Production of frambinone by a recombinant fungal microorganism

Abstract

The invention relates to a genetically modified fungal microorganism for the production of frambinone, the microorganism having the following characteristics: the capacity to produce frambinone from tyrosine; and a limited capacity or no capacity to break tyrosine down into tyrosol, p-hydroxyphenylacetaldehyde and/or p-hydroxyphenylacetate; and to the use of same for producing frambinone.

Claims

1. A genetically modified fungal microorganism for the production of frambinone, said microorganism having the following characteristics: a capacity to produce frambinone from tyrosine; and a limited capacity or no capacity to break tyrosine down into tyrosol, p-hydroxyphenylacetaldehyde and/or p-hydroxyphenylacetate, wherein said genetically modified fungal microorganism has hydroxyphenyl pyruvate decarboxylase (HPPDC) activity less than or equal to 210.sup.6 kat per g of protein, and comprises at least one mutation or deletion in at least one of the genes encoding the following enzymes: deaminase Aro8, deaminase Aro9, decarboxylase Aro10, decarboxylase Pdc5, decarboxylase Pdc6, or alcohol dehydrogenase (ADH).

2. The microorganism according to claim 1 characterized in that the microorganism belongs to the phyla ascomycetes or basidiomycetes.

3. The microorganism of claim 1 characterized in that the microorganism includes at least one heterologous sequence encoding the enzyme 4-coumarate: CoA ligase (4CL) or benzalacetone synthase (BAS).

4. The microorganism according to claim 3 characterized in that the microorganism additionally comprises at least one heterologous or supernumerary sequence encoding the enzyme tyrosine ammonia lyase (TAL) or benzalacetone reductase (BAR).

5. The microorganism according to claim 3 characterized in that the microorganism additionally comprises at least one heterologous or supernumerary sequence encoding the enzyme phenylalanine ammonia lyase (PAL) or cinnamate 4 hydroxylase (C4H).

6. The microorganism according to claim 1 which is the strain Saccharomyces cerevisiae RK8, registered with the CNCM (National Collection of Microorganism Cultures, Institut Pasteur, 25 rue du Docteur Roux, 75724 Paris Cedex 15) on Apr. 26, 2017 under number I-5200.

7. The microorganism according to claim 1 characterized in that the microorganism belongs to the genus Yarrowia, Debaryomyces, Arxula, Scheffersomyces, Geotrichum, Pichia, or Saccharomyces.

8. The microorganism according to claim 1 characterized in that the microorganism is belongs to the species Yarrowia lipolytica, Debaryomyces hansenii, or Saccharomyces cerevisiae.

9. The microorganism of claim 3 characterized in that the microorganism includes at least one heterologous sequence encoding 4CL and BAS, at least one heterologous or supernumerary sequence encoding TAL and BAR, and/or at least one heterologous or supernumerary sequence encoding PAL or C4H.

10. The microorganism according to claim 1 characterized in that the microorganism comprises at least one mutation or deletion in at least one of the genes encoding Aro10, Pdc5, and/or Pdc6.

11. A process for the production of frambinone comprising culturing the fungal microorganism according to claim 1 in a medium comprising tyrosine.

12. A process for the production of frambinone comprising culturing the fungal microorganism according to claim 1 with a capacity to produce frambinone from tyrosine in a medium comprising tyrosine and a repressor of the path for breaking tyrosine down into tyrosol, p-hydroxyphenylacetaldehyde and/or p-hydroxyphenylacetate.

13. The process for the production of frambinone according to claim 12 characterized in that the repressor is glutamate.

14. The process for the production of frambinone according to claim 12 characterized in that the microorganism belongs to the phyla chosen from among the ascomycetes or basidiomycetes, includes at least one heterologous sequence encoding the enzyme 4-coumarate: CoA ligase (4CL) or benzalacetone synthase (BAS), comprises at least one heterologous or supernumerary sequence encoding the enzyme tyrosine ammonia lyase (TAL) or benzalacetone reductase (BAR), and comprises at least one heterologous or supernumerary sequence encoding the enzyme phenylalanine ammonia lyase (PAL) or cinnamate 4 hydroxylase (C4H).

15. The process for the production of frambinone according to claim 12 characterized in that the microorganism is the industrial strain Saccharomyces cerevisiae RK4 registered with the CNCM on Jun. 1, 2016 under number I-5101.

16. The process for the production of frambinone according to claim 12 characterized in that the microorganism is the industrial strain Saccharomyces cerevisiae RK5 registered with the CNCM on Apr. 26, 2017 under number I-5199.

Description

LEGENDS FOR THE FIGURES

(1) FIG. 1: Path for the biosynthesis of frambinone

(2) (1) L-tyrosine (2) p-coumaric acid (3) coumaroyl-CoA (4) p-hydroxybenzalacetone (5) frambinone (6) malonyl-CoA (7) phenylalanine (8) cinnamic acid

(3) TAL: tyrosine ammonia-lyase, 4CL: 4-coumarate-CoA ligase, BAS: benzalacetone synthase, BAR: benzalacetone reductase, PAL: phenylalanine ammonia-lyase, C4H: cinnamate 4 hydroxylase

(4) FIG. 2: Path for the biosynthesis of tyrosine

(5) This consists of the deamination (Aro8/9) and decarboxylation (Aro10/PDC5/PDC6) steps leading to the formation of p-hydroxyphenylacetaldehyde. The latter may be reduced using an alcohol dehydrogenase (ADH) into tyrosol or oxidized to p-hydroxyphenylacetate. Possible inactivation of the path is indicated by a cross.

(6) FIG. 3: Diagram of the HO locus modified in the genome of strains S. cerevisiae RK4, RK5 and RK8 and the cassette allowing modification of locus ARO10 A) Diagram of integrated cassettes at HO locus allowing the synthesis of frambinone. Five gene expression cassettes and a marker cassette have been integrated at the level of HO locus by using different regions of overlapping recombination (RR1-5) between the cassettes for the in vivo assembly. In strains RK5 and RK8 the cassette marker (pTEF-KanMX-tTEF) is absent. B) Diagram of the cassette allowing inactivation of gene ARO10 and conferring resistance to hygromycin.

(7) FIG. 4: Production of frambinone by strain S. cerevisiae RK4 after 7 days

(8) The concentration of frambinone (mg/L) was determined after 7 days of culture as a function of the substrate (tyrosine synthesized by the cell from glucose and ammonium sulfate (called de novo process), tyrosine or coumaric acid added to the culture medium).

(9) FIG. 5: Production of tyrosol by strain S. cerevisiae RK4 after 7 days

(10) The concentration of frambinone (mg/L) was determined after 7 days of culture as a function of the substrate (tyrosine synthesized by the cell from glucose and ammonium sulfate (called de novo process), tyrosine or coumaric acid added to the culture medium).

(11) FIG. 6: HDPPC activity in cellular extracts (strain S. cerevisiae RK4, RK5 and RK8) after 16 hours of fermentation in the synthetic medium containing tyrosine with or without glutamate

(12) The hydroxy-phenylpyruvate (HDPPC) activity (expressed in nmol/min/mg protein) was determined in the cell extract after 16 hours of fermentation in a synthetic medium containing 0.3 g/L of tyrosine and optionally 2 g/L of glutamate.

EXAMPLE EMBODIMENTS

(13) The present invention will be further illustrated with respect to a strain of Saccharomyces cerevisiae genetically modified to express 4 heterologous genes encoding enzymes TAL, 4CL, BAS and BAR integrated at its HO locus and effectively producing frambinone from tyrosine in a medium enriched by glutamate, or with a derived strain showing an inactivated ARO10 gene. These examples are in no way limiting.

(14) I) Material and Methods

(15) Generation of Expression Cassettes and Recombinant Strains

(16) To synthesize frambinone from tyrosine, it was chosen to express four heterologous genes in Saccharomyces cerevisiae as shown in the Table 4 below:

(17) TABLE-US-00004 TABLE 4 Heterologous genes used to establish a path for synthesizing frambinone with S. cerevisiae Sequence (codon-optimized for Enzyme Source Reference expression with S. cerevisiae) TAL Rhodotorula Uniprot SEQ ID NO: 5 glutinis P11544 4CL Arabidopsis Uniprot SEQ ID NO: 8 thaliana Q42524 BAS SA Rheum Uniprot SEQ ID NO: 11 palmatum Q94FV7 BAR Rubus idaeus Uniprot SEQ ID NO: 2 G1FCG0

(18) The encoding sequences were codon-optimized with S. cerevisiae. The corresponding sequences are shown in Table 4 above.

(19) They were cloned between the proponents and terminators to ensure their expression. Five gene expression cassettes were built (Table 5 below) and, in addition to a cassette marker, were integrated into the genome of industrial strain S. cerevisiae filed with the CNCM Sep. 4, 2008 under number I-4071, at the HO locus using the modular cassette integration technique (FIG. 3A). The resulting strain, RK4, was registered with the CNCM (Collection Nationale de Cultures de Microorganismes, Institut Pasteur, 25 rue du Docteur Roux, 75724 Paris Cedex 15) on Jun. 1, 2016, under number I-5101.

(20) TABLE-US-00005 TABLE 5 Gene expression cassettes and marker used in strain RK4 position of relative Coding cassette integration Promoter* SEQ ID NO: sequence SEQ ID NO: Terminator* SEQ ID NO: BAR 1 TDH3 1 RiRZS1 2 CYC1 3 TAL 2 PFK2 4 RgTAL 5 PFK2 6 4CL 3 PGI1 7 At4CL-1 8 PGI1 9 BAS1 4 PMA1 10 RpBAS 11 ZWF1 12 BAS2 5 PYK1 13 RpBAS 11 PYK1 14 marker 6 TEF1 15 kanMX 16 TEF1 17 *S. cerevisiae with the exception of TEF1 (Ashbya gossypii) Ri = Rubus idaeus Rg = Rhodotorula glutinis At = Arabidopsis thaliana Rp = Rheum palmatum

(21) From the RK4 strain described above, the RK5 strain was obtained by elimination of the kanMX cassette marker. This was done by expressing the Cre recombinase that leads to excising the kanMX marker which is flanked by loxP sites (Steensma and Linde, 2001, Yeast, 18(5): 469-72). This is the strain registered with the CNCM on Apr. 26, 2017 under number I-5199.

(22) An inactivation cassette of the gene ARO10 was then constructed (FIG. 3B). It is composed of a cassette marker in hygromycin, pTEF1short-HPH-tTEF1 (pTEF1 short: SEQ ID NO: 22; hph: SEQ ID NO: 23; tTEF1: SEQ ID NO: 24), flanked by sequences homologous to the promoter (pARO10: SEQ ID NO: 25) and the terminator (tARO10; SEQ ID NO: 26) of ARO10, both upstream and downstream, respectively. A RNA sequence guide specific to the ARO10 gene has been cloned into a plasmid to express Case9p and said RNA in S. cerevisiae. The RK5 strain described above has been co-transformed with this plasmid and the inactivation cassette of gene ARO10. Positive clones (resistant to hygromycin) have been selected and verified for the inactivation of all alleles of ARO10.

(23) The resulting strain RK8 was registered with the CNCM on Apr. 26, 2017 under number I-5200.

(24) HPLC Measurements of the Path's Intermediates and Products

(25) Frambinone, tyrosol and other intermediates of the path were analyzed and quantified by two HPLC methods called long method and ACE_29, respectively. HPLC devices and their parameters are summarized in Table 6 below and allow the separation of the frambinone and tyrosol from other compounds. For quantification, calibration was done with standard solutions between 0.1 and 300 mg/L. The samples of the yeast cultures were centrifuged (>15,000g, 10 min), and the supernatant filtered through a 0.45 m filter before injection into the HPLC.

(26) TABLE-US-00006 TABLE 6 HPLC devices and parameters Long method ACE_29 HPLC system Shimadzu LC20AD, PDA detector Column Polar Advantage II 3 150 mm 3 m C18-PFP [MK] (reverse phase) Eluent A Water + 0.1% formic acid Eluent B Methanol Elution Isocratic: Gradient 82% eluent A and 18% eluent B Temperature of 30 C. the column Flow Rate 0.4 mL/min 0.35 mL/min Injection volume 20 L; Temperature of 10 C. the automatic sampler PDA detection 200-300 nm 200-400 nm Quantification at 280 nm Quantification at 280 nm
Enzymatic Detection of Hydroxyphenyl Pyruvate Decarboxylase (HPPDC) Activity

(27) To quantify the enzymatic activity of the decarboxylation of hydroxyphenyl pyruvate (HPPDC activity), an enzymatic test paired with a crude cell extract has been developed. The cell extract was prepared from an overnight culture (16 h) of the strain of interest in a synthetic medium composed of 1.7 g/L of YNB (Difco), 5 g/L of ammonium sulfate, 2.7 g/L of potassium phosphate and 20 g/L of dextrose. The environment was also supplemented by 300 mg/L L-tyrosine and optionally different nitrogen sources (for example, L-glutamate). After aerobic growth at 30 C., the cells were harvested by centrifugation (5000g, 4 min, 4 C.) and washed twice in a wash buffer (10 mM phosphate potassium, 2 mM EDTA, pH 6.8). The packed cell was taken up in an extraction buffer (100 mM potassium phosphate, 2 mM magnesium chloride, 1 mM DTT, 1 cOmplete proteinase inhibitors, pH 6.8) and the cells were broken with a FastPrep disruptor (with 0.45 mm glass beads; four 30 s to 6 m/s cycles, and 1 min on ice). The cell debris was removed by centrifugation and the supernatant used as crude cell extract. The protein concentration was determined using the Uptima BC Assay Protein Quantification Kit according to the manufacturer's instructions.

(28) The enzyme assay was carried out as described by Kneen et al. (2011, FEBS J. 278, 1842-53), with minor modifications. The assay couples the HPPDC reaction (decarboxylation of hydroxyphenyl pyruvate (HPP) in hydroxyphenylacetaldehyde) with a second reaction (oxidation of hydroxyphenylacetaldehyde in hydroxyphenyl ethanol/tyrosol) catalyzed by the auxiliary enzyme alcohol dehydrogenase (ADH). The ADH activity leads to reduction of NADH to NAD.sup.+ which can be followed thanks to the decrease of absorption at 340 nm in a spectrophotometer. The reaction mixture (1 mL) contained 100 mM potassium phosphate, 1 mM magnesium chloride, 0.5 mM thiamine pyrophosphate, 0.1 mM of NADH, 0.5 U of ADH with horse liver, 4 mM of HPP and the crude cell extract equivalent to approximately 200 g of total protein. Reactions were measured at 32 C. and pH 6.8. The reaction was initiated by the addition of the substrate HPP.

(29) II/Results

(30) 1/Production of Frambinone and other Metabolites from Tyrosine or Coumaric Acid as Substrate by Strain RK4.

(31) Fermentation tests were conducted with strain RK4, cultivated in an inorganic medium composed of 1.7 g/L of YNB (Difco), 5 g/L of ammonium sulfate, 2.7 g/L of potassium phosphate and 20 g/L of dextrose. Optionally, the medium can contain 300 mg/L of tyrosine or 100 mg/L of coumaric acid. As suggested by Ayuso et al. (2016, Microb. Cell Factories 15. doi:10.1186/s12934-016-0446-2), fermentation tests were conducted under aerobic conditions to optimize the production of frambinone.

(32) As shown in FIG. 4, strain RK4 can synthesize approximately 6 mg/L of frambinone from tyrosine. When coumaric acid is used as substrate, the concentration of frambinone in mediums reaches approximately 14 mg/L. These results are in agreement with those of the prior art and confirm the fact that the production of frambinone from aromatic amino acids is less efficient than from coumaric acid. However, in view of the price of the substrates used for bioconversion, industrial applications are only profitable when tyrosine is used rather than coumaric acid.

(33) A last point of these results concerns the de novo synthesis of frambinone by the constructed strain RK4. Note that the frambinone concentration is approximately the same that observed in the presence of tyrosine. It has been hypothesized that this is probably linked to the regulation of the biosynthesis of tyrosine by extracellular tyrosine, and also to diverting this amino acid by a degradation pathway.

(34) 2/Breaking Tyrosine Down into Tyrosol in Strain RK4

(35) Breaking tyrosine down into tyrosol is a well-known method (FIG. 2) (Hazelwood et al., 2008, Appl. Environ. Microbiol. 74, 2259-66). The first step is transamination of tyrosine resulting in the formation of hydroxyphenyl pyruvate. This compound is then decarboxylated into hydroxyphenylacetaldehyde (EC: 4.1.1.80). Finally, the aldehyde function is reduced to form a hydroxylated molecule called tyrosol (EC: 1.1.1.90).

(36) The production of tyrosol during the fermentation of the strain RK4 was followed. FIG. 5 shows the concentrations of tyrosol in different mediums after 7 days of fermentation. Results confirm that the major part of the tyrosine provided is used for the production of tyrosol.

(37) 3/Inhibition of Hydroxyphenyl Pyruvate Decarboxylase (HPPDC) Activity.

(38) To reduce production of tyrosol in strain RK4, it was decided to inhibit the activity of the enzyme involved in the decarboxylation of hydroxyphenyl pyruvate by adding into the fermentation compounds that reduce HPPDC activity.

(39) Glutamate was selected to be added to the medium to reduce the HPPDC activity. To be certain that adding this supplemental amino acid yields a reduction in enzyme activity, the enzymatic test described above was performed. FIG. 6 shows activity in the cell extraction of RK4 after culture in the fermentation medium containing tyrosine with or without 2 g/L of glutamate. Results confirm reduced HPPDC activity in cells fermenting in the presence of glutamate.

(40) HPPDC activity of strain RK8 was measured at the same time. FIG. 6 states that the deletion of the ARO10 gene encoding for decarboxylase permits the reduction of HPPDC activity in the same proportions as strain RK4 in the presence of glutamate.

(41) FIG. 6 also reveals that, as expected, strains RK4 and RK5 (from which strain RK8 derives) have the same level of HPPDC activity in the presence of tyrosine.

(42) 4/Impact of Inhibition of Breaking Tyrosine Down on the Production of Frambinone and Tyrosol

(43) In the first of a series of experiments, the production of tyrosol and frambinone by strain RK4 was followed by HPLC (Long Method) under both conditions (with or without glutamate). The concentrations of tyrosol observed demonstrate that the reduction of HPPDC activity also reduces the formation of tyrosol: a reduction of 27% in the tyrosol concentration was observed in response to glutamate, in line with the reduction of the HPPDC activity. At the same time, the production of frambinone increased by 40%. This data strongly suggests that the reduction of HPPDC activity reduces the production of tyrosol, thus rendering the tyrosine more available for the frambinone path.

(44) In a second series of experiments, determination of the levels of frambinone (Table 7) and tyrosol (Table 8) produced by the three strains of S. cerevisiae built (RK4, RK5 and RK8) was assessed by HPLC according to the two methods described (long method and ACE_29). The ACE_29 method uses a new generation column with increased separation and compound resolution abilities, thus reducing the risk of co-elution, notably of frambinone, with other compounds as compared to the long method.

(45) The results obtained with the long method from fermentation in a culture medium containing glutamate show an increase in the frambinone synthesis by the strain RK5 (+58%) compared with fermentation in a culture medium without glutamate, equivalent to strain RK4 under the same conditions. The increases observed equally approach the values of the first series of experiments.

(46) The production of frambinone by strain RK8, cultivated without glutamate, was compared to strain RK5 under the same conditions. A +36% increase in the synthesis of frambinone, around the same observed for strains RK4 and RK5 cultivated with glutamate, was observed.

(47) Determination of the levels of synthesis of frambinone by strains RK4, RK5 and RK8 by using the ACE_29 method similarly shows an increase in the production of frambinone, in proportions, however in higher proportions, that is, +100%, +95% and +129%, respectively.

(48) TABLE-US-00007 TABLE 7 Determination by HPLC of the frambinone concentration at the end of fermentation by various strains built into the mediums without glutamate (Glu) and with glutamate (Glu+). Frambinone concentration Long method ACE_29 (mg/L) Glu MD Glu+ MD Increase Glu MD Glu+ MD Increase RK4 4.60 0.32 7.23 0.18 +57% 1.09 0.17 2.17 0.02 +100% RK5 4.57 0.27 7.23 0.21 +58% 1.14 0.02 2.21 0.03 +95% RK8 6.20 0.85 n.d. n.d. +36% 2.60 0.29 n.d. n.d. +129% n.d.: undetermined MD: Standard Deviation (SD)

(49) The results obtained using the long method from fermentation in a culture medium containing glutamate show an increase in the frambinone synthesis by the strain RK5 (20%) compared with fermentation in a culture medium without glutamate, equivalent to strain RK4 under the same conditions. The decreases observed equally approach the values of the first series of experiments.

(50) The production of tyrosol by strain RK8, cultivated without glutamate, was compared to strain RK5 under the same conditions. A 49% decrease in the synthesis of tyrosol, around the same observed for strains RK4 and RK5 cultivated with glutamate, was observed.

(51) Determination of the levels of synthesis of frambinone by strains RK4, RK5 and RK8 using the ACE_29 method similarly shows a decrease in the production of tyrosol in similar proportions, that is, 19%, 19% and 52%, respectively.

(52) TABLE-US-00008 TABLE 8 HPLC determination of the tyrosol concentration at the end of fermentation by various strains built into the mediums without (Glu) and with (Glu+) glutamate. Concentration In tyrosol Long method ACE_29 (mg/L) Glu MD Glu+ MD Decrease Glu MD Glu+ MD Decrease RK4 158.8 0.3 128.7 2.2 19% 184.2 4.4 149.2 2.5 19% RK5 158.8 0.6 127.4 0.5 20% 183.9 1.0 149.4 1.9 19% RK8 80.3 4.3 n.d. n.d. 49% 88.3 5.9 n.d. n.d. 52% n.d.: undetermined MD: standard deviation (SD)
III/Conclusions:

(53) In conclusion, introduction of the frambinone path as shown in FIG. 1 is sufficient to produce frambinone from tyrosine by means of S. cerevisiae. However, when tyrosine is added to the fermentation medium, the bioconversion is not very effective and leads to the production of a large quantity of by-products, essentially tyrosol as reported above. To reduce hijacking the substrate, the HPPDC activity involved in breaking tyrosine down into tyrosol was successfully reduced by adding an inhibitor of said enzyme activity or the inactivation of gene ARO10. As reported, this allowed the production of frambinone to be increased. Alternatively, other strategies can be used, such as the selection of a microbial strain with very low HPPDC activity, based on a natural genetic background or via the introduction of other genetic modifications such as the deletion of the genes encoding the HPPDC enzymes.