S. cerevisiae
In S. cerevisiae , a novel sakuranetin biosynthetic pathway was designed and constructed to synthesize sakuranetin from glucose, which contained two parallel branching pathways (Fig. 1). One is the formation of p -coumaric acid from the L-tyrosine branch catalyzing by tyrosine ammonia lyase, codon-optimizedHaTAL from Hemiphyllodactylus aurantiacus . Another is from the L-phenylalanine branch, which is catalyzed by phenylalanine ammonia-lyase from Arabidopsis thaliana (AtPAL2) to generate cinnamic acid, and then cinnamic acid is catalyzed by cinnamate-4-hydroxylase from Arabidopsis thaliana (AtC4H), P450 reductase (AtATR2), and yeast native cytochrome b5 (CYB5), to synthesizep -coumaric acid.
Then, for sakuranetin synthesis starting from p -coumaric acid, 4-coumarate-CoA ligase from A. thaliana (At4CL1), chalcone synthase from Petunia hybrida (PhCHS), chalcone isomerase fromMedicago sativa (MsCHI) and naringenin-7-O-methyltransferase fromOryza sativa (OsNOMT) were chosen as the target genes to integrate into S. cerevisiae . The gene cluster of the abovementioned four enzymes under four constitutive promoters (Fig. 2A), was introduced into strain YT02 (Xiao et al., 2022) (Table 1). The resulting strain YHS01 (YT02, X3 ::PGK1p-At4CL1-HXT7t-TPI1p-PhCHS-TPI1t-ENO2p-MsCHI-PGK1t-TEF1p-OsNOMT-TEF1t ) produced 4.28 mg/L sakuranetin (Fig. 2D), indicating that the synthesis pathway of sakuranetin was successfully constructed inS. cerevisiae. To improve the expression of target genes responsible for the transformation process from p -coumaric acid to sakuranetin and the yield of sakuranetin, the effect of galactose promoters instead of constitutive promoters was evaluated. However, it is necessary to eliminate the dependence of galactose promoter on expensive inducer galactose, while maintaining a high expression of target genes under a GALregulation system (Xie et al., 2015). Thus, GAL80 was deleted in strain YT02. Then, a series of expression cassettes were constructed and introduced into the YT02 mutant strain which GAL80 was deleted for the expression of the above four genes (At4CL1, PhCHS, MsCHI, OsNOMT ) under the control of GAL promoters (Fig. 2B). The resultant strain YHS02 (YT02,GAL80 ::ADH1t-At4CL1-GAL10p-GAL7p-PhCHS-TPS1t-PGK1t-MsCHI-GAL2p-GAL1p-OsNOMT-CYC1t ) exhibited increased productivity of sakuranetin (Fig. 2C, D, E), which titer (9.19 mg/L) was twice higher than that of strain YHS01. These results suggested that the yield of sakuranetin modified by GALregulation system is higher than that by constitutive regulation system.
Although ΔGAL80 does not require galactose for induction, these strains still require the utilization of galactose, ΔGAL1 /7 /10 can eliminate the utilization of galactose (Westfall et al., 2012). Here, the resulting strain YHS05 (YHS04, ΔGAL1/7/10 ) produced 13.59 mg/L sakuranetin (Fig. 3B) without improvement.
3.2 Enhanced sakuranetin biosynthesis by adjusting the number of gene copies
Although the constructed strain YHS02 can de novo synthesize sakuranetin from glucose, the yield is low. To improve the yield of sakuranetin, increasing the number of gene copies of heterologous genes is a feasible strategy for pathway optimization (Lyu et al., 2017). Firstly, one more gene copy of phenylalanine ammonia-lyase (AtPAL2) was introduced into strain YHS02. The resultant strain was named strain YHS03 (YHS02, HO-1 ::GPM1p-AtPAL2-ADH1t ), in which sakuranetin titer was 8.24 mg/L. YHS03 exhibited no significant improvement in sakuranetin production. Meanwhile, more carbon flux can flow to p -coumaric acid. However, when one more copy of the tyrosine ammonia lyase (HaTAL) was integrated into strain YHS03, the resultant strain YHS04 (YHS03, XVI1 :: GAL7p-HaTAL-TPS1t ) produced 17.97 mg/L of sakuranetin (Fig. 3A), which increased by 95.5% compared with YHS02.
After one more copy of chalcone synthase (PhCHS) and naringenin-7-O-methyltransferase (OsNOMT) was introduced into YHS04, the resultant strain YHS07 (YHS04, X5 ::TPI1p-PhCHS-TPI1t-TEF1p-OsNOMT-TEF1t ) produced 25.37 mg/L sakuranetin, 41.2% higher than strain YHS04(17.97 mg/L)(Fig. 3A), which indicated increasing copy number of PhCHS and OsNOMTcan improve the yield of sakuranetin efficiently.
3.3 Optimizing the synthetic pathway of aromatic amino acids(L-Phe/L-Tyr)
Phenylalanine and tyrosine, are the direct precursor of sakuranetin biosynthesis. Therefore, relieving feedback inhibition of tyrosine synthesis may release more carbon flux into the sakuranetin pathway.ARO4 and ARO7 are feedback inhibition genes in the tyrosine synthesis pathway, it has been reported that feedback-insensitive DAHP synthase (ARO4K229L) and chorismate mutase (ARO7G141S) can alleviate the feedback inhibition regulation (Liu et al., 2019). Unexpectedly, the resulting strain YHS09 (YHS07, TRP1 ::TEF1p-ARO4K229L-TEF1t-PGK1p-ARO7G141S-HXT7t ) produced 20.51 mg/L sakuranetin by enhancing the expression level ofARO4K229L andARO7G141S , decreased by 19.1% compared with YHS07 (25.37 mg/L) (Fig. 3B), we speculated that there may be a bottleneck in the synthesis of sakuranetin due to the weak metabolic flux in the downstream pathway. Meanwhile, knocking out of the bypass metabolic flux genes of phenylpyruvate decarboxylase (ARO10) and pyruvate decarboxylase (PDC5), can increase carbon flux into aromatic amino acids (Koopman et al., 2012; Rodriguez et al., 2015). Then,ARO10 and PDC5 were deleted respectively in YHS09 and YHS10. The resultant strains, YHS10 (YHS09, △ARO10 ) and YHS11 (YHS10, △PDC5 ) produced 24.91 mg/L and 26.65 mg/L of sakuranetin after 72 h culture (Fig.3B), which indicated no significant improvement in production.
Endogenous genes ofARO2/ARO1/PHA2 , MtPDH1 (tyrosine prephenate dehydrogenase fromMedicago truncatula ), and EcaroL (shikimate kinase from E. coli) have been demonstrated beneficial to release more carbon flux into aromatic amino acid pathway (Liu et al., 2019). These genes were introduced into strain YHS11 one by one via a series of the expression cassette, the sakuranetin production was improved to 33.05 mg/L, 25.96 mg/L, 26.23 mg/L, 23.89 mg/L, and 43.82 mg/L in the resultant strains YHS12 (YHS11, XII5 :: GPDp-ARO2-CYC1t ), YHS13 (YHS11,ARO1p :: GPDp-ARO2-CYC1t-ENO2p ), YHS14 (YHS13,III1 :: GPM1p-PHA2-ADH1t ), YHS15 (YHS14, X2 ::FBA1p-MtPDH1-CYC1t ), and YHS16 (YHS15, HO-2 ::PDC1p-EcaroL-ADH3t ), respectively (Fig. 3B). The strain YHS16 produced 43.82 mg/L sakuranetin, 64.4% higher than that of YHS11 (26.65 mg/L), which suggested that introducing more carbon flux into the downstream pathway of sakuranetin synthesis could significantly improve the production of sakuranetin.
3.4 Enhancing malonyl-CoA biosynthetic pathway
In a previous publication, it was reported that the supply of malonyl-CoA is of great importance for flavonoid production in S. cerevisiae (Leonard et al., 2007; Zhang et al., 2021). Thus, the effect of malonyl-CoA supply on sakuranetin synthesis was evaluated by two methods. One method is to delete YPL062W in S. cerevisiae , which can decrease the transcription of downstream geneALD6 and increase acetyl-CoA accumulation (Chen et al., 2019; Leonard et al., 2007). However, when YPL062W was deleted, the resultant strain YHS17 (YHS16, △YPL062W ) produced 35.38 mg/L sakuranetin, which decreased by 19.2% compared to YHS16 (Fig. 4). This result suggested that the excessive accumulation of acetyl-CoA and the weak transformation to downstream malonyl-CoA might occur. Another method is to convert more acetate-CoA to malonyl-CoA. Acetyl-CoA carboxylase (ACC1) can catalyze acetyl-CoA to malonyl-CoA in S. cerevisiae , expression of a double-point mutant at Ser659 and Ser1157 (ACC1S659A, S1157A ) has been verified to increase manlony-CoA production (Ferreira et al., 2018; Sun et al., 2019). Thus, one copy of ACC1S659A, S1157A was further introduced into the biosynthetic pathway to convert more acetate-CoA to malonyl-CoA, YHS18 (YHS16, △YPL062W ::PGK1t-ACC1S659A, S1157A-GAL2p ) produced 50.62 mg/L of sakuranetin, increase by 15.5% compared with YHS16 (Fig. 4). Our results confirmed that enhancing the supply of the precursor malonyl-CoA was beneficial for sakuranetin production.
3.5 Production of sakuranetin in a 1-L bioreactor fermentation
To scale up sakuranetin production, the engineered strain YHS16 was selected for fermentation in a 1-L bioreactor with 500 mL YPD medium. As shown in Fig. 5, the cell densities increased gradually with the prolongation, and the broth OD660 reached a maximum of 25.7 after 46 h fermentation, while sakuranetin titer increased continuously and finally reached 158.65 mg/L after 70 h. To our knowledge, it is the highest titer of sakuranetin via microbial cell factories method among all in the past publications.