Cadmium is a well known non-essential element that causes oxidative stress in plants. Most attention so far has been paid to the effects of Cd exposure on seedlings or adult plants, and we know little about the oxidative stress and antioxidant mechanisms in germinating seeds. The present study indicated that, challenged by Cd, MDA content and proline level of germinating soybean seeds responded in a exposure time, treatment concentration and exposure time × treatment concentration interacting manner; at Cd200 GST responded to both the exposure time and treatment concentration, while FRAP and GSH responded to exposure time (Table 2). Although soybean seeds germinated under all treatment concentrations, seed germination at higher concentration, i.e., Cd100 and Cd200, was delayed (Yang et al., personal observation). In addition, Cd concentration in germinating seeds increased with increasing exposure time and treatment concentration (Table 1).

FRAP is a semi-quantitative method developed to quantify the enzymatic and non-enzymatic antioxidants present in plants and animals [12,13,14]. An exposure time-dependent increasing in FRAP concentration was observed in our experiment among all treatments, but the increasing tendency was depressed after 48 h of exposure though insignificant; the interaction between exposure time and treatment concentration was not found (Table 2; Figure 1). In germinating mustard seeds exposed to the same Cd concentration, Szőllősi et al., [15] observed that FRAP responded markedly to exposure time, treatment concentration and their interactions. Szőllősi et al. also found that at Cd50, FRAP activities tended to decrease during 12–24 h exposure, and increase later on, though this effect was insignificant [15], but at Cd100 and Cd200, FRAP activities decreased significantly during 12–96 h exposure. The general declining tendency of FRAP’s response to Cd exposure is similar to the results of present study. The difference of sensitivity of FRAP toward Cd exposure may be species-specific.

Malondialdehyde (MDA) is a product of lipid peroxidation and the most honest marker of oxidative stress in plants in response to stress conditions. Profound decrease in lipid peroxidation has been observed in cadmium-treated germinating seeds of Brassica juncea L. [15]. As expected, MDA content increased significantly with increasing Cd concentration and prolonged exposure time (Figure 2; Table 2). The highest MDA content was observed at Cd200 and 96-h exposure (Figure 2), suggesting that higher Cd levels induced lipid peroxidation, resulting in irreversible damage to membrane. A possible reason may be that in the context of higher Cd concentration and longer exposure time, the antioxidant load e.g., proline and GSH/hGSH content, is being depleted (Figure 3 and Figure 4), rendering less antioxidative capacity [16,17,18,19]. Seed coat also functions as a barrier to prevent the entry of Cd into germinating seeds [20]. But with prolonging exposure time and increasing Cd concentration, seeds imbibe Cd-containing solution, expand and break, more Cd enters the seeds, causing cytotoxicity and leading to peroxidation of lipid membrane (Figure 2). Similar findings were also observed in tomato seedlings [21] and a Cd-hyperaccumulator, Solanum nigrum L. [22] challenged by Cd exposure. However, the general changing tendency of MDA in our experiment was increasing with increasing Cd concentration and exposure time, which was opposite to that reported by Szőllősi et al. [15] in germinating Indian mustard seeds. The reasons behind the differences between both research remains to be elucidated.

GSH plays an important role in regulating the antioxidative mechanism and redox homeostasis of plant cells [16,17,20,21]. Homologous GSHs (hGSHs) are usually found alongside GSH in many legumes, which indicate the signaling in optimal and stress conditions [23]. GSH/hGSH content decreased slightly after 48 h exposure, but it declined significantly after 72 h and 96 h exposure (Figure 3). The treatment concentration of Cd affected insignificantly GSH content (Table 2). Similarly, by using germinating pea seeds as study model, Smiri et al. [6] also observed a two-fold lower GSH/hGSH levels in Cd-treated germinating pea seeds. The subsequent depletion of GSH (Figure 3) may be attributed to the involvement of GSH in the synthesis of phytochelatins induced by Cd [16,17] and the formation of Cd-GSH complexes [16]. Challenged by 100 mg/L Cd, the GSH content in germinating mustard seeds increased markedly after 24 h exposure, but decreased afterwards [15]. By using the same Cd concentration, depressed GSH content was observed after 48 h exposure in the present study (Figure 3), and Gallego et al. [24] observed a depletion of GSH pool after 96 h exposure in sunflower seeds treated by 100 and 200 μM Cd as compared to the control seeds.

It has been demonstrated that plants accumulate proline to counteract osmotic stress. Proline accumulation is not only an indicator of abiotic stress, but also an important protecting agent against heavy metal stress [23,24] since exogenously proline can improve the antioxidative capacity and confer cultured tobacco cell’s tolerance to Cd exposure [25]. Free proline has been found to chelate Cd ion in plants by forming non-toxic Cd-proline complex [26]. In addition, accumulating results also suggested that proline accumulation in response to heavy metal may be related to its antioxidative nature [27,28]. An exposure time- and treatment concentration-dependent production of proline was observed in our experiment (Table 2; Figure 4), and there was also exposure time × dosage interaction for proline content (Table 2), suggesting that with increasing Cd concentration and protracted exposure time, the antioxidant mechanisms in germinating seeds were potentiated.

Glutathione-S-transferases can catalyze glutathione-dependent isomerization and the reduction of toxic reactive oxygen species [29] by acting as glutathione peroxidase [30]. Dixit et al. also observed a significant increase of GST activity in pea plants subjected to Cd treatment [31]. The highest increase in GST activity was observed in pea plants after three-day exposure to 40μM Cd, suggesting that GST might have played some role in the metal detoxification process [31,32]. A similar change was also observed in our experiments. After Cd200 treatment, the highest GST activity was observed after 72 h and 96 h exposure (Figure 5). In germinating Indian mustard seeds, however, Szőllősi et al. found an initial increase, but a later on declining trend of GST activity with the same treatment [15]. In Phragmites australis plants treated chronically with 50 μM CdSO₄, enhanced GST activities were found in leaves, roots and stolons as compared to control [33]. Whether Cd-induced increase in GST activity is a detoxification response is the subject of future study.

Soybean (Glycine max. L.) seeds with uniform size were sterilized in 10% NaClO for 10 min and then were germinated in Petri dishes containing vermiculite. The vermiculite was saturated with different concentrations of CdCl₂·2.5H₂O solution prepared with distilled water. Cd concentrations used in this experiment were 0 mg/L, 50 mg/L, 100 mg/L and 200 mg/L, hereafter indicated as Cd0, Cd50, Cd100 and Cd200, respectively. Twenty soybean seeds were germinated for each treatment. The Petri dishes were then kept in dark for 24, 48, 72 and 96 h according to natural temperature rhythm in July in Wuhu, China. There were four replicates for each treatment.

Cadmium content was determined by atomic absorption spectrophotometer (AAS, Shimazadu, AA-6300C, Kyoto, Japan). Dried seeds were homogenized and 0.20 g seed powder was digested with mixture of HClO₄ and HNO₃ (v/v = 1:4) at 200 °C until white smoke disappeared. The digested colorless liquid was washed by 2 mL diluted HNO₃ (distilled water/concentrated HNO₃ = 1:1) and then by distilled water for three times. The liquid collected was then transferred to 25 mL volumetric flasks. Four replicates were taken for each concentration and exposure time. Values of Cd content were expressed as μg/g dry weight.

Germinating seeds (1.0 g) were homogenized with 4 mL cold sodium phosphate butter solution (0.1 M, pH 7.6) containing 0.1 mM EDTA, 0.5% (v/v) Triton X-100 and 0.5% (w/v) PVPP and centrifuged at 10,000 g for 15 min at 4 °C. The supernatant was used for the determination of oxidants and antioxidants as well as proline content. All samples are four replicates.

A simple and quick method called ferric reducing ability of plasma (FRAP) [34] was used for the determination of the antioxidant power of the extracts. FRAP reagent was prepared by mixing 10 mM 2,4,6-tripyridyl triazine (TPTZ; Aladdin, Shanghai, China), 20 mM ferric chloride (FeCl₃·6H₂O; BBI) and sodium acetate (BBI) buffer (pH 3.6) at a ratio of 1:1:10 (v/v/v). Reaction mixture contains 50 μL seed extract and 2.95 mL FRAP reagent and was incubated at 37 °C for five minutes. The absorbance was read spectrophotometrically (P General, TU-180AP, Beijing, China) at 593 nm. The reaction mixture containing 3 mL FRAP serves as control. The total antioxidant capacity was expressed as units of μmol/mL seed extract.

Malonyldialdehyde (MDA) is one of the end products of lip peroxidation and it was determined according to Szőllősi et al. [15] with some modification. Seed extract (150 μL) was mixed with a mixture of 10% (w/v) trichloroacetic acid (Aladdin, AR) and 0.5% (w/v) thiobarbituric acid (BBI, AR). The mixture was heated for 15 min at 96 °C in water bath and then cooled down in cold water to room temperature. The resulting solution was then centrifuged for 5 min at 4,000 g. Absorbance of the supernatant was read at 532 nm and 600 nm, respectively and MDA content was calculated using an extinction coefficient of 155 mM⁻¹ cm⁻¹. The value of MDA was expressed as μmol.g⁻¹ FW.

The non-proteinic thiols (glutathione and functionally homologous glutathione, GSH/hGSH hereafter) was quantified spectrophotometrically by utilizing the Ellman’s regent [35]. Briefly, 250 μL seed extract was added to 3.0 mL reaction mixture containing 2.6 mL sodium phosphate buffer solution (0.1 M, pH 7.6) and 150 μL DTNB (10 mM). The absorbance at 340 nm was read after five-minute incubation at 30 °C. GSH content was expressed as ng·g⁻¹ FW.

Proline content was estimated according to Ringel et al. [36] with some modification. Seed extract (0.5 mL) was mixed with 3% (w/v) aqueous sulfosalicylic acid (BBI, AR), 1 mL acetic acid (BBI, AR) and 2.5% acidic ninhydrin reagent. The reaction mixture was heated in boiling water for 30 min and proline was extracted with toluene. Absorbance of the supernatant was read at 520 nm, and proline content was calibrated by a standard curve. Unit of proline content is μg·g⁻¹ FW.

GST activity was measured following the method of Cui et al. [37] with small modification. Briefly, 10 μL seed extract was mixed with 2.95 mL sodium phosphate buffer (0.1 M, pH 7.4) containing 10 μM reduced glutathione (Sigma-Aldrich, AR). The reaction was started by adding 40 μL 1-chloro-2,4-dinithrobenzene (BBI, AR). The change of absorbance was followed spectrophotometrically at 340 nm for 2.5 min (P-general, TU1810AP, Beijing, China). One unit (U) is the amount of enzyme producing 1 μmol conjugate produced in one minute. The activity was calculated from the extinction coefficient of 9.6 mM⁻¹ cm⁻¹ and expressed as U mg⁻¹ FW.

All data were analyzed using SPSS 15.0 software package. Two-way analysis of variance (ANOVA) was applied to investigate the effects of Cd treatment and exposure time on the parameters examined. Data are expressed as mean values ± standard deviation (SD) and calculated by fresh weight. Cd content in germinating soybean seeds was calculated by dry weight.