DISCUSSION
Knock-down of γECS resulted in a drastic limitation of biothiol concentrations under control and Hg-stress conditions, particularly in the cad2-1 and pad2-1 mutants (Fig. 1), confirming previous results in these GSH-depleted genotypes (Parisy et al., 2007; Ball et al., 2004; Cobbett, 2000) and in Arabidopsis leaf discs and plants treated with similar doses of Hg and Cd (Sobrino-Plata et al., 2014a; 2014b). Interestingly, the mildly-affected knock-downrax1-1 γECS mutant treated with 3 µM Hg also accumulated PC2 and PC3 (but not PC4) in roots, but to a lower extent than did wild-type plants as observed in Cd-treated plants (Sobrino-Plata et al., 2014b). We were unable to detect PCs in shoots, organs that accumulated much less Hg than roots (by two orders of magnitude), since a certain Hg concentration threshold may be required to trigger synthesis of PCs. In fact, numerous PCs appeared in Col-0 and rax1-1 leaf discs subjected to direct infiltration with 3 and 30 µM Hg; behaviour that was accentuated at longer exposure times (48 h) when PC2 and PC3 also appeared in cad2-1 (Sobrino-Plata et al., 2014a). On the other hand, the inability to synthesize PCs incad1-3 led to a significantly higher GSH accumulation in when compared to Col-0 (Fig. 1, Table 1). Interestingly, this increase became larger under Hg stress, in agreement with our previous observations incad1-3 leaf discs infiltrated with 3 µM Hg for 24 h (Sobrino-Plata et al., 2014a). It has been proposed recently that PCS functions as a transpeptidase important for GSH and conjugated GSH turnover, which may explain the high GSH levels found in cad1-3plants (Kühnlenz, Westphal, Schmidt, Scheel, & Clemens, 2015).
Depletion of GSH resulted in the elicitation of a severe oxidative stress with 3 µM Hg, with a marked inhibition of GR activity in the roots of γECS mutants in comparison with Col-0 and cad1-3 , without any changes in enzyme amount (Fig. 3). The mutant cad1-3lacked the ability to form Hg-PC complexes, but the Hg-induced damage was similar to that found in Col-0, possibly due to enhanced GSH levels in this mutant. Strong GR inhibition occurred in roots of cad2-1 ,pad2-1 and rax1-1 treated with 10 µM Hg for 72 h, which also suffered extensive alterations in membrane proteins (i.e., degradation of H+-ATPase and strong inhibition of NADPH-oxidase; Sobrino-Plata et al., 2014b). The GR inhibition appears to be triggered specifically by Hg over certain concentrations inMedicago sativa or Silene vulgaris , whereas other toxic elements usually lead to an enhanced activity (Sobrino-Plata et al., 2013; 2009), as can be used as a marker of Hg-stress. Besides the strong GR inhibition, there were minor and non-consistent changes in the proportion of GSSG in the analysed Arabidopsis genotypes (Table 1). This concurs with the minimal oxidation of homoglutathione (hGSH) (less than 15%) found in 30 µM Hg-treated alfalfa seedlings (Ortega-Villasante et al., 2007). Therefore, even though GSH synthesis was compromised in γECS mutants, much severe and chronic cellular damage would be required to observe relevant GSH oxidation. On the other hand, the poorer tolerance to Hg caused by limited GSH also led to alterations of chlorophyll fluorescence parameters, with a remarkable NPQ decrease (Fig. 2), in accordance with results obtained in Arabidopsistreated with Hg, Cd and Cu over 72 h (Maksymiec, Wójcik, & Krupa, 2007; Sobrino-Plata et al., 2014a). GSH plays a central role in chloroplast redox balance, keeping ASA and xanthophyll pools reduced at optimal levels to sustain NPQ under stress (Yin et al., 2010), which may be hampered in γECS mutants. Interestingly, the increase in ASA shoot concentrations under Hg stress was particularly intense in γECS mutant genotypes. Similar response was found in Cd-treated cad2-1mutants, where ASA concentration was higher than in wild-type plants, particularly in roots (Jozefczak et al., 2015). In this respect, recent experiments showed that increases in ASA concentrations are a common response of plants to metal stress, especially in shoots where this antioxidant metabolite helps protecting the photosynthetic apparatus, which may be hampered by both the lack of GSH and the oxidative stress induced by Hg (Bielen, Remans, Vangronsveld, & Cuypers, 2013).
Mercury is thought to bind strongly to cell walls of epidermal and xylem root cells, possibly bound to the Cys thiol residues of proteins, thus preventing translocation to shoots (Carrasco-Gil et al., 2011; 2013), as found in roots of different plant species (Carrasco-Gil et al., 2011; Sobrino-Plata et al., 2009; 2013; 2014b). Interestingly, γECS mutants roots had significant lower Hg concentration than Col-0, with no effects in shoots, whereas stronger Hg-induced damages appeared in the mutants. Similarly, metal accumulation in shoots did not change in Cd- and Hg-treated cad2-1 plants (Li, Dankher, Carreira, Smith, & Meagher, 2006), in line with the view that cellular biothiol levels have little impact on overall plant metal distribution (Lee et al., 2003). On the other hand, it is known that transpiration is strongly impaired by Hg (Moreno, Anderson, Stewart, & Robinson, 2008), a toxic metal that drastically reduces metabolic-driven water conductance in roots (Lovisolo, Tramontini, Flexas, & Schubert, 2008). Toxic effect that impelled us to use the Schölander pressure chamber to collect enough xylem sap under Hg stress, particularly in γECS mutant plants. Therefore, it is feasible that the strong Hg-stress in γECS mutants caused poorer water flow to shoots, limiting Hg uptake and translocation to the aerial part of Hg-exposed plants.
Xylem conforms, along with phloem, the major long-distance transport system for movement and distribution of water, ions and metals throughout the plant (Álvarez-Fernández et al., 2014). Cadmium transport by the xylem determines Cd accumulation in shoots, which depends on loading driven by metal transporters (Wu et al., 2015), while biothiols have been suggested as long distance carriers for Cd in the phloem ofBrassica napus (Mendoza-Cózatl et al., 2008). The high stability of Hg-PC complexes found in plant roots could provide a basis for Hg long-distance transport, as it was suggested by the association of Hg with sulphur in stems and leaf veins of alfalfa plants exposed to Hg (Carrasco-Gil et al., 2013). HPLC-ESI-MS(TOF) analysis revealed for the first time that [HgPC2-H]+ indeed occurs in the xylem sap of Col-0 (Fig. 5), identity that was confirmed by MSn analysis, with daughter molecular ions in the MS2 and MS3 spectra matching those of standards. We also detected free [PC2-H] and [PC2oxd-H] in xylem sap, confirming our preliminary findings in the xylem sap of Col-0 plants treated with 10 µM Cd for 72 h (Supplementary Fig. 2). Oxidised PC2 was also found in the xylem sap of Brassica napus plants subjected to Cd (Mendoza-Cózatl et al., 2008) andArabidopsis seedlings treated with As (Liu et al., 2010), but metal(loid)-PC complexes were not found in those cases. Moreover, a very low concentration of As was found in xylem sap of the metallophyte castor bean, which was accompanied again with oxidised GSH and PC2 (Ye et al., 2010), probably as a result of the oxidative stress and redox imbalance triggered by metal(loid)s. As(III)- and Cd-biothiols complexes may be less stable than those formed with Hg in our conditions, able to withstand even acidic extraction.
Plants treated with metals experience alterations in sulphate uptake and assimilation (Na & Salt, 2011; Nocito et al., 2006), which prompted us to analyse the expression of twenty genes involved in the sulphur assimilatory pathway under Hg-stress. Our results revealed in allA. thaliana genotypes tested different responses to Hg in roots and shoots, indicating that both organs had independent stress responses as found with other metals (Jozefczak et al., 2014). In general, we observed a modest response of genes with fold-changes generally not larger than three (significant at p < 0.05), following the same pattern of recent transcriptomic analyses performed after short-term Hg treatments in Medicago (Montero-Palmero et al., 2013; Zhou et al., 2013), barley (Lopes et al., 2013), rice (Chen et al., 2014) and tomato (Hou, Liu, Wang, Zhao, & Cui, 2015).
With regard to sulphur metabolism regulation, several transcription factors have been reported to be overexpressed under S-starvation, such as the central hub SLIM1 regulator and several R2R3-MYBs, including MYB28 and MYB51 (Frerigmann & Gigolashvili, 2014). However, we only observed MYB28 upregulation in cad2-1 and pad2-1 shoots under Hg stress. Incidentally, a rice R2R3-MYB (OsARM1) has been found to be upregulated in stems and leaves upon As exposure (Wang et al., 2017), and several R2R3-MYBs control response to Cd-stress viaABA signalling (Zhang et al., 2019). However, we found marked MYB28, MYB51 and SLIM1 down-regulation in roots of Hg-stressed γECS andcad1-3 Arabidopsis mutants, which can likely explain the low expression of several sulphur assimilatory pathway genes. Little is known about how SLIM1 may operate under abiotic stress, which may undergo post-transcriptional redox imbalance regulation occurring in Hg-treated γECS mutants (Koprivova & Kopriva, 2014).
Sulphate uptake is a bottleneck in plant sulphur incorporation, which were upregulated under metal stress, such as SULTR1;1 in roots of maize (Nocito et al., 2006) and Arabidopsis (Ferri et al., 2017). However, other members of the SULTR transporter gene family in Chinese cabbage plantlets and sorghum responded in different manner in leaves and roots under metal stresses (Shahbaz et al., 2014; Akbudak, Filiz, & Kontbay, 2018). We found that sulphate transporter SULTR1;2 was up-regulated in shoots in Arabidopsis γECS and PCS mutants under Hg-stress, response was also found for SULTR3;5 in roots ofMedicago just after 6 h exposure to 3 µM Hg (Montero-Palmero et al., 2013). Conversely, SULTR1;2 was down-regulated in shoots of Col-0 and roots of all Arabidopsis mutants, following the same pattern of SULTR2;1 and SULTR3;5 (Figs. 6, 7), in agreement with the short-term down regulation of SULTR3;3 in rice seedlings treated with 25 µM Hg for 3 h (Chen et al., 2014). Cadmium exposure and sulphate limitation revealed differences in the transcriptional control of three sulphate transporter (SULTR1;2 ) genes in Brassica juncea (Lancilli et al., 2014). Similarly,SULTR1 and SULTR2 expression decreased in roots and shoots of Cd-treated Arabidopsis at high Cd doses (over 40 µM) (Yamaguchi et al., 2016). Therefore, SULTR expression under metal stress changed depending on the plant organ, supplied metal and doses, implying a complex regulation and specific responses. Time-course experiments to monitor the metal induced expression of SULTR1;2showed that in roots it peaked a few hours after metal exposure but subsided subsequently (Jobe et al., 2012). It is feasible that the GSH depletion promoted SULTR1;2 expression in shoots under Hg stress, where we observed significant redox alterations, whereas under acute cellular damage there might be a general transcriptional down-regulation in roots (Montero-Palmero et al., 2013).
APRs are key enzymes of sulphur assimilatory pathway, that produce sulphite from adenosine 5′ phosphosulphate (Kopriva, 2006), genes that were up-regulated in Arabidopsis γECS and PCS mutants shoots treated with Hg, in agreement with the overexpression found in short-term Hg-treated Medicago (Montero-Palmero et al., 2013). However, the rest of S-assimilatory pathway genes in shoots and roots of γECS and PCS mutants were modestly affected or down-regulated by Hg (Figs. 6, 7). It must be emphasized that until now none of the transcriptomic analyses carried out in plants treated with Hg showed significant changes in gene expression of enzymes involved in Cys, γEC, GSH or PCs synthesis (Chen et al., 2014; Hou et al., 2015; Lopes et al., 2013; Montero-Palmero et al., 2013; Zhou et al., 2013). In consequence, despite the several significant changes in S-assimilatory gene expression, occurring mainly in GSH deprived plants, we cannot rule out that the process can be post-transcriptionally controlled. Several stress hormones and the redox cellular balance can contribute to altered enzymatic activities that modify biothiol pools (Kopriva et al., 2019); mechanisms that should be the matter of future research.
In conclusion, depletion of GSH led to stronger Hg toxicity visualised by strong inhibition of GR activity, a poor accumulation of Hg-PC complexes and a limited translocation of HgPC2 to shoots via xylem transport. Sulphur metabolism and accumulation of biothiols help withstanding Hg-induced oxidative stress, but the mechanisms of regulation remain to be characterised in detail. Although some responses at the transcriptional level were detected, we cannot rule out post-transcriptional regulation, which probably play a relevant role to procure sufficient biothiols to limit Hg induced damage. In this sense, transcriptional sulphur-assimilation regulation could be independent of GSH cellular levels, in spite of being an essential factor to maintain the cellular redox balance that was compromised by Hg.