Soybean (Glycine max L.cv. Jiyu70) seeds were surface sterilized for 5 min in 1.0% (v/v) NaClO, rinsed thoroughly with tap water and germinated in peat-moss for 3 days at 25°C in the dark. The seedlings were transferred to plastic containers filled with 1.0 L nutrient solution (pH 4.5) (Horst et al., 1992). The medium was refreshed every 3 days. The climate conditions in the chamber were set as follows: day/night temperature 25/22°C, photoperiod 14 h, light intensity 300 μmol photons m⁻² s⁻¹ (sodium lamp Philips SON, 400 W), and relative humidity 70%.

After 7 days growth in the nutrient solution, soybean seedlings were transferred to 0.5 mM CaCl₂ solution (pH 4.5) for 12 h in the dark. Then, the seedlings were subjected to different treatment solutions. In the time course experiment, the roots of seedlings were exposed to 0.5 mM CaCl₂ solution (pH 4.5) containing 0 or 30 μM AlCl₃ with or without 10 μM SA for 3, 6, 9, or 12 h. In the pharmacological experiment, the roots of seedlings were subjected to 0.5 mM CaCl₂ solution (pH 4.5) containing 0 or 30 μM AlCl₃ with or without 10 μM SA or 100 μM PAC. Each experiment was repeated at least three times. Root apices (0–3 cm) were excised and stored in a −80°C freezer until the determination of the SA content and an enzyme activity assay.

The total SA content (free plus bound) in fresh root tips was determined according to the method used by Liu et al. (2012). The soybean roots (about 1.0 g) were ground in liquid nitrogen, 4 mL of 5% trichloroacetic acid (TCA), 16 mL of double distilled H₂O and 30 mL of ethyl ether were added to the homogenate and transferred to 50 mL centrifuge tube, fully shaken, and extracted for 12 h. After 5 min centrifugation at 10,000 g, the ethyl ether supernatants was partitioned, the aqueous phase was extracted two times in succession with ethyl ether. Three combined extracts were condensed by rotary evaporation at 40°C and the residue was dissolved in 1 mL methanol and acetate buffer (pH 3.2) (1:1, v/v) for free-SA analysis. Bound SA content was indirectly quantified by acid hydrolysis of the compounds remaining in the aqueous phase after organic extraction. Acid hydrolysis was performed by adding 18.5% HCl to the aqueous phase in order to make the final concentration 3.2%. The samples were incubated at 80°C in a water bath for about 1 h. After cooling, they were extracted with ethyl ether for three times. The combined ethyl ether phase supernatants were condensed by rotary evaporation at 40°C and 1 mL of the mixture of methanol and acetate buffer (pH 3.2) (1:1, v/v) was added to 50 ml rotary evaporator flask, and dissolved fraction was transferred to 1.5 mL Eppendorf vials containing the bound SA. Quantitative analysis of free and bound SA were performed by high-performance liquid chromatography (HPLC, LC 20 AT, Shimadzu, Tokyo, Japan) equipped with a reverse-phase C18 column (VP-ODS, 150 mm × 4.6 mm) and a fluorescence detector (excitation wavelength = 310, emission wavelength = 415 nm). All the samples were filtered through 0.3 μm microporous filters before injection.

Root apices (0–3 cm) were excised for the PAL activity assay, following Ferrarese et al. (2000) with some modifications. Samples (1.0 g) were ground at 4°C in 0.1 M sodium borate buffer (pH 8.8), containing 15 mM β-mercaptoethanol, 5 mM EDTA, 1 mM phenylmethanesulfonyl fluoride, and 0.15% polyvinyl pyrrolidone (w/v). Homogenates were centrifuged (15,000 g, 4°C, 20 min), and the supernatant was used for enzyme preparation. The reaction mixture contained 0.1 M of borate buffer (pH 8.8), 15 mM L-phenylalanine and 1 mL enzyme extract in a 5 mL volume. The reaction proceeded for 1 h at 37°C and was interrupted by the addition of 0.5 mL 6 M HCl. The activity of PAL was determined spectrophotometrically at a wavelength of 290 nm.

BA2H activity was determined using the method presented by Qian et al. (2009). Samples (1.0 g) were frozen in liquid nitrogen and ground in a chilled mortar. The frozen powder was added to 0.03 M potassium phosphate buffer (pH 7.5) containing 1 mM EDTA, 1 mM sodium benzoate and 0.5 mM dithiothreitol. The homogenate was centrifuged at 12,000 g for 15 min at 4°C. The supernatant was used for determination of BA2H enzyme activity. The reaction mixture (final volume of 4 mL) contained 1.2 mL 1 M Tri-HCl buffer at pH 7.5, 0.4 mL H₂O, 0.6 mL 150 μM NADPH, 0.6 mL 20 μM FAD, and 1.2 mL enzyme extract. The reaction mixture was partitioned into two parts. One part was added to 1 mL 1 mM benzoic acid to start the enzymatic reaction; the other part was added to 1 mL H₂O. The product of the reaction catalyzed by NADPH was determined spectrophotometrically at 340 nm. Protein content was determined according to the method of Bradford (1976), using bovine serum albumin as a standard.

O2- production was determined according to the method of Shi and Zhu (2008).

The malondialdehyde (MDA) content was measured as described by Heath and Packer (1968). Briefly, root tips (1.0 g) were ground in 3 mL 0.1% trichloroacetic acid (TCA) solution. The homogenate was centrifuged at 15,000 g for 10 min, and 0.5 mL the supernatant was assayed for MDA content.

The H₂O₂ concentration in crude extracts from soybean roots was measured by spectrofluorometry, according to the method used by Jana and Choudhuri (1982), and the absorbance of the titanium-peroxide complex was monitored at 410 nm. Using a standard calibration plot, the A₄₁₀ readings were converted to corresponding H₂O₂ concentrations.

For H₂O₂ detection by fluorescence microscopy according to Rodríguez-Serrano et al. (2009). Soybean root tips (0–1 cm) were excised in centrifuge tube (1.5 mL). The 2′,7′-dichlorofluorescin diacetate (DCF-DA, 25 μM) was used as a fluorescent probe to stain the root tips for 20 min at 37°C in the dark. Then, the root tips were washed thrice with Tris-HCl (pH 7.2) buffer. The stained root tips were observed under a Nikon Eclipse 80i upright fluorescence microscope equipped with a mercury lamp (100 W) and a C-FL-C 30 mm Epi-Fluorescence Filter Block (MBE44725, FITC) consisting of excitation filter EX465-495, Dichroic Mirror DM505, Barrier filter BA512-558. The excitation and emission wavelengths of H2DCF-DA are 485 and 530 nm. The CCD camera (Nikon Fi2) and the NIS-Elements F software were used to capture the images.

Soybean root tips (1.0 g) were frozen in liquid nitrogen and homogenized in 2 mL 50 mM sodium phosphate buffer (pH 7.8) containing 0.1 mM EDTA and 1% (w/v) PVP. The homogenate was centrifuged at 15,000 g for 20 min at 4°C, and the supernatant was used for the following enzyme assays.

The activities of SOD, CAT, and APX were determined as described by Jiang and Zhang (2001). Total SOD activity was assayed by monitoring the inhibition of photochemical reduction of nitro blue tetrazolium (NBT). One unit of SOD activity was defined as the amount of enzyme required to cause 50% inhibition of the reduction of NBT as monitored at 560 nm. CAT activity was determined by measuring the rate of decomposition of H₂O₂ at 240 nm for 5 min. The enzyme activity was calculated using an extinction coefficient of 39.4 mM⁻¹ cm⁻¹. APX activity was determined by measuring the decrease rate of the H₂O₂-dependent oxidation of ascorbate at 290 nm for 5 min. The enzyme activity was calculated using an extinction coefficient of 2.8 mM⁻¹ cm⁻¹. Peroxidase (POD) activity was monitored by oxidation of guaiacol using H₂O₂. The enzyme activity was calculated using an extinction coefficient of 26.6 mM⁻¹ cm⁻¹ at 470 nm.

Total RNA was isolated using RNA isoreagent (TaKaRa Bio Inc., Shinga, Japan) according to the manufacturer's protocol. cDNA was synthesized from 1 μg total RNA in a 10 μL aliquot using Moloney Murine Leukemia Virus reverse transcriptase (M-MLV, TaKaRa Bio Inc.) at 42°C for 1 h with the oligo-dTadaptor primer (TaKaRa Bio Inc.) following the manufacturer's protocol. Real-time quantitative PCR was performed using 0.5 μL of reverse transcription production in a 25-μL reaction volume with SYBR Premix Ex Tag (TaKaRa Bio Inc.) on an ABIPRISM 7500 Sequence Detection System (Applied Biosystems, Foster City, CA, USA). Real-time PCR was performed using gene-specific primers: β-tubulin was used as a house-keeping gene according to Xu et al. (2010), and the GmNPR1 Gene-Bank accession number is FJ418595; upstream primer 5′-ATGGCAAGGTTGGATATGAAGC-3′ and downstream primer 5′-TGGCAGGTCTACACGCATCA-3′; the amplification size of the product was 132 bp. Each sample was run in triplicate. The relative transcript abundance was calculated using the 2^−ΔΔCT method.

All data were statistically analyzed using SPSS 13.0 software (SPSS Inc., Chicago, IL, USA), including variance analysis and multiple mean comparison using Tukey's test. The figures were drawn using Kyplot software (OriginLab Corporation, Northampton, MA, USA).

The root elongation rate was reduced by 51.7% after 12 h' Al treatment (Figure 1A). The root elongation rate only decreased by 15.5% when roots were treated with the solution containing both Al and 10 μM SA. This showed that SA alleviated Al-induced inhibition of root elongation.

The alleviating effect of SA on Al-induced inhibition of root elongation suggested that the inhibition of root elongation by Al may be caused by the reductions in endogenous SA concentration in soybean roots. To test this hypothesis, the SA biosynthesis inhibitor PAC was used in the absence and presence of Al and SA. Treatment with 100 μM PAC for 12 h aggravated Al-induced root elongation inhibition, which could be reversed by exogenous SA supply (Figure 1A). In addition, SA decreased, while PAC increased Al content in the root tips of the soybean seedlings under Al stress (Figure 1B).

To confirm whether Al-induced inhibition of root elongation was related to endogenous SA concentration, the concentrations of total SA were investigated at 3-h intervals over a period of 12 h (Figure 2A). The effect of SA on Al-induced endogenous SA concentrations in soybean roots was also examined under PAC treatment (Figure 2B). In the presence of Al, both SA and PAC caused higher SA concentration as compared to the corresponding controls without Al. Exogenous SA markedly enhanced the Al-induced endogenous SA concentrations by 88.7% compared with the Al treatment without SA. However, the application of PAC decreased the SA content regardless of SA application.

PAL and BA2H serve as two key enzymes responsible for SA biosynthesis in plants. To determine the contribution of PAL and BA2H enzymatic pathways to SA production under Al stress, we analyzed the activities of the two enzymes at 3 and 12 h. As shown in Figure 3, the PAL activity was increased after 3 h of Al treatment, while it was decreased slightly at 12 h. Al stress elevated the activity of BA2H at 3 and 12 h, i.e., by 2.8-fold and 4.7-fold, respectively, in relation to the control. Application of exogenous SA under Al stress further enhanced the PAL and BA2H activities at 3 and 12 h. It has been reported that PAC is an effective inhibitor of BA2H during SA biosynthesis. As expected, the activity of BA2H, but not PAL, was markedly inhibited in the PAC treatment under Al stress.

To test SA accumulation under Al stress at the transcriptional level, we examined the expression of GmNPR1 at 3, 6, 9, and 12 h (Figure 4). In soybean roots, the transcriptional level of GmNPR1 was strikingly increased by Al treatment with the elongation of treatment duration. This result was consistent with the endogenous SA accumulation pattern during Al treatment. In the SA plus Al treatment, the gene expression could be rapidly induced and reached a maximum at 6 h, which was over 2-fold higher than that in response to Al alone. However, the high expression of GmNPR1 could not be maintained until 12 h.

To assess the effect of SA on ROS and lipid peroxidation under Al stress, the accumulation of O2- and MDA was detected in soybean roots after 12 h of Al treatment (Figure 5). Adding exogenous SA reduced the contents of O2- and MDA by 44 and 20%, respectively, compared with Al treatment alone, which indicated SA mitigated oxidative damage caused by Al stress.

To determine how H₂O₂ changed in response to exogenous SA under Al stress, we detected H₂O₂ production using spectrophotometry (Figure 6A). Compared with the control, SA application significantly increased H₂O₂ content, and an oxidative burst occurred at the 6th h; Al application also increased H₂O₂ content, but the oxidative burst occurred at the 9th h (Figure 6A). Compared to only Al treatment, the concomitant application of Al and exogenous SA caused an oxidative burst at the 6th h, which was even stronger than with only SA; after that, the H₂O₂ content promptly declined to levels significantly lower than in the only Al-treated samples. To provide extra information of the H₂O₂ accumulation in different treatments, we used a H₂O₂-specific fluorescent probe, i.e., H₂DCF-DA (Figures 6B,C), and the results were similar with spectrophotometry methods.

We also measured the activities of the antioxidant enzymes POD, CAT, SOD, and APX to analyze the source of H₂O₂ formation and protection against oxidative stress. As shown in Figure 7, Al-induced POD, SOD, and APX activities increased gradually with the elongation of treatment duration. SOD activity at 9 h was the highest and roughly paralleled the changes in the H₂O₂ content under Al stress. However, an Al-induced variation tendency in CAT activity occurred with different patterns under Al stress. CAT activity was repressed rapidly at 3 h and was then maintained at a relatively stable state until the end of the Al treatment. Application of exogenous SA further increased the activities of Al-induced POD and APX within 12 h; SOD activity peaked at 6 h, and CAT activity decreased 3 h after Al treatment.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.