen electrode with a temperature-controlled chamber containing 1.4 ml suspension. Cells were treated at 26 or 37°C for 120 min in the presence or absence of inhibitors, and allowed to recover at 26°C for 30 min. Oxygen consumption rates were measured in nanomoles of O2 per minute per milligram FW at 26°C.
Determination of PCD execution
Cells collected after 0, 6, 24, 48 and 54 h incubation, at 26°C following heat exposure at 50°C, were used for determination of their ability to reduce TTC (as described above) and DNA-fragmentation assay. Samples were ground in liquid nitrogen and resuspended in extraction buffer (100 mm Tris-HCl pH 8.0, 50 mm EDTA, 500 mm NaCl, 10 mmβ-mercaptoethanol pH 8.0) followed by the addition of 1 ml 20% SDS. DNA isolation and agarose gel electrophoresis were performed as described by Balk and Leaver (2001).
Isolation of mitochondrial and cytosolic protein fractions
The cells were carefully ground in mortar and then disintegrated in a Dounce grinder tube in the homogenization buffer (10 mm Tris–HCl pH 7.5, 2 mm EDTA, 0.3 m sucrose, 0.1% 2-mercaptoethanol). The unbroken cells were discarded by centrifugation (4000 g, 3 min), then mitochondrial fractions were sedimented (10 000 g, 3 min). Soluble cytosolic proteins were precipitated by cold acetone from the supernatant and dissolved in the sample loading buffer. Mitochondrial proteins were extracted in the sample loading buffer by 5 min incubation in boiling water bath. Mitochondrial and cytosolic proteins were separated by SDS–PAGE, transferred onto nitrocellulose membrane and probed with antibodies to cytochrome c (Biosan, http://www.biosan.com) and mitochondrial isocitrate dehydrogenase (gift of Dr Oliver, Iowa State University, USA).
Assay of mitochondrial membrane potential
For the measurement of mitochondrial membrane potential, 5 μm TMRM (Molecular Probes, http://probes.invitrogen.com) was added to cell cultures. After incubation for 15 min at 26 or 37°C, cells were washed, transferred to a microscope slide and observed with a fluorescence microscope (Carl Zeiss, Axiovert 200) with optical filters (546 nm excitation, 590 nm emission) to visualize the red fluorescent probe. Quantitative images were captured and data were analysed.
Reproducibility of the results
All experiments were repeated a minimum of three times. The data obtained were analysed statistically and means and SE were determined.
This work was supported by the Russian Foundation for Basic Research (projects 05-04-48966a, 07-04-01177a, 07-04-01055a), by the Basic Research Program of the Russian Academy of Sciences ‘Dynamics of gene pools of plants, animals, and humans’, and by a Grant of the President of Russia for support to leading science schools (4812.2006.4). The authors are grateful to Professor E. Vierling and Dr M. Escobar (University of Arizona, USA) for the kind gift of antibodies to Hsp101 and Hsp17.6 (class I and II) and Dr D.J. Oliver (Iowa State University, USA) for antibodies to mitochondrial isocitrate dehydrogenase. The authors are indebted to Professor I.M. Møller (Department of Agricultural Sciences, The Royal Veterinary and Agricultural University, Denmark) for critical reading of the manuscript and helpful comments) and to Ms J. Sutton for checking the English language.