A glutathione S-transferase regulated by light and hormones participates in the modulation of Arabidopsis seedling development

Glutathione S-transferases have been well documented to be involved in diverse aspects of biotic and abiotic stresses, especially detoxification processes. Whether they regulate plant development remains unclear. Here, we report on our isolation by RT-PCR of a plant glutathione S- transferase, AtGSTU17 , from Arabidopsis thaliana and demonstrate that its expression is regulated by multiple photoreceptors, especially phytochrome A (phyA) under all light conditions. Further physiological studies indicated that AtGSTU17 participates in various aspects of seedling development, including hypocotyl elongation, anthocyanin accumulation, and far-red (FR) light-mediated inhibition of greening with a requirement of functional phyA. The loss-of-function mutant of AtGSTU17 ( atgstu17 ) resulted in reduced biomass of seedlings and number of lateral roots in the presence of auxin, as well as insensitivity to ABA-mediated inhibition of root elongation, with similarity to different phyA mutant alleles. Moreover, the root phenotype conferred by atgstu17 was reflected by histochemical GUS staining of AtGSTU17 promoter activity with the addition of auxin or abscisic acid. Further microarray analysis of wild-type Columbia and atgstu17 seedlings treated with FR irradiation or ABA revealed that AtGSTU17 might modulate hypocotyl elongation by positively regulating some light signaling components and negatively regulating a group of auxin-responsive genes and modulate root development by negatively controlling an auxin transport protein in the presence of abscisic acid. Therefore, our data reveal that AtGSTU17 participates in light signaling and might modulate various aspects of Arabidopsis development by affecting glutathione pools via a coordinated regulation with phyA and phytohormones.

The large numbers of phi and tau classes are plant specific and have major roles in herbicide detoxification (Dixon et al., 2002b;Edwards et al., 2000;DeRidder and Goldsbrough, 2006). Recent evidence indicates that GSTs are also involved in endogenous metabolism, including oxidative stress, flavonoid binding, and regulation of apoptosis (Alfenito et al., 1998;Kilili et al., 2004;Marrs et al., 1995;Mueller et al., 2000). Classes Z and T are plant homologs of mammals and other organisms and function in primary metabolism such as isomerization of maleylacetoacetate and detoxification of hydroperoxides formed during oxidative stress (Dixon et al., 2000;Kampranis et al., 2000).
Classes L and DHARs are newly found in plants and function in redox homeostasis (Dixon et al., 2002a;Edwards et al., 2005). Thus, the functions of plant GSTs are diverse and might be due to the ability to conjugate glutathione (GSH) to various targets involved in biotic and abiotic stress.
In addition to having functions in various stresses, plant GSTs appear to be involved in plant growth and development (Gong et al., 2005;Moons, 6 and abscisic acid (ABA) (Moons, 2003;Smith et al., 2003;Wagner et al., 2002). That all these hormones regulate many aspects of plant development implies that plant GSTs may play vital roles in plant growth and development as well. However, evidence to substantiate this role has been limited. Several plant GSTs have been shown to be induced by different qualities of light (Chen et al., 2007;Loyall et al., 2000;Tepperman et al., 2001).
PcGST1 isolated from parsley cell cultures by fluorescent differential display was induced by UV-B light and involved in a UV light-dependent signaling pathway to chalcone synthase (Loyall et al., 2000). Tepperman et al. (2001) used a DNA microarray approach to examine the gene expression profiles induced rapidly by far-red (FR) light irradiation. The expression of one GST, AAD32887, increased rapidly but was inhibited by phyA mutation (Tepperman et al., 2001). Our previous studies have shown that AtGSTU20 (At1g78370/FIN219-interacting protein 1 [FIP1]) can interact in vitro and in vivo with FIN219 (Chen et al., 2007), which functions in a phytochrome A (phyA)-mediated FR signaling pathway (Hsieh et al., 2000). Moreover, transgenic Arabidopsis seedlings overexpressing or reducing FIP1 expression exhibited a hyposensitive long-hypocotyl phenotype under continuous FR (cFR) light (Chen et al., 2007). These data indicate that some plant GSTs are regulated by light. However, the functional roles of these GSTs involved in light signaling remain to be elucidated.
To further understand the functional mechanisms of plant GSTs in light signaling pathways, we focused on several candidates affected by phyA or FIN219. Here, we report on functional studies of the GST AAD32887/At1g10370/AtGSTU17 previously detected by microarray assay www (Tepperman et al., 2001) and downregulated by fin219 mutation in FR (our unpublished microarray data). Our data presented here using transgenic plants and molecular genetic approaches provide further insight into possible functions of AtGSTU17 involved in light signaling, especially phyA-mediated photomorphogenesis, and in the integration of various phytohormones to modulate glutathione homeostasis in the regulation of Arabidopsis development.

Expression of AtGSTU17 Is Regulated by Multiple Photoreceptors.
To further confirm the expression patterns of FR-regulated AtGSTU17 transcripts, we performed dark-light transition experiments. Wild-type and phyA mutant seedlings were grown in the dark for 2 days, then transferred to FR light for various times; AtGSTU17 expression was examined by RNA gel blot analysis. AtGSTU17 was induced in 2-day-old wild-type seedlings transferred from the dark to FR light for 1 h, and the level peaked with 6-h FR light; the expression was gradually reduced to constant levels for the remaining FR irradiation periods (Fig. 1A, left panel). However, in the phyA mutant seedlings, AtGSTU17 induction by 1-and 6-h FR irradiation was substantially reduced (Fig. 1A, right panel), which indicates that AtGSTU17 is indeed induced rapidly by FR, and its expression depends on PHYA.  1, B-F), which implies that AtGSTU17 expression depends strictly on functional PHYA. Moreover, AtGSTU17 expression was also reduced in cry1, cry2 and cry1cry2 (cry1/2) double mutants in the dark (Fig. 1, B and F). In contrast, the level of AtGSTU17 transcripts in fin219 remained comparable to that in Columbia (Col) and was slightly reduced in the phyB mutant as compared with its ecotype Landsberg erecta (Ler) in the dark (Fig. 1, B and F).
In the transition of dark-grown seedlings to 6-h FR irradiation, the level of AtGSTU17 transcripts was decreased in fin219, cry1, cry2 and cry1/2 mutants ( Fig. 1, C and F) but substantially increased in the phyB mutant under the same conditions (Fig. 1, C and F). In the transition from dark to red light, the level of AtGSTU17 transcripts appeared to be reduced in cry1, cry2 and cry1/2 mutants (Fig. 1, D and F) but remained largely the same in the phyB mutant as in the Ler ecotype, which implies that PHYB under red light may play a lesser role in the regulation of AtGSTU17 expression. As well, the AtGSTU17 transcript level was slightly lower in fin219 than in Col (Fig. 1, D and F). In the transition from dark to blue light, the AtGSTU17 transcript level was decreased in cry1, cry2 and cry1/2 double mutants, with a substantial reduction in cry1 and cry1/2 (Fig. 1, E and F). Intriguingly, the level was markedly higher in the phyB mutant than in the Ler ecotype, in which AtGSTU17 transcripts were barely detected. In addition, AtGSTU17 expression was not affected by fin219 under the blue light transition (Fig. 1 and F). Taken together, these expression data reveal that AtGSTU17 is differentially regulated by various photoreceptors and highly controlled by functional phyA.

AtGSTU17 Regulates Hypocotyl Elongation Mainly in FR Light.
To further understand the functions of AtGSTU17 involved in light signaling, we cloned it by RT-PCR on the basis of sequence information deposited in the NCBI database. The full-length cDNA encodes a 227 amino acid protein and shares 74% identity at the amino acid level with AtGSTU18 (At1g10360) within the same tau class (Fig. S1). The recombinant proteins generated from an E. coli expression system showed enzymatic activities to the substrates GSH and 1-chloro-2,4-dinitrobenzene (CDNB), with Km value 0.285 mM for GSH at 1.0 mM CDNB and 1.400 mM for CDNB at 1.0 mM GSH ( Fig. S1), which indicates that AtGSTU17 has higher affinity to both substrates than do other plant GSTs reported in the literature (23). Thus, AtGSTU17 has GST activities.
To further elucidate the functional regulation between AtGSTU17 and phyA, we used two T-DNA-inserted null mutants, SALK_139615 and SALK_025503, named atgstu17-1 and atgstu17-2 from the Arabidopsis Biological Resource Center (ABRC) (Fig. S2), a new phyA allele in a Col background (Fig. S4), the double mutant gstu17-2phyA, and AtGSTU17overexpressed transgenic lines in Col (GSTU17OE-2) for phenotypic examination under various light conditions. The atgstu17 mutants atgstu17-1 and atgstu17-2 exhibited a hyposensitive long-hypocotyl phenotype as compared with Col under low fluences of FR light ( Fig. 2A  other light conditions . In contrast, GSTU17OE-2 displayed a short-hypocotyl phenotype comparable to that of Col under all light conditions examined ( Fig. 2 and 3A). The double mutant atgstu17phyA showed slightly longer hypocotyls than did phyA in the low FR fluence (< 5 μ mol/m 2 sec) and was much longer than atgstu17 under FR light ( Fig phenotype comparable to that of Col, which implies that AtGSTU17 function requires functional phyA to regulate hypocotyl elongation in response to FR light. Moreover, atgstu17 showed reduced anthocyanin accumulation and defects in FR-mediated blockage of greening regulated by phyA (Fig. 3, B and C). In contrast, GSTU17OE-2 in Col showed opposite effects on anthocyanin levels and chlorophyll content (Fig. 3, B and C). The double mutant atgstu17phyA showed levels of anthocyanin and chlorophyll comparable to that of the respective single mutants atgstu17 and phyA (Fig. 3, B and C).
GSTU17OE-2phyA showed defects in chlorophyll content and anthocyanin accumulation that were substantially different from those of GSTU17OE-2 in Col (Fig. 3, B and C), which suggests that the effects of AtGSTU17 on FR blockage of greening and anthocyanin accumulation require functional phyA.
In addition, atgstu17 showed a delayed flowering phenotype similar to that of phyA (Fig. 3D). The double mutant atgstu17phyA showed a flowering time similar to that of atgstU17. In contrast, GSTU17OE-2 exhibited a flowering time similar to that of wild-type Col, but GSTU17OE-2phyA showed a delayed flowering phenotype under long-day conditions (Fig. 3D). Thus, AtGSTU17 participates in the control of hypocotyl elongation, anthocyanin accumulation, FR blockage of greening, and flowering in a phyA-dependent manner.

Development.
GSTs are induced by stresses from biotic and abiotic factors, including herbicides, chemical toxins, osmotic stress, and plant hormones (DeRidder and Goldsbrough, 2006;Marrs, 1996;Smith et al., 2003). Our studies by RNA gel blot analysis also revealed AtGSTU17 transcripts upregulated by ABA, 2,4-D, and jasmonic acid (JA) treatment (Fig. S3). Intriguingly, the induction of the presence of auxin (Fig. 4A); however, AtGSTU17 overexpression (OE1 and OE2) resulted in an opposite phenotype, with increased biomass and chlorophyll content under the same condition (Fig. 4A), which implies that AtGSTU17 is able to affect seedling development in response to auxin stimulation. Moreover, AtGSTU17 affected the generation of lateral roots in the presence of auxin, with reduced number of lateral roots in atgstu17 mutants and more lateral roots in OE1 and OE2, but did not largely influence the length of primary roots (Fig. 4B).
Because AtGSTU17 was highly induced at 1 h after ABA treatment (

Auxin and FR Light Treatment.
AtGSTU17 proteins showed enzymatic activity to GSH (Fig. S1), which indicates that AtGSTU17 may regulate GSH pools in Arabidopsis under light and various hormone conditions. The reduced (GSH) and oxidized forms (GSSG) of glutathione, as well as its redox state, play critical roles in the control of plant development in response to changes in light and environmental signals (Belmonte et al., 2005;Foyer and Noctor, 2009;Loyall et al., 2000). To associate the GSH/GSSG ratio with the atgstu17 mutant phenotype under cFR light and auxin treatment, we determined the GSH/GSSG ratio in seedlings of atgstu17 (m1 and m2) and GSTU17OE (OE1 and OE2). The GSH/GSSG ratio was higher in the atgstu17 mutant than in the wild-type Col in the absence of auxin; however, the GSH/GSSG ratio was decreased in atgstu17 in the presence of 0.5 μM 2,4-D but was still higher than that in Col under the same condition ( Fig. 5A), which suggests that AtGSTU17 may act as a key component to regulate a glutathione pool for fine-tuning the GSH/GSSG ratio in cells. In contrast, the GSH/GSSG ratio in GSTU17OE (OE1 and OE2) was similar to that in Col in the control treatment and in the presence of auxin. Thus, AtGSTU17 is able to affect GSH/GSSG homeostasis in response to the phytohormone auxin.
In addition, wild-type seedlings grown in cFR versus the dark control showed a significant reduction in GSH/GSSG ratio, which corresponds to a short-hypocotyl phenotype; however, atgstu17 (m1) showed slightly increased GSH/GSSG ratio as compared to Col in cFR. GSTU17OE (OE1 and OE2) showed a substantial reduction of GSH/GSSG ratio under the same condition www.plantphysiol.org on August 31, 2017 -Published by Downloaded from Copyright © 2010 American Society of Plant Biologists. All rights reserved. (Fig. 5B). These data are consistent with a role of AtGSTU17 induced by FR irradiation and leading to a decrease in GSH/GSSG ratio. By contrast, the GSH/GSSG ratio for atgstu17 grown in the dark did not differ from that of Col, but GSTU17OE showed a reduction in GSH/GSSG ratio (Fig. 5B). Thus, AtGSTU17 is highly regulated by the FR light photoreceptor phyA ( Fig. 1) and may play a more prominent role in the regulation of glutathione pools in FR light than in the dark.

Response to ABA.
Further examination of the promoter activities of AtGSTU17 by histochemical GUS staining revealed that AtGSTU17 was mainly expressed in the maturation zone of roots and the basal region of hypocotyls (Fig. 6, A-C).
In the presence of auxin, the staining pattern was extended to the meristematic zone with enhanced stains in the initiation sites of lateral roots . Surprisingly, the expression of phototropismrelated genes such as Non-phototropic Hypocotyl 3 (NPH3; At3g08660) and NPH1 (PHOT1; At3g45780) was affected in atgstu17 under cFR (see Table   S1). As well, more than 20 auxin-responsive genes, including IAA5 (At1g15580), IAA6 (At1g52830) and SAURs (At3g03830; At4g13790; At4g38850) were upregulated in atgstu17 under cFR light, which implies that AtGSTU17 negatively regulates the expression of these auxin-responsive genes under cFR light. In addition, atgstu17 showed altered positive or negative expression of more than 15 transcription factors (Table S1) (Table   S2A). In contrast, SPA1 was upregulated in atgstu17 under ABA treatment (Table S2B). Some auxin-responsive genes, including auxin transport protein PIN7 (At1g23080), were also affected in ABA-treated atgstu17. Intriguingly  (Table S2B). Intriguingly, 13% of the genes were the common genes affected by atgstu17 mutation in both  a loss of auxin-regulated phenotypes such as apical dominance, vasculature and secondary root production, and auxin transport and auxin levels, which implies that the glutathione pool within cells is linked to auxin homeostasis.
Moreover, alterations of the glutathione redox state can improve apical meristem structure and somatic embryo quality in white spruce (Belmonte et al., 2005). AtGSTU17, encoding a tau class member of GST, has GST activity ( Fig. S1). It can be induced by ABA (Fig. 6, G-K and S3), and the loss-offunction mutant exhibits a less sensitive root phenotype to ABA-mediated inhibition of root elongation than does the wild type (Fig. 4, C and D). This phenotype might be due to high levels of GSH accumulating in the roots of the loss-of-function atgstu17 mutant as compared with the wild type under ABA treatment (Fig. S5), which is consistent with high levels of endogenous GSH enhancing cell division in the root meristematic region leading to root elongation (Vernoux et al., 2000). Because promoter activity assay revealed AtGSTU17 expression in roots during seedling development (Fig. 6, A-C), AtGSTU17 expression was even enhanced in the meristematic zone and the initiation sites of lateral roots under auxin and ABA treatments (Fig. 6, D-F and I-K), which is consistent with high levels of GSH in atgstu17 seedlings treated with auxin or ABA (Fig. S5). Intriguingly, microarray data revealed that atgstu17 showed substantially increased levels of PIN7 transcript (Wiśniewska et al., 2006) (Table S2), thus leading to more auxin flow toward the initiation sites of lateral roots under ABA treatment, which may lead to more lateral roots in the atgstu17 mutant in the presence of ABA (Fig. 4, C and D). This speculation is consistent with very recent reports indicating that the glutathione level is linked to PIN proteins such as PIN7, and accordingly www.plantphysiol.org on August 31, 2017 -Published by Downloaded from Copyright © 2010 American Society of Plant Biologists. All rights reserved. auxin transport (Bashandy et al., 2010;Koprivova et al., 2010). Surprisingly, the phyA mutant also showed an insensitive long-root phenotype in response to ABA-mediated inhibition of root elongation. Other phyA alleles, including phyA-211 and SALK_014575 (T-DNA insertion line) and SALK_020360, were also less sensitive to ABA inhibition of root elongation (Fig. 4, C and D).
Moreover, phytochrome mutants were reported to differentially affect lateral root production and showed upregulated expression of PIN3 and PIN7 transcripts (Devlin et al., 2003;Salisbury et al., 2007). Intriguingly, Sung et al. (2007) reported that a genetic screen identified an arsenic tolerant double mutant with a single T-DNA insertion in the PHYA gene, which exhibited elevated levels of the thiols cysteine, γ-glutamylcysteine and glutathione. This finding indicates that PHYA negatively regulates thiol synthesis (Sung et al., 2007), which is consistent with our data showing that AtGSTU17 transcripts are highly downregulated in the phyA mutant (Fig. 1), corresponding to increased levels of glutathione in the atgstu17 mutants. Thus, phyA might regulate root development by integrating with AtGSTU17 to fine-tune glutathione homeostasis in response to ABA.
As well, atgstu17 exhibited a long-hypocotyl phenotype under cFR and showed a slight increase of GSH/GSSG ratio compared to wild-type Col; furthermore, AtGSTU17 overexpression resulted in a significant reduction of GSH/GSSG ratio (Fig. 5B), which implies that AtGSTU17 can contribute to the regulation of the GSH/GSSG ratio; a high GSH/GSSG ratio is essential for cell division (Belmonte et al., 2005). However, AtGSTU17 in the dark may have a redundant effect with other GST members on GSH/GSSG ratio for hypocotyl elongation, which led to atgstu17 having a GSH/GSSG ratio similar to that of www.plantphysiol.org on August 31, 2017 -Published by Downloaded from Copyright © 2010 American Society of Plant Biologists. All rights reserved. Col (Fig. 5 B). In addition to the FR effect on GSH/GSSG ratio, application of exogenous auxin could decrease the GSH/GSSG ratio in the wild-type Col (Fig. 5A), which is consistent with data showing that AtGSTU17 is induced by auxin and has an enzymatic activity to the substrate GSH ( Fig. S3 and S1). In contrast, the atgstu17 mutant showed an increase of GSH/GSSG ratio in the absence or presence of auxin (Fig. 5A), which resulted in a significant decrease of biomass and chlorophyll content (Fig. 4A). This finding is probably due to excess oxidant status caused by higher concentration of auxin, such as 0.5 μM 2,4-D, which results in a redox imbalance towards oxidation. This speculation is supported by our observation of the atgstu17 mutant with more anthocyanin accumulation under the same condition (data not shown).
In addition, the atgstu17 mutant in the presence of ABA showed downregulated levels of 12 MYB genes (Table S2)  method (Clough and Bent, 1998). Transgenic seedlings were selected on 1% GM plates containing 25 µg/ml hygromycine. The resulting homozygous transgenic lines were used for phenotype examination.

RNA Gel Blot Analyses.
Total RNA from 3-day-old seedlings under different light conditions, including darkness, or 3-day-old dark-grown seedlings transferred to various light conditions was isolated as described (Hsieh et al., 1998). RNA samples were separated in a 1.2% denaturing agarose gel, transferred onto a positivecharged nylon membrane (Roche) and UV cross-linked and then hybridized with probes. The riboprobe was synthesized by in vitro transcription according to Dig-labeling procedures (Roche). The gene-specific riboprobe was derived from a BamH1-digested construct of the 3' UTR region of AtGSTU17. All hybridization and washing conditions followed standard methods (Sambrook et al., 1989).

Measurement of Hypocotyl Length, Anthocyanin, Chlorophyll and
Flowering.
The wild type, atgstu17 mutants, phyA, double mutant atgstu17phyA, and AtGSTU17-overexpressing lines GSTU17OE and GSTU17OE-2phyA were grown for 3 days under cFR or other light conditions, then hypocotyl lengths and anthocyanin accumulation were determined as described (Hsieh et al., 2000), or were grown for another 2 days of white light after FR light treatment for measurement of chlorophyll content as described (Hsieh et al., 2000). 25 under long-day (16-h day, 8-h dark) conditions. Flowering time was recorded by use of 2 different indexes: days from seedlings sown in the soil to appearance of the first inflorescence, and leaf number at bolting. More than 10 plants for each condition were recorded.

Assays of Hormone-mediated Inhibition of Root Elongation.
Seeds of the wild type, atgstu17 mutants, GSTU17OE and phyA mutants were germinated on MS medium for 7 days in white light, then transferred to medium plates containing 0.01 μM 2,4-D or 30 μM ABA for vertical growth of roots. The plates were incubated for 5 days under white light. The elongation lengths of new roots of seedlings were measured and recorded. For chlorophyll content and biomass of the wild type, atgstu17 mutants and GSTU17OE under 0.5 μM 2,4-D treatment were also recorded. All these physiological responses to auxin and ABA were examined in at least 2 biological replicates.

GST Activity Assay
GST activity of AtGSTU17 as purified recombinant proteins His-6-GSTU17 was determined as described by Chen et al. (2007).

Analysis of Glutathione.
About 200 mg of seedlings was ground in liquid nitrogen. Subsequently, 2 ml of 1 mM EDTA and 6% (v/v) metaphosphoric acid, pH 2.8, were added and mixed, then centrifuged at 15000 g for 20 min; the supernatant used for www.plantphysiol.org on August 31, 2017 -Published by Downloaded from Copyright © 2010 American Society of Plant Biologists. All rights reserved. determination of total glutathione and oxidized glutathione (GSSG) by a spectrophotometric assay was as described by Griffith (1980).

Histochemical GUS Staining of AtGSTU17 Promoter Activity.
Transgenic seedlings harboring the pAtGSTU17::GUS construct were selected on 1% MS medium containing 25 mg/l hygromycin. The resulting transgenic plants were selected to homozygous lines and underwent histochemical GUS staining as described (Chen et al., 2007).

Microarray Experiment.
Seedlings of the wild type and atgstu17 mutants were grown in the dark for 2 days, then transferred to FR for 6 h; or germinated on MS medium for 7 days in white light, then transferred to medium plates containing 100 μM ABA for another 5 days under white light. Total RNA was extracted from treated plant materials as described (Hsieh et al., 1996). cDNA and cRNA synthesis and hybridization to 22K Affymetrix Gene Chips (ATH1) were performed according to Affymetrix instructions. Genes with 2-fold expression difference between the wild type and atgstu17 mutants were selected for further analysis and validated by RT-PCR.

Supplemental Data
The following materials are available in the online version of this article.  Table S1. A list of selected genes misregulated in atgstu17 under far-red light. Table S2. A list of selected genes affected in atgstu17 in the presence of ABA.

ACKNOWLEDGEMENTS.
We thank the Arabidopsis Biological Resource Center (ABRC; Ohio State University, Columbus) for providing AtGSTU17 T-DNA inserted mutant seeds and Ramanarayan Krishnamurthy for reading and critically commenting on the manuscript. We also thank Min-Yan Kuo at Academia Sinica, Taiwan, for assistance in microarray experiments.