The Arabidopsis gene ATST4a is not a typical brassinosteroids catabolic gene

Abstract

Brassinosteroid (BR) homeostasis is maintained in part by this hormone’s catabolism. The presence of multiple BR-catabolic pathways in Arabidopsis demonstrates the importance of this process in growth and development. Previous biochemical analyses suggest that ATST4a has BR catalytic activity. We have used both overexpression and loss-of-function genetic approaches to further explore the role of ATST4a in Arabidopsis. Up to 1000-fold overexpression of the ATST4a gene did not result in any characteristic BR-deficient phenotypes. In addition, the T-DNA insertion null mutant atst4a1–1 did not display enhanced seedling hypocotyl growth in the presence or absence of the active BR brassinolide when grown in white light. This lack of hallmark characteristics for BR-inacitivion genes suggests that ATST4a encodes an atypical BR catabolic enzyme.

Introduction

Catabolism is an important aspect of brassinosteroid (BR) homeostasis. The discovery of multiple independent BR-catabolic pathways and their enzymes in Arabidopsis demonstrates the importance of this process in plant growth and development.^1-10 These BR catabolic pathways were mainly discovered and latter characterized using genetic manipulation of the genes involved. In all cases, in planta overexpression of BR catabolic genes leads to characteristic BR-deficient phenotypes at both seedling and adult stages of Arabidopsis.^1-10 These BR-deficient phenotypes typified by a BR biosynthetic mutant de-etiolated 2–1 (det2–1), include small organ size, small round and dark green leaves, and delayed flowering.^11-14 Genetic analyses of knockout or knockdown mutants of most of these BR catabolic genes show a reduction in hypocotyl-elongation inhibition in white light in Arabidopsis seedlings.^1,3,6,7,9 The knockout or knockdown approach may in some cases fail to provide any evidence of involvement in BR catabolism due to genetic redundancy or yet other unknown reasons.⁶ In vivo and in vitro biochemical assays of the enzymes encoded by some of these genes were also used to confirm their role in BR catabolism.^2,6,8,10,15

Members of the sulfotransferase protein superfamily are known to perform various metabolic functions in plants^16,17 and some family members are also known to be metabolically active in hormonal pathways.^18,19 ATST4a (AT2G14920) encodes a sulfotransferase member in Arabidopsis.¹⁶ ATST4a was cloned on the basis of homology to ATST1.²⁰ ATST1 is an ortholog of BNST (steroid sulfotransferase from Brassica napus) proteins and displays similar catalytic activities and substrate specificity. Recent evidence shows that ATST1 is also involved in the sulfonation of salicylic acid in response to pathogen stress.¹⁹ In vitro protein expression and sulfotransferase assay studies demonstrate that ATST4a has catalytic activity with BRs.²⁰ Catalytic activity of ATST4a is specific for biologically active end products of the BR biosynthetic pathway, including castasterone, brassinolide, related 24-epimers, and the naturally occurring (22R, 23R)-28-homobrassinosteroids. The addition of a sulfate group to the steroid molecules may cause a change in their activity and therefore play a role in steroid homeostasis.²¹ These observations suggest that ATST4a is involved in BR homeostasis via catabolism in Arabidopsis.

Although biochemical analysis can provide important clues, use of genetic approaches can potentially further our understanding of endogenous gene functions. In this study, we have used both overexpression and loss-of-function genetic approaches to further explore the role of ATST4a in Arabidopsis. This study shows that overexpression of the ATST4a gene in transgenic lines did not result in any characteristic BR-deficient phenotypes, and did not cause a predicted feedback-regulation change in the transcript accumulation of DWF4 and BAS1.²² Additionally, atst4a1–1, a T-DNA insertion null mutant of ATST4a did not display enhanced seedling hypocotyl growth in the presence or absence of the active BR brassinolide when grown in white light. These observations suggest that ATST4a gene encodes an atypical BR catabolic enzyme.

Results

To test the hypothesis that the constitutive overexpression of ATST4a will lead to BR-deficient phenotypes, multiple transgenic lines overexpressing the ATST4a cDNA were generated. These ATST4a-OX transgenic lines were characterized by quantitative RT-PCR analysis. Three independent transgenic lines displaying high levels of ATST4a expression (Fig. 1A) were selected for further phenotypic analysis. Adult ATST4a-OX transgenic lines did not display any BR-deficit phenotypes (Fig. 1B). The ben1-D (bri1–5 enhanced 1-D) mutant was included as a BR-deficit-phenotype control for comparison.⁷ ben1-D mutant plants were grown along with and under the same growth conditions as the ATST4a-OX lines for comparison of their leaf phenotypes. The ben1-D mutant conferred small and round leaves with short petioles, whereas the ATST4a-OX plants displayed a wild-type leaf phenotype (Fig. 1B).

Figure 1. Overexpression of ATST4a does not lead to BR-deficient phenotypes in Arabidopsis. (A) The ATST4a overexpression lines show ~1000-fold levels of ATSt4a transcript as compared with the Col-0. Error bars indicate SE (B) Adult phenotypes of the 3-week-old Arabidopsis plants overexpressing ATST4a cDNA.

A T-DNA insertion line (atst4a1–1) with an insertion in ATST4a gene was obtained from GABI-Kat.²³ Genetic analysis by gene- and T-DNA-specific primers indicated that the T-DNA is inserted in the coding sequence of the ATST4a gene (Fig. 2A). Histochemical analysis of the ATST4a-GUS translational fusion lines indicated that ATST4a is expressed in both white-light and dark grown seedlings (Fig. 2B). This observation suggested that ATST4a may play a role in hypocotyl growth. To test this hypothesis, hypocotyl growth was studied in the wild type and the atst4a1–1 lines. Hypocotyl growth in dark and 2 white-light fluence rates showed no significant differences between the wild type and atst4a1–1 lines (Fig. 2C). This suggests that ATST4a does not play a role in light-induced hypocotyl growth inhibition.

Figure 2. Genetic analysis of the atst4a1–1 T-DNA insertion line. (A). Graphical representation of the location of T-DNA insertion in the ATST4a ORF. Arrows show the location of the gene and T-DNA specific primers used for genetic screening. (B) ATST4a is expressed in the hypocotyls of the white-light and dark grown seedlings as shown by the histochemical analysis of the pATST4a: ATST4a-GUS expressing lines. Scale bar = 2mm. (C) The atst4a1–1 mutant does not show an aberrant hypocotyl-elongation phenotype in both darkness and continuous white light conditions. Fluence rate analysis of hypocotyl growth of atst4a1–1 in darkness and at 2 different fluence rates, WL-3.5 (white light = 3.5 µmol m^–2 s^–1) and WL-10 (white light = 10 µmol m^–2 s^–1) did not show any altered hypocotyl-elongation response when compared with Col-0. Three replications of 5-d-old seedlings were used to measure hypocotyl growth. Error bars indicate standard error (SE).

Hypocotyl-growth was also studied in the homozygous ATST4a-OX, wild-type and the atst4a1–1 seedlings in both darkness and white-light (Fig. 3A and B). The hypocotyl length in the atst4a1–1 line was not significantly different than the wild type (p > 0.5). The 3 overexpression lines displayed slightly shorter hypocotyl lengths than the wild type (p < 0.01). The hypocotyl-growth response of wild type, ATST4a-OX, and atst4a1–1 seedlings was not altered by BL treatment in the dark (Fig. 3C and D). In white light, the hypocotyl length in the atst4a1–1 line was not significantly different than the wild type (p > 0.5) (Fig. 3D). The T3–52 and T3–64 overexpression lines displayed slightly shorter hypocotyl lengths than the wild type (p < 0.01) (Fig. 3D).

Figure 3. (A and B) Hypocotyl-elongation response of atst4a1–1 and ATST4a-OX T3 lines to dark and continuous white light conditions. Fluence rate analysis of hypocotyl growth of ATST4a-OX T3 lines in dark (A) and in continuous white light (10 µmol m^–2 sec^–1) (B) did not show dramatic differences in hypocotyl-elongation response when compared with Col-0. (C and D) Hypocotyl-elongation response of atst4a1–1 and ATST4a-OX T3 lines to BL treatment. Seedlings were grown in continuous light for 5 d on media containing BL at 100 nM concentration in dark (C) and white light (25 µmol m^–2 sec^–1) (D). Hypocotyl-elongation response of atst4a1–1 and ATST4a-OX T3 lines do not show dramatic differences when compared with Col-0 in both dark and light conditions. Three replications of 5-d-old seedlings were used to measure hypocotyl growth. Error bars indicate SE.

Overexpression of BR catabolic genes often causes alterations in the transcript levels of BR-feedback regulated genes. In these conditions, the transcript accumulation of BR biosynthesis genes, such as DWF4, is enhanced whereas that of catabolic genes, such as BAS1, is reduced. This response is often vectorially opposite in the loss-of-function lines.²² To test this hypothesis, gene expression of the BR-feedback regulated genes BAS1 and DWF4 was studied in wild type, atst4a1–1 and ATST4a-OX seedlings. Quantitative RT-PCR analysis showed that transcript accumulation of both BAS1 and DWF4 was not significantly different in any these lines (p > 0.05 for all pair-wise comparisons) (Fig. 4A and B).

Figure 4.ATST4a-OX T3 lines did not display feedback regulation of BR biosynthesis and catabolic genes DWF4 and BAS1. Quantitative RT-PCR analysis of the DWF4 (A) and BAS1 (B) transcript in Col-0, atst4a1–1, and 3 single locus insertion ATST4a-OX T3 lines show that ATST4a overexpression lines did not show the predicted altered levels of DWF4 and BAS1 transcript accumulation when compared with Col-0. Error bars indicate SE.

Discussion

Biochemical in vitro analysis of ATST4a suggests its role in BR catabolism.²⁰ Not enough genetic data are available, however, to support this role of ATST4a in BR catabolism in Arabidopsis. This study uses both gain-of-function and loss-of-function genetic approaches to further analyze the role of ATST4a in Arabidopsis.

Overexpression of the ATST4a cDNA in Arabidopsis did not result in BR-deficient phenotypes for seedlings or adult plants (Fig. 1A and B). Moreover, seedlings of the transgenic lines overexpressing the ATST4a cDNA did not show any dramatic increase in the light-induced inhibition of hypocotyls (Fig. 3). This observation is significant since all the known BR catabolic genes characterized so far result in dramatic BR-deficient phenotypes both at seedling and at adult stages when overexpressed in Arabidopsis.^1-10 This suggests that ATST4a overexpression did not reduce BR content in the transgenic lines. In addition, quantitative transcript analysis of the feedback regulated genes DWF4 and BAS1 also suggests that BR content is not reduced in the ATST4a-OX lines when compared with wild type (Fig. 4). Overall, this study shows that despite its in vitro preference for BR substrates, ATST4a does not behave like a typical BR catabolic gene in planta.

Mutants of BR catabolic genes in most cases are compromised in light-mediated hypocotyl growth inhibition.^1,3,6,7,9 However, despite the observation that ATST4a is expressed in the seedling hypocotyl (Fig. 2B), atst4a1–1 did not display any defect in light-mediated hypocotyl inhibition (Fig. 2C). Single-mutant loss-of-function analysis can, however, lead to inconclusive results based on multiple factors, including gene regulation and genetic redundancy.⁶ The latter is unlikely since biochemical analysis shows that the closest homologs of ATST4a (ATST4b and ATST4c) do not display any affinity for brassinosteroids.²⁰ In vitro biochemical analysis can give important clues to the function of an enzyme. However enzyme activity and function at a tissue or cellular level may be dependent on certain in vivo conditions and therefore difficult to assess. For example, the unique metabolic environment within a cell or a tissue may provide crucial cofactors, or in another case the substrate preference may be different in an in vivo vs the in vitro environment.

Sulfotransferases can display preference for a diverse range of substrates under in vitro conditions. This, however, may not fully reflect their endogenous functions.¹⁷ For example, biochemical analysis of BNST3 suggests that it is involved in catabolizing 24-epibrassinosteroids.²⁴ It was also shown that sulfonation of 24-epibrassinolide at 22-C position abolishes its biological activity.²⁴ Overexpression of BNST3, however, do not lead to BR-deficient phenotypes in Arabidopsis, suggesting that BR catabolism is not its endogenous function.²⁵ It is also possible that in planta, sulfonation of brassinosteroids is a reversible conjugation process, and that the role of sulfonation is to generate a reserve pool of inactive brassinosteroids, which can be readily activated by a reverse reaction. The overexpression of conjugating enzymes, it may be argued in some cases, still lead to some partial symptoms of an active hormone deficiency.^8,26 This is not the case, however, for ATST4a-OX, suggesting that it may not also be a conjugating enzyme. Another explanation for lack of BR-deficient phenotypes in overexpression lines is that of post-translational regulation of ATST4a. It is also possible that the overexpressed ATST4a protein is not properly localized in the plant cells. Given these observations, more studies are required to assign biological function to ATST4a. The Brassica napus steroid sulfotransferases are inducible by ethanol, xenobiotics, and low-oxygen stress and may be involved in plant stress responses.²⁵ Therefore, it is possible that ATST4a also has a similar role in Arabidopsis. This can be studied by observing growth of wild type, ATST4a loss-of-function, and overexpression lines in response to various stresses. Another direction would be to identify the transcriptional regulators of ATST4a by using its promoter in a yeast-one-hybrid screen. They may help identify the biological pathways in which ATST4a might be involved.

Material and Methods

Plant material

All plant material used in this study was in the Col-0 background. For loss-of-function analysis of ATST4a, a T-DNA insertion line GK-177E08 with an insertion in the ATST4a ORF was obtained from GABI-Kat.²³ The T-DNA line was backcrossed twice to the Col-0 to clean the background of any unwanted mutations. The T-DNA insertion was followed through the crosses using the following set of primers: gene specific primers- 5′-ATGGATGAAA AAGATAGACC-3′ and 5′-TTAGAATTTC AAACCGGAAC C-3′ and T-DNA specific primer 5′-ATATTGACCA TCATACTCAT TG-3′. For overexpression analysis, the ATST4a coding sequence including the start and stop codon were cloned in the pENTR/D/TOPO vector (Invitrogen). The cloned DNA was sequenced to make sure that there were no mutations. The coding sequence containing fragment was further subcloned into a binary vector pEARLYGATE100 by gateway^® cloning using LR clonase (Invitrogen). The binary vector was transformed into atst4a1–1 to increase the frequency of transgenic overexpressers by avoiding RNAi interference from the endogenous transcript. Single locus insertion lines based on a 3:1 ratio of resistant to susceptible were selected for further analysis. For generating ATST4a:GUS translational fusion lines, the 3.0 Kb genomic fragment containing the ATST4a promoter and gene was cloned in frame with uidA gene in the pCAMBIA1305.1 vector. This construct was transformed into atst4a1–1 plants. Multiple transgenic lines segregating at a 3:1 ratio (hygromycin resistant/sensitive ratio) in the T2 generation were identified as single insertion lines.

Transcript analysis

Total RNA was isolated from 5-d-old seedlings grown in continuous white light (25 µmol m^–2 sec^–1) using the RNeasy Plant Kit (Qiagen). On-column DNase digestion was performed using the RNase-Free DNase Set (Qiagen), to eliminate genomic DNA contamination. Total cDNA was synthesized using SuperScriptIII First-Strand Synthesis System (Invitrogen). ATST4a transcript was amplified using primers 5′-ATGGATGAAA AAGATAGACC-3′ and 5′-TTAGAATTTC AAACCGGAAC C-3′. ACTIN2 was used as an internal control in RT-PCR. ACTIN2 transcript was amplified using primers 5′-GGTCGTACAA CCGGTATTGT GCTGG-3′ and 5′-CTGTGAACGA TTCCTGGACC TGCC-3′. The linear range of amplification for each gene transcript was determined by comparing samples obtained using different numbers of cycles. Lack of genomic and foreign DNA contamination was determined by using all RNA samples and water as a template in a PCR reaction.

Real-time quantitative analysis of transcript levels

Applied Biosystems 7500 Fast Real-Time PCR System (Applied Biosystems; Foster City, CA) was used for real-time quantitative RT-PCR analysis. ATST4a transcript was amplified using primers 5′-TCACTCGAGC GGAGGATT AC-3′ and 5′-GAGTACGAGG CTCCGCTTT-3′. Internal control ubiquitin gene (AT5G15400) was amplified using PCR Primers 5′-GAAATGCATG GAGACGGATT-3′ and 5′-TTGGTCTCTG CTCCCACTCT-3′. PCR thermocycling program profile used was as following: initial denaturation at 95 °C for 20 s, followed by 40 cycles of 95 °C for 3 s, 60 °C for 30 s. The melt curve profile, which was used for melt curve analysis, was as follows: 95 °C for 15 s, 60 °C for 1 min, 95 °C for 30 s, and 60 °C for 15 s.

Exogenous hormone treatment

The stock solution of brassinolide (BL) was dissolved in 95% ethanol (v/v). BL treatment was given by adding BL stock solution to the seedling growth media to a final concentration of 100 nM.

Hypocotyl measurement

Procedures for seed sterilization, plating, growth conditions, and hypocotyl measurement were done as described in Turk et al.²

Histochemical GUS analysis

Histochemical GUS staining was performed as described in Sandhu et al.²⁷

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.