3.1 The physiological mechanism of ginger in response to the toxicity of pendimethalin
3.1.1 The effect of pendimethalin on ginger biomass
Pendimethalin has a dose-dependent effect on the growth of ginger. Under PM1 and PM2, there is no significant difference in plant height, stem diameter or rhizome weight compared with CK (Figure 1, P>0.05). As the concentration of pendimethalin increases, the growth of ginger is inhibited. Under PM4, the plant height, stem diameter, and rhizome weight are reduced by 28.73%, 8.78%, and 15.89%, respectively, compared with CK (Figure 1, P<0.05). This indicates that with an increased concentration of pendimethalin, the development of ginger is inhibited and could ultimately affect agricultural yield. However, pendimethalin has no significant effect on the number of shoots. There are known dose-dependent effects of other herbicides on various crops. Stephenson et al. [26] found that S-metolachlor damaged cotton, while pendimethalin had no such effect. Smith [27] found that low-dose pendimethalin had no significant effect on the biomass of Basella alba.
Root length is routinely used as a marker for toxicity of substances [28]. The root length of ginger under PM1 and PM2 is not significantly different from the control (Table 1, P>0.05), similar to the development of the aboveground tissues. Under the two higher doses, toxicity of pendimethalin to ginger root is demonstrated, and the root length decreases under PM3 and PM4. Under PM4, the root length decreases by 17.97% compared with CK (Table 1, P<0.05). A similar study was reported by Smith [29], who found that the root lengths of Corchorus olitorius and Abelmoschus esculentus seedlings treated with pendimethalin were significantly reduced. Similarly, under PM4, the root surface area, diameter, tips, and root weight of treated ginger plants are reduced by 28.68%, 8.69%, 6.91% and 19.17% compared with CK (Table 1, P<0.05), indicating that high-dose pendimethalin causes significant toxicity to ginger roots.
3.1.2 The effect of pendimethalin on the photosynthetic efficiency of ginger
Pendimethalin has an influence on ginger photosynthesis. Figure 2 shows that, except for PM4, pendimethalin has no significant effect on the net photosynthetic rate (Pn) of ginger (P>0.05). The Pn of ginger under PM4 is significantly reduced by 11.37% compared with the control (Figure 2, P<0.05), indicating that the root of ginger is more sensitive to pendimethalin than the leaves. On the other hand, pendimethalin accumulates only in the root, so that the ginger root system is more susceptible to pendimethalin poison than the leaves. Farhoudi and Lee [30] showed that pendimethalin could significantly reduce the photosynthetic rate of sunflower, while Jursík et al. [31] established that pendimethalin had no significant effect on the photosynthetic rate of lettuce. It may be that different crops have different tolerant thresholds for the toxicity of pendimethalin, or the decrease in the photosynthetic rate of ginger caused by pendimethalin is related to the inhibition of root development. In addition, the effects of pendimethalin on ginger transpiration rate, stomatal conductance, and intercellular CO2 concentration follow the same trend as Pn (Figure 2), indicating that pendimethalin causes a stress response. The closure of stomata results in decreased transpiration rate and assimilated substrate (CO2). This presumably leads to a decrease in Pn, mainly due to stomatal limitation. Wang et al. [32] pointed out that under stress, the closure of rice stomata is the main reason for the reduced photosynthetic rate. In addition, Li et al. [33] found that ABA-induced H2O2 production is related to the closure of stomata, which is also related to that PM4 increases the content of H2O2 in ginger leaves (Figure 4).
Chlorophyll fluorescence reflects the photosynthetic efficiency of plants and is measured to determine the degree of damage to plants [34]. The maximum quantum yield Fv/Fm of photosystem II (PSII) is correlated to the degree of damage to plant leaves [35, 36]. As shown by chlorophyll fluorescence images (Figure 3A), pendimethalin causes no obvious damage to ginger leaves. Under PM4 only, the Fv/Fm is 3.32% lower than that of the control (P<0.05), indicating that the critical toxicity concentration value for ginger to pendimethalin is between PM3 and PM4. Li et al. [37] found that pendimethalin did not affect the Fv/Fm of soybeans, but Shabana et al. [38] found that pendimethalin treatment significantly reduced the Fv/Fm of Protosiphon botryoides.
φPSII represents the non-cyclic electron transport efficiency of PSII, and qP reflects the reduction status of QA in the PSII reaction center. The changing trends of φPSII and qP in ginger leaves are similar to Fv/Fm, and under PM4, φPSII and qP decrease by 6.26% and 4.59%, respectively, compared with that of the control CK (Figure 3, P<0.05). The effect of pendimethalin on the PSII of soybeans is consistent with the results of this study [37]. Pendimethalin reduces the photosynthetic efficiency of ginger, and the reduction in linear electron transfer efficiency leads to the formation of ROS by processed light energy, which is related to the higher ROS content of ginger under PM4.
Non-photochemical quenching (NPQ) reflects the degree of heat dissipation of crops. This usually increases under abiotic stress [39]. Figure 3 shows that NPQ exhibits an opposite trend compared to Fv/Fm. Pendimethalin has no significant effect on NPQ of ginger leaves, except for PM4, which significantly increases NPQ by 6.99% compared with the control (P<0.05). Studies have found that herbicides could improve NPQ in plants [40, 41].
3.1.3 Effect of pendimethalin on the antioxidant system of ginger leaves
The formation and elimination of ROS in plants are usually in a balanced state, and the content of ROS increases when exposed to exogenous toxicity [42, 43]. The production of ROS can affect plants by damaging proteins and cells [44], causing membrane peroxidation [45], and affecting various metabolic pathways. Pendimethalin has no significant effect on the O2- content of ginger leaves (P>0.05), except that under PM4, H2O2 and O2- are significantly higher than the control by 27.89% and 19.58% (Figure 4, P<0.05). This indicates that low-dose pendimethalin does not reach the critical toxicity level for inducing ginger response, but under PM4, ginger begins to respond to the toxicity of pendimethalin, and the ROS produced is closely related to the closure of ginger stomata. Similar studies have shown that pendimethalin induced the production of ROS [46, 2].
Plants maintain a balance of ROS through generating antioxidant enzymes including superoxide dismutase (SOD), catalase (CAT) and peroxidase (POD) [47]. SOD, which catalyzes the conversion of free radicals to H2O2, is the first line of defense against ROS. Figure 4 shows that when the concentration of pendimethalin is increased, the SOD activity gradually increases. The accumulation of O2− in ginger is the main reason for increased SOD activity, and ginger can resist oxidative stress caused by antibiotics through increased SOD activity [48]. POD can catalyze the oxidation of substrates by H2O2 [49, 61], and it plays an important role in cell wall biosynthesis, lignification, and other cell functions [50, 62]. This study found that the change of POD is similar to SOD, and it is significantly different from the control under PM2 (P<0.05). It shows that POD responds earlier than SOD in defending the toxicity of pendimethalin. In addition, during the early stages when pendimethalin affects ginger, the increased level of antioxidant enzyme activity could normalize the ROS content. Lehman et al. [59] used ROS as the indicator of cell toxicity. CAT can directly degrade H2O2 to H2O, thereby reducing H2O2 damage [48]. This study found that pendimethalin has no significant effect on CAT activity (P>0.05), which may be related to the insensitivity of CAT to pendimethalin toxicity. It has been found that CAT behaved differently in different plants against external stress [51, 65, 66].
3.1.4 The effect of pendimethalin on root activity, MDA and relative conductivity
Root activity is the expression of plant absorption capacity. Under PM1 and PM2, no significant difference is found in root activity between the treatment and the control (P>0.05). Under PM3 and PM4, the root activity is significantly reduced, that of PM4 is 22.62 % lower than the control (Figure 5, P<0.05). A similar study was reported by Song et al. [52], who found that 742 g/hm2 pendimethalin had no significant effect on the root activity of cotton, while 1113 g/hm2 pendimethalin significantly inhibited its root activity, destroying the membrane structure.
Malondialdehyde (MDA) content reflects the degree of cell membrane peroxidation, and electrical conductivity reflects the degree of ion release after the cell membrane is damaged. Generally, when the plant is under external stress or toxicity, MDA content and conductivity will increase [53, 63]. In this study, pendimethalin has no significant effect on MDA and relative conductivity of ginger (P > 0.05). It has been demonstrated that herbicides can increase MDA and relative conductivity of plants [24, 54, 64], mainly because that herbicides damage plants. However, in this study, the critical value of pendimethalin toxicity to ginger is around PM3, and the effects of the toxicity have just begun to manifest. In order to resist the toxicity of pendimethalin, ginger reduces the excessive production of ROS by increasing the activity of antioxidant enzymes. The ROS content in this experiment is in a dynamic balance, the plasma membrane of ginger is not damaged, and the relative conductivity does not change significantly.
3.1.5 Correlation analysis shows that root development and the antioxidant system could be used as a sensitive indicator of the toxicity of pendimethalin in ginger.
First, the gradient length of each axis was estimated by detrended correspondence analysis (DCA), and then the canonical correspondence analysis (CCA) or redundancy analysis (RDA) was further selected. According to the principle of gradient length, >4.0 corresponded to CCA; 3.0 < gradient length < 4.0 corresponded to RDA or CCA; and gradient length < 3.0 corresponded to RDA. Based on the results of DCA analysis (maximum gradient length was 0.064), RDA was selected in this study. RDA analysis shows that the growth indexes (plant height, stem diameter and weight) of ginger are positively correlated with root structure (root length, root weight, surface area and tips) and photosynthetic indexes (Pn, Ci, E, Gs), negatively correlate with ROS, antioxidant enzyme (SOD, POD), MDA and relative conductivity, and had no significant correlation with CAT (Figure 6). It has been shown that root length can be used as a performance indicator of plant toxicity [28]. In this study, root length is also the most relevant indicator with among all ginger root growth configurations. In addition, the SOD and POD of ginger are more sensitive to the toxicity of pendimethalin, which is the same as the results of this section. Therefore, the degree of ginger poisoning by pendimethalin can be assessed by the root development and the activities of SOD and POD of ginger.
3.2 Accumulation of pendimethalin in ginger and its optimal selection for weeding in ginger fields
Detection of pendimethalin residues in ginger roots, stems, leaves and rhizomes reveal that pendimethalin only accumulates in the roots. Jursík et al. [55] found that the residue of pendimethalin in lettuce increased with higher concentrations. This study has similar results. Pendimethalin is not detectable in the roots of PM1. When the concentration of pendimethalin is increased, the accumulation of pendimethalin in ginger roots gradually increases, and the highest residue identified is under PM4 (289.22 μg Kg-1) (Figure 7). The octanol partition coefficient (LogKow = 5.18) of pendimethalin is higher, indicating that its water solubility is low, and it is not easily transferred from the root to the stem and leaves through passive transport. The accumulation of pendimethalin is mainly in the root. European Medicines Agency (EMA) defines the critical value of LogKow ≥ 4.5 for the phenomena of persistence, bioaccumulation, and the analysis of toxicity [56]. However, within a proper concentration range, there is no accumulation of pendimethalin in the edible organs (rhizome) of ginger. Therefore, the use of pendimethalin for weeding in ginger fields may have no health risk to humans.
There are significant differences in the weed removal effect of different concentrations of pendimethalin. Table 2 shows that there are no significant differences in herbicidal effects between different concentrations of pendimethalin at 5 days of treatment (P>0.05). After 10 days, there is no significant difference between PM2, PM3 and PM4 (P>0.05), but PM2 is significantly higher than PM1 (P<0.05). After 30 days, there is no significant difference in weed removal rate between PM2 and PM3 (P>0.05), but PM2 is significantly higher than PM1 and lower than PM4 (P<0.05). In view of the effect of pendimethalin on the growth of ginger, PM2 can be used as the optimal concentration of ginger for pre-emergent weed prevention, which has no significant effect on the development of ginger.
3.3 Bioinformatic analysis of ginger α-tubulin and its response to pendimethalin
Musa acuminata is a plant of Zingiberales and is the closest related species to ginger in the known genome database. We downloaded all α-tubulin gene sequences of Musa acuminata from NCBI, and then selected the sequences with P<10-50 by local Blast from ginger transcriptome data (unpublished). Through conservative domain prediction, it is found that the selected genes have a predicted typical α-tubulin structure (Figure S1). Also, the α-tubulin gene of Arabidopsis thaliana, Musa acuminata and the selected ginger genes were examined by phylogenetic tree analysis. It was found that the genes were divided into 7 categories. CL17489.Contig1 was classified as Class I, with high homology to AtTUA2; Unigene28871 was classified as Class III; Unigene1894, Unigene39213, CL17215.Contig2 and Unigene33980 were classified as Class Ⅳ, there was no predicted homologous α-tubulin gene of Arabidopsis thaliana and Musa acuminata; Unigene38694 and Unigene39214 are classified as Class VI, with higher homology to AtTUA1; CL7006.Contig1 is classified as Class Ⅶ with high homology to α-tubulin3 of Musa Acuminata (Figure S2).
Comparative analysis of nine selected ginger α-tubulin and AtTUA1 sequences (Table S3) found that AtTUA1 codes for a predicted 450 amino acid protein. From the nine predicted genes, α-tubulin amino acids in ginger are 188-448, the least is CL17215.Contig2, and the most is Unigene38694. The change in the trend of molecular weights is identical to the number of amino acids. The isoelectric point of AtTUA1 is 4.92, and the isoelectric points of the nine predicted ginger α-tubulin proteins are between 4.69 and 6.00.
Aliphatic index is often considered a measure of thermal stability of a protein [57], in this study, while the molecular weight of the protein increases, the aliphatic index tends to decrease. GRAVY indicates the hydrophilicity of a protein; a negative value indicates that the protein is a hydrophilic protein and a positive value indicates a hydrophobic protein. This study found that the predicted GRAVY of AtTUA1 is -0.194, and among the nine ginger α-tubulin sequences screened, eight had a negative predicted GRAVY, ranging from -0.204 to -0.028; only Unigene39213 has a predicted positive GRAVY (0.021).
Instability index indicates the stability of the protein; generally, the protein is predicted to be stable when the value is less than 40 and may be unstable when it is greater than 40 [58]. In this study, AtTUA1 has a predicted instability index of 40.92, which is an unstable protein. Unigene28871 (41.37), CL7006.Contig1 (42.43) and Unigene33980 (45.84) were all predicted to be unstable proteins; and Unigene1894 (34.40), Unigene39213 (33.24), CL17489.Contig1 (34.19), CL17215.Contig2 (23.14), Unigene38694 (38.26), and Unigene39214 (37.07) were all predicted as stable proteins.
Analysis of protein transmembrane structure (Figure S3) found that there is one predicted transmembrane structure in AtTUA1, same as Unigene38694. Unigene28871 and Unigene33980 had no predicted transmembrane structure; the remaining ginger with predicted α-tubulin protein had one predicted transmembrane structure. The predicted secondary structure analysis (Table S4) of ginger α-tubulin and AtTUA1 protein found that only alpha helix, extended strand and random coil existed, and random coil accounts for the highest proportion, followed by alpha helix, with extended strand the least. Saboury et al. [60] found that alpha helix was positively correlated with protein stability. The alpha helix of the 9 predicted ginger α-tubulin proteins screened in this study are higher than those of Arabidopsis, indicating that ginger α-tubulin has a higher stability, and at the same time providing evidence that ginger has a high resistance to pendimethalin.
According to the evolutionary relationship of ginger α-tubulin, one gene in each Class was selected for qRT-PCR analysis. Except for Unigene39213, pendimethalin has no significant effect on the transcription of α-tubulin mRNA in ginger (Figure 8, P>0.05), indicating that the effect of pendimethalin on ginger development is not caused by the different levels of α-tubulin. In addition, the determination of α-tubulin content in the root shows that pendimethalin has no significant effect (Figure 9, P>0.05), further indicating that the toxicity of pendimethalin to ginger is mainly due to oxidative stress, but not the effect of α-tubulin.