Glucosinolate Distribution in the Aerial Parts of sel1-10, a Disruption Mutant of the Sulfate Transporter SULTR1;2, in Mature Arabidopsis thaliana Plants

Plants take up sulfur (S), an essential element for all organisms, as sulfate, which is mainly attributed to the function of SULTR1;2 in Arabidopsis. A disruption mutant of SULTR1;2, sel1-10, has been characterized with phenotypes similar to plants grown under sulfur deficiency (−S). Although the effects of −S on S metabolism were well investigated in seedlings, no studies have been performed on mature Arabidopsis plants. To study further the effects of −S on S metabolism, we analyzed the accumulation and distribution of S-containing compounds in different parts of mature sel1-10 and of the wild-type (WT) plants grown under long-day conditions. While the levels of sulfate, cysteine, and glutathione were almost similar between sel1-10 and WT, levels of glucosinolates (GSLs) differed between them depending on the parts of the plant. GSLs levels in the leaves and stems were generally lower in sel1-10 than those in WT. However, sel1-10 seeds maintained similar levels of aliphatic GSLs to those in WT plants. GSL accumulation in reproductive tissues is likely to be prioritized even when sulfate supply is limited in sel1-10 for its role in S storage and plant defense.


Introduction
Sulfur (S) is an essential macronutrient for all organisms. Plants take up inorganic sulfate as the major S source and assimilate it into a variety of S-containing organic compounds [1,2]. As animals are unable to assimilate sulfate, the role of plants in the global S cycle on the earth is extremely important [2]. In addition, many of the S-containing compounds biosynthesized in plants are beneficial to health, such as methionine (an essential amino acid for animals), glutathione (a redox controller), and various secondary compounds with specific functions [2]. Glucosinolates (GSLs) are the major S-containing secondary compounds biosynthesized in Brassicaceae, that act as defense compounds against insects and pathogens [3][4][5]. Depending on their amino acid precursors, most GSLs accumulated in Arabidopsis are classified into aliphatic and indolic GSLs (iGSLs) synthesized from methionine and tryptophan, respectively [3][4][5]. Among them, some aliphatic GSLs (mGSLs) are known to be beneficial for humans as cancer-preventive chemicals [6,7]. Thus, understanding GSL accumulation in plant tissues would contribute to improved food quality in Brassica crops.

Growth Phenotypes of WT and sel1-10 Plants
To investigate the metabolic changes occurring in mature sel1-10 plants, we initially observed the growth phenotypes of sel1-10 plants (Figure 1). WT and sel1-10 plants were grown for six weeks in vermiculite. Although visible differences in shoot phenotype were not observed between WT and sel1-10 plants (Figure 1a,b), a significant decrease was observed in the primary stem diameters of sel1-10 plants compared to those of the WT, while the plant heights were similar between WT and sel1-10 plants (Figure 1c). Correlated with the decrease in primary stem diameter in sel1-10, dry weight of primary stems (PS) was decreased in sel1-10 to 70% of that in WT plants (Figure 1d). Dry weights of rosette leaves (RL), cauline leaves (CL), lateral stems (LS), and siliques (Si) were not significantly lower but tended to be lower in sel1-10 plants relative to those in WT plants (Figure 1d). Asterisks indicate significant differences (Student's t-test; ** p < 0.05) between WT and sel1-10.

Concentrations of Sulfate and Selected Sulfur-Containing Metabolites in Different Parts of WT and sel1-10 Plants
We harvested RL, CL, PS, LS, and Si separately and analyzed sulfate, cysteine, glutathione (GSH), and GSL in different parts of the sel1-10 and WT plants ( Sulfate content in the RL of sel1-10 plants was 26% higher than that in the WT plants. Both WT and sel1-10 plants accumulated a similar level of sulfate in CL, PS, and LS. In Si, the sulfate content of sel1-10 plants was 61% of that in the WT plants. These results indicated that the distribution of sulfate was modulated in sel1-10 plants. To examine the effects of modulated sulfate distribution in sel1-10, cysteine and GSH contents in WT and sel1-10 plants were analyzed ( Figure 3). Cysteine content was not significantly different between WT and sel1-10 plants in all examined parts. The GSH content in Si of sel1-10 plants was 29% lower than that in WT plants, suggesting that the dysfunction of SULTR1;2 affects GSH accumulation in reproductive tissues as observed in the seedlings [28]. GSH content in other parts of sel1-10 plants was similar to that in the WT plants. , and siliques (Si) in WT and sel1-10 plants. White and green bars represent WT and sel1-10, respectively, in (c) and (d). Data are shown as the averages with error bars denoting SEM (n = 5). Asterisks indicate significant differences (Student's t-test; ** p < 0.05) between WT and sel1-10.

Concentrations of Sulfate and Selected Sulfur-Containing Metabolites in Different Parts of WT and sel1-10 Plants
We harvested RL, CL, PS, LS, and Si separately and analyzed sulfate, cysteine, glutathione (GSH), and GSL in different parts of the sel1-10 and WT plants (Figures 2-4).
Sulfate content in the RL of sel1-10 plants was 26% higher than that in the WT plants. Both WT and sel1-10 plants accumulated a similar level of sulfate in CL, PS, and LS. In Si, the sulfate content of sel1-10 plants was 61% of that in the WT plants. These results indicated that the distribution of sulfate was modulated in sel1-10 plants.
To examine the effects of modulated sulfate distribution in sel1-10, cysteine and GSH contents in WT and sel1-10 plants were analyzed ( Figure 3). Cysteine content was not significantly different between WT and sel1-10 plants in all examined parts. The GSH content in Si of sel1-10 plants was 29% lower than that in WT plants, suggesting that the dysfunction of SULTR1;2 affects GSH accumulation in reproductive tissues as observed in the seedlings [28]. GSH content in other parts of sel1-10 plants was similar to that in the WT plants.  The cysteine and GSH contents of different parts were measured using HPLC-fluorescence detection. WT and sel1-10 seedlings were grown for 6 weeks in vermiculite, after which each part was harvested. Rosette leaves (RL), cauline leaves (CL), primary stems (PS), lateral stems (LS), and siliques (Si). White and green bars represent cysteine and GSH levels in WT and sel1-10 plants, respectively. Data are shown as the averages with error bars denoting SEM (n = 3). Asterisks indicate significant differences (Student's t-test; * 0.05 < p <0.1, ** p < 0.05) between WT and sel1-10 plants.
GSL levels were generally lower in RL, CL, PS, and LS of sel1-10 relative to the same parts of WT plants ( Figure 4). However, in Si, GSL levels did not significantly vary between sel1-10 and WT plants, and some GSL levels were even higher in sel1-10 plants relative to the WT plants, that is, the levels of MSOX GSLs and I3M were similar between sel1-10 and WT plants, but the levels of MTX GSLs were higher in sel1-10 plants relative to the WT plants ( Figure 4).  The cysteine and GSH contents of different parts were measured using HPLC-fluorescence detection. WT and sel1-10 seedlings were grown for 6 weeks in vermiculite, after which each part was harvested. Rosette leaves (RL), cauline leaves (CL), primary stems (PS), lateral stems (LS), and siliques (Si). White and green bars represent cysteine and GSH levels in WT and sel1-10 plants, respectively. Data are shown as the averages with error bars denoting SEM (n = 3). Asterisks indicate significant differences (Student's t-test; * 0.05 < p <0.1, ** p < 0.05) between WT and sel1-10 plants.
GSL levels were generally lower in RL, CL, PS, and LS of sel1-10 relative to the same parts of WT plants ( Figure 4). However, in Si, GSL levels did not significantly vary between sel1-10 and WT plants, and some GSL levels were even higher in sel1-10 plants relative to the WT plants, that is, the levels of MSOX GSLs and I3M were similar between sel1-10 and WT plants, but the levels of MTX GSLs were higher in sel1-10 plants relative to the WT plants ( Figure 4). The cysteine and GSH contents of different parts were measured using HPLC-fluorescence detection. WT and sel1-10 seedlings were grown for 6 weeks in vermiculite, after which each part was harvested. Rosette leaves (RL), cauline leaves (CL), primary stems (PS), lateral stems (LS), and siliques (Si). White and green bars represent cysteine and GSH levels in WT and sel1-10 plants, respectively. Data are shown as the averages with error bars denoting SEM (n = 3). Asterisks indicate significant differences (Student's t-test; * 0.05 < p < 0.1, ** p < 0.05) between WT and sel1-10 plants.
GSL levels were generally lower in RL, CL, PS, and LS of sel1-10 relative to the same parts of WT plants ( Figure 4). However, in Si, GSL levels did not significantly vary between sel1-10 and WT plants, and some GSL levels were even higher in sel1-10 plants relative to the WT plants, that is, the levels of MSOX GSLs and I3M were similar between sel1-10 and WT plants, but the levels of MTX GSLs were higher in sel1-10 plants relative to the WT plants ( Figure 4).
Because GSL levels in Si were not affected in sel1-10 plants except for 4MTB and 7MTH, GSL levels in mature dried seeds were analyzed to determine the effects of reduced sulfate uptake ( Figure 5). Seeds contained much higher levels of MTX GSLs and 8MSOO and lower levels of I3M compared to other vegetative tissues in both plant lines, which is consistent with previous studies [8,9]. In seeds, GSL levels did not significantly vary between sel1-10 and WT plants. Because GSL levels in Si were not affected in sel1-10 plants except for 4MTB and 7MTH, GSL levels in mature dried seeds were analyzed to determine the effects of reduced sulfate uptake ( Figure 5). Seeds contained much higher levels of MTX GSLs and 8MSOO and lower levels of I3M compared to other vegetative tissues in both plant lines, which is consistent with previous studies [8,9]. In seeds, GSL levels did not significantly vary between sel1-10 and WT plants.  . White and green bars represent the relative GSL content in WT and sel1-10 seeds, respectively. Data are shown as averages with error bars denoting SEM (n = 3). Statistical analysis was performed with Student's t-test between WT and sel1-10 plants, but any significant differences were not detected.

Discussion
Growth phenotypes of mature sel1-10 plants have not been well studied as regards their aerial part. The sulfate uptake rate in sel1-10 plants was almost half of that in WT plants under both +S and -S conditions at the seedling stage [27,29]. In addition, the biomass and the levels of sulfate and GSH in sel1-10 seedlings were significantly lowered relative to those in the WT plants under both +S and -S conditions [27][28][29]. Similar growth retardation in mature sel1-10 plants observed in Figure 1 is assumed to be due to the reduction in sulfate uptake in sel1-10 plants.
It is known that GSL accumulation is differentially regulated in plant parts and S status [8,9,13,21,22]. In our analysis, the levels of MSOX GSLs and I3M in the leaves and stems of sel1-10 plants were significantly lower than those in the WT plants (Figure 4), in agreement with the -Sinduced-like phenotypes observed in sel1-10 seedlings [27,28]. In contrast, the GSL levels in Si and Se of sel1-10 plants were similar or higher than those in WT plants (Figures 4 and 5). Considering the previous theory that GSLs accumulated in the seeds provide an S source for seedling growth [8][9][10]21], GSL accumulation should be prioritized in reproductive tissues even when the S supply is limited in sel1-10 plants. Plants should have adapted to fluctuations in S availability by using GSLs as S storage substances in reproductive tissues. GSLs can be considered as a beneficial S storage compounds because of the relatively high molecular weight, that enable them to reduce osmotic pressure in the seeds. Additionally, GSLs can be a source of carbon and nitrogen, especially in the case of long-chain mGSLs highly accumulated in the seeds, and can also act as the defense compounds to protect the seeds from diseases or predators [11,21].
Unexpectedly, the levels of MTX GSLs were higher in Si of sel1-10 plants compared to those in the WT plants, while MSOX GSL and iGSL levels in Si were similar between sel1-10 and WT plants . White and green bars represent the relative GSL content in WT and sel1-10 seeds, respectively. Data are shown as averages with error bars denoting SEM (n = 3). Statistical analysis was performed with Student's t-test between WT and sel1-10 plants, but any significant differences were not detected.

Discussion
Growth phenotypes of mature sel1-10 plants have not been well studied as regards their aerial part. The sulfate uptake rate in sel1-10 plants was almost half of that in WT plants under both +S and -S conditions at the seedling stage [27,29]. In addition, the biomass and the levels of sulfate and GSH in sel1-10 seedlings were significantly lowered relative to those in the WT plants under both +S and -S conditions [27][28][29]. Similar growth retardation in mature sel1-10 plants observed in Figure 1 is assumed to be due to the reduction in sulfate uptake in sel1-10 plants.
It is known that GSL accumulation is differentially regulated in plant parts and S status [8,9,13,21,22]. In our analysis, the levels of MSOX GSLs and I3M in the leaves and stems of sel1-10 plants were significantly lower than those in the WT plants (Figure 4), in agreement with the -S-induced-like phenotypes observed in sel1-10 seedlings [27,28]. In contrast, the GSL levels in Si and Se of sel1-10 plants were similar or higher than those in WT plants (Figures 4 and 5). Considering the previous theory that GSLs accumulated in the seeds provide an S source for seedling growth [8][9][10]21], GSL accumulation should be prioritized in reproductive tissues even when the S supply is limited in sel1-10 plants. Plants should have adapted to fluctuations in S availability by using GSLs as S storage substances in reproductive tissues. GSLs can be considered as a beneficial S storage compounds because of the relatively high molecular weight, that enable them to reduce osmotic pressure in the seeds. Additionally, GSLs can be a source of carbon and nitrogen, especially in the case of long-chain mGSLs highly accumulated in the seeds, and can also act as the defense compounds to protect the seeds from diseases or predators [11,21].
Unexpectedly, the levels of MTX GSLs were higher in Si of sel1-10 plants compared to those in the WT plants, while MSOX GSL and iGSL levels in Si were similar between sel1-10 and WT plants (Figure 4). Taking into account that GSL levels in Si were the sum of the levels in silique tissues, including the developing seeds, and all samples were collected on the same date, the increase of MTX GSLs could be because of the acceleration of seed maturation in sel1-10 plants. In general, nutrient stress accelerates bud appearance and subsequent development of the siliques and seeds in Arabidopsis [30,31]. Considering that MTX GSLs in the seeds are continuously increased during the seed maturation period [8,9], the timing of flowering and seed development may occur earlier in sel1-10 than in the WT plants. Lower levels of sulfate and GSH in Si of sel1-10 plants relative to those in the WT plants also support this assumption (Figures 2 and 3).
Several maternal tissues have been suggested as source tissues for seed GSLs, including the leaves and siliques [10,11]. Although the GSL transport machinery in whole plants is not fully understood [10][11][12]32], GSL transporters, GTR1 and GTR2, that belong to the NRT/PTR family have been characterized for their roles [32,33]. In the double disruption lines of GTR1 and GTR2, most GSLs were not found in the seeds, whereas, mGSLs were highly accumulated in rosette leaves and siliques [32,33]. This suggested that seed GSLs are mostly transported from the source tissues. Decreased GSL levels in vegetative tissues and the maintenance of GSL levels in the seeds suggested that GSL transport to the seeds was not restricted or was even accelerated in sel1-10 plants.
In conclusion, we found that GSL levels of the MSOX group were decreased in the leaves and stems, whilst all GSL were found to be maintained in the seeds in sel1-10 plants. This shows that accumulation of mGSL characterizes the reproductive tissues, thus indicating that mGSL are destinated to store in the seeds in order to support the initial growth of the next generation.

Plant Materials and Growth Conditions
Arabidopsis thaliana were cultured in a growth chamber controlled at 23 ± 2 • C under constant illumination (40 µmol m −2 s −1 ). The sel1-10 mutant, carrying a T-DNA insertion in the ninth exon of SULTR1;2 (At1g78000) [28], and the background Wassilewskija (Ws-0) wild-type plants (WT) were used as plant materials. Seeds of WT and sel1-10 were sown on vermiculite as growth substrate supplemented with MGRL mineral nutrient media in 5 × 5 × 5 cm plastic pots [34,35]. After germination, the number of plants was adjusted to three plants per pot. Plants were grown for 6 weeks and the different parts of the plants, rosette leaves (RL), cauline leaves (CL), primary stems (PS), lateral stems (LS), and siliques (Si) were harvested separately from each pot and weighed for the fresh weights. Mature dried seeds collected from the former generation were used for the analysis. Right after harvest, plant tissues were frozen in liquid nitrogen, freeze-dried, ground into a fine powder using a Tissue Lyser (Retsch, Germany), and used for each metabolite analysis. Three independent samples for each part were used for metabolite analysis.

Measurement of Glucosinolates
Three milligrams of the plant powder was extracted with 300 µL of ice-cold 80% methanol containing 2 µM L(+)-10-camphor sulfonic acid (10CS, internal standard, Tokyo Kasei, Japan) using a Tissue Lyser. After homogenization, cell debris was separated by centrifugation (15,000 rpm, 10 min, 4 • C), and the supernatants were evaporated with a centrifugal evaporator (CVE-3110, EYELA, Japan) connected to a high vacuum pump (DAH-60, ULVAC, Japan) and a cold trap (UNI TRAP UT-1000, EYELA). Dried supernatants were dissolved into water, filtered with Millex-GV filter units (Millipore, USA), and analyzed by a high-performance liquid chromatograph connected to a triple quadrupole (LC-QqQ)-MS (LCMS8040, Shimadzu, Kyoto, Japan) using L-column 2 ODS (pore size 3 µm, length 2.1 × 150 mm, CERI, Japan). The mobile phase consisted of solvent A (0.1% formic acid, Wako Pure Chemicals, Osaka, Japan) and solvent B (0.1% formic acid in acetonitrile, Wako Pure Chemicals, Osaka, Japan). The gradient elution program was as follows with a flow rate of 0.3 mL/min, 0-0.1 min, 1% B; 0.1-15.5 min, 99.5% B; 15.5-17 min, 99.5% B; 17-17.1 min, 1% B; and 17.1-20 min, 1% B as described previously [36]. For the MS, electrospray ionization mass spectrometry technique in negative ionization mode was used. The ionization parameters were as follows: the nebulizer gas flow was 1.5 L/min, the CDL temperature was 250 • C, heat block temperature was 400 • C. All GSLs were detected with optimized selective reaction monitoring transitions in negative ionization mode as follows ( MRM transitions were determined by using standard compounds (Cfm Oskar Tropitzsch GmbH, Marktredwitz, Germany) or a database (MassBank, http://www.massbank.jp). The relative quantities of GSLs were calculated as the ratio of peak height to the height of 10CS.

Measurement of Sulfate, Cysteine and Glutathione
One mg of the plant powder was extracted with 200 µL of 10 mM HCl. The cell debris was removed by centrifugation, and the supernatant was used for the analysis. The extracts were diluted 100 fold with extra pure water and analyzed by ion chromatography as described previously [29], using an eluent containing 1.9 mM NaHCO 3 and 3.2 mM Na 2 CO 3 .
Cysteine and GSH contents were determined by monobromobimane (Invitrogen) labeling of thiol bases after reduction of the extracts with dithiothreitol (Nacalai Tesque) as described [13,28,29]. The labeled products were then separated by HPLC (JASCO, Tokyo, Japan) using the TSKgel ODS-120T column (150 × 4.6 mm, TOSOH) and detected with a fluorescence detector FP-920 (JASCO), monitoring for fluorescence of thiol-bimane adducts at 478 nm under excitation at 390 nm. GSH and Cys standards were purchased from Nacalai Tesque (Kyoto, Japan).

Statistical Analysis
The data were statistically analyzed using Student's t-test with Microsoft Excel. Significant differences between WT and sel1-10 in biological replicates are shown in each Figure.