Heat Suppresses Activation of an Auxin-Responsive Promoter in Cultured Guard Cell Protoplasts of Tree Tobacco^,

Abstract

Cultured guard cell protoplasts (GCP) of tree tobacco (Nicotiana glauca) comprise a novel system for investigating the cell signaling mechanisms that lead to acquired thermotolerance and thermoinhibition. At 32°C in a medium containing an auxin (1-naphthaleneacetic acid [NAA]) and a cytokinin (6-benzylaminopurine), GCP expand, regenerate cell walls, dedifferentiate, and divide. At 38°C, GCP acquire thermotolerance within 24 h, but their expansion is limited and they neither regenerate walls nor reenter the cell cycle. These putative indicators of auxin insensitivity led us to hypothesize that heat suppresses induction of auxin-regulated genes in GCP. Protoplasts were transformed with BA-mgfp5-ER, in which the BA auxin-responsive promoter regulates transcription of mgfp5-ER encoding thermostable green fluorescent protein (GFP) or with a similar 35S-cauliflower mosaic virus constitutive promoter construct. Heat suppressed NAA-mediated activation of BA. After 21 h at 32°C in media with NAA, 49.0% ± 3.9% of BA-mgfp5-ER transformants strongly expressed GFP; expression percentages were similar to those of 35S-mgfp5-ER transformants at 32°C or 38°C. After 21 h at 38°C in media with NAA, 7.9% ± 1.6% of BA-mgfp5-ER transformants weakly expressed GFP, similar to GCP cultured at 32°C in media lacking NAA. Expression at 38°C was not increased by incubating for 48 h or increasing NAA concentrations 20-fold. After 9 to 12 h at 38°C, BA was no longer activated when cells were transferred to 32°C. Heat-stressed cells accumulate reactive oxygen species, and hydrogen peroxide (H₂O₂) suppresses auxin-responsive promoter activation in Arabidopsis (Arabidopsis thaliana) mesophyll protoplasts. H₂O₂ did not suppress BA activation at 32°C, nor did superoxide and H₂O₂ scavengers prevent BA suppression at 38°C.

Worldwide, the three major, interrelated abiotic plant stresses affecting plant growth and development are rising atmospheric CO₂ concentration, heat, and drought (Mittler, 2006). Models (Scholze et al., 2006; Yesson and Culham, 2006) and experimental studies (Jump et al., 2006; Marchand et al., 2006; Qaderi et al., 2006; Sato et al., 2006; Walker et al., 2006; White et al., 2006) suggest that rising global temperatures will have important consequences both for crop production (Mittler, 2006; Qaderi et al., 2006; Sato et al., 2006; White et al., 2006) and for natural selection of noncrop plants (Marchand et al., 2006; Walker et al., 2006).

Heat can trigger cell-autonomous mechanisms that enable plants to survive extended periods of high temperature (Francis and Barlow, 1988; Hong et al., 2003; Larkindale et al., 2005), but little is known about how the network of cell signaling pathways for acquired thermotolerance is activated or regulated (Larkindale et al., 2005). Moreover, plants that tolerate heat usually exhibit reduced growth and/or delayed development. Heat damages cells of young leaves (Pareek et al., 1997), delays cell cycling in meristems (Francis and Barlow, 1988), inhibits cell division in endosperm of developing seeds (Commuri and Jones, 2001), interferes with anthesis in corn (Zea mays; Warrington, 1983) and rice (Oryza sativa; Matsui and Omasa, 2002), and inhibits fruit set in tomato (Solanum lycopersicum) by interfering with sugar metabolism at a critical stage of male reproductive development (Sato et al., 2006). In Brassica napus, even a modest increase in growth temperature reduces plant biomass under all but ideal growth conditions, and whole-plant concentrations of indole-3-acetic acid are reduced at high temperature under all growth conditions (Qaderi et al., 2006). Despite these observations, the cell signaling mechanisms that underlie thermoinhibition of plant cell expansion and division are largely unexplored.

Cultured guard cell protoplasts (GCP) of tree tobacco (Nicotiana glauca) ‘Graham’ are a novel system for investigating the cell signaling mechanisms that lead to acquired thermotolerance and thermoinhibition (Roberts et al., 1995; Gushwa et al., 2003; Tallman, 2005). At 32°C, in a medium containing an auxin (1-naphthaleneacetic acid [NAA]) and a cytokinin (6-benzylaminopurine [BAP]), GCP survive in high percentages (70%–80%) for up to 1 month (Roberts et al., 1995). Under the influence of NAA and BAP, GCP expand 20- to 30-fold, regenerate cell walls, dedifferentiate, and divide to form callus from which plants can be regenerated (Sahgal et al., 1994). Both NAA and BAP must be supplied exogenously before GCP will survive and undergo the extensive cellular remodeling that accompanies dedifferentiation and division (Gushwa et al., 2003). 1-Aminoethoxyvinyl-glycine, an inhibitor of ethylene biosynthesis, and abscisic acid (ABA) are each lethal at 32°C (Roberts et al., 1995).

GCP cultured at 38°C survive in the same high percentages as at 32°C, but neither exogenous NAA nor BAP is required for, nor do 1-aminoethoxyvinyl-glycine or ABA reduce, cell survival at 38°C (Roberts et al., 1995; Gushwa et al., 2003). GCP cultured at 38°C in NAA and BAP expand only 5- to 6-fold (Roberts et al., 1995). They do not regenerate cell walls, nor do they make the G1-to-S phase cell cycle transition (Gushwa et al., 2003). ABA prevents even the limited cell growth that occurs in its absence (Roberts et al., 1995; Gushwa et al., 2003), effectively maintaining GCP in their differentiated state in vitro (Taylor et al., 1998).

Plant cell expansion, cell wall deposition, and cell division are all regulated, in part, by auxin (Kende and Zeevaart, 1997). Because each of these processes is inhibited in GCP cultured at 38°C (Roberts et al., 1995; Gushwa et al., 2003), we hypothesized that heat might suppress induction of auxin-regulated genes in cultured GCP. To test the plausibility of this hypothesis, we transformed GCP with the surrogate reporter BA-mgfp5-ER, in which the Pisum sativum BA early auxin-responsive promoter regulates transcription of mgfp5-ER encoding a thermostable form of GFP (Siemering et al., 1996; Aspuria et al., 2002). We then exposed transformed GCP to NAA and after various periods of culture at 32°C or 38°C examined the cells for GFP accumulation microscopically and with a fluorescence-activated cell sorter (FACS). GCP from the same cell isolates were transformed with a similar construct containing the 35S-cauliflower mosaic virus (CaMV) constitutive promoter and cultured in parallel as controls.

In Arabidopsis (Arabidopsis thaliana) mesophyll protoplasts, exogenously supplied hydrogen peroxide (H₂O₂) activates a mitogen-activated protein (MAP) kinase, ANP1, which suppresses activation of the soybean (Glycine max) GH₃ auxin-responsive promoter in a transient gene expression assay (Kovtun et al., 1998, 2000). Tobacco (Nicotiana tabacum) plants overexpressing NPK1, the tobacco ortholog of ANP1, are better able to withstand heat shock (Kovtun et al., 2000). While a plant's capacity to metabolize reactive oxygen species (ROS; ¹O₂, O₂⁻, H₂O₂, HO^•) generally increases when it acquires thermotolerance (Suzuki and Mittler, 2006), ROS may accumulate transiently in cultured plant cells in response to heat (Vacca et al., 2004). Furthermore, GCP cultured at 38°C retain a 41-kD MAP kinase that is lost at 32°C just prior to the G1-to-S phase transition (Gushwa et al., 2003). Therefore, we also examined whether exogenously supplied H₂O₂ or ROS scavengers would alter induction of the BA promoter by NAA in tree tobacco GCP cultured at 32°C or 38°C.

Our results indicate that heat suppresses capacity for BA promoter activation in GCP but that neither superoxide nor H₂O₂ is required to suppress cellular capacity for BA activation.

RESULTS

Heat Suppresses NAA-Mediated Activation of the BA Early Auxin-Responsive Promoter

Based on putative indicators of auxin insensitivity at high temperature, we tested the hypothesis that heat would suppress activation of an auxin-responsive promoter in cultured GCP. After 21 h at 32°C in medium with NAA, 49.0% ± 3.9% of GCP transformed with BA-mgfp5-ER expressed GFP (mean; se; n = 3; Figs. 1 and 2, A and B). This expression percentage was similar to those of cells from the same isolates that were transformed with 35S-mgfp5-ER and cultured for 21 h at 32°C (Figs. 1 and 2E) or at 38°C (Figs. 1 and 2F). When GCP from the same batches of BA-mgfp5-ER transformants were cultured for 21 h at 38°C in medium with NAA (Fig. 2D), the mean percentage of cells expressing GFP was 7.9% ± 1.6% (Fig. 1), similar to the 7.5% ± 0.9% of cells expressing GFP after 21 h at 32°C in medium lacking NAA (Fig. 2C). When viewed through the microscope, GFP fluorescence intensities of BA-mgfp5-ER transformants cultured in medium with NAA were visibly greater at 32°C than at 38°C (compare Fig. 2A to 2D). There was no visible expression of GFP in BA-mgfp5-ER transformants cultured at 38°C for up to 24 h in medium lacking NAA (Table I).

Figure 1.

Mean percentages of cultured GCP of tobacco expressing GFP from mgfp5-ER regulated by the BA early auxin promoter (▪, ♦) or by the 35S CaMV promoter (•, ▴) over 24 h at 32°C (▪, •) or 38°C (♦, ▴). All cells were cultured in a medium containing NAA. Each value is the mean and se from three separate experiments. At each time point, approximately 300 cells were scored microscopically for GFP accumulation in each replicate experiment.

Figure 2.

Expression of GFP from mgfp5-ER regulated by the BA early auxin promoter (A–D) or by the 35S CaMV promoter (E and F) in GCP of tree tobacco cultured for 21 h at 32°C (A–C and E) or at 38°C (D and F). Cells were cultured with (A, B, D, and E) or without (C and F) NAA as shown. In D, GFP fluorescence was visible to the eye but may be difficult to distinguish on the printed segment. Red in A, Chlorophyll autofluorescence from guard cell chloroplasts. B shows localization of GFP to the endoplasmic reticulum = image in A minus red channel. Bar in A = 15 μm.

Table I.

Effect of extended incubation, increased NAA concentration, and ROS treatments on the percentage of cultured GCP of tree tobacco expressing GFP under control of the BA early auxin promoter from P. sativum at 32°C versus 38°C

Values are means and se from three experiments performed on separate days. Approximately 300 cells were scored microscopically for GFP accumulation in each replicate. All media contained BAP. *, Significantly different from corresponding 18 h, 1× NAA control at the same temperature (unpaired t test; n = 3; P = ≤0.05).

Treatment	Mean Percent Expressing GFP, 32°C	Mean Percent Expressing GFP, 38°C
+NAA for 24 h	48.0 ± 4.8	6.3 ± 2.7
+NAA for 24 h + 24 additional h	51.5 ± 6.0	10.7 ± 2.2
−NAA for 24 h	12.9 ± 1.4	0.0 ± 0.0
20× increased concentration of NAA for 18 h	43.6 ± 1.5	2.3 ± 0.3
1× NAA for 18 h; same batches of transformants as 20×	45.9 ± 1.5	2.8 ± 0.8
200 μm H2O2 for18 h	42.1 ± 1.9	2.3 ± 1.2
50 μmN-acetyl-cysteine + 50 μm Tiron for 18 h	43.8 ± 1.5	1.1 ± 0.7
0.03 mg mL−1 catalase for 18 h	36.8 ± 2.0*	0.7 ± 0.2

We used a FACS to quantify relative GFP intensities in transformed GCP cultured at the two temperatures. In GCP transformed with BA-mgfp5-ER, after 15 h of culture at 32°C in medium containing NAA, a distinct population of GFP-expressing cells (Fig. 3A) could be distinguished from a sham-transformed cell population from the same isolate (Fig. 3F) that was used to set the background fluorescence threshold. When GCP from the same batch of transformants were cultured at 32°C without NAA (Fig. 3B) or at 38°C with NAA (Fig. 3C), the level of GFP expression was below the threshold level required to distinguish GFP-containing cells from those of sham-treated controls. In GCP transformed with 35S-mgfp5-ER, after 15 h of culture in medium containing NAA, the median GFP intensity at 32°C (Fig. 3D) was approximately twice that of the median GFP intensity of cells from the same batch of transformants cultured at 38°C (Fig. 3E). Median GFP intensities of GCP transformed with 35S-mgfp5-ER and cultured at both temperatures were substantially greater than median GFP intensities of cells transformed with BA-mgfp5-ER and cultured at 32°C in a medium containing NAA (Figs. 2, B, E, and F, and 3, A, D, and E).

Figure 3.

Population distribution analysis of cultured GCP of tree tobacco expressing GFP from mgfp5-ER regulated by the BA early auxin promoter (A–C) or by the 35S CaMV promoter (D and E). Cells were cultured for 15 h at 32°C (A, B, and D) or 38°C (C and E) in media containing or lacking NAA as indicated and then analyzed with a FACS. In each segment, cell distributions to the right of the vertical line are expressing GFP; those to the left are not expressing GFP. Because GCP exhibit intense red chloroplast autofluorescence that obscures weaker GFP fluorescence, protoplasts subjected to transformation protocols without plasmid DNA (F, sham control) were used to establish conservative threshold levels of autofluorescence beyond which only those cells unequivocally expressing GFP were scored by the machine. Note that vertical scales of various segments differ.

We tested the hypotheses that GCP accumulate GFP more slowly at 38°C and that higher concentrations of NAA are required to activate BA at 38°C. Extending culture times to 48 h did not increase significantly the mean percentage of GCP transformed with BA-mgfp5-ER that expressed GFP at 38°C compared to the corresponding 24-h temperature control (Table I; unpaired t test; n = 3; P = 0.22). Increasing NAA concentrations 20-fold did not increase mean GFP expression percentages over those of 1× NAA controls from the same batches of transformants either at 32°C (Table I; unpaired t test; n = 3; P = 0.57) or at 38°C (Table I; unpaired t test; n = 3; P = 0.6).

After 9 h at 38°C, the BA Promoter Is No Longer Activated by Cell Transfer to 32°C

We examined whether heat-mediated suppression of BA becomes irreversible with time by preincubating GCP for various times at 38°C and then transferring them to 32°C for 18 h before examining them for GFP accumulation. After 3 or 6 h at 38°C in medium containing NAA, on average, approximately three to four times as many GCP transformed with BA-mgfp5-ER expressed GFP upon subsequent transfer to 32°C for an additional 18 h as did controls from the same batches of transformants maintained for an equal period at 38°C (Fig. 4). After ≥9 h at 38°C, the mean percentages of cells expressing GFP after transfer to 32°C for an additional 18 h were similar to those of parallel transformant controls (approximately 7%–8%) maintained for an equal period at 38°C (Fig. 4).

Figure 4.

Percentage of cultured GCP of tree tobacco transformed with BA-mgfp5-ER expressing GFP after preincubation for various periods at 32°C or 38°C in a medium with NAA and subsequent incubation at the alternate temperature for 18 h. As controls, at every time point, the percentage of cells expressing GFP was also estimated in BA-mgfp5-ER-transformed protoplasts incubated continuously at 32°C or at 38°C for the same accumulated time periods as treatments. Each value is the mean and se from three separate experiments. At each time point, approximately 300 cells were scored microscopically for GFP accumulation in each replicate experiment.

In reciprocal experiments, culture for ≤12 h at 32°C before an additional 18 h of culture at 38°C reduced the mean percentages of GCP expressing GFP below those of controls from the same batches of transformants maintained at 32°C for an equal period of time (Fig. 4). After 12 h at 32°C, transfer to 38°C diminished only slightly the mean percentage of GCP expressing GFP compared to parallel transformant controls maintained at 32°C for an equal period (Fig. 4).

When GCP transformed with BA-mgfp5-ER were transferred after 18 h at 32°C to 38°C for an additional 18 h, there was no visible decrease in fluorescence intensity at the end of the 36-h period (data not shown). When cells from the same batches of transformants were transferred after 18 h at 38°C to 32°C for an additional 18 h, there was no visible increase in fluorescence intensity at the end of the 36-h experiment (data not shown).

H₂O₂ Does Not Suppress NAA-Mediated BA Activation at a Lower Culture Temperature

Cultured plant cells can transiently accumulate ROS under heat stress (Vacca et al., 2004), and H₂O₂ suppresses auxin-responsive promoter activation in Arabidopsis mesophyll protoplasts (Kovtun et al., 1998, 2000). Addition of 200 μm H₂O₂ to the culture medium did not decrease or increase significantly the mean percentages of GCP expressing GFP at 32°C (unpaired t test; n = 3; P = 0.25) or at 38°C (unpaired t test; n = 3; P = 0.77) over those of their corresponding 18-h, 1× NAA temperature controls (Table I). In medium with NAA, the superoxide scavenger, Tiron (50 μm), and the H₂O₂ scavenger, N-acetyl-cysteine (50 μm) in combination did not affect the mean percentage of GCP transformed with BA-mgfp5-ER that expressed GFP after 18 h at 32°C compared to untreated 18-h, 1× NAA 32°C controls (unpaired t test; n = 3; P = 0.44). Nor did these compounds alter significantly the mean percentage of cells that expressed GFP at 38°C (Table I; unpaired t test; n = 3; P = 0.23) compared to untreated 18-h, 1× NAA 38°C controls. Addition of catalase (0.03 mg L⁻¹) to the cell culture medium reduced mean expression percentages at 32°C and at 38°C compared to their corresponding 18-h, 1× NAA temperature controls (Table I). The reduction at 32°C was statistically significant (Table I; unpaired t test; n = 3; P = 0.03), but the reduction at 38°C was not (Table I; unpaired t test; n = 3; P = 0.068).

Cultured GCP Have the Capacity to Rapidly Degrade H₂O₂

H₂O₂ is a signaling molecule in ABA-induced stomatal closure (Murata et al., 2001; Bright et al., 2006; Kwak et al., 2006), and it is plausible that low H₂O₂ concentrations might be maintained in guard cells to protect ABA signaling capacity. When cell culture medium containing 23.5 μm H₂O₂ was exposed for 5 min to various dilutions of culture extracts, the number of relative light units (RLU) emitted per second in a chemiluminescent H₂O₂ assay decreased in inverse exponential proportion to extract dilutions (Fig. 5). Extract from the equivalent of 1 × 10⁴ freshly isolated cells was sufficient to decrease the number of RLU s⁻¹ to approximately 4.0% ± 2.2% of that of an untreated control in 5 min (Fig. 5; mean; se; n = 3). Extracts from freshly isolated GCP and GCP cultured for 12 h at 32°C or 38°C in media containing NAA had similar capacities to degrade H₂O₂ (Fig. 5). In assays containing the peroxidase inhibitors sodium azide (10 mm), KCN (10 mm), and 3-amino-1,2,4-triazole (10 mm), the percentage of RLU s⁻¹ remaining after 5 min compared to those of controls containing a single inhibitor but no extract were 6.3% ± 4.4%, 7.6% ± 3.4%, and 21.7% ± 2.2%, respectively (mean; se; n = 3). In a similar experiment employing the NADPH oxidase inhibitor diphenylene iodonium chloride (0.2 mm), the percentage of RLU s⁻¹ remaining after 5 min compared to those of controls was 18.8% ± 7.4% (mean; se; n = 3). Only the percentage of RLU s⁻¹ remaining after 5 min in an assay containing 10 mm 3-amino-1,2,4-triazole was significantly greater than the 4.0% ± 2.2% remaining after 5 min in an assay containing extract but no inhibitor (unpaired t test; n = 3; P = 0.004).

Figure 5.

Percentage of RLU s⁻¹ remaining in a chemiluminescent H₂O₂ assay after 5 min of exposure of a 23.5-μm solution of H₂O₂ to various dilutions of extracts of cultured tree tobacco GCP. Extracts were prepared from freshly isolated protoplasts (□) or from protoplasts cultured for 12 h at 32°C (▵) or at 38°C (○) by freezing 2 × 10⁵ cells in culture medium at −80°C, thawing, and centrifuging. Extract dilutions are expressed as cell equivalents in the fixed-volume assay mixture. Inset, Expanded version of region between 0 and 2,500 cell equivalents. Values are means and se from three replicate experiments on separate days. The mean initial RLU s⁻¹ of untreated solutions in all treatments was 3.2 ± 0.4 × 10⁶ (n = 8); the mean value after 5-min exposure to culture medium containing no cells was 3.2 ± 0.4 × 10⁶ RLU s⁻¹ (n = 8).

DISCUSSION

Heat Suppresses NAA-Mediated Activation of the BA Early Auxin-Responsive Promoter in Cultured GCP

From microscopic (Figs. 1 and 2) analyses the mean percentage of BA-mgfp5-ER transformants expressing GFP after 18 to 24 h at 32°C in medium containing NAA was approximately 45% to 50% (Fig. 1). This value was not significantly lower than those of 35S-mgfp5-ER transformants from the same isolates that were cultured similarly at 32°C or 38°C (Fig. 1). In cells cultured at 32°C in medium containing NAA, GFP could be detected readily with a FACS (Fig. 3A), a CCD camera (Fig. 2B), and the eye.

By contrast, <10% of GCP from the same isolates transformed with BA-mgfp5-ER expressed GFP after 18 to 24 h at 38°C in a medium containing NAA (Fig. 1). In cells cultured at 38°C in medium containing NAA, GFP fluorescence could not be detected with a FACS (Fig. 3C) or imaged effectively with a CCD camera (Fig. 2D); fluorescence levels were so low that GFP could only be detected with the eye. No GFP expression could be detected visually in BA-mgfp5-ER transformants cultured at 38°C in a medium lacking NAA at any time point (see 24-h time point, Table I; time course not shown).

In addition to suppression of promoter activation, reductions in GFP accumulation could be envisioned to result from heat-induced gene modification (e.g. methylation), global reductions in transcription rates, global reductions in translation rates, accelerated degradation of GFP mRNA, accelerated GFP degradation, or (a) combination(s) of these processes. Reductions in GFP accumulation would not be expected due to gene silencing in this system because a cryptic intron has been removed from this particular version of the GFP gene to thwart plant gene silencing mechanisms (Haseloff et al., 1997). However, results of experiments with 35S transformants cultured at 38°C suggest that the effects of heat on promoter-independent processes are not sufficient to explain the observed reduction in GFP expression percentages and GFP intensities in BA transformants cultured at 38°C in media with NAA.

In three separate experiments with 35S-mgfp5-ER transformants, median GFP intensities after 15 to 18 h at 38°C were approximately one-half those of cells cultured at 32°C (compare Fig. 3D to 3E). However, this decrease in intensity did not translate to a corresponding reduction in the percentage of cells that could be observed microscopically to express GFP (Figs. 1 and 2). Indeed, GFP was still expressed strongly at 38°C under the control of the constitutive 35S promoter (Figs. 2F and 3E), indicating that individually, or in combination, the effects of heat on promoter-independent processes could not account for the magnitude to which heat suppresses GFP accumulation when mgfp5-ER is under control of the auxin-responsive BA promoter. In short, if the eye can still detect GFP in 7% to 10% of cells when it can be detected neither by FACS nor a CCD camera, it would still be possible to see through the microscope GFP in the GCP shown in Figure 2B even if mean GFP intensities were reduced by 50%. At most, a heat-induced promoter-independent reduction in median GFP intensities at 38°C in BA-mgfp5-ER transformants exposed to NAA might have been expected to reduce the percentage of cells observed to express GFP from approximately 45% to 50% to approximately 20% to 25%. However, at every time point, the percentage of cells expressing GFP at 38°C in a medium with NAA was <10% (Fig. 1), and GFP intensities were low enough that this small population of cells could only be detected microscopically (Fig. 3C). Thus, we conclude that heat suppresses activation of the BA promoter.

GFP-based reporters were chosen for these experiments for a number of reasons. GFP has been shown to be a quantitative reporter of gene expression in eukaryotic cells (Soboleski et al., 2005) and is highly stable once it is produced in plant cells (Chiu et al., 1996). In addition, the mutated GFP employed in these experiments is thermostable (Siemering et al., 1996) and has been mutated to resist silencing (Haseloff et al., 1997). After cells cultured at one temperature for 18 h were switched to the alternate temperature for another 18 h, we could not detect visually any changes in intensity of GFP fluorescence. We have not, however, determined directly the extent to which GFP intensities reflect GFP mRNA levels under all of the conditions described here. Thus, these results do not rule out entirely the possibility that sustained heat has an effect on overall equilibrium rates of GFP turnover in GCP cultured at 38°C. They indicate, however, that heat-induced alterations to turnover rates alone would be inadequate to explain the magnitude of the reduction in GFP accumulation observed at 38°C in a medium containing NAA.

The percentage of BA-mgfp5-ER transformants expressing GFP at 38°C in a medium with NAA was not increased by extending incubations to 48 h (Table I), suggesting that heat-treated transformants did not simply accumulate GFP more slowly.

Heat-Induced Suppression of NAA-Mediated BA Activation Is Not Reversed by a Higher NAA Concentration

Increasing NAA concentrations 20-fold did not increase the percentage of BA-mgfp5-ER transformants that expressed GFP at 38°C (Table I). These data suggest that GCP harbor at least one heat-sensitive auxin-mediated promoter activation pathway that is fully suppressed at high temperature rather than just reduced in its sensitivity to NAA. This observation is consistent with our previous result showing that neither 20- nor 50-fold increases in NAA concentration enabled cultured GCP to reenter the cell cycle at 38°C (Gushwa et al., 2003). BA-mgfp5-ER was developed as a reporter for use in intact Arabidopsis plants where it is expressed in various cell layers of the root. The 1× NAA concentration of 0.81 μm used in our culture experiments with GCP is the optimum concentration required for GCP to survive and reenter the cell cycle at 32°C (Tallman, 2005), but it is considerably lower than the 10-μm concentration used to activate the promoter in Arabidopsis plants (Aspuria et al., 2002). Even so, GCP reached maximum GFP accumulation within the same 21- to 24-h time period as Arabidopsis roots (Aspuria et al., 2002), and increasing concentrations 20-fold did not visibly increase final steady-state GFP intensities (data not shown).

After 9 to 12 h at 38°C, Heat-Induced Loss of Cell Capacity for NAA-Mediated BA Activation Is Irreversible

When BA-mgfp5-ER transformants were preincubated at 38°C in an NAA-containing medium for <9 h prior to being cultured for 18 h at 32°C, substantial percentages (20%–40%) of cells still accumulated GFP (Fig. 4). When BA-mgfp5-ER transformants were preincubated at 38°C for 9 to 18 h prior to being cultured at 32°C, cells did not accumulate GFP in percentages greater than those of controls maintained at a constant 38°C (approximately 7%–8%; Fig. 4). Reciprocal controls preincubated at 32°C for 9 to 12 h before transfer to 38°C did accumulate GFP. These data indicate that 9 to 12 h are required to irreversibly suppress GCP capacity for NAA-mediated BA activation. This period is much longer than those typical of heat shock experiments (Vacca et al., 2004), raising the question of whether plant cell mechanisms used to respond to extended periods of high temperature may be distinct from, or augment those used by, plant cells to respond to transitory heat shock.

In previous experiments, returning GCP to the lower culture temperature of 32°C after 9 to 12 h at 38°C resulted in a high percentage of cell death, but GCP incubated for >12 h survived in higher percentages upon transfer to 32°C (Gushwa et al., 2003). Indeed, after 24 h at 38°C, GCP became thermotolerant and survival at 32°C nearly equaled that of unheated controls (Gushwa et al., 2003). At 32°C, exogenous auxin (NAA) and cytokinin (BAP) and endogenously produced ethylene are all required for cell survival (Roberts et al., 1995; Gushwa et al., 2003). We speculate that NAA and BAP are both required for survival at 32°C because cytokinin stabilizes the 1-aminocyclopropane-carboxylic acid synthase (ACS) protein (Chae et al., 2003) that results from auxin-mediated induction of ACS genes (Abel et al., 1995; Abel and Theologis, 1996; Merritt et al., 2001; Tanaka et al., 2006), thereby optimizing ethylene production required for protoplast survival at 32°C. After 24 h of preincubation at 38°C, none of these growth regulators are required for GCP to survive in culture at 32°C (Gushwa et al., 2003). Our combined data suggest that after 9 to 12 h at 38°C, GCP have become auxin insensitive but have not yet developed the mechanism for auxin/cytokinin/ethylene-independent survival; i.e., development of auxin insensitivity at high temperature precedes development of thermotolerance.

Unlike GCP cultured at 32°C, GCP cultured at 38°C are thermoinhibited. Their expansion is limited and they fail to regenerate cell walls and make the G1-to-S phase transition required for cell cycle reentry. While the data presented here do not establish definitively a causal linkage between development of auxin insensitivity and thermoinhibition, all of the processes above have been linked to auxin signaling for gene expression (Meyer et al., 1984; Himanen et al., 2002; Mockiatis and Estelle, 2004).

Lack of NAA Does Not Fully Eliminate GFP Accumulation at 32°C; Heat Does Not Fully Eliminate NAA-Mediated BA Activation at 38°C

At 32°C, a small percentage (approximately 5%–10%) of GCP accumulate GFP in medium without auxin. GFP accumulation in the absence of auxin could be due to promoter leakiness, to the action of endogenous auxin, or both. In no experiment did culturing BA-mgfp5-ER transformants at 38°C in media containing NAA completely suppress BA-mediated GFP expression. Even when GCP were preincubated at 38°C for 18 h before they were cultured at 32°C in a medium containing NAA, nearly 10% of cells expressed GFP (Fig. 4). These results could be due to leakiness in the promoter, a heat-insensitive promoter activation pathway, or both, but the observation that GFP does not accumulate in BA-mgfp5-ER transformants at 38°C in a medium lacking NAA indicates that heat-insensitive GFP accumulation at 38°C is NAA dependent.

Low Levels of NAA-Mediated Gene Activity Are Probably Not Required for Cell Survival at 38°C

The percentage of BA-mgfp5-ER transformants expressing GFP at 38°C in a medium with NAA was similar to that of BA-mgfp5-ER transformants cultured at 32°C in a medium lacking NAA. However, GCP do not survive at 32°C in a medium lacking NAA (Gushwa et al., 2003), whereas GCP cultured at 38°C GCP survive in high percentages with or without NAA (Gushwa et al., 2003). Because there was no detectable GFP expression at 38°C in media lacking NAA, we conclude that heat-insensitive NAA-mediated gene regulatory mechanisms are probably not required for survival of GCP at high temperature. This result is consistent with our previous observation that removing auxin and/or cytokinin and/or inhibiting ethylene production are all lethal to GCP cultured at 32°C but not to GCP cultured at 38°C (Roberts et al., 1995; Gushwa et al., 2003).

Neither H₂O₂ nor Superoxide Is Required for, or Adequate to Suppress, NAA-Induced BA Activation

In mesophyll protoplasts isolated from Arabidopsis leaves, exogenous H₂O₂ activates a MAP kinase, ANP1, which suppresses activation of the soybean GH₃ auxin-responsive promoter (Kovtun et al., 1998, 2000). Our data indicate that H₂O₂ and superoxide are not required for development of auxin insensitivity in GCP cultured at high temperature. Neither scavengers of superoxide and H₂O₂ in combination, nor exogenous catalase, rescued GCP from heat-induced suppression of NAA-mediated BA activation (Table I). Nor did 200 μm H₂O₂ suppress BA activation by NAA in GCP cultured at 32°C (Table I). Furthermore, both at 32°C and 38°C, cultured GCP exhibited a high capacity to scavenge and/or degrade H₂O₂ for up to 12 h (Fig. 5). The capacity to scavenge/degrade H₂O₂ was inhibited to a statistically significant but limited extent by 3-amino-1,2,4-triazole (Havir, 1992), suggesting that the mechanism is largely independent of catalases (Apel and Hirt, 2004). Song et al. (2006) have shown that when stomatal opening is induced by cytokinin and auxin, cytokinins initiate H₂O₂ scavenging mechanisms while auxin restrains or inhibits H₂O₂ production in Vicia faba guard cells.

H₂O₂ is a component of a guard cell ABA signaling pathway(s) that prevents stomatal opening/triggers stomatal closure (Murata et al., 2001; Bright et al., 2006; Kwak et al., 2006). ABA is lethal to GCP cultured at 32°C (Roberts et al., 1995), perhaps because it acts as an antagonist to ethylene produced by coordinate auxin and cytokinin regulation (above; Tanaka et al., 2006). However, once GCP become thermotolerant, they no longer require auxin, and/or cytokinin, and/or ethylene for survival. GCP cultured at high temperatures readily tolerate ABA, and the hormone simply prevents any cell volume expansion as it does in intact leaves (Roberts et al., 1995; Gushwa et al., 2003). Whereas our data suggest that thermotolerance develops after loss of auxin sensitivity, these experiments do not rule out the possibility that ROS might be involved in signaling for development of thermotolerance at the point at which GCP develop an auxin/cytokinin/ethylene-independent survival mechanism at 38°C. However, our data do not provide any evidence for a role of these particular ROS in a ROS-mediated mechanism leading to development of auxin insensitivity in GCP cultured at high temperatures.

CONCLUSION

Within 9 to 12 h of exposure of cultured GCP of tree tobacco to high temperature (38°C), heat irreversibly suppresses their capacity for NAA-mediated activation of the BA early auxin-responsive promoter. Previous studies show that when GCP are returned after 9 to 12 h at 38°C to 32°C, a large percentage die (Gushwa et al., 2003), probably because at the lower temperature they still require auxin, cytokinin, and endogenously produced ethylene and therefore must be auxin responsive to survive (Gushwa et al., 2003). If cells are left at 38°C for 24 h, their survival no longer depends on exogenous auxin or cytokinin or endogenously produced ethylene at either temperature (Gushwa et al., 2003). However, as they become thermotolerant they also become thermoinhibited. Cell expansion is limited and GCP fail to regenerate cell walls and make the G1-to-S cell cycle transition as they do at lower culture temperatures (Gushwa et al., 2003). While these experiments do not establish a causal linkage between heat-induced suppression of NAA-mediated auxin-responsive promoter activation and thermoinhibition, cell expansion, cell wall regeneration, and/or the G1-to-S transition would all be expected to require auxin-mediated gene activity. Taken together, our time course data suggest that at high temperature, development of thermotolerance is subsequent to development of auxin insensitivity and that auxin insensitivity could be responsible, at least in part, for thermoinhibition of cell expansion, cell wall regeneration, and cell cycle reentry. Whether signaling for development of thermotolerance is coordinately and/or obligatorily linked to loss of auxin sensitivity at high temperature awaits investigation.

MATERIALS AND METHODS

Plants

Tree tobacco (Nicotiana glauca) ‘Graham’ was grown from seed (Garden Makers) as described (Tallman, 2005).

Plasmids

Escherichia coli (DH5-α) harboring BA-mgfp5-ER/pPZP121 (Aspuria et al., 2002) or 35S-mgfp5-ER/pPZP 121 were cultured at 37°C for 21 to 22 h at 200 rpm in Terrific broth (modified; Sigma T0918) containing glycerol (0.8% w/v) and chloramphenicol (25 mg L⁻¹). To generate 35S-mgfp5-ER/pPZP121, the HindIII-SacI fragment of the 35S promoter plus mgfp5-ER and the SacI-EcoRI fragment of the nos terminator were obtained from pBIN35S-mgfp5-ER (Siemering et al., 1996) and introduced into the HindII-EcoR1 site of the pPZP121 vector (Hajdukiewicz et al., 1994). Maxiprep kit protocols (Quantum Prep Plasmid Maxiprep kit, Bio-Rad 732–6130) were followed to isolate constructs. To improve purity, fresh, cold wash buffer was made daily. To maximize the yield of DNA from spin basket elution, sterile deionized water was used at 70°C instead of at 25°C. A one-tenth volume of sodium acetate (3 m, pH 5.7) instead of a one-eighteenth volume of NaCl followed by two volumes of ice-cold 100% (w/w) ethanol was used to precipitate DNA; the use of sodium acetate improved the solubility of DNA for the subsequent transformation protocol. The construct/sodium acetate/ethanol mixture was incubated for 30 min at −20°C before the DNA precipitate was collected by centrifugation for 20 min at 3,000g. After a clean up with 70% ethanol as described in the kit protocol, precipitated plasmid DNA was dried in open centrifuge tubes in a laminar flow cabinet until no odor of ethanol could be detected but the DNA pellet still glistened.

Protoplast Isolation and Culture

Tree tobacco GCP were isolated and cultured as described (Tallman, 2005). After the second enzyme digest, the suspension of epidermal segments was filtered and then the epidermis captured on the nylon net was washed with 8 mL of S&T medium, pH 6.8 (Tallman, 2005) to free GCP trapped among segments. The filtrate was divided evenly among four 15-mL centrifuge tubes instead of the two tubes described (Tallman, 2005). The decrease in volume per tube facilitated sedimentation of GCP during centrifugation at 40g for 8 min, resulting in higher yields. After centrifugation, supernatants were aspirated and cells were resuspended in S&T medium, pH 6.8, and combined in two centrifuge tubes in a volume of 8 mL/tube. Following centrifugation, supernatants were aspirated and GCP were washed by centrifugation a third time with 8 mL/tube of S&T medium, pH 6.1. After supernatants were removed, GCP were resuspended in 0.5 mL S&T medium, pH 6.1, combined in a single tube in a total volume of 1 mL, and counted in a hemacytometer (Tallman, 2005). S&T medium, pH 6.1 was added to a total volume of 8 mL and cells were again collected by centrifugation. The supernatant was carefully aspirated in preparation for cell transformation.

Protoplast Transformation

GCP were transformed with either BA-mgfp5-ER containing a truncated version of the early auxin BA promoter from Pisum sativum or with 35S-mgfp5-ER containing the 35S CaMV constitutive promoter. The GFP gene in these reporter constructs, mgfp5-ER, has been modified to yield a thermostable protein (Siemering et al., 1996) that localizes to the endoplasmic reticulum.

Arabidopsis (Arabidopsis thaliana) mesophyll protoplast transformation protocols (Chiu et al., 1996) were adapted to transform tree tobacco GCP. All cell and solution transfers were performed in a laminar flow cabinet. BA-mgfp5-ER or 35S-mgfp5-ER DNA (200 μg) was resolubilized in 0.6 mL modified MaMg (0.4 m mannitol, 19.35 mm MgCl₂×6H₂O, 4 mm MES, pH 5.7; Chiu et al., 1996). If DNA was difficult to resolubilize, solutions were vortexed and pipetted several times before transformation. In experiments requiring more than 200 μg of BA-mgfp5-ER or 35S-mgfp5-ER, DNA in MaMg from multiple tubes was combined in one 50-mL centrifuge tube prior to cell transformation to decrease the potential for variation in transformation rates associated with multiple transformation events. MaMg was added to isolated GCP (see above) so that the cell density was 4 × 10⁵ GCP mL⁻¹. Resuspended GCP were then added to solubilized DNA (0.5 mL MaMg containing 2 × 10⁵ GCP/0.6 mL DNA in MaMg) and the mixtures were incubated on ice for 30 min. During incubation, a fresh polyethylene glycol (PEG) solution was prepared by mixing 4 g PEG4000 (Fluka no. 81240; Sigma), 2.5 mL of 0.8 m mannitol, 1 mL of 1 m CaCl₂×6H₂O, and 3 mL of deionized water (Chiu et al., 1996). At the end of the incubation, an equal volume of filter-sterilized PEG solution was added to GCP and DNA in MaMg. The PEG and fragile GCP were mixed by slowly tilting centrifuge tubes back and forth. The mixture was then incubated at 23°C for 30 min. After incubation, the mixture was divided between two 15-mL centrifuge tubes and cells were washed by centrifugation three times each with 8 mL S&T medium, pH 6.1, as described above. After the final wash, enough medium was aspirated to give a final density of 2 × 10⁵ GCP/1.5 mL for use in the experiments described below.

Temperature Effects on BA- or 35S-Regulated GFP Accumulation

Because tree tobacco GCP had not been transformed previously, neither the time required for maximum GFP accumulation nor the transformation percentages to be expected were known. Therefore, the effects of culture at 32°C versus 38°C on GFP accumulation regulated by the BA promoter (Ballas et al., 1995) were measured every 3 h over a 24-h time course in media containing or lacking NAA. For each experiment, GCP from the same isolates were transformed with 35S-mgfp5-ER and cultured in parallel. These control cultures were used to estimate the maximum expected percentage of GCP transformed and evaluate whether heat might reduce globally overall rates of transcription and/or translation.

Because heat greatly reduced GFP expression in GCP transformed with BA-mgfp5-ER (above), some experiments were extended to 48 h to determine whether GFP was simply accumulating more slowly or with different kinetics at 38°C than at 32°C. In other attempts to increase GFP accumulation at 38°C, NAA concentrations were increased 20-fold to 16.2 μm and GCP were cultured for 18 h.

To initiate experiments, 0.3 mL of S&T medium, pH 6.1, containing 4 × 10⁴ transformed GCP was pipetted into each well of an 8-well chamber slide. NAA (0.012 g) and/or BAP (0.003 g) was dissolved in 10 mL of ethanol. NAA and/or BAP in ethanol (12.5 μL) was added to 25 mL of S&T medium at pH 6.1, and 0.1 mL of the appropriate hormone-containing medium was added to each well to final concentrations of 0.81 μm NAA and/or 0.166 μm BAP. For 20× stocks, 0.24 g NAA and/or 0.06 g BAP was dissolved in 10 mL of ethanol. In each experiment, cells in some wells were treated with a combination of NAA and BAP while GCP in other wells were treated with BAP alone. Chamber slides were incubated at 32°C or at 38°C in water-jacketed incubators.

The number of GCP expressing GFP was estimated microscopically every 3 h for 24 h using an Olympus inverted microscope (Gushwa et al., 2003). GCP were first examined at 400× in a bright field of view and viable cells in the field were identified and counted. The brightfield illuminator was then turned off, and the same field was illuminated with blue actinic light from an epifluorescence illuminator to excite GFP (for illuminator filter specifications, see Poffenroth et al., 1992). For imaging, GFP fluorescence was filtered through an RGB color slider (Q-Imaging, model RGB-HM-S) situated between the side port of the microscope and a Q-Imaging camera (Retiga 2000R) driven by IPLab software (v.3.9.4 r3). All images were captured using the same camera settings (exposure time 500; color range 24; gain 1660; offset 0; normalization autoenhance contrast; magnification 1/2; bin size 1 × 1). As expected, BA-mgfp5-ER and 35S-mgfp5-ER localized to the endoplasmic reticulum. Some GCP exhibited a weak, dull, yellow-green autofluorescence that was easily distinguishable from the brilliant green color of GFP. This autofluorescence was not localized within the cell and usually appeared in cells that were not counted as viable. Approximately 300 GCP were counted and scored for visible GFP accumulation at each time point for each treatment. Images were captured in fields where the number of fluorescing cells and their intensities were the greatest to allow a more objective visual comparison of GFP intensities among treatments and experiments.

FACS

To confirm that the manual counting methods described above accurately reflected GFP accumulation levels in GCP transformed with BA-mgfp5-ER or 35S-mgfp5-ER, a FACS was used to measure the fraction of cells in each treatment expressing GFP and the intensity distribution of GFP fluorescence in cell populations. GCP transformed with either BA-mgfp5-ER or 35S-mgfp5-ER were incubated for 15 h at 32°C or 38°C in media containing BAP with or without NAA. Thirty-thousand events were then machine scored using a FACS Aria (Becton-Dickinson) at low pressure setting and a flow rate of 6 to 7. Voltage settings were: 150 forward scatter, 320 side scatter, 500 fluorescein isothiocyanate (GFP). Because they exhibit intense red chloroplast autofluorescence, GCP subjected to transformation protocols without plasmid DNA (sham controls) were used to establish conservative threshold levels of autofluorescence beyond which only GCP that were unequivocally expressing GFP were scored by the machine.

Reciprocal Temperature Transfer Experiments

In all experiments with BA-mgfp5-ER, fluorescence intensities were substantially lower in cells cultured at 38°C than in those cultured at 32°C (above). Reciprocal temperature transfer experiments were conducted to determine: (1) whether GCP preincubated at 38°C would develop a higher fluorescence intensity upon transfer to 32°C; (2) whether the higher fluorescence intensity observed in cells cultured initially at 32°C would be diminished upon transfer to 38°C; and (3) the approximate time period at 38°C beyond which cellular capacity for BA promoter activation was irreversible. Nine chamber slides with BA-mgfp5-ER-transformed cells in media containing NAA and BAP (above) were incubated at 32°C or at 38°C. The percentage of cells expressing GFP was estimated (above) in one slide from each temperature treatment after 0, 3, 6, 9, 12, 15, or 18 h. Each culture was then incubated at the alternate temperature for an additional 18 h before the percentage of GCP expressing GFP was reestimated. As a control, at every time point the percentage of cells expressing GFP was also estimated in BA-mgfp5-ER-transformed GCP incubated continuously at 32°C or at 38°C for the same accumulated time periods as treatments.

ROS Experiments

To determine whether endogenously produced superoxide and/or H₂O₂ might be required to suppress activation of the BA promoter at high temperature, GCP were cultured at 32°C or at 38°C in a medium containing both the superoxide scavenger, 4,5,dihydroxy-1,3-benzenedisulfonic acid (Tiron; 50 μm), and N-acetyl-l-Cys (50 μm), a scavenger of H₂O₂. To achieve final concentrations, 5 μL of a 4-mm stock solution containing both compounds prepared in cell culture medium were added to 400-μL chamber-slide cultures. To determine whether H₂O₂ accumulating in the culture medium might be required to suppress BA activation at high temperature, bovine liver catalase (EC 1.11.1.6; 0.03 mg mL⁻¹, Sigma; Lee et al., 1999) was dissolved directly in cell culture medium. Because 200 μm H₂O₂ suppresses activation of the soybean (Glycine max) GH₃ auxin-responsive promoter in mesophyll protoplasts of Arabidopsis (Kovtun et al., 1998, 2000), we cultured tree tobacco GCP at 32°C or 38°C in a medium containing 200 μm H₂O₂ created by adding 1 μL of an 80-mm stock solution of H₂O₂ to 400-μL cultures.

The endogenous capacity of GCP to metabolize and/or scavenge H₂O₂ was evaluated using cell extracts prepared by freezing 2 × 10⁵ GCP at −80°C for 5 min in 1 mL of cell culture medium, thawing, centrifuging for 2 min at top speed in a microcentrifuge to clear insoluble debris, and transferring supernatants to a fresh tube. Extracts were stored on ice while their effects on H₂O₂ metabolism and/or scavenging were determined.

In preparation for H₂O₂ assays, extracts were diluted with cell culture medium, pH 6.1, over the range of 0.04% to 20% of their full concentration, and an H₂O₂ solution was prepared by diluting a 1 × 10⁴ ng mL⁻¹ stock 1:9 with culture medium, pH 6.1, lacking hormones (final concentration 1 × 10³ ng mL⁻¹).

To evaluate the capacity of extracts to metabolize and/or scavenge H₂O₂, 50 μL of medium (control) or each diluted or undiluted extract was added to 200 μL of diluted H₂O₂ (above) so that the final H₂O₂ concentration was 800 ng mL⁻¹ (23.5 μm). After exactly 5 min at room temperature, 50 μL of the mixture was transferred to a luminometer tube and H₂O₂ remaining was measured with a chemiluminescent assay (catalog no. 907–102; Assay Designs). The luminometer (Zylux Femtomaster FB15, Zylux) was programmed to inject 20 μL of chemiluminescent substrate and then 20 μL of trigger after a 2-s delay. After injection of the trigger, samples were counted for 10 s. To determine whether loss of H₂O₂ was due to enzymatic or nonenzymatic processes, assays were repeated with undiluted extracts in media containing the peroxidase inhibitors sodium azide (final concentration 10 mm), KCN (10 mm), or 3-amino-1,2,4-triazole (10 mm), or the NADPH oxidase inhibitor, diphenylene iodonium chloride (final concentration 0.2 mm from a 32-mm dimethyl sulfoxide stock). Assays containing inhibitors but with cell culture medium in place of extracts were used as controls. Experiments were repeated three times on separate days.