2.1 SA application enhances the tolerance to Cd in rice
Preliminary experiments showed that SA played an important role at the reproductive growth period (data not shown). Therefore, our experiments were conducted mainly at the reproductive growth stages (heading stage, flowering stage and mature stage). Cd stress obviously reduced seedlings growth, which mainly reflected in the following aspects: root length, plant height, shoot Dw plant-1, and1000-grain weight. SA application could relieve the Cd toxicity obviously, increased the plant height and shoot Dw plant-1 evidently of the Cd-exposed rice plants, but there was no significant difference between Cd and AOPP+Cd groups (Table 2 and 3), the same trends were observed at all the three stages of rice plants. In addition, application of SA or AOPP to non-stressed seedlings did not result in any difference in agronomic traits.
2.2 SA application alleviates the Cd accumulation in rice grain by the reduction of Cd content in leaves at flowering stage
2.2.1 SA application changed the physical and chemical properties the soil
From the analysis of the effects of SA and AOPP application on the total Cd content, available Cd content and pH value of the soil, we found the total Cd and the available Cd contents of the soil were decreased continuously during the growth of rice. And the exogenous SA or AOPP treatment did not affect the total Cd and the available Cd contents at the four different stages obviously except there was a significant difference between SA+Cd group and AOPP+Cd group at the mature stage. Furthermore, the pH value of the Soil also was not changed after the treatment with exogenous SA or AOPP (data not shown).
2.2.1 SA application induces low Cd accumulation in rice grain
Cd accumulation of rice grain (0.29mg/kg) in Cd-exposed rice plants was significantly greater than the regulatory Cd limit (0.2mg/kg), but SA application could reduced the accumulation of Cd to 0.12mg/kg. Meanwhile, the AOPP+Cd group increased Cd accumulation (0.34mg/kg) significantly which was higher than that of the Cd group (Figure 1).
2.2.3 SA application decreases the Cd contents in other parts of rice
The rice plants accumulated significant amount of Cd in different parts during Cd stress, and the roots accumulated the most Cd by nearly 80%. Compared with the Cd-exposed rice plants, application of SA could reduce the Cd accumulation of leaves and flag leaves remarkable, especially the leaves at flowering stage (around 45% reduction), resulted in around 53% reduction of Cd accumulation in panicle. However, SA application had no significant impact on Cd accumulation in roots, stems and nodes. Furthermore, AOPP application significantly enhanced the total Cd accumulation in panicle, and the main reason was that the Cd contents in leaves at flowering stage were increased (Figure 2). Therefore, we indicated the main reason of SA treatment could effectively reduce Cd content in brown rice and panicles was SA could significantly reduce Cd contents in rice leaves at flowering stage.
Different treatments could affect the mobility of Cd in rice. From the analysis of effect of the three groups (Cd, SA+Cd and AOPP+Cd) on the Cd mobility in rice, we only found the Cd mobility at flowering stage (stem-to-leaf and leaf-to-panicle) and mature stage (leaf-to-panicle) were shown obvious significance (Table 4). SA application could markedly reduce the Cd mobility, as well as AOPP application could markedly increase the Cd mobility in stem-to-leaf at flowering stage. Furthermore, the different Cd mobility resulted in the different Cd accumulation in rice panicles, SA application both reduced the Cd mobility in leaf-to-panicle approximately both 41% at flowering stage and mature stage, and AOPP application increased the Cd mobility in leaf-to-panicle about 47% at flowering stage and 29% at mature stage separately. Therefore, leaves at flowering stage were the critical parts for SA regulate Cd accumulation in rice grain.
2.3 SA application boosts endogenous SA accumulation and activates the SA-signaling pathway
2.3.1 Application of SA boosts endogenous SA accumulatin in rice
As leaves at the flowering stage were the key parts for SA regulating Cd accumulation in rice grain, we determined the content of SA in rice leaves at flowering stage. SA and AOPP applications both had a significant impact on SA contents in leaves. Without Cd stress, SA treatment could obviously increase the SA contents in rice and AOPP treatments could obviously reduce the SA contents in rice. During Cd stress, SA application showed a time dependent increase of the SA contents in leaves, reached the maximum on the 3rd day by 94% than control, and then began to decrease on the 5th day; AOPP treatment reduced by about 80%than control (Figure 3).
Moreover, the differently biosynthesis of SAin different plant species were contribute to the activities of PAL and ICS enzymes (Xu et al. 2017).We found Cd stress could decrease the expression of PAL enzyme, and SA application could induce higher PAL expression (Figure 4). The expression of OsPAL in SA group were higher than that of CK group, it could arrived at more than 2 times at 6 hours after SA application at the flowering stage, as well as the expression level of OsPAL in SA+Cd group were more than 2.5 times than that of Cd group at 6 hours. In contrast to the changes of OsPAL expression level, there were no significance of the expression levels of OsICS between CK and Cd group, but the expression levels of OsICS in SA (or SA+Cd) group could increase by more than 4 times than that of WT (or Cd) group, respectively.
2.3.2 SA application regulates the expression of key genes in SA-signaling pathway
It was generally known that SA signaling in rice was mediated by NPR1 and the WRKY sub-pathway (Xu et al. 2017; Nakayama et al. 2013). We found the expression of OsNPR1, OsWRKY45, OsWRKY7, and OsWRKY70 were all modulated by Cd stress, and the SA application could increase the expression of OsNPR1 and OsWRKY45 obviously (Figure 5). The trends of the OsWRKY45expressionwhich represented the SA level in rice in different experimental groups were consistent with the SA levels (Amit et al., 2015; Shimono et al., 2007; Ryu et al., 2006), these results suggest that SA enhancedOsWRKY45expression resulted in an raise of endogenous SA content. OsWRKY7 and OsWRKY70 were also found related with the SA level in rice seedlings (Xu et al., 2017). Our results indicated that the expression levels of OsWRKY7 and OsWRKY70 were negatively correlated with the SA contents in rice.
2.4 SA application sustains a high level of H2O2and boostsantioxidant enzymes’ activities under Cd stress
In order to examine whether SA induces H2O2 production in rice leaves during Cd stress, the contents of H2O2 and enzyme activities of CAT, POD, and APX in the rice leaves at the flowering stage were determined. Results showed that Cd stress could result in a high H2O2 accumulation. The H2O2 content in SA group was reduced in a short period of time after SA application, the H2O2 content was lowest on the 1st day, then increased continuously, and returned to the same level as the control on the 5th day, similar trend was observed in SA+Cd group. Furthermore, the H2O2content in SA+Cd group was higher than that of the Cd group temporary, but after SA application it decreased with time to the lowest level on the 3rd day, and then increased to the level similar with the Cd group on the 5th day (Figure 6).
Furthermore, compared with the Cd group, the activities of antioxidant enzymes (such as SOD, POD and CAT) in SA group and Cd+SA group showed a similar enhancement moderated by SA treatment (Figure 6). It was obvious that SA application on rice leaves during Cd stress could increase the H2O2 content in a short-term and improve the Cd tolerance of rice.
2.5 SA application modulates the expression level of genes related to Cd accumulation
Two genes related to Cd accumulation of rice grain in rice were detected in this experiment. OsLCT1which was reported could mediate Cd transport in the phloem of rice (Uraguchi et al., 2014). As shown in figure 7, Cd group could decrease the expression level of OsLCT1 by around 80% than that of the control, and SA group enhanced the OsLCT1 expression significantly in comparison to control. Compared with the Cd group, SA+Cd group could obviously enhance the OsLCT1 expression though the level was lower than the control.
OsLCD was found related to the Cd tolerance and accumulation (Shimo et al., 2011), there was no significant change of the OsLCD expression in Cd group in comparison to the control, except for SA application (SA group or SA+Cd group) (Figure 7). SA group and SA+Cd group could increase the OsLCD expression in a short time, and then quickly dropped to its original level.