Discussion
2,4-DNT is an important intermediate to produce polyurethane foams and dyes. It is also found in the environment as a mixture with TNT since 2,4-DNT is a byproduct of TNT biosynthesis. Although DNT and TNT are broadly used in civilian and military activities, both are highly toxic xenobiotics that threaten human health and environmental safety. Phytoremediation is a low-cost technique to clean up these nitroaromatic compounds (Doty, 2008). Previous studies have suggested that engineering bacterial NfsI in plants can detoxify TNT efficiently (Van Dillewijn et al., 2008; Zhang et al., 2017; Hannink et al., 2001). However, it remains unknown if this strategy could be employed for detoxification of DNT and its derivatives. It had been shown thatEnterobater cloacae NfsI, a member of the group B type I NRs, can convert both TNT and DNT in vitro . A recent study also indicates that the cleanup rate of N. tabacum plants for 2,4-DNT is significantly increased after inoculation with 2,4-DNT degrading P. putida strains (Akkaya, 2020). These works shed light on the effectiveness of genetically modifying plants for DNT remediation. Switchgrass is a perennial tall grass with great potential for phytoremediation of nitroaromatic compounds (Zhang et al., 2019). Unfortunately, overexpression of Enterobater cloacae NfsIin switchgrass plants did not improve their resistance to TNT. Thus, challenges remain for engineeringNfsI in switchgrass for the detoxification of DNT and its derivatives.
It has been suggested that the expression levels of genes driven by 35S promoter in switchgrass are lower than maize ubiquitin (ZmUbq) promoter (Mann et al., 2011). Therefore, we re-overexpressedNfsI in switchgrass under the control of ZmUbq promoter to increase NfsI transcript abundance. Moreover, we employed a codon optimization technique to improve transcription and translation efficiency of NfsI in switchgrass since NfsI is a bacterial derived NR . Finally, we generated NfsI overexpressing transgenic switchgrass plants and improved their tolerance and detoxification to 2,4-DNT. The uptake rate of 2,4-DNT increased from ~56% in the control plants to ~94% in transgenic lines NfsI_OE-02 and -14 (i.e., a relative increase of 67%). Compared with 2,4-DNT, NfsI exhibited little conversion capacity to DNTS neither in vitro norin vivo , implying that sulfonate has a crucial effect on nitro reduction of 2,4-DNT. Therefore, the enzyme specificity to nitroaromatic compounds should be fully considered before engineering NRs in plants for detoxification of these highly toxic pollutants.
NfsI reduces TNT to 2-hydroxyl-4,6-dinitrotoluene (HADNT) through two-electron reduction and then transforms HADNT to 2-amino-4,6-dinitrotoluene (ADNT). The reduction pathway of 2,4-DNT is similar to that of TNT, and both HAMNT and AMNT were produced during 2,4-DNT reduction catalyzed by NfsI. Most strikingly, a novel compound that likely formed by an autonomous polymerization was detected by LC-PDA/ESI-MS/MS, suggesting HAMNT and its derivatives are the greatest sources of 2,4-DNT reduction. Compared with 2,4-DNT, DNTS were only transformed to HAMNTS by NfsI in vitro . Furthermore, the enzyme kinetics analysis indicates that NfsI has much higher affinity and catalytic efficiency for 2,4-DNT than DNTS. Given the fact that DNTS is a sulfonate of 2,4-DNT, we speculated that the presence of the sulfonate may block NfsI from binding to its substrate. Although the crystal structures of many bacterial NRs have been determined (Chauviac et al., 2012), the selectivity mechanism of NRs for their nitrated substrates warrants further investigation.
Overexpressing NfsI in switchgrass alleviated the impact of 2,4-DNT on root length and ROS production significantly. This is similar with the previous observation that overexpressing NfsI in Arabidopsis , tobacco, and poplar can partially restore the inhibition of TNT on root elongation and reduce ROS accumulation (Van Dillewijn et al., 2008; Hannink et al., 2001). Although the morphological and physiological responses of genetically modifying plants to nitroaromatic pollutants have been assessed widely, there are few studies involved in their transcriptome responses. One study indicated that TNT treatment can induce more than 500 genes that are differently expressed (at least 5-fold change) in wild type Arabidopsis roots. Serial analysis of gene expression further reveals multiphase mechanisms of TNT detoxification inArabidopsis including oxidative and reductive processes, conjugation reactions, and sequestering within the vacuole and/or cell wall (Reference, Ekman 2003). Genes involved in these detoxification processes were also differentially expressed in our control switchgrass treated with 2,4-DNT. It suggests that other plant species might employ a similar mechanism for detoxification of nitroaromatic pollutants. Furthermore, we found that overexpression of NfsI in switchgrass was able to restore approximately 22.9% of DEGs induced by 2,4-DNT treatment by different extents. Among them, genes involved in reactions of oxidation and reduction, conjugation, and sequestering are worthy of functional characterization in the future.
In conclusion, the bacterial NfsI was successfully overexpressed in switchgrass, which is a multiple purpose crop used for forage and biofuel production as well as phytoremediation. Overexpression of NfsI in switchgrass can remove 94.1% of 2,4-DNT from liquid culture medium after 6 days which is 1.7-fold higher than that of control plants. Moreover, our transcriptome analysis suggests that approximately 22.9% of switchgrass DEGs induced by 2,4-DNT might be involved in NfsI-mediated detoxification in switchgrass. Therefore, these genes are potential candidates for deciphering molecular mechanisms underlying switchgrass responses to nitroaromatic compounds. Further investigation will lead to more novel targets being discovered and engineered to improve switchgrass tolerance and detoxification to nitroaromatic pollutants in the future.