Desensitization of GSTF8 Induction by a Prior Chemical Treatment Is Long Lasting and Operates in a Tissue-Dependent Manner^,

Abstract

The Arabidopsis (Arabidopsis thaliana) GSTF8 gene is a member of the glutathione S-transferase (GST) family whose expression is induced by defense signals, certain chemical stresses, and some pathogens. Here, we have used transgenic plants and an in vivo imaging system to demonstrate that GSTF8 expression is subject to a distinct desensitization phenomenon because prior chemical treatment significantly reduces reactivation of the GSTF8 promoter by hydrogen peroxide, auxin, and salicylic acid. A GSTF8 null line had similar desensitization properties to wild type, demonstrating that GSTF8 protein levels are not responsible for desensitization. The resulting refractory period is unusually long lasting, with full recovery taking 4 d. Expression of the GSTF8 promoter following a second treatment occurred predominantly in newly formed tissue at the root tip, suggesting that desensitization is lost upon cell division. Expression of the endogenous GSTF8 gene and another GST gene, GSTF6, is also desensitized following treatment with hydrogen peroxide. The desensitization phenomenon can be activated by a very low concentration of inducer that is not sufficient to activate the GSTF8 promoter. These results demonstrate that activation of the GSTF8 promoter is not essential for eliciting desensitization. A key promoter sequence within the GSTF8 gene, the ocs element, is also affected by desensitization. Treatment with a phosphatase inhibitor prevents desensitization of GSTF8 expression and ocs element activity, suggesting that dephosphorylation of one or more proteins is required for desensitization to occur.

Glutathione S-transferases (GSTs) are a diverse group of multifunctional proteins that catalyze the conjugation of glutathione to a range of electrophilic substrates. In plants, GSTs appear to play critical roles in the protection of tissues against oxidative damage as well as other roles, for example, in flavanoid synthesis (Marrs, 1996; Edwards et al., 2000; Frova, 2003). The regulation of expression of the Arabidopsis (Arabidopsis thaliana) GSTF8 gene, previously called GST6, has been well studied. GSTF8 promoter activity, RNA expression, and protein levels are under tissue-specific control and induced following treatment with auxin, salicylic acid (SA), hydrogen peroxide (H₂O₂), and by specific strains of the necrotropic fungus, Rhizoctonia solani (Chen and Singh, 1999; Perl-Treves et al., 2004; Sappl et al., 2004; Uquillas et al., 2004). The ocs element, a well-characterized promoter element (Bouchez et al., 1989; Ulmasov et al., 1994) within the GSTF8 promoter is responsible for part, but not all, of the induction by SA and H₂O₂ (Chen and Singh, 1999). In the case of SA, GSTF8 is considered to be an SA early-responsive gene whose expression is independent of NPR1, a key regulator of some SA-mediated responses, including pathogenesis-related gene expression (Uquillas et al., 2004).

Whereas considerable research has been directed at understanding how gene expression is regulated in plants in response to various stimuli, less work has been undertaken to understand what happens after gene induction has peaked. A number of signal transduction pathways in animals are regulated by desensitization, a process by which components in the pathway, in many cases the receptor, become refractory to prolonged or repeated exposure to a signal (for review, see Oppermann, 2004; Gainetdinov et al., 2004). Desensitization can be the result of repeated treatments with the same signal (homologous desensitization) or by different signals (heterologous or cross desensitization).

Desensitization has also been observed in plants, for example, with cell cultures in response to light, chitin, and other elicitors (Bowler et al., 1994; Felix et al., 1998; Chandra et al., 2000). However, there have been technical limitations to these studies, particularly in whole plants and when studying the effects of desensitization on gene expression, because it has been hard to study both gene induction and postgene induction phases in the same plant. The use of in vivo imaging systems that permit noninvasive approaches can help circumvent these problems and allow detailed temporal and spatial studies to be performed on individual plants during all phases of gene induction. In this article, we have used such an approach with transgenic Arabidopsis plants containing different GSTF8 promoter constructs fused to the luciferase (LUC) reporter gene. Using this system, we have uncovered additional complexity in the regulation of GSTF8 expression. We demonstrate that the GSTF8 promoter is subject to a distinct desensitization phenomenon with novel properties, which significantly reduces reactivation of GSTF8 expression following treatment with an initial stimulus. Furthermore, we present evidence that activation and desensitization of the GSTF8 promoter are not directly linked, that the GSTF8 protein is not involved, and that dephosphorylation of one or more proteins is required for desensitization.

RESULTS

GSTF8 Expression Is Desensitized following H₂O₂ Treatment and Is Regulated Spatially in Roots

We were interested in analyzing whether there was a difference in the kinetics of the response of the GSTF8 promoter to SA and H₂O₂, which may help elucidate whether these plant defense signals were working through similar or distinct mechanisms to induce GSTF8 expression. For these experiments, we used transgenic Arabidopsis plants containing a 792-bp GSTF8 promoter linked to the LUC reporter gene (GSTF8∷LUC) and a noninvasive imaging system, as previously reported (Perl-Treves et al., 2004), to perform detailed time course experiments. As shown in Figure 1A, we observed differences both in the magnitude and the induction kinetics of the GSTF8 promoter in response to 1 mm SA versus 1 mm H₂O₂.

Figure 1.

Response of the GSTF8 promoter to SA and H₂O₂ treatments. A, Four-day-old GSTF8∷LUC seedlings were treated with 1 mm H₂O₂ or 1 mm SA for 40 min and the bioluminescence over a 19-h time course was measured using a CCD camera. The average light units per seedling for each treatment for each hour, with ses, are presented. B, Four-day-old GSTF8∷LUC seedlings were treated at time 0 and then again at 13 h with 1 mm H₂O₂ for 40 min and then decanted. The average relative light units per GSTF8∷LUC seedling with ses is presented for each hour over a 19-h time course. C, In vivo imaging of a representative, 4-d-old GSTF8∷LUC seedling treated with 1 mm H₂O₂ at time 0 and 13 h and visualized using a CCD camera. The white signal comes from the cotyledons and represents the transient chlorophyll fluorescence, which was imaged prior to measuring the bioluminescence. Bioluminescence was monitored after 10 min in the dark and is imaged in blue. Bioluminescence images were captured at 0, 3, 6, and 13 h after the first treatment and 3 and 6 h after the second treatment. The fluorescence and bioluminescence images were superimposed in the figure. The left- and right-most sections are black and white photographs of the same seedling taken before the first and second H₂O₂ treatment, respectively.

Because the response of the GSTF8 promoter to H₂O₂ peaked earlier than the SA response, we were interested in analyzing whether this was because H₂O₂ was more labile than SA. If this was the case, one may expect that the induction would be restored by a second treatment of H₂O₂. However, as shown in Figure 1B, we observed that a second H₂O₂ treatment, administered after the expression following the first treatment, had returned close to basal levels, resulting in a much smaller induction of the GSTF8 promoter compared to the first treatment. The graph in Figure 1B shows the relative light units averaged from 22 4-d-old seedlings after treatment with 1 mm H₂O₂ at 0 h and a second treatment at 13 h. The second induction peaks at 4 h after the second treatment and is only about 30% as strong as the induction seen with the first treatment. These results bear analogy to the refractory phenomenon described for some signaling pathways in plants and animals, where transient desensitization of the pathway occurs following an initial stimulus (e.g. Bowler et al., 1994; Felix et al., 1998; Oppermann, 2004). We cannot rule out a reduction in H₂O₂ concentration after the first treatment. If so, this may contribute to GSTF8 expression returning to basal levels by 13 h after the first chemical treatment. However, what these experiments demonstrate is that the expression of the GSTF8 promoter and the endogenous gene are subject to desensitization following treatment with H₂O₂ because restimulation after the expression levels from an initial induction had returned to a basal level, resulted in a much smaller induction than the initial treatment.

We also found that desensitization of the GSTF8 promoter caused by H₂O₂ also reduced the response to other stimuli such as 1 mm SA and 50 μm of the synthetic auxin, 2,4-dichlorophenoxyacetic acid (2,4-D; data not shown). To rule out the possibility that LUC activity itself was being affected by treatment with any of these stimuli, we used plants expressing the cauliflower mosaic virus 35S promoter fused to LUC and found that LUC activity was not affected by any of these stimuli (see Supplemental Fig. 1).

We used the in vivo imaging system to examine the temporal expression patterns of the GSTF8 promoter in young seedlings in response to consecutive H₂O₂ treatments and these results are shown in Figure 1C. For both the first and second H₂O₂ treatments, expression is predominantly in the root. Interestingly, the second H₂O₂ treatment only resulted in GSTF8 expression in newly formed root tissue at the root tip or in regions of the root where GSTF8 expression was no longer visible by the end of the first H₂O₂ treatment, as observed for the middle of the root at the 3- and 6-h time points after the second treatment.

To determine whether the desensitization phenomenon observed with the GSTF8 promoter also occurred for the native GSTF8 gene, we looked at GSTF8 RNA levels in 8-d-old seedlings following consecutive H₂O₂ treatments. Figure 2 shows the results of reverse transcription (RT)-PCR performed with specific primers for GSTF8 and a second Arabidopsis GST gene called GSTF6, originally called GST1, also linked to defense responses (Grant et al., 2000). β-Tubulin RNA levels were used as a control to normalize the RT-PCR results. RNA levels for both GSTF6 and GSTF8 increased after treatment with 1 mm H₂O₂ and were at their highest levels at the 6-h time point. After 13 h, they were back to basal levels. A second H₂O₂ treatment at 13 h resulted in only a small induction of GSTF6 and GSTF8 expression, demonstrating that the regulation of these GST genes by H₂O₂ also involves desensitization.

Figure 2.

Expression of the endogenous GSTF8 gene is also regulated by desensitization. Eight-day-old GSTF8∷LUC seedlings were treated with 1 mm H₂O₂ and RNA was extracted at 0, 3, and 6 h after the first H₂O₂ treatment and 0, 3, and 6 h after the second H₂O₂ treatment. RT-PCR was performed using GSTF6, GSTF8, and β-tubulin-specific primers and the results for GSTF6 and GSTF8 were standardized to the β-tubulin control. The average and se of two biological repeats are graphed.

Feedback Regulation Is Not Involved in Desensitization of GSTF8 Expression

To explore whether GSTF8 protein levels played a role in desensitization of the GSTF8 promoter, we used a T-DNA insertion SALK line (039887), developed by Alonso et al. (2003), which contains a T-DNA insertion in the coding region of GSTF8 (At2g47730). A line homozygous for the T-DNA insertion, as determined by PCR, was selected. This line (called ΔGSTF8) had no detectable expression of either AtGSTF8 mRNA (Fig. 3A) or AtGSTF8 protein (Fig. 3B). ΔGSTF8 was crossed into the GSTF8∷LUC background and F₃ plants homozygous for ΔGSTF8 and GSTF8∷LUC were examined for desensitization of GSTF8 expression. For these experiments, 4-d-old GSTF8∷LUC seedlings and homozygous ΔGSTF8 × GSTF8∷LUC lines were treated for 18 h with water or 1 mm H₂O₂. The seedlings were then treated with 1 mm H₂O₂ and bioluminescence was measured for each seedling 6 h after the second treatment (approximately 10–20 seedlings/treatment). As shown in Figure 3C, there was no significant difference between the desensitization of GSTF8 promoter activity in wild-type plants versus the ΔGSTF8 plants, suggesting that feedback regulation through changes in GSTF8 protein levels are not involved in regulating desensitization.

Figure 3.

GSTF8 protein does not regulate desensitization. A, AtGSTF8 mRNA is undetectable in ΔGSTF8 plants. Eight-day-old seedlings were mock treated with water or 10 mm H₂O₂ for 30 min. Whole seedlings were harvested 7 h after commencing the treatments and gene expression was measured using real-time PCR. Relative AtGSTF8 mRNA expression levels were normalized to cyclophilin. The average of three biological replicates is shown along with ses. B, AtGSTF8 protein is undetectable in ΔGSTF8 plants. Ten-day-old seedlings were mock treated with water or 1 mm SA for 30 min. Whole soluble cell extracts were prepared from whole seedlings harvested 24 h after commencing the treatments. GSTs were purified by glutathione affinity chromatography and resolved by isoelectric focusing/SDS-PAGE over a pI range of 3 to 10. The resulting two-dimensional gels were silver stained and GSTs were identified based on their relative mobilities according to our previous mass spectrometry analysis of Arabidopsis GSTs (Sappl et al., 2004). C, Plates containing 4-d-old wild-type (columns 1 and 2) and ΔGSTF8 (columns 3 and 4) seedlings in the GSTF8∷LUC background were pretreated with water (columns 1 and 3) or 1 mm H₂O₂ (columns 2 and 4). The seedlings were treated again with 1 mm H₂O₂ 18 h later. For each treatment, 3 mL were added to a 55-mm Murashige and Skoog plate containing 50 μm luciferin and decanted after 40 min. Bioluminescence was measured for each seedling (approximately 20 seedlings/treatment) 6 h after the second treatment using a CCD camera and the average light units per seedling with ses are presented.

Desensitization of GSTF8 Expression Lasts for about 4 d

In other systems where desensitization was reported, the refractory period is transient and often lasts for minutes and rarely more than a few hours (Gainetdinov et al., 2004; Oppermann, 2004). We examined the length of the desensitization period for the GSTF8 promoter following an initial treatment with H₂O₂. Ten plates containing approximately 28 seedlings (4-d-old) were either treated with water or 1 mm H₂O₂ at day 0, and bioluminescence was measured at 0 and 4 h. At days 1, 2, 3, and 4, two plates (one previously treated with water and one previously treated with H₂O₂) were given a second 1 mm H₂O₂ treatment and bioluminescence was measured 4 h later. As shown in Figure 4, the second H₂O₂ treatment resulted in only a small induction at days 1 and 2 compared to the first H₂O₂ treatment. By day 3, the response of the GSTF8 promoter to a second H₂O₂ treatment had been restored to 70% of the original induction and, by day 4, the response had been fully restored.

Figure 4.

Desensitization of GSTF8 expression is regulated temporally. Four-day-old GSTF8∷LUC seedlings were pretreated with either water (white bars) or 1 mm H₂O₂ (black bars) on day 0. Bioluminescence was measured 4 h after a second 1 mm H₂O₂ treatment delivered either on days 1, 2, 3, or 4 after the first treatment. Separate seedlings were used for each treatment and discarded after each day's measurements. The average bioluminescence values of seedlings are presented with ses.

Desensitization of GSTF8 Expression Can Occur without Detectable Preinduction

We were interested in examining whether lower concentrations of an inducer could also cause desensitization of GSTF8 expression. In the case of H₂O₂, we found that desensitization only occurred at H₂O₂ concentrations that were able to induce the GSTF8 promoter (see Supplemental Fig. 2). The synthetic auxin 2,4-D has been shown to induce the GSTF8 promoter at concentrations of 100 μm (Chen and Singh, 1999). However, as shown in Figure 5, lower concentrations that failed to induce the GSTF8 promoter (Fig. 5A) could still activate the desensitization response (Fig. 5B). For these experiments, 4-d-old GSTF8∷LUC seedlings were treated for 4 h with water or 1 μm auxin, an auxin concentration that does not induce GSTF8 promoter activity. The seedlings were then treated with 1 mm H₂O₂ the following day and bioluminescence was measured for each seedling (approximately 20 seedlings/treatment). As shown in Figure 5B, low concentrations of auxin (1 μm) also caused desensitization. Therefore, regulation of GSTF8 expression appears to be complex and activation and desensitization of GSTF8 promoter activity are not directly linked.

Figure 5.

Low concentrations of auxin can activate the desensitization phenomenon. A, Four-day-old plants were treated with 1 or 300 μm auxin and plant extracts were assayed biochemically for LUC activity and graphed as LUC activity per milligrams protein with sds. Samples were tested with technical and biological repeats. The averages of the biological means and sd are graphed. B, Four-day-old GSTF8∷LUC seedlings were pretreated with water or 1 μm auxin and 13 h later all seedlings were treated with 1 mm H₂O₂. Bioluminescence was measured for each seedling (approximately 20 seedlings/treatment) using a CCD camera and the average light units per seedling with ses are presented.

Activity of the ocs Element Is Also Regulated through Desensitization

The GSTF8 promoter contains an ocs element (TTATGTCATTGATGACGACC), also referred to as an as-1 site, that is responsive to SA, H₂O₂, auxin, and R. solani (Lam et al., 1989; Xiang et al., 1996; Chen and Singh, 1999; Perl-Treves et al., 2004). It was of interest to see whether the ocs element is involved in the desensitization phenomenon. For these experiments, we used an Arabidopsis transgenic line in which a tetramer of the GSTF8 ocs element was fused to a LUC reporter gene and transformed into Arabidopsis (Perl-Treves et al., 2004). As shown in Figure 5, pretreatment with 1 mm H₂O₂ (Fig. 6A) or 1 μm auxin (Fig. 6B) resulted in a substantial reduction in the response to a subsequent treatment with H₂O₂, compared to plants that had been pretreated with water. Therefore, activity of the ocs element within the GSTF8 promoter, which plays a major role in defense signal induction, is also regulated by desensitization.

Figure 6.

Ocs element activity is also desensitized following treatment with 1 mm H₂O₂ or 1 μm auxin. Four-day-old 4 × ocs∷LUC seedlings were pretreated with water (white columns), 1 mm H₂O₂ (black columns in A), or 1 μm auxin (black columns in B) and 14 h later the seedlings were treated with 1 mm H₂O₂. Bioluminescence was measured for each seedling (approximately 20 seedlings/treatment) using a CCD camera. The relative bioluminescence for each seedling was recorded at 6 and 12 h after the second treatment. The average light units per seedling with ses are presented.

Treatment with a Phosphatase Inhibitor Prevents Desensitization of GSTF8 and ocs Element Expression

In a number of mammalian systems, phosphorylation of a receptor is one of the early steps in desensitization (e.g. Gainetdinov et al., 2004; Oppermann, 2004). To begin to analyze the mechanisms underlying the desensitization of GSTF8 expression, we asked whether phosphorylation/dephosphorylation events played any role in this system. For the experiments shown in Figure 7A, we used 1 μm auxin, which is unable to induce the GSTF8 promoter, but is able to desensitize the response of the promoter to a subsequent treatment with H₂O₂. Treatment with 1 μm okadaic acid, an inhibitor of type 1 and type 2A protein phosphatases, is not able to induce the GSTF8 promoter (see Supplemental Fig. 3) or lead to desensitization of the GSTF8 promoter because a subsequent treatment with H₂O₂ gives a normal level of induction (Fig. 7A, compare columns 1 and 3). However, when okadaic acid is administered together with 1 μm auxin in the first treatment, desensitization of the GSTF8 promoter is prevented (Fig. 7A, compare columns 2 and 4). Okadaic acid was also able to prevent desensitization of the ocs element (Fig. 7B, compare columns 2 and 4). These results suggest that dephosphorylation of one of more proteins is required for desensitization of GSTF8 and ocs element expression.

Figure 7.

Okadaic acid releases desensitization of the GSTF8 promoter and ocs element activity. A, Plates containing 4-d-old GSTF8∷LUC seedlings were pretreated with water (column 1), 1 μm auxin (column 2), 1 μm okadaic acid (column 3), or both 1 μm auxin and 1 μm okadaic acid (column 4). The seedlings were treated with 1 mm H₂O₂ 17 h later. For each treatment, 1 mL was added to a 55-mm Murashige and Skoog plate containing 50 μm luciferin and decanted after 40 min. Bioluminescence was measured for each seedling (approximately 20 seedlings/treatment) 6 h after the second treatment and the average light units per seedling with ses are presented. B, Plates containing 4-d-old 4 × ocs∷LUC seedlings were pretreated as in Figure 7A. The seedlings were treated with 1 mm H₂O₂ 24 h later. Bioluminescence was measured for each seedling (approximately 20 seedlings/treatment) 12 h after the second treatment and the average light units per seedling with ses are presented.

DISCUSSION

The Arabidopsis GSTF8 promoter is induced by a range of stimuli, including defense signals, auxin, and some pathogens. In this article, we report the use of an in vivo imaging system coupled with promoter/reporter constructs to study desensitization. We demonstrate that activity of the GSTF8 promoter and expression of the endogenous GSTF8 gene are regulated by desensitization, as reactivation is significantly reduced following pretreatment with a stimulus. A T-DNA knockout line for GSTF8 was used to demonstrate that failure to reactivate is not due to feedback regulation via GSTF8 protein levels.

Desensitization of GSTF8 expression occurred following repeated treatment with the same or different stimuli and is therefore an example of both homologous and heterologous desensitization. Desensitization assays have been used as a research tool by others to determine whether different plant signaling pathways share components, for example, UVB, oligosaccharide elicitors, and systemin (Yalamanchili and Stratmann, 2002; Holley et al., 2003). Our results suggest that different inducers of the GSTF8 promoter share one or more components in their respective signal transduction pathways and are consistent with the fact that many inducers of plant GSTs cause oxidative stress (Marrs, 1996). These results are also consistent with our previous findings that a specific promoter sequence, the ocs element, mediates in part the response of the GSTF8 promoter to different inducers (Chen and Singh, 1999; Perl-Treves et al., 2004) and the observations reported here that the ocs element is also regulated by desensitization.

There are a number of interesting features associated with the desensitization of GSTF8 expression. One is the length of the desensitization period, which lasts several days, in contrast to the desensitization periods observed in many other signaling pathways. For example, in some light responses, desensitization ranges from seconds to minutes (e.g. Govorunova et al., 1997; Mathieu et al., 1998; Baum et al., 1999), whereas desensitization of the mitogen-activated protein kinase response in tomato (Lycopersicon esculentum) to elicitors lasted less than 120 min (Holley et al., 2003). In soybean (Glycine max), cultured cells pretreated with a chemical elicitor were able to be quickly restimulated by the same or different elicitors, with the refractory period having a half-life of 20 min (Chandra et al., 2000).

Another interesting feature associated with desensitization of GSTF8 expression involved the site of reinduction; whereas a second induction of the GSTF8 promoter was about 30% as strong as the first induction, spatial analysis revealed that much of the initial response to a second treatment occurred in newly formed tissue at the root tip. This demonstrates that the desensitization of GSTF8 expression is more pronounced in older parts of the plant and suggests that desensitization may be lost upon cell division. Perhaps upon cell division the concentration of a key component involved in desensitization falls below a critical level in the daughter cells and consequently desensitization no longer occurs. We also observed that specific regions of the root responded differently to repeated treatments. The response of the middle region of the root to an initial stimulus was short lived compared to the rest of the root. Consequently, the middle region was able to respond much more rapidly to a second treatment (Fig. 1C). The reason for such an altered desensitization response for GSTF8 expression in this part of the root compared to other parts is unknown, but highly specific patterns of GST expression in different parts of the root have been reported previously (Smith et al., 2003).

It was also interesting to observe that very low concentrations of auxin (1 μm), which were not able to detectably induce GSTF8 expression, were able to desensitize the GSTF8 promoter. These results bear some similarity to the response of tomato cells with chitin treatment, which results in a rapid and transient alkalinization of the medium and involves a desensitization phenomenon (Felix et al., 1998). The concentration of chitin oligomers required for desensitization was about 20 times lower than that required for alkalinization. In contrast to the results obtained with low concentrations of auxin, we were not able to identify a concentration of H₂O₂ that could cause desensitization of GSTF8 expression while also not acting as an inducer (see Supplemental Fig. 2), and similar results were observed with SA (data not shown). At this stage, the reasons for these differences between auxin and H₂O₂/SA are unclear, but they raise the interesting possibility that desensitization of GSTF8 expression can occur by more than one mechanism. From the auxin results, it is clear that detectable activation of the GSTF8 promoter is not essential for eliciting desensitization.

Whereas desensitization of signaling pathways has been observed in both animals and plants, much of our knowledge of the underlying mechanisms comes from animal studies. A well-studied example of desensitization in animals is called agonist-induced desensitization and involves phosphorylation of G-protein-coupled receptors by Ser-Thr protein kinases. Subsequent binding of a class of β-arrestins prevents G protein from binding to the receptors and helps initiate their endocytosis (for review, see Gainetdinov et al., 2004; Oppermann, 2004). In some cases, receptors can be subjected to both homologous and heterologous desensitization via receptor phosphorylation (Xin et al., 2004; Avendano-Vazquez et al., 2005) and, in a number of studies, treatment with okadaic acid blocks the resensitization response (e.g. McLaughlin et al., 2003). In a few cases, heterologous desensitization in animal systems involves second messenger-dependent kinases that phosphorylate a range of proteins and lead to suppression of signaling pathways induced by different stimuli (Freedman and Lefkowitz, 1996).

In plants there have been only limited studies on the mechanisms involved in desensitization. In the case of chitin signaling, desensitization was not associated with inactivation of the stimulus or with the disappearance of high-affinity chitin-binding sites from the cell surface (Felix et al., 1998). In oligogalacturonide-activated signaling of soybean cells, there is evidence that phosphorylation events are involved in establishing a refractory state following an initial treatment with oligogalacturonides (Navazio et al., 2002). We examined desensitization of GSTF8 expression in different SA, jasmonic acid, or ethylene signal transduction mutant backgrounds, but found no evidence for a role of the SA, jasmonic acid, or ethylene pathways (data not shown). However, we did find evidence for a role of phosphorylation in the desensitization of GSTF8 expression with okadaic acid treatment, suggesting that dephosphorylation of one of more proteins is required for desensitization of the ocs element and the GSTF8 promoter. Whereas these results are in contrast with those reported for oligogalacturonide-activated desensitization, where okadaic acid acted to prolong desensitization (Navazio et al., 2002), they are consistent with results from the wound-induced SAMK pathway, where a Ser-Thr protein phosphatase 2C is one of the factors responsible for resetting the SAMK pathway in preparation for subsequent signaling (Meskiene et al., 1998). One possibility is that okadaic acid, which does not induce GSTF8 expression, may be acting to inhibit the dephosphorylation of an activator and/or repressor that is required for desensitization. For example, the CK2 enzyme phosphorylates and affects the DNA binding of several transcription factors (Klimczak et al., 1992; Tjaden and Coruzzi, 1994; Ciceri et al., 1997; Sugano et al., 1999), including TGACG-sequence-specific binding protein (TGA) family members that bind to the ocs element (Kang and Klessig, 2005).

Desensitization of signaling pathways has been suggested to allow an increase in the dynamic range of the sensory system (Armitage, 1992). Desensitization may also play a protective role in containing the extent of a response and preventing potentially deleterious effects of overstimulation such as can occur via the overproduction of toxic compounds during the oxidative burst associated with the plant defense response (Chandra et al., 2000). Because a number of the stimuli that induce GSTF8 expression are linked to defense, namely, the defense signals H₂O₂ and SA, a similar explanation may hold for the desensitization reported here.

MATERIALS AND METHODS

Plant Material

Arabidopsis (Arabidopsis thaliana) transgenic lines contained the −783 GSTF8 promoter fused to the LUC reporter gene (GSTF8∷LUC) in the Columbia ecotype background as previously described (Chen and Singh, 1999). All data presented are from a single T₄ line homozygous for the transgene. The 4 × ocs element promoter construct was made by ligating four copies of the GSTF8 ocs element in front of the −58 GSTF8 minimal promoter (Perl-Treves et al., 2004) and a homozygous T₄ line was used. For both the GSTF8 and ocs element promoters, all experiments were repeated with a second independent line and similar results were obtained. The T-DNA insertion line, SALK_039887, was obtained from the Salk Institute Genomic Analysis laboratory and contained a T-DNA insertion in the third exon of GSTF8 (A2g47730). This T-DNA insertion was confirmed by using PCR on genomic DNA with the primers 5′-GGGCGAGAGTCAAAGAGCACC-3′ and 5′-TGCAATAAATAAAGGGGACCCAA-3′. The homozygous seed line, SALK_039887C, will be available through the Arabidopsis Biological Resource Center at Ohio State University.

Growth of Arabidopsis on Agar Plates

Agar plates contained 1× Murashige and Skoog salts (4.3 g/L; Gibco) and 0.8% agar, with 3% Suc, pH adjusted to 5.7, with 1 m KOH. Plates for the LUC assay were supplemented with 50 μm luciferin (Biosynth AG) added after autoclaving the medium. Arabidopsis seeds were surface sterilized for 15 min with 70% ethanol followed by 2 d at 4°C. Seeds were germinated on Murashige and Skoog plates containing 50 μm luciferin and grown vertically for 4 d. Approximately 30 seeds were plated on round 55-mm plates, sealed with Micropore tape, and incubated vertically in the growth room (22°C, 16-h light/8-h dark photoperiod).

Plant Treatment

All auxin treatments used the synthetic auxin, 2,4-D (Sigma). Okadaic acid was from Sigma. Unless specified otherwise, 3 mL of treatment were pipetted onto the agar plates containing 4-d-old seedlings grown vertically. The solution was drained off after 40 min. The plates were left in the dark chamber in the EG & G Berthold molecular light imager during the time course experiments.

In Vivo Bioluminescence Assay

Luminescence was measured in an EG & G Berthold molecular light imager using a 5-min exposure after a 10-min fluorescent decay delay. Plants were superimposed onto a fluorescent exposure for orientation as previously described (Perl-Treves et al., 2004). The luminescence from each plant was measured against background and the average reading and se were graphed. Results are presented with the bioluminescence (blue) image superimposed on the fluorescence (white) image or graphically as average light units per seedling as determined by the EG & G Berthold software, WinLight 32. Figure 1, A and B, shows different light units per seedling as the camera settings were modified to capture the different number of plates in each experiment.

Biochemical LUC Assay

Eight-day-old seedlings were treated with 1 μm auxin or 300 μm auxin. After 0, 3, and 6 h, seedlings were homogenized in 0.2 m KPO₄, pH 7.2, 1 mm dithiothreitol, and 0.5 mm phenylmethanesulfonyl fluoride (Sigma). Samples were then tested for LUC activity on a FLUOstar optima microplate reader (BMG Labtech) using the manufacturer's protocol. LUC activity was standardized against protein concentrations in the extracts as determine by the Bio-Rad DC protein assay.

RT-PCR

Eight-day-old seedlings were harvested from Murashige and Skoog agar plates. RNA isolation, cDNA synthesis, and quantitative RT-PCR were performed as previously described (Kang et al., 2003; Lister et al., 2004), using the iCycler program described in Sappl et al. (2004). The gene-specific primer pairs used for real-time PCR were AtGSTF6 (At1g02930): 5′-CCCCGTCGATATGAGAGC-3′ and 5′-GAGAGAGGGTCACTACTGCTTCTGG-3′; ATGSTF8 (A2g47730): 5′-CACAGGCTTGGTGAGTCCAAG-3′ and 5′-CAAATCAAACACTCGGCAGCAG-3′; β-tubulin (At5g23860): 5′-CACAGCAATACAGAGCCTTAACC-3′ and 5′-CTGTTGTTATTGCTCCTCCTGCA-3′; and Cyclophilin (At2g29960): 5′-TCTTCCTCTTCGGAGCCATA-3′ and 5′-AAGCTGGGAATGATTCGATG-3′.

Two-Dimensional Gel Electrophoresis

Whole soluble cell extracts were prepared from 10-d-old seedlings (grown on Murashige and Skoog plants) treated with 1 mm SA for 24 h. GSTs were purified by glutathione affinity chromatography and resolved by isoelectric focusing/SDS-PAGE over a pI range of 3 to 10 as previously described (Sappl et al., 2004). The resolved proteins were visualized by silver staining according to standard protocols. AtGSTU19 (At1g78380) and AtGSTU20 (At1g78370) were identified by mass spectrometry as described previously (Sappl et al., 2004). The identities of the other GST proteins were inferred from their relative mobilities as described (Sappl et al., 2004).

Sequence data from this article can be found in the GenBank/EMBL data libraries under accession numbers NM180148, NM100174, NM106485, NM106484, AY081473, and NM128550.