Discussion

Combined with our previous and current research, a total of four EGT synthetases were mined in P. eryngii , which were PeEgt1 ,PeEgt2a , PeEgt2b and PeEgt2c . The amino acid sequence of PeEgt1 was significantly similar to EgtD andEgtB in M. smegmatis . In the structural prediction, it also showed that PeEgt1 contained two domains, which were methyltransferase superfamily domain and sulfatase superfamily domain, corresponding to histidine specific methyltransferase and histidine cysteine sulfoxide synthase, respectively. PeEgt2a ,PeEgt2b and PeEgt2c all contain CsdA domain, which are selenocysteine lyases or cysteine desulfurases. According to the current reports on EGT synthetase genes in fungi, EGT synthetase genes can be divided into Egt1 and Egt2 . Egt1 has two domains, methyltransferase and histidine cysteine sulfoxide synthase; Egt2has only one domain, which is a cysteine desulfurase. InFlammulina velutipes , both FvEgt2 and FvEgt3 are cysteine desulfurases, which convert Cys-HER into EGT[27].
EGT synthesis genes mined fromP. eryngii were heterologous expressed in S. cerevisiaeseparately or in combination, and the EGT yield was shown in Figure 4. During the construction of engineered strains, it was found that the engineered strains only integrating PeEgt1 could synthesize EGT, probably because S. cerevisiae contains a cysteine desulfurization enzyme NFS1, which is a PLP-dependent enzyme withEgt2 , and can break the C-S bond of Cys-HER to obtain EGT. However, the engineered strains that only integrated PeEgt2 could not synthesize EGT, possibly because the strain itself could not synthesize the intermediate Cys-HER.PoEgt1 from Pleurotus ostreatus , PtEgt1 fromPleurotus nebrodensis [30], GfEgt1from G. frondosa [28], and EgtA(equivalent to Egt1 ) from Aspergillus fumigatus[33] were respectively introduced into S. cerevisiae , and all of them could synthesize EGT, which confirmed the above speculation. Zhihui Chen et al. [34] found that recombinant E. coli engineered strains containing onlytrEgt1 (derived from Trichoderma reesei ) could also synthesize EGT, probably because cysteine desulfurization enzymes also exist in E. coli . Therefore, it is speculated that only one synthase Egt1 is needed to be introduced into the host containing such proteins to synthesize EGT. PeEgt2a and PeEgt2b were integrated into MX581-PeEgt1 strain respectively, and the yield of EGT increased significantly, indicating that PeEgt2a andPeEgt2b had enzyme activity. It was speculated that Egt2contained at least two or more isoenzymes in the EGT synthesis pathway of P. eryngii . However, the yield of EGT did not increase significantly when PeEgt2c was integrated into MX581-PeEgt1 strain, suggesting that it had no enzyme activity. Finally, two PeEgt2 genes with enzyme activity were simultaneously integrated into the MX581-PeEgt1 strain to obtain a triple-gene engineered strain. The EGT yield was not significantly higher than that of the double-gene engineered strains, suggesting that the substrate or intermediate products may have been consumed; it may also be because there are three metabolic pathways for C-S bond cleavage in the engineered strain at this time, and there is only one synthesis pathway for the intermediate product Cys-HER. The catalytic efficiency of PeEgt1 is far less than the efficiency of the three Egt2 , resulting in insufficient intermediate products.PeEgt1 synthetase may be the key rate-limiting step; it may also be that the excessive production of an intermediate product in the pathway leads to feedback inhibition. Van der Hoek et al.[24] combined and expressed NcEgt1 fromN. crassa and CpEgt2 from Claviceps purpurea inS. cerevisiae , and then overexpressed them. It was found that the EGT yield did not increase when the second CpEgt2 was integrated, but the EGT yield increased when the second NcEgt1 was integrated. According to the research team’s later study, it was speculated that overproduction of SAM became the key rate-limiting step[35]. In Aspergillus oryzae , overexpression of NcEgt1 and NcEgt2 from N. crassa increased EGT production, but HER accumulation was found in the medium, suggesting that the Egt1 catalytic reaction may be a bottleneck[25]. Therefore, if we want to further improve the yield of EGT, we can consider integrating an Egt1 on the basis of triple-gene engineered strain. Inspired by the combined expression, in the later research, the EGT synthases from different edible fungi can also be combined to express, and the higher activity can be screened to construct high-yield EGT engineered strain. In summary, it is speculated that the synthesis pathway of EGT in P. eryngii is shown in Figure 5.
In order to continue to increase the yield, only expressing exogenous genes is often not enough. The fed-batch fermentation or fermentation optimization can effectively increase the yield, for example, fermentation conditions can be optimized: temperature, time, rotational speed, etc.; optimization of medium: carbon and nitrogen source and ratios, addition of elicitors and EGT synthesis precursors. Adding an appropriate amount of histidine could significantly increase the yield of EGT in submerged fermentation of P. eryngii [36, 37]. In the engineered strain of S. cerevisiae , the researchers found that the production of EGT increased when the right amount of histidine, methionine or pyridoxal was added, because pyridoxal is the precursor of PLP, and methionine can be converted into the precursor SAM and cysteine[24].
However, in industry, the supplementation of certain compounds may be expensive, so metabolic pathways of engineered strains can be modified to reduce production costs. van der Hoek et al.[35] selected 28 genetic targets at different levels of amino acid metabolism for experiments, and found that nine of them could significantly increase EGT yield by 10%-51%. In S. cerevisiae , the precursor SAM has a variety of physiological functions such as transsulfuration, transaminopropyl-group and as a methyl donor, and is an essential substance for the growth of yeast. It is mainly synthesized through SAM1 and SAM2. Previous studies have shown that the regulation of SAM2 gene can regulate the SAM content in yeast[38]. Naoyuki Tanaka et al.[39] constructed a high-yield cysteine system inE. coli engineered strain, and the final yield of EGT was as high as 1.31 g/L. Therefore, by modifying the metabolic pathways in S. cerevisiae , enhancing the synthesis of EGT precursors (methionine, histidine, PLP, SAM, etc.) or enhancing the enzyme activity in related metabolic pathways may be a potential method to increase EGT production.