Biosynthesis of Hesperetin, Homoeriodictyol, and Homohesperetin in a Transcriptomics-Driven Engineered Strain of Streptomyces albidoflavus

Abstract

Flavonoids exhibit various bioactivities including anti-oxidant, anti-tumor, anti-inflammatory, and anti-viral properties. Methylated flavonoids are particularly significant due to their enhanced oral bioavailability, improved intestinal absorption, and greater stability. The heterologous production of plant flavonoids in bacterial factories involves the need for enough biosynthetic precursors to allow for high production levels. These biosynthetic precursors are malonyl-CoA and l-tyrosine. In this work, to enhance flavonoid biosynthesis in Streptomyces albidoflavus, we conducted a transcriptomics study for the identification of candidate genes involved in l-tyrosine catabolism. The hypothesis was that the bacterial metabolic machinery would detect an excess of this amino acid if supplemented with the conventional culture medium and would activate the genes involved in its catabolism towards energy production. Then, by inactivating those overexpressed genes (under an excess of l-tyrosine), it would be possible to increase the intracellular pools of this precursor amino acid and eventually the final flavonoid titers in this bacterial factory. The RNAseq data analysis in the S. albidoflavus wild-type strain highlighted the hppD gene encoding 4-hydroxyphenylpyruvate dioxygenase as a promising target for knock-out, exhibiting a 23.2-fold change (FC) in expression upon l-tyrosine supplementation in comparison to control cultivation conditions. The subsequent knock-out of the hppD gene in S. albidoflavus resulted in a 1.66-fold increase in the naringenin titer, indicating enhanced flavonoid biosynthesis. Leveraging the improved strain of S. albidoflavus, we successfully synthesized the methylated flavanones hesperetin, homoeriodictyol, and homohesperetin, achieving titers of 2.52 mg/L, 1.34 mg/L, and 0.43 mg/L, respectively. In addition, the dimethoxy flavanone homohesperetin was produced as a byproduct of the endogenous metabolism of S. albidoflavus. To our knowledge, this is the first time that hppD deletion was utilized as a strategy to augment the biosynthesis of flavonoids. Furthermore, this is the first report where hesperetin and homoeriodictyol have been synthesized from l-tyrosine as a precursor. Therefore, transcriptomics is, in this case, a successful approach for the identification of catabolism reactions affecting key precursors during flavonoid biosynthesis, allowing the generation of enhanced production strains.

Keywords: flavonoid; l-tyrosine feeding; methyltransferase; substrate flexibility.

Figure 1

Biosynthetic pathway for the heterologous biosynthesis of hesperetin and homoeriodictyol. Tyrosine ammonia-lyase (TAL); 4-Coumaroyl-CoA ligase (4CL); Chalcone synthase (CHS); Chalcone isomerase (CHI); Flavone synthase (FNS); 4′-O-methyltransferase (4′OMT); 3′-O-methyltransferase (3′OMT); S. albidoflavus endogenous O-methyltransferase (3′OMT).

Figure 2

Volcano plot highlighting differentially expressed (DE) genes of the l-tyrosine-fed S. albidoflavus J1074 cultivation versus the control cultivation. The blue and red dots indicate the genes that are significantly down-regulated and up-regulated, respectively. The two vertical grey lines indicate the boundaries of genes with |log2FC| > 1, p-value < 0.05. The horizontal grey line indicates the significance threshold of 1.3, calculated using −log10 of the p-value.

Figure 3

GSA of the gene expression in the l-tyrosine feeding condition. Gene sets were defined by pathway terms. The top 10 pathway terms are shown based on the significant DE genes. The number of genes with changes in the relative gene expression within each category is shown. The percentage of genes with significant DE genes (p-adjusted < 0.05) is shown within the bars in dark blue (repression) and dark red (overexpression). Pathway term annotations can be redundant, and the same genes could belong to different pathway terms.

Figure 4

Box plots showing the normalized counts of the different genes belonging to l-tyrosine metabolism either in the l-tyrosine-fed or the non-fed cultures of S. alibidoflavus. XNR_RS01500: NADP-dependent succinic semialdehyde dehydrogenase; XNR_RS07370: NAD-dependent succinate-semialdehyde dehydrogenase; XNR_RS08935: 4-hydroxyphenylpyruvate dioxygenase (HppD); XNR_RS14570: histidinol-phosphate transaminase; XNR_RS15390: NAD(P)-dependent alcohol dehydrogenase; XNR_RS18055: fumarylacetoacetase; XNR_RS18340: pyridoxal phosphate-dependent aminotransferase; XNR_RS18910: succinate-semialdehyde dehydrogenase (NADP(+)); XNR_RS23090: NAD-dependent succinate-semialdehyde dehydrogenase; XNR_RS23895: histidinol-phosphate transaminase; XNR_RS24325: maleylpyruvate isomerase family mycothiol-dependent enzyme; XNR_RS24600: Rv2231c family pyridoxal phosphate-dependent protein CobC; XNR_RS25335: homogentisate 1,2-dioxygenase; XNR_RS25365: zinc-binding dehydrogenase; XNR_RS27215: FAD-dependent oxidoreductase; XNR_RS27770: alcohol dehydrogenase; XNR_RS28285: zinc-binding alcohol dehydrogenase family protein. The asterisks indicate statistically significant differences in the only gene, showing these differences between the two conditions (**** p <0.0001).

Figure 5

Enzymatic steps of tyrosine catabolism. TAT: tyrosine aminotransferase; HPPD: 4-hydroxyphenylpyruvate dioxygenase; HGO: homogentisate dioxygenase; MAAI: maleylacetoacetate isomerase; FAH: fumarylacetoacetate hydrolase.

Figure 6

Effect of the hppD knock-out in the biosynthesis of naringenin in S. albidoflavus. The asterisks indicate statistically significant differences (** p < 0.005).

Figure 7

(A) Production titers of different flavonoids in the S. albidoflavus UO-FLAV-005-∆hppD-HES/HOM strain. (B) Production titers of different flavonoids in the control S. albidoflavus UO-FLAV-005-∆hppD-ERI strain.