Construction of plasmids and strains
Construction of argA plasmids
To maximize the arginine production, plasmids pTYT-argA and pTYT-argA^fbr are constructed to express argA and argA^fbr respectively under the control of promoter Ptac which can be induced by IPTG by GoldenGate assembly (fig.7 A). Plasmids pTYT-argA and pTYT-argAfbr were successfully constructed according to the PCR verification (fig. 8).
Construction of cysE plasmids
In order to maximize the cysteine production, plasmids pTYT-cysE, pTYT-cysE-256, pTYT-cysE-5, pTYT-cysE-11-2, pTYT-cysE-5-11-2, pTYT-cysE-256-5, pTYT-cysE-256-11-2, pTYT-cysE-256-5-11-2 were constructed to express cysE genes with mutants at position 256, 5 and 11-2 respectively under the control of IPTG-induced promoter Ptac by GoldenGate assembly (fig.7 B). Plasmids pTYT-cysE, pTYT-cysE-256, pTYT-cysE-5, pTYT-cysE-11-2, pTYT-cysE-5-11-2, pTYT-cysE-256-5, pTYT-cysE-256-5-11-2 were verified by PCR as successful constructions (fig. 8), except pTYT-cysE-256-11-2, which was failed to be constructed after several trials.
Construction of myrcene production plasmids
Three plasmids were constructed to produce myrcene in the E.coli DH10b. Among them, pMevT and pMBIS were extracted from Jay Keasling's engineered Saccharomyces cerevisiae strain to produce arteannuic acid. pMevT contains promoter Plac and genes AtoB, HMGS, tHMGR (fig.7). pMBIS contains promoter Plac and genes ERG12, ERG8, MVD1, idi, ispa (fig.7 C). Plasmid pTYT-GPPS-MS was constructed to express genes GPPS and MS under the control of promoter Ptac by overlap PCR and Gibson assembly (fig.7 C). The successful construction of pTYT-GPPS-MS was verified by PCR (fig. 8).
Figure. 7 Genetic engineering pathways design. (A) Genetic engineering pathway for converting NH3 into arginine. (B) Genetic engineering pathways for converting H2S into cysteine. (C) Genetic engineering pathways for producing myrcene.
Figure. 8 PCR verification of plamid construction. Line 1, 2, 3 are pTYT-GPPS-MS; Line 4, 5, 6, 7 are pTYT-cysE-5-11-2; Line 8, 9, 10 are pTYT-argAfbr; Line 11 is the marker. All of the plasmid's lengths are correct.
argR knockout in DH10b genome
To prevent inhibitory feedback regulation, related gene argR in the genome was knocked out by CRISPR-cas9 in DH10b. The knockout of argR failed on EcN but succeed in DH10b, so we used DH10b instead(Fig.9 A). argR was successfully deleted with the verification of PCR. (Fig. 9 B&C)
**Figure.9 Contrast of E.coli DH10b genome before and after the deletion of argR. (A)DH10b with argR. (B) DH10bgenome without argR. (C) The electrophoresis graph of DH10b genome with and without argR. Line 1 is marker (RB-MKS); Line 2 is wildtype; Line 3 is DH10b ∆argR.
Improvement on Ammonia Absorbtion
To test the engineered bacteria's ability to absorb ammonia, we cultivated the bacteria with 5mM ammonia overnight, and then Nessler's method was used to test the concentration of the remaining NH4+. By testing the standard samples, the least-squares regression line was drawn and the mathematical formula of it was gained by computer (fig.10 A). Based on the least-squares regression line, the concentration of each group's remaining NH4+ was calculated (fig.10 B). According to the result, all groups, including wildtype showed a significant consumption in ammonia for at least 14.00%(pTYT-argA ∆argR) and at most 52.71% (pTYT-argAfbr ∆argR), compared to the original ammonia concentration 0.06287 mM in M9 medium (which serves as blank control, fig.10 B). The overexpression of gene argA has improved the conversion of ammonia for about 26.34%, while the overexpression of gene argA along with argR knocking out didn't show improvement (even slightly decreased) of ammonia absorbing compared with wildtype, and this result should be further confirmed by more replicated experiments. The DH10b with plasmid pTYT-argAfbr and ∆argR showed the most significant increase in conversion of ammonia, 40.47% more conversion rate compared to the wildtype.
In conclusion, most of our engineered bacteria have shown a satisfying ability to absorb the ammonia, especially when we overexpress argAfbr and knock out argR at the same time.
Figure 10 Detection results of the effect of the ammonia pathway. (A) The least-squared regression line of NH4+. **(B)**The concentration of NH4+. (C) The percentage of decrease in NH4+ compared with M9 medium.
Improvement on Cysteine Synthesis
According to the hydrogen sulfide conversion pathway, the production of cysteine can be represented as the consumption of H2S. To test the engineered bacteria's ability to absorb H2S, we cultivated the bacteria overnight, and then acid ninhydrin reagent was used to determine the concentration of amino acid cysteine, to see the absorption of H2S sidelong. The least-squares regression line was drawn according to the standard samples (fig.11 A). Based on the least-squares regression line, the concentration of each group's cysteine was calculated (fig.11 B). From the result, we can see that compared with M9 media, the EcN wildtype didn't produce any cysteine nearly (fig.11 B). In contrast, all the mutated groups and overexpression of cysE group displayed a larger concentration of cysteine for at least 82.66% larger than EcN wildtype and M9 medium (fig.11 C). Within the high production groups (pTYT-cysE, pTYT-cysE-5, pTYT-cysE-11-2, pTYT-cysE-5-11-2, pTYT-cysE-256, pTYT-cysE-256-5, and pTYT-cysE-256-5-11-2), as expected, the mutated groups(pTYT-cysE-5, pTYT-cysE-11-2, pTYT-cysE-5-11-2, pTYT-cysE-256, pTYT-cysE-256-5, and pTYT-cysE-256-5-11-2) showed larger production of cysteine (fig.11 B, C, D), proving that these mutants delined the feedback inhibition. EcN with plasmid pTYT-cysE-5-11-2 had the best effect, 98.72% batter than EcN wildtype and 8.79% better than overexpression of cysE (pTYT-cysE).
In conclusion, all of our engineered bacteria have shown great improvement in the ability to produce cysteine, which represents that they can absorb H2S to a larger extent.
Figure.11 Detection results of the effect of the hydrogen sulfide pathway. (A) The calibration line of cysteine. (B) Concentration of cysteine detected in each group. (C) The ratio of increased cysteine concentration in each group compared to wildtype. (D) Cysteine production compared to pTYT-cysE.
A fragrant scent can be perceived after the E.coli was cultivated over one night. Gas chromatography−mass spectrometry (GC−MS) was used to determine the production of myrcene. The least-squares regression line was drawn according to standard myrcene samples (Fig.12 A D). After calculation, the concentration of myrcene was 0.06947 mg/ml in the sample tube, so the concentration of myrcene of E.coli production was 0.034735 mg/ml since the concentration is doubled by extraction. It showed that DH10b with plasmids pMevT, pMBIS, and pTYT-GPPS-MS could produce myrcene, but in relatively low productivity compared to the previous research (Kim, 2015).
Figure.12 Detection result of the production of myrcene by the engineered bacteria. (A) GC-MS result of standard myrcene sample, blue arrow indicates the peak of myrcene. (B) GC-MS result of the culture solution of pMevT pMBIS pTYT-GPPS-MS transformed E.coli DH10b, blue arrow indicates the peak of myrcene. (C) The Calibration Line of myrcene. (D) The myrcene peak area and myrcene concentration of each group.
Kim EM, et al. Microbial Synthesis of Myrcene by Metabolically Engineered Escherichia coli. J Agric Food Chem 2015,13;63(18):4606-12.