GENETICALLY IMPROVED MICROORGANISM STRAIN CAPABLE OF PRODUCING LARGER QUANTITIES OF VOLATILE ORGANIC COMPOUNDS OF INTEREST AND SCREENING METHOD FOR OBTAINING SUCH STRAINS

Abstract

The invention relates to the field of natural flavors used in the food industry. More particularly, the invention relates to a genetically modified microorganism strain capable of producing larger quantities of volatile organic compounds of interest, in particular α-unsaturated esters. It also relates to a screening method for selecting such strains, and a method for preparing a genetically modified strain to produce satisfactory quantities of volatile organic compounds (VOCs).

Description

Title of the invention: GENETICALLY IMPROVED MICROORGANISM STRAIN CAPABLE OF PRODUCING LARGER QUANTITIES OF VOLATILE ORGANIC COMPOUNDS OF INTEREST AND SCREENING METHOD FOR OBTAINING SUCH STRAINS

[0001] The invention relates to the field of natural flavors used in the food industry. More particularly, the invention relates to a genetically modified strain of microorganism capable of producing larger quantities of volatile organic compounds of interest, in particular α-unsaturated esters. It also relates to a screening method for selecting such strains, and a process for preparing a genetically modified strain to produce satisfactory quantities of volatile organic compounds (VOCs), defined as organic compounds with molecular masses of less than 400 Da. Field of the invention

[0002] The perception of aromas results from a complex mixture of volatile organic compounds (VOCs) which are organic compounds of low molecular weight (<400 Da), in defined proportions, which confer the originality of the taste of food. Until now, some of these aromatic compounds, such as 2-phenylethanol which gives the smell of roses, are chemically synthesized due to their low production cost. However, some of the chemically synthesized aromas can be modified during the integration of human metabolic pathways and be implicated in chronic diseases such as allergies, diarrhea or cancer. These potential side effects as well as the unsustainability of chemical processes have raised concerns among consumers who are gradually becoming more aware of the risks associated with these synthetic compounds, as well as the issues of process sustainability and food safety.Direct extraction of natural flavors from fruits or plants could meet consumer demand. However, the production of these flavors from their natural source depends on seasonal weather conditions, requires large areas of land due to their low plant concentration, which can reduce the space for other food crops. In addition, the use of petroleum-based solvents is often necessary to extract these flavors, which can undermine the "natural" qualification of these products.

[0003] To overcome these limitations, the production of aromas by microbial activity is a promising solution, considering that microbial cultures are carried out in controlled bioreactors and independently of the season. However, this microbial approach has significant bottlenecks. While some microbial systems naturally produce aromas of interest, it is generally at low yield, low titer and low productivity, resulting in a high production cost. This challenge can now be addressed thanks to the well-mastered tools and techniques of metabolic engineering. In addition, not all microbial systems have the metabolic pathways to produce aromas. To solve this problem, the metabolic pathway(s) that produce these specific products can be modified or reconstructed in an industrially produced microorganism such as S. cerevisiae or E.coli by applying systems and synthetic biology tools. As this solution may pose problems of societal acceptance due to the fact that the product would be produced by genetically modified organisms, an alternative to this solution would therefore be to search, in ecosystems that are still largely unexplored, for new microorganisms capable of producing flavoring compounds in quantity and diversity.

[0004] In this regard, unconventional yeasts have interesting aroma production capabilities, such as Pichia kluyveri, Torulaspora delbrueckii, Candida stellate, Hanseniaspora delbrueckii, which have been reported to produce high amounts of several aroma compounds such as 2-phenylethanol, ethyl octanoate, butanoic acid or phenolic esters. In a previous work, the inventors isolated from dragon fruits of Reunion Island, an aroma-producing filamentous yeast strain, identified as Saprochaete suaveolens.This yeast species produces a wide variety of volatile organic compounds (VOCs), including many α-unsaturated esters (isobutyl tiglate, isoamyl tiglate, butyl tiglate, ethyl tiglate, ethyl 2-methylpropanoate, ethyl 3-methylbut-2-enoate, ethyl but-2-enoate) that are rarely found in the aromatic bouquet of other Saccharomyces and non-Saccharomyces species. The inventors also demonstrated that the production of these α-unsaturated esters by S. suaveolens originates from the catabolism of branched-chain amino acids via the [3-oxidation (BOP) pathway. A simplified representation of the metabolic pathways of these branched-chain amino acids is shown in [Fig.l]. Briefly, these branched-chain amino acids are first converted to their corresponding branched-chain α-ketoacid in a reaction catalyzed by a transaminase (TA).Then, in both Saccharomyces and non-Saccharomyces species, the α-ketoacid enters the Ehrlich pathway (EP) which involves decarboxylation followed by either oxidation to produce carboxylic acids under the action of an aldehyde dehydrogenase (A1DH) or a . reduction to higher alcohols by alcohol dehydrogenase (ADH). Esterification of higher alcohols with acetyl-CoA can generate acetates under the catalytic action of ester synthase or alcohol-O-acetyltransferase (AFT). A second pathway that has been identified in filamentous fungi but is absent in Saccharomyces cerevisiae is the [3-oxidation of branched-chain α-ketoacid to acetyl-CoA. In this pathway, the α-ketoacid intermediate is oxidatively decarboxylated to an acyl-CoA by branched-chain α-ketoacid dehydrogenase (BCKAD), the enzyme regulating this pathway. The produced acyl-CoA can be completely degraded to acetyl-CoA (+ propionyl COA in the case of L-leucine and L-isoleucine) by the [3-oxidation] pathway.While yeasts can produce ethyl esters by condensation of short- or medium-chain acyl-CoA derived from fatty acid degradation with ethanol or higher alcohols (see [Fig.l]) in a process called "alcoholysis" catalyzed by an acyl-CoA:ethanol-acyltransferase, a novelty noted in Saprochaete suaveolens is that the enoyl-CoA intermediate derived from amino acid oxidation can be esterified with alcohols to give rise to various α-unsaturated esters. Overall, the combination of the Ehrlich pathway (EP) and the [3-fatty acid oxidation (BOP) pathway leads to the production of a new variety of VOCs, α-unsaturated esters, which characterize the aromatic bouquet of Saprochaete and Geotrichum species.

[0005] The identification of the capacity of the Saprochaete suaveolens strain to produce VOCs of interest opens new perspectives in the production of complex natural aromas with high added value. However, the quantities produced by the natural strain are very low, which does not allow industrial exploitation to be envisaged.

[0006] Food industry professionals are looking for new sources of natural flavors to be able to offer consumers quality products in terms of both taste and food safety. Statement of the invention

[0007] The present invention is based on a common inventive concept, namely the demonstration of the existence of a functional link between, on the one hand, the capacity / incapacity of strains of microorganism, in particular yeasts, to grow on a medium using only branched amino acids as a carbon source, and on the other hand their capacity to produce α-unsaturated esters.

[0008] On this basis, the inventors first implemented an original screening method based on the rapid growth capacity of the strains on media containing branched-chain amino acids (isoleucine, valine and leucine; strains called "ILV+") as the sole carbon source. Then, they studied the ability of ILV+ strains to produce α-unsaturated esters in order to identify natural strains capable of producing aroma molecules of interest. Twelve strains producing α-unsaturated esters were isolated, including two producing α-unsaturated esters not produced by the reference strain S. suaveolens, validating the interest of this screening method.

[0009] As a logical continuation of this initial work, the inventors sought to increase the amount of α-unsaturated esters produced by the S. suaveolens strain. Considering the metabolic pathways of the Ehrlich pathway and the [3-oxidation pathway] described previously, the inventors hypothesized that the inability of S. suaveolens to grow on branched-chain amino acids could be used as an approach to improve VOC production. In the absence of detailed knowledge of the genome of this species, the raw nucleotide sequence of which has only recently been published, a screening approach based on random mutagenesis was chosen to screen S. suaveolens mutants unable to grow on media containing only branched-chain amino acids (leucine, isoleucine and valine, ILV) as the sole carbon source. Nine mutants exhibiting this phenotype were isolated and their VOCs were quantified.One of them was found to exhibit the expected phenotype of higher VOC producer. Determination of the activities of enzymes involved in the catabolism of branched-chain amino acids by the Ehrlich and [3-oxidation] pathways showed that the higher VOC accumulation in this mutant coincided with an almost complete loss of enoyl-CoA hydratase activity and a 2-fold increase in acyl-CoA hydrolase.

[0010] This result establishes that the combination of these two characteristics - (i) absence of enoyl-CoA hydratase activity and (ii) sustained acyl-CoA hydrolase activity - can be used to prepare strains of microorganisms capable of increasing their production of α-unsaturated esters.

[0011] By these two screening methods, the inventors propose to select strains capable of producing α-unsaturated esters and which are of interest either for the diversity of the aroma molecules produced, or for their level of production. These strains can then be exploited as is or genetically modified to improve their yield performance.

[0012] The invention relates in particular to a strain of microorganism capable of producing at least 400 mg / L of alpha-unsaturated esters

[0013] It also relates to the screening of yeast strains capable of producing α-unsaturated esters consisting of:

[0014] - select, in natural isolates, the strains capable of growing on a medium containing only branched-chain amino acids (isoleucine, leucine, and valine) as a carbon source

[0015] - quantify the production of α-unsaturated esters in the isolated strains

[0016] - select strains capable of producing the required quantity of a- esters unsaturated.

[0017] It also relates to a method for screening strains of microorganisms obtained by UV mutagenesis of S. suaveolens exhibiting increased production of α-unsaturated esters compared to the non-mutated reference parent strain, comprising the steps of:

[0018] - Random mutagenesis by UV irradiation of isolates of the “mother strain” containing glucose as a carbon source

[0019] - Transfer of isolates to:

[0020] o a medium containing glucose (glucose medium) as the sole carbon source

[0021] o a medium containing isoleucine as the sole carbon source

[0022] o a medium containing leucine as the sole carbon source

[0023] o a medium containing valine as the sole carbon source

[0024] - Identification of mutants developing on glucose medium but incapable of grow on medium containing isoleucine, leucine or valine as the sole carbon source, among the initially irradiated strains

[0025] - Quantification of the production of α-unsaturated esters of each of the strains corresponding to strains unable to grow on medium containing isoleucine, leucine or valine

[0026] - Selection of strains incapable of growing on medium containing isoleucine, leucine or valine which produce a greater amount of α-unsaturated esters than the parent strain.

[0027] Thus, the present invention also relates to a process for preparing a genetically modified strain of microorganism capable of producing at least 400 mg / L of α-unsaturated esters consisting of: a. Have a strain producing a basal level of α-unsaturated esters b. Reduce enoyl-CoA hydratase activity c. Increase acyl-CoA hydrolase activity,

[0028] steps b. and c. can be carried out by mutagenesis of the gene(s) coding for these enzymes.

[0029] Finally, it relates to a process for preparing a genetically modified strain of microorganism capable of producing at least 400 mg / L of α-unsaturated esters consisting of: a. Have a strain that does not produce α-unsaturated esters b. Reduce enoyl-CoA hydratase activity if present in said strain c. Increase acyl-CoA hydrolase activity if present, or provide acyl-CoA hydrolase activity if absent,

[0030] steps b. and c. can be carried out by mutagenesis of the gene(s) coding for these enzymes. Advantages of the invention

[0031] The present invention has the advantage of describing a new method for screening strains naturally producing VOCs of interest, in particular α-unsaturated esters with aromatic notes, based on their ability to grow on a medium containing only branched amino acids as a carbon source. This method is simple and rapid.

[0032] The invention also proposes to screen among VOC-producing strains of interest, mutant strains having acquired the capacity to produce VOCs in greater quantity. This screening method is simple to analyze and quick to implement. It consists of random mutagenesis followed by selection on a medium containing only branched amino acids as a carbon source. Random mutagenesis techniques generate a very large number of strains potentially carrying the desired mutation and therefore require a subsequent screening method that is rapid. However, the screening of microorganisms for the production of VOCs is today essentially based on chromatographic methods which can prove very time-consuming when the number of strains to be analyzed is high.The inventors propose an innovative approach based on established physiological screening by hypothesizing that the ability of strains to grow on a medium containing the branched amino acids isoleucine, leucine or valine (ILV medium) is the cause of a decrease in VOC production yield. It is therefore a question of selecting, from a parent strain producing VOCs of interest, mutant strains having lost the ability to grow on ILV medium as the sole carbon source. These mutant strains have potentially increased their VOC production by promoting the flow of branched amino acids towards the aroma production pathway. The relevance of this approach is confirmed by the experimental results described below and the identification of a mutant strain meeting these selection criteria.

[0033] The present invention also describes a genetically modified S. suaveolens yeast strain capable of producing at least 400 mg / L of α-unsaturated esters. This level of production is significantly higher than that of currently available strains. It makes it possible to envisage production on an industrial scale. The target production thresholds are specific to each farm.

[0034] Finally, the present invention provides a method for preparing a microorganism capable of producing VOCs, in particular α-unsaturated esters, consisting of: reducing or eliminating the enoyl-CoA hydratase activity if this activity is present in the targeted microorganism and increasing the activity of the acyl-coA hydrolase enzyme if it is present or providing it if it is not present in the targeted microorganism. This method is applicable to a large number of cells, in particular to cells validated for industrial production, such as S. cerevisae. DETAILED DESCRIPTION OF THE INVENTION

[0035] A first subject of the invention relates to a strain of microorganism capable of producing at least 400 mg / L of α-unsaturated esters.

[0036] Such a strain produces VOCs of aromatic interest. The quantity produced differentiates it from the VOC-producing strains described to date.

[0037] The strains according to the invention produce at least 400 mg / L of alpha-unsaturated esters, and preferably at least 800 mg / L, even more preferably at least 1000 mg / L.

[0038] By "microorganism" within the meaning of the invention, we mean preferably a yeast, a bacterium or a fungus.

[0039] By "α-unsaturated esters" is meant in particular isobutyl tiglate, isoamyl tiglate and ethyl tiglate.

[0040] In a first embodiment, a strain according to the invention is a genetically modified strain obtained from an initial strain (before genetic modification) producing a basal level of α-unsaturated esters and in which the genetic modifications consist of a reduction in the activity of the enoyl-CoA hydratase enzyme and an increase in the acyl-CoA hydrolase activity compared to the initial strain.

[0041] In a second embodiment, a strain according to the invention is a genetically modified strain obtained from an initial strain (before genetic modification) not producing α-unsaturated esters and in which the genetic modifications consist of (i) a reduction or elimination of the activity of the enoyl-CoA hydratase enzyme if it exists and (ii) an expression or overexpression of the acyl-CoA hydrolase activity depending on whether this activity is absent or present in the initial strain.

[0042] By way of example, a strain of microorganism according to the invention is a strain in which the activity of the enoyl-CoA hydratase enzyme is reduced and the activity of the acyl-CoA hydrolase enzyme is increased by at least 50% compared to the activity of the initial strain.

[0043] A second object of the invention relates to a method for screening strains of microorganism capable of producing α-unsaturated esters consisting of:

[0044] - isolate in natural environments, strains capable of growing on environments culture containing at least one branched-chain amino acid (isoleucine, leucine or valine) as the sole carbon source

[0045] - quantify the production of α-unsaturated esters in the isolated strains

[0046] - select strains capable of producing the required quantity of a- esters unsaturated.

[0047] This method can be used either to directly select strains producing at least 400 mg / L of α-unsaturated esters, or to select strains producing α-unsaturated esters of interest at a level below 400 mg / L of α-unsaturated esters and which can be modified to increase this production.

[0048] The natural environments in which the strains are sought may be soil samples, plant microbial flora or any other natural environment likely to contain microorganisms.

[0049] A third subject of the invention relates to a method for screening a mutated microorganism strain exhibiting increased production of α-unsaturated esters compared to the non-mutated reference parent strain capable of producing α-unsaturated esters, comprising the steps of:

[0050] a. Random mutagenesis by UV irradiation of isolates of microorganism strains “mother strains” on medium containing glucose as the sole carbon source

[0051] b. Transfer of each of the irradiated strains so as to obtain four series of identical strains

[0052] c. Transfer onto culture medium in parallel of said four series of irradiated strains, • said first series of strains on a medium containing glucose (glucose medium) as the sole carbon source • said second series on a medium containing isoleucine as the sole carbon source • said third series on a medium containing leucine as the sole carbon source • said fourth a medium containing valine as the sole carbon source

[0053] d. Identification of mutants growing on glucose medium but unable to grow on media containing only isoleucine, leucine or valine as the sole carbon source, among the initially irradiated screening strains

[0054] e. Quantification of the production of α-unsaturated esters of each of the strains corresponding to the strains incapable of growing on medium containing only isoleucine, leucine or valine as the sole carbon source

[0055] f. Selection of strains incapable of growing on medium containing only isoleucine, leucine or valine as the sole carbon source, which produce a greater quantity of α-unsaturated esters than the parent strain.

[0056] Preferably, the quantification of the production of α-unsaturated esters is carried out by chromatography, in particular gas chromatography.

[0057] In a preferred embodiment of the invention, the selected strains produce at least 400 mg / L of α-unsaturated esters.

[0058] In a particular embodiment of the invention, the non-mutated reference mother strains capable of producing α-unsaturated esters are selected by the method of screening strains from natural isolates as described previously.

[0059] A fourth subject of the invention relates to a process for preparing a genetically modified strain of microorganism capable of producing at least 400 mg / L of α-unsaturated esters consisting of: a. Have a strain producing a basal level of α-unsaturated esters b. Reduce enoyl-CoA hydratase activity c. Increase acyl-coA hydrolase activity,

[0060] steps b. and c. can be carried out by mutagenesis of the gene(s) coding for these enzymes.

[0061] A fifth subject of the invention relates to a process for preparing a genetically modified strain of microorganism capable of producing at least 400 mg / L of α-unsaturated esters consisting of: a. Have a strain that does not produce α-unsaturated esters b. Reduce or eliminate enoyl-CoA hydratase activity if present in said strain c. Increase acyl-coA hydrolase activity if present or provide acyl-coA hydrolase activity if absent,

[0062] steps b. and c. can be carried out by mutagenesis of the gene(s) coding for these enzymes.

[0063] In both of the preparation methods described above, the reduction of enoyl-CoA hydratase (ECH) activity can be achieved by mutation or deletion of the gene responsible for ECH activity or of a gene regulating the expression of the ECH gene.

[0064] Similarly, the increase in acyl-CoA hydratase (ACH) activity can be obtained by adding one or more copies of the ACH gene or by modifying the system regulating its expression (promoter, enhancer, regulatory proteins, etc.).

[0065] These two modifications are preferably obtained by mutagenesis of the gene(s) coding for these enzymes.

[0066] The present invention will be better understood from reading the following examples, provided for illustration purposes and in no way to be considered as limiting the scope of the present invention. DESCRIPTION OF FIGURES

[0067] [Fig. 1] [Fig. 1]: Simplified diagram of branched-chain amino acid catabolism involving the Ehrlich pathway (EP) and the p-oxidation pathway (BOP) in Saprochaete suaveolens.

[0068] Abbreviation: TA = transaminase, BCKAD: branched-chain ketoacid dehydrogenase; ADH: alcohol dehydrogenase; DC: decarboxylase; ACyD: FAD-dependent acyl-CoA dehydrogenase; ACyH: acyl-CoA hydrolase; ECH: enoyl-CoA hydratase; ACyDH: NAD-dependent acyl-CoA dehydrogenase, -KT: -keto acyl-CoA thiolase; AAT: alcohol acyl transferase.

[0069] [Fig.2] [Fig.2]: Representative curve of PC2 scores compared to PCI according to the aroma production of the strains and their origins. Component analysis was performed using XLStat Applied Sensory software (2020.1.3). Group 1 included S13, S18, S21, S37, S38, S70, S74, S82, S84, S105, and S106. Group 2 consisted of only S88. Group 3 included S3, S12, S64, S68, S75, and S91. Group 4 included S27, S99, and S100.

[0070] [Fig.3] [Fig.3]: Clustering analysis of isolated strains producing VOC. Strains were grouped based on their aroma production and origin using Ward's method and XL Stat Applied Sensory software (2020.1.3). Automatic truncation (dotted line) identified four coherent groups of strains. Most strains were classified into group 1 (11 strains) and group 3 (6 strains). Group 4 included three strains. Strain S88 was found alone in group 2. The dendrogram is more flattened for group 4, suggesting that this group of strains is more homogeneous than the other two groups.

[0071] [Fig.4] [Fig.4]: Survival rate of S. suaveolens as a function of time UV exposure. Values ​​are from eight independent experiments, with the standard value shown as a vertical bar.

[0072] [Fig.5] [Fig.5]: Total VOCs determined by gas chromatography coupled to a mass spectrum detector (HS-SPME-GC / MS) produced by the wild-type strain of Saprochaete suaveolens and 9 ILV- mutants isolated. Data shown are the mean ± SD of three independent cultures.

[0073] [Fig.6] [Fig.6]: Principal component analysis (PCA) of the produced VOCs by the wild strain of Saprochaete suaveolens and its 9 ILV mutants generated by UV irradiation.

[0074] [Fig.7] [Fig.7]: Color-coded map representation of VOCs determined in the wild type and in 9 ILV- mutants of S. suaveolens obtained by UV mutagenesis. Data are represented as the ratio of the VOC value in the mutant relative to the wild type. VOCs were classified as originating mainly from the Ehrlich pathway (EP), acyl-COA or enoyl-COA intermediates of the oxidation pathway. EXPERIMENTAL PART

[0075] EXAMPLE 1: Screening of yeast biodiversity from South African wild animal feces for the production of volatile α-unsaturated esters I. Materials and methods # Biological materials

[0076] The wild type Saccharomyces cerevisiae CEN.PK 112-2N (van Dijken et al., 2000) was used in this study. A strain of Saprochaete suaveolens (formerly Geotrichum fragrans) previously isolated from Pitaya (Hylecereus polyrhisus') fruits in Réunion Island, France (Grondin et al., 2015a) was also used.

[0077] Fresh wildlife feces were collected aseptically from various locations in zoos, reserves, and national parks in South Africa. Collection was carried out under sterile conditions, and samples were then stored at 4°C until use. In addition, freeze-dried samples of wildlife feces were obtained from an anonymous donor. A total of 118 samples were studied.

[0078] Isolation and selection of ILV+ yeast strains.

[0079] For swab samples, the swabs were added directly to 10 ml of sterile Ringer's solution. For tube samples, 1 g of feces was pre-weighed and dissolved in Ringer's solution. Serial dilutions were then prepared and 100 pL of each dilution was spread on the surface of Yeast-Peptone-Dextrose agar medium (Biolab diagnostics LTD, South Africa) supplemented with 0.1 gL 1 of chloramphenicol (EMDQ Millipore Corp. Billerica, MA USA) (YPD-chloramphenicol) to promote yeast growth and selection. The Petri dishes were then incubated at 30°C for 48 hours until the Cell colonies were fully formed. Colonies were then isolated by repeating the subculture on YPD-chloramphenicol agar medium and stored at 4°C. For lyophilized samples, 1 g of feces was resuspended in 10 ml of sterile nutrient broth (Merck, South Africa) and vortexed before being incubated at 25°C for 5 days before being inoculated onto YPD-chloramphenicol medium.

[0080] Selection of strains capable of growing on media containing only branched-chain amino acids (isoleucine, leucine, and valine) as the sole carbon source (ILV+ strains) was carried out as follows. 5 ml of YPD broth was inoculated with each isolated strain, and the culture was incubated overnight at 30°C with shaking. After centrifugation of the culture (13,000 rpm, 5 min), the cell pellet was washed twice with 5 mL of sterile physiological saline and resuspended in 10 mL of sterile physiological saline. Then, a culture of 106 cells.mL 1 was prepared in sterile physiological saline, and cultures of 105 and 104 cells.mL 1 were prepared by serial dilution. 5 μL of each suspension were deposited twice on a YNB medium (yeast-nitrogen base) supplemented with 1g.L 1 of isoleucine, leucine or valine or with 2 g.L 1 of glucose (YNB-Ile, YNB-Leu, YNB-Val YNB-Glc respectively). The strains S. suaveolens and S.cerevisiae were prepared according to the same protocol and were used as positive and negative controls, respectively. The Petri dishes were then incubated for 48 hours at 30°C before reading.

[0081] Qualitative and semi-quantitative analysis of volatile organic compounds (VOCs)

[0082] The strains were inoculated into a 20 mL screw-cap tube containing 15 mL of slanted YPD-agar medium and incubated at 30°C for 24 hours. Then, the flasks were sealed and incubated for 24 hours at 30°C. For qualitative analysis, the headspace of the slanted cultures was directly subjected to solid-phase microextraction (HS-SPME) using a 2 cm long fiber coated with 50 / 30 µm divinylbenzene / Carboxene on polydimethylsiloxane bonded to a flexible fused silica core (Supelco). For semi-quantitative analysis, prior to microextraction, 10 μL of octan-l-ol (0.5 gL in dichloromethane) was added to the sealed tubes as an internal standard. The fiber was then exposed to the headspace of the strains for 15 min at 30°C and inserted into the injection port at 250°C for 2 min.The metabolites were separated by gas chromatography (GC) using a ZB-5MSI column (30m * 0.32mm * 0.25 pm film thickness), coupled to a mass spectrometer (Shimadzu GCMS-QP2010 Ultra). The carrier gas (H2) was set at a flow rate of 1.4 mL.min *. The column temperature was maintained at 45 °C for 2 min and was increased to 230 °C at a rate of 4 °C.min1. This temperature . was maintained for 5 min before termination. Volatile organic compounds were identified by comparing their mass spectra and experimental Kovats index with the NIST database (www.chemdata.nist.gov). Colony PCR

[0083] Colony PCR was performed using fresh cells as an amplifiable template. Cells were picked directly from a fresh yeast colony using a loop of IpL. Cells were suspended in 100pL of PCR reaction mix containing 0.5pM of primers ITS1 (5' TCCGTAGGTGAACCTGCGG 3') or NL1 (5'-GCATATCAATAAGCGGAGGAAAAG), 0.5 pM of primer ITS4 (5'-TCCTCCGCTTATTGATATGC 3') or NL4 (5'-GGTCCGTGTTTCAAGACGG), 200 pM of each deoxynucleotide, 1 x buffer and 2.5 units of DNA polymerase (Qiagen). PCR conditions were as follows: initial denaturation at 95 °C for 5 min; 35 cycles of denaturation at 94°C for 1 min, 1 min of primer annealing at 55.5°C for 1 min, 1 min of extension at 72°C, and a final extension at 72°C for 10 min.

[0084] PCR products were analyzed by electrophoresis on 2% agarose gels, with 1 x TAE buffer at 100V for 20 min. Gels were stained with Midori Green (Nippon Genetics Europe) and visualized under UV light. Sizes were estimated by comparison with a 1 kbp DNA ladder (1 kbp ladder, Gene O'ruler, ThermoFischer). Strain identification was performed by sequencing PCR products obtained by Eurofins Genomics (Germany). Sequences were analyzed using BLAST at NCBI (http: / / www.ncbi.nlm.nih.gov / blast). Statistical analysis

[0085] The experimental data were subjected to factor analysis (principal component analysis method) and cluster analysis (Ward's method) using XLStat software Applied Sensory (2020.1.3) II - Results and discussion

[0086] Selection of strains with high VOC production capacity

[0087] In 2015, Grondin and colleagues showed that α-unsaturated esters were produced by S. suaveolens by esterification of the enoyl-CoA intermediate of the [3-oxidation] pathway of branched-chain amino acids (Grondin et al2015). Catabolism of the enoyl-CoA intermediate by the [3-oxidation] pathway leads to the production of acetyl-CoA which can then be metabolized via the TCA cycle and the glycoxyl shunt. This specific property of S. suaveolens metabolism was used as a basis for the development of a selection method using the ability of yeasts to grow on media containing only amino acids branched-chain amino acids (isoleucine, leucine, and valine) as the sole carbon source (ILV+ strains). Hypothesizing that this method would select strains overproducing α-unsaturated esters, we searched for ILV+ strains among the 119 strains isolated from the feces of wild animals, while using S. suaveolens and S. cerevisiae as positive and negative controls, respectively. Surprisingly, we found that 43 of the 119 strains tested met the ILV+ criteria.

[0088] The volatilomes of the selected strains, i.e., the volatile organic compounds produced by all ILV+ strains (43 strains isolated from feces and the positive control S. suaveolens S0) were extracted by headspace solid-phase microextraction and analyzed by gas chromatography coupled with a mass spectrometry detector (HS-SPME-GC / MS) after 48 h of growth on YPD medium at 30°C. Among the 43 ILV+ strains, 21 strains, namely S4, S17, S22, S28, S33, S40, S44, S58, S60, S62, S63, S65, S72, S78, S92, S97, SI 17, S121, S123, S124 and S125, did not produce any VOCs. This suggests that for these strains, the Ehrlich pathway is inactive. Furthermore, it demonstrates that the assimilation of isoleucine, leucine, and valine into the [3-oxidation] pathway cannot be strictly linked to aroma production. A total of 50 different VOCs were detected in the volatilome of the remaining 22 strains and S. suaveolens.These VOCs have been classified into four categories, namely acids, alcohols, esters and α-unsaturated esters.

[0089] Multivariate analysis of isolates based on their production of volatile organic compounds.

[0090] To better visualize the behavior of the strains with respect to their VOC production performance (number of VOCs) a principal component analysis (PCA) was performed on the 22 selected strains that produced VOCs (Figures 2 and 3).

[0091] The number of acids was not correlated with any of the other parameters, probably because they represent only a small part of the detected VOCs compared to other VOC categories. However, we noted a correlation between the other VOC parameters. For example, the total number of VOCs was highly correlated with the number of esters (Pearson correlation r of 0.990) and the number of α-unsaturated esters (r=0.938) but it was negatively correlated with the number of alcohols (1, corresponding to 2-phenylethanol, r=-0.514), which means that esters and α-unsaturated esters contribute significantly to the total number of VOCs.Thus, we also identified a number of esters and α-unsaturated esters that were highly correlated with each other (r= 0.903), but they were both negatively related to the number of alcohols (r= -0.560 and r=-0.551 respectively), meaning that the volatilomes of strains that contained a lot of esters (unsaturated or not), did not show alcohol. We finally concluded that there was no significant correlation between VOC parameters and strain origin.

[0092] Analysis of VOCs produced by all aroma-producing strains

[0093] The majority of VOCs originate from amino acid catabolism. For example, ethyl 2-methylbutanoate and ethyl 3-methylbutanoate are probably produced in the Ehrlich pathway via the catabolism of isoleucine and leucine respectively (Hazelwood et al., 2008). In addition, ethyl 2-methylbut-2-enoate and ethyl 3-methylbut-2-enoate are probably also produced by the catabolism of isoleucine and leucine, but via the [3-oxidation] pathway (Grondin et al., 2015). According to the KEGG metabolic pathway database (www.genome.jp), the main metabolic pathways involved in the production of other VOCs detected in this study are the fatty acid [3-oxidation] pathway (ethyl hexanoate, ethyl octanoate, etc.), the butanoate pathway (butyl acetate, butyl butanoate, ethyl butanoate, etc.), the propanoate pathway (butyl propanoate), and the pentanoate pathway (ethyl pentanoate).Finally, some esters probably result from the esterification of two units generated by two different metabolic pathways. For example, the formation of 2-methylbutyl butanoate is probably generated by the esterification of 2-methylbutanol (synthesized by the Ehrlich pathway) and butanoic acid (synthesized by the butanoate pathway).

[0094] Esters account for 37 of the 50 VOCs detected in this study. They are present in the volatilome of all strains, except for strains S74 and S82, which produce only 2-phenylethanol. Strains S0 (S. suaveolens), S12, and S91 had the greatest diversity of esters in their volatilome. The volatilome of S0 had the greatest diversity of esters, with 30 different compounds in this category detected. Of these, 13 esters were not identified in the other strains investigated in this study. However, pentyl butanoate, which releases an apricot-pineapple aroma frequently used in food and perfumery, was not detected in the volatilome of S0, but was in strains S12 and S91. This result is interesting because this compound is mainly present in fruits (apple and cocoa bean) but rarely detected in the volatilome of microorganisms.

[0095] Finally, 9 different α-unsaturated esters were detected in strains S0 and S12 of the 22 VOC-producing strains. The remaining 10 strains did not show any α-unsaturated esters in their volatilome, meaning that the ILV+ criterion cannot be strictly linked to the production of α-unsaturated esters. The most frequently detected α-unsaturated esters were ethyl 3-methylbut-2-enoate (also called ethyl 3-methylcrotonate) and ethyl 2-methylbut-2-enoate (or ethyl tiglate), produced via the [3-oxidation of leucine and isoleucine respectively. Two α-unsaturated esters were produced by strains from the South African fauna but not by strain SO. The first was 3-methylbutyl 3-methylbut-2-enoate (flavor description unknown), produced only by strain S12 while the second was ethyl hex-2-enoate (fruity flavor) which was produced by both strains S12 and S91. Since these compounds are both rarely described in the yeast volatilome, it was decided to study in more detail strains S12 from black-footed cat ( F elis nigripes ) feces and S91 from serval ( Leptailruus serval).

[0096] Semi-quantitative analysis of VOCs produced by strains S12, S91 and SO showed that SO produces more than 50 times more VOCs than S12 and S91. This highlighted that the aromatic metabolism of SO is much more active than that of S12 and S91 for the production of alcohol, esters and α-unsaturated esters.From a quantitative point of view, strains S12 and S91 presented a similar overall VOC production (4.9 mg / L) but also in terms of esters (4.0 mg / L) and α-unsaturated esters (0.6 mg / L), suggesting that these two strains have a very similar genetic background. This was confirmed by the amplification and sequencing results of the non-coding and variable region of the ribosomal internal transcribed spacer (ITS) and the D1 / D2 domain. Indeed, analyses and BLAST (NCBI) of the sequences of the ITS or D1 / D2 regions did not allow to distinguish strains S12 and S91 from each other since these two strains presented nucleotide sequences 100% identical to that of the strain Galactomyces candidus (Geotrichum candidum MK381259.1).

[0097] Literature data already mention the production of α-unsaturated esters by Galactomyces candidus (Grondin et al., 2017; Simon Amoikon et al., 2020) but none report the production of 3-methylbutyl 3-methylbut-2-enoate and ethyl hex-2-enoate for this strain. In our study, these compounds were produced at very low concentrations by strains S12 and S91 (“3 pg / L and 12 pg / L respectively) but the production of ethyl hex-2-enoate reached 110 pg / L for strain S91. As molecular data did not allow to distinguish strain S12 from strain S91, the difference in their respective volatilome suggests that they belong to different subspecies. Conclusion

[0098] This work aimed to isolate strains producing α-unsaturated esters from the feces of South African wildlife by a strategy based on the metabolism of S. suaveolens, taken as a model for its ability to grow on media containing only branched-chain amino acids as a carbon source (ILV +) coupled with its production of α-unsaturated esters. 12 strains producing α-unsaturated esters were isolated.

[0099] Among the isolated strains, two strains, namely S12 and S91, both identified as Galactomyces candidus (Geotrichum candidum) produced esters a-unsaturated esters, namely 3-methylbutyl 3-methylbut-2-enoate and ethyl hex-2-enoate, which are not produced by the model strain of S. suaveolens. Therefore, it can be concluded that the strain selection method based on the search for the ILV+ phenotype is a promising way to isolate strains producing new a-unsaturated esters.

[0100] EXAMPLE 2: Screening by UV mutagenesis and physiological characterization of mutant strains of the yeast Saprochaete suaveolens having a greater capacity to produce aromatic compounds I - Materials and methods Yeast strains, UV mutagenesis and screening

[0101] The prototrophic diploid strain Saccharomyces cerevisiae CEN.PK 112-2N (van Dijken et al., 2000) and a wild strain of Saprochaete suaveolens (formerly Geotrichum fragrans) previously isolated from Pitaya (Hylecereus polyrhisus) in Réunion Island, France (Grondin et al., 2015a) were used in this study. UV mutagenesis of S. suaveolens was performed according to the protocol described by Winston in 2008 established for UV mutagenesis of the yeast Saccharomyces cerevisiae, which was modified for S. suaveolens. This strain was incubated in 10 mL of yeast extract peptone dextrose (YPD; 20 gL 1 peptone, 20 gL 1 dextrose, and 10 gL 1 yeast extract) medium at 30°C overnight, collected, and washed twice with 10 mL of sterile water. A 100 pL aliquot at 5.104 cells s. mL 1 was plated on yeast nitrogen base (YNB) medium supplemented with 2 g.L 1 of glucose (YNB-Glc) before exposure to UV irradiation (254 nm, 30 cm high) at times ranging from 0 to 540 sec.To stop photoreactions, plates were kept in the dark and incubated at 30°C for 15 h. Using the replication method, irradiated colonies were replicated onto YNB supplemented with 1 gL of isoleucine, leucine, or valine or 2 gL of glucose (YNB-Ile, YNB-Leu, YNB-Val, YNB-Glc respectively) and incubated for 5 days at 30°C. Colonies growing on YNB-Glc and not on YNB-Ile, YNB-Leu, or YNB-Val were then subcultured onto YPD agar medium. Confirmation of selected mutant clones was performed by an additional growth test on YPD medium followed by YNB-Ile. Thus, clones were suspended in 5 ml of YPD and incubated overnight at 30°C. The culture was then centrifuged (13,000 rpm, 5 min) and resuspended in 5 mL of sterile water. Then, 5 μL of 106, 105 and 104 cells.mL 1 prepared in sterile water were plated in duplicate onto a YNB-agar plate containing glucose, leucine, valine, or isoleucine as the carbon source. The plates were incubated for 48 hours at 30°C before being read. This procedure was performed two additional times to ensure mutation stability. Wild-type strains of . S. suaveolens and S. cerevisiae were used as positive and negative controls, respectively. Analysis of volatile organic compounds (VOCs)

[0102] The strains were inoculated into a 20 ml crimp-top flask containing 15 ml of YNB agar medium supplemented with 2 gL 1 of glucose and 1 gL 1 of isoleucine (YNB-Glc-Ile) and incubated at 30°C for 24 h. Then, the flask was sealed and reincubated for 24 h at 30°C. Before analysis, 10 μL of octan-1-ol (at 1 gL 1 in dichloromethane) was added into the sealed flasks as an internal standard. The headspace of the slant cultures was subjected to SPME analysis using a 2 cm long fiber coated with 50 / 30 μm divinylbenzene / carboxen on polydimethylsiloxane bonded to a flexible fused silica core (Supelco). The fiber was exposed to the headspace for 15 minutes at 30°C and inserted into the injection port at 250°C for 2 minutes. The metabolites were separated by GC, on a ZB-5MSI column (30m * 0.32mm * 0.25pm ​​film thickness), coupled to a mass spectrometer (Shimadzu GCMS-QP2010 Ultra).The carrier gas (H2) was set at a flow rate of 2 mL.min *. The column temperature was maintained at 40°C for 2 min, increased to 150°C at 10°C.min *, and then increased to 240°C at 30°C.min 1 before the end. Volatile organic compounds were identified by comparing their mass spectra and experimental Kovats index with the NIST database (www.chemdata.nist.gov).

[0103] Preparation of crude extracts and determination of enzymatic activities

[0104] Yeast cells were cultured in 250 ml Erlenmeyer flasks containing 50 ml of YNB-Glc or YNB-Glc-Ile for 24 h at 30°C. Then, the equivalent of 100 OD units at 600 nm of the culture was collected in 50 ml Falcon tubes by centrifugation (2000 rpm, 4°C, 1 min). The cell pellets were resuspended in 1 ml of cold water, transferred to Eppendorf tubes, washed once more with cold water. The pellets obtained after centrifugation (2 min at 10,000 rpm) were stored at -20°C until use.

[0105] For S. cerevisiae, glass beads (0.5 mm diameter) and 500 pL of extraction buffer (50 mM potassium phosphate buffer, pH 7.4 containing 2 mM EDTA, 100 mM KCl and 1 mM DTT) were added to the cell pellet. The cells were disrupted in an MP BiomedicalsTM FastPrep-24TM 5G instrument using 6 cycles of 30s at 6.5 ms-1 and 1 min in ice between each cycle. As this procedure did not work for S. suaveolens, these cells were disrupted using a Qiagen Retsh Tissue Lyser II (3 min, 30Hz) with a cold tungsten bead introduced into the tube in the presence of 500 pL of the same extraction buffer as above. After centrifugation of the Eppendorf tubes (1000 g, 5 min, 4°C), the Supernatant corresponding to the crude extracts of both yeast species was passed through Amicon Ultra-05 10K centrifugal filters according to the manufacturer's protocol to remove small molecules. The filtered extracts were then used for enzymatic assays and protein determination.

[0106] All enzymatic assays were performed using an Agilent 8453 UV-Visible spectrophotometer in 1 ml final volume cuvettes with an appropriate amount of crude extract. The specificity (or blank) of the enzymatic reaction was verified by performing the reaction both in the presence of the substrate without crude extract and in the presence of crude extract without substrate. For branched-chain α-ketoacid dehydrogenase (BCKAD), the reaction corresponded to the oxidative decarboxylation of 2-oxo-4-methylpentanoic acid measured by the reduction of NAD+ to NADH at 340 nm. The reaction mixture containing 50 mM potassium phosphate buffer pH 7.4, 2 mM EDTA, 0.2 mM DTT, 0.5 mM TPP, 5 mM MgSO4, 0.5 mM CoA-SH and 1.5 mM NAD+ was initiated by the addition of 0.2 mM 2-oxo-4-methylpentanoic acid. Enoyl-CoA hydratase (ECH) activity was determined according to

[24] using 0.225 mM crotonyl-CoA as substrate.The conversion of this substrate to 3-hydroxy-butanoyl-CoA was measured at 263 nm at pH 8.0 in 45 mM Tris-HCl pH 8.0. Decarboxylase (DC) activity was determined using 2-oxo-4-methylpentanoic acid as a substrate in a reaction mixture containing 50 mM Na+citrate at pH 6.2, 100 mM KCl, 0.2 mM DTT, 0.5 mM TPP, 5 mM MgSO, 0.2 mM NADH, and 5 U / mL yeast alcohol dehydrogenase. The reaction was initiated by the addition of 0.2 mM 2-oxo-4-methylpentanoic acid, and NAD+ oxidation was monitored at 340 nm. Alcohol dehydrogenase (ADH) was assayed with 2 mM 2-methylpropanal or 100 mM ethanol. In the former, the reduction of 0.2 mM NADH to NAD+ was monitored at 340 nm in a mixture containing 50 mM Na+citrate buffer pH 6.2, 100 mM KCl, 0.2 mM DTT, 0.5 mM TPP, and 5 mM MgSO4. In the latter, the reaction was performed in 50 mM potassium phosphate pH 7.4, 100 mM KCl, 0.2 mM DTT, 0.5 mM TPP, 5 mM MgSO4, and 0.2 mM NADH.The reaction was initiated with 100 mM ethanol. For the assay of aldehyde dehydrogenase (A1DH) activity, the reaction was carried out in the presence of 50 mM Na+-citrate buffer pH 6.2, 100 mM KCl, 0.2 mM DTT, 0.5 mM TPP, 5 mM MgSO4 and 1.5 mM NAD+. The reaction was initiated by the addition of 2 mM 2-methylpropanal and monitored at 340 nm by the reduction of NAD+ to NADH. Acyl-CoA dehydrogenase (ACyD) activity was determined after the reduction of DCIP to DCIPH2 at 655 nm. The reaction was carried out in 50 mM potassium phosphate buffer pH 7.4, 2 mM EDTA, 100 mM KCl, 0.1 mM FAD+, 0.5 mM DCIP and 0.2 mM DTT and was initiated by the addition of 0.2 mM 2-methylbutanoyl-CoA. Alcohol acyltransferase (AAT) was tested in . the hydrolytic direction (esterase), which was carried out according to

[25] using p-nitrophenyl butyrate which is hydrolyzed to butyrate and p-nitrophenol, which absorbs at 415 nm. The reaction was carried out in a mixture containing 50 mM potassium phosphate pH 7.4, 2 mM EDTA and 100 mM KCl and started with the addition of 2 mM p-nitrophenyl butyrate. Finally, the acyl-CoA hydrolase (ACyH) activity assay was carried out in 10 mM potassium phosphate buffer (pH 6.5), 50 mM MgCl2 and 0.1 mM DTNB. The reaction was initiated by the addition of 0.1 mM 2-methylbutanoyl-CoA and the release of CoASH was monitored at 415 nm, which corresponds to the production of reduced DTNBH. Protein concentration was determined at 550 nm by the Bradford method

[26] , with bovine serum albumin (BSA, Sigma-Aldrich) as a standard. All assays were performed in triplicate from independent cultures and for each culture, assays were performed in duplicate. Statistical analysis

[0107] Means and standard deviations (SD) were determined based on triplicate fermentations and represented as mean ± SD. Experimental data were subjected to one-way analysis of variance (ANOVA) using XLStat Applied Sensory software (2020.1.3) at the 95% confidence level. Principal component analysis (PCA) and hierarchical cluster analysis (HCA) were applied to discriminate the means of total volatile organic compound production and volatile compound profile of S. suaveolens wild type and ILV- mutants. PCA was performed using XLStat Applied Sensory software (2020.1.3) and the data are presented as a biplot graph. A heatmap was created based on the ratio of each VOC produced by the mutants relative to those produced by the wild type of S. suaveolens using NG-CHM Builder: Interactive heatmap online software (www.build.ngchm.net). II - Results and discussion

[0108] Condition of UV mutagenesis and screening strategy

[0109] The UV irradiation duration adequate to accumulate stable non-lethal mutations in S. cerevisiae has been empirically described as the duration that results in a population survival rate ranging from 70 to less than 40%. First, we performed a dose-response study of the effect of UV irradiation on our wild-type strain S. suaveolens to determine the optimal UV exposure duration of this unconventional strain. As shown in [Fig.4], the population survival rate immediately dropped by 25% after the first 10 seconds of irradiation and remained at this value even after 110 seconds of irradiation. With a longer irradiation time, a linear decrease (R2 = 0.96) in the survival of The population with complete mortality after 420 s of irradiation was clearly observed. A duration of 180 s of UV irradiation was chosen, which corresponds to approximately 50% of the population survival. As a result, 15,393 colonies were isolated after this UV irradiation time from wild-type S. suaveolens on YNB-Glc agar plates (medium with only glucose as a carbon source).

[0110] To screen for mutants that lost the ability to grow on media containing isoleucine, leucine, or valine as the sole carbon source, these 15,000 irradiated clones were replicated on YNB-Ile, YNB-Leu, and YNB-Val agar plates. Only 10 failed to grow on all three branched-chain amino acids. These 10 clones were then regrown on YNB-Glc, and retested for growth / lack of growth on non-permeable YNB-Leu / Ile / Val medium. With this procedure, only one of the 10 UV-irradiated clones was lost, leaving 9 potentially stable ILV- mutants, designated mutants M2 to M10.

[0111] Multivariate analysis of wild type S. suaveolens and its 9 ILV- mutants as a function of their VOC production

[0112] Volatile organic compounds (VOCs) produced by wild type S. suaveolens and the 9 ILV- mutants were extracted by solid phase microextraction from space and then analyzed by gas chromatography coupled with a mass spectrum detector (HS-SPME-GC / MS) after 48h of growth on YNB agar tubes containing 20 gL-1 of glucose and 1 gL-1 of isoleucine (YNB-Glc-Ile) at 30°C. Assuming similar growth of mutants and wild-type S. suaveolens on agar plates, we found that mutants M9 and M10 had total VOC production levels 1.8 and 8 times higher, respectively, than wild-type S. suaveolens, whereas VOCs were two times lower in mutants M6 and M7 and more than 4 times lower in mutants M2 and M4 ([Fig. 5]). The remaining two mutants, M3 and M5, showed a VOC level roughly similar to that of the wild-type.

[0113] Using the list of all VOCs identified and quantified in the wild type and the 9 ILV- mutants of S. suaveolens , a principal component analysis (PCA) was performed by the Pearson method to look for possible aggregation or discrimination between the mutants and the wild type based on their VOC production profile. As shown in [Fig.6], the first PCI axis, which accounts for 66.64% of the total variance, clearly isolated the M10 mutant from a group including the wild type and mutants M2, 3, 4, 6, 7 and 8. The second axis (PC2), which accounts for about 10% of the total variance, separated the M9 mutant from the wild type (WT) and the M5 mutant, which were roughly clustered in the same group, indicating that this mutant had a VOC profile similar to that of the wild type. On the other hand, the VOC profiles of mutants M9 and M10 were the most different from each other and from the wild type, which is also consistent with the higher amount of VOCs produced by these mutants. As mutants M2, M3, M4, M6, M7, and M8 were clustered together and not far from M5 and the WT, these data also suggest that these mutants have a VOC profile very similar to each other and not very different from the wild type. In conclusion, this PCR analysis showed that UV mutagenesis was successful in generating at least two interesting mutants with a VOC profile very different from the wild type.

[0114] Comparative analysis of VOC production between S. suaveolens and the 9 ILV- mutants.

[0115] As previously reported, the VOCs produced by wild-type and mutant S. suaveolens are primarily ester-type compounds. Two types of esters have been detected, namely acetate esters which are formed by condensation of a higher alcohol typically produced in the Ehrlich pathway with acetyl-CoA and ethyl esters which are formed by condensation of an acyl-CoA with ethanol or another alcohol such as methanol, propanol or butanol

[19] . While these two types of esters have also been found in the VOCs of mutants and wild-type S. suaveolens, other types of esters have been identified. Some of these most likely result from the condensation of an alcohol derived from the catabolism of isoleucine in the Ehrlich pathway (i.e., 2-methylbutanol) with an acyl-CoA (i.e., propanoyl-CoA, butanoyl-CoA), giving rise to 2-methylbutyl butanoate and 2-methylbutyl propanoate.The origin of these short-chain acyl-CoAs can come either from the degradation of the carbon chain of fatty acids in the case of butanoyl-CoA or from the oxidation of isoleucine which leads to propanoyl-CoA and acetyl-CoA. Other esters can be produced from an acid such as 2-methylbutanoate acid which must be activated as a CoA intermediate to condense with ethanol to give ethyl 2-methylbutanoate. These esters have been classified as Ehrlich pathway (EP) compounds because of the higher alcohols and acids that originate from this pathway. Another category is that of esters which are formed by the condensation of an acyl-COA with ethanol or a higher alcohol such as octanol to give octyl acetate, octyl propanoate and octyl butanoate. We have assigned this category to acyl-COA esters derived from the [3-oxidation (BOP) pathway.Finally, a third group, presumably original to Saprochaete species, included all VOCs whose acid precursor is an enoyl-COA that can be esterified with ethanol (i.e., ethyl tiglate, ethyl but-2-enoate) or with a higher alcohol from the Ehrlich pathway such as 2-methylbutanol and 2-methylpropanol. to give 3-methylbutyl-2-methylbut-2Z-enoate (isoamyl angelate) and 2-methylpropyl-2-methylbut-2E-enoate (isobutyl tiglate). We classified these esters as compounds derived from enoyl-CoA from the [3-oxidation (BOP-enoyl-COA) pathway. [Fig. 5] clearly illustrates the greater capacity of M10 to produce esters from these three categories, including a 5-fold increase in EP esters, reaching the exceptional level of more than 1 gL 1 and about 10 times more esters of enoyl-CoA intermediates.

[0116] As indicated above, our strategy for selecting high VOC producers, based on mutants unable to grow on a synthetic medium in which the branched amino acid is the sole carbon source, led to the isolation of two mutants M9 and M10 that presented both a different VOC profile and a higher amount of these VOCs compared to the wild type of S. suaveolens. To better highlight these differences between the strains, the content of each VOC in the mutants was divided by that of the wild type. Then, the resulting ratio value was expressed by a color code ranging from white (absence), green (~ 1), brown (> 10) to dark red (> 100), resulting in a heat map presented in [Fig.7]. This representation clearly reinforced our multivariate analysis indicating that the VOC profile of the M10 mutant and, to a lesser extent, that of M9, differs from that of the wild type strain.In particular, the VOC profile of the M10 mutant exhibited two important features. First, the levels of 17 of the 22 VOCs categorized in the EP pathway were 5- to 50-fold higher than in the wild-type strain. Specifically, esters produced from alcohol (2-methylbutanol) or acid (2-methylbutanoate) from isoleucine catabolism were approximately 10-fold more abundant than in the wild-type. This result could suggest increased activity of isoleucine catabolism in the Ehrlich pathway. On the other hand, the production of esters obtained by condensation of acyl-CoA with ethanol and enoyl-CoA with ethanol or EP-derived alcohols (i.e., ethyl tiglate, butyl tiglate, isobutyl tiglate, etc.) was 5 to 100 times higher in this M10 mutant than in the wild type.Overall, the production of total α-unsaturated esters was thus 5 times greater in the M10 mutant than in the wild-type strain of S. suaveolens (Table 1).

[0117] [Tableauxl] S. suaveolens Mutant M10 Propyl 2-methylbut-2£'-enoate 0.1+0.0 1.4+1.2 Butyl 2-methylbut-2£'-enoate 28.5 + 2.2 90.6 + 10.5 Ethyl 2-methylbut-2£'-enoate 48.4 + 3.0 115.9 + 30.7 Ethyl 2-but-2£'-enoate 0.1+0.1 155.8 + 136.2 2-methyl propyl 2-methylbut-2£'-enoate - 37.4 + 20.0 2-methyl propyl 2-methylbut-2£'-enoate 1.1+0.6 1.1+0.7 2-methylbut-2£'-enoate 3-methyl butyl 0.2 + 0.2 - 2-methylbut-2Z-enoate 3-methyl butyl 0.4 + 0.5 0.6 + 1.1 TOTAL ALPHA-INS ATURE ESTERS 78.9 ± 2.4 402.8 ± 112.1

[0118] Table 1. Alpha-unsaturated esters produced (mg / L) by the parent strain of S. suaveolens and the mutant Ml.

[0119] These data strongly indicate that an enzymatic reaction downstream of enoyl-COA oxidation was impaired by UV mutagenesis, leading to greater availability of acyl-CoA and enoyl-COA intermediates to be esterified with alcohols derived either from isoleucine catabolism or from glycolysis (ethanol).

[0120] The loss of Enoyl-C o A hydratase activity in the oxidation pathway may explain the higher VOC production in the M10 mutant.

[0121] The fact that the M10 mutant showed an 8-fold increase in total VOCs and, more specifically, that esters derived from the condensation of acyl-CoA and enoyl-CoA with alcohols were dramatically increased compared to the wild type, suggests that the activity of oxidation enzymes that catalyze reactions downstream of enoyl-CoA was altered. On the other hand, the higher levels of EP-derived aromas in the M10 mutant may also suggest that the activities of EP enzymes were increased. To answer these questions, we decided to determine the activity of key enzymes in the Ehrlich pathway and the [3-oxidation] pathway. The Ehrlich pathway involves a decarboxylase (DC) and an alcohol dehydrogenase (ADH) for the reductive pathway or an aldehyde dehydrogenase for the oxidative pathway.The catabolism of branched-chain amino acids by oxidation requires a branched-chain α-keto dehydrogenase (BCKAD) that produces an acyl-COA. This acyl-COA intermediate can lose CoA by an acyl-CoA hydrolase (AcyH) to give the corresponding acid, to be reduced to enoyl-CoA by a FAD+-dependent acyl-CoA dehydrogenase (ACyD) or even esterified to ethyl esters by a medium-chain fatty acid ester synthase

[28] . Specifically in S. suaveolens, enoyl-COA can be esterified to esters by an alcohol acetyltransferase whose nature remains to be identified.

[0122] The enzymes of these two metabolic pathways were therefore determined in the crude extracts of wild-type and M10 mutant S. suaveolens. Overall, the enzyme activities measured in the crude extracts of wild-type and mutant S. suaveolens strains were highly variable between cultures, which could be explained by the unexpected difficulty in breaking cells and achieving reproducible cell disruption. However, statistical analysis of the data allows us to be confident about the enzymatic differences, when they occurred, between the mutant and wild-type. Thus, the DC activity of wild-type S. suaveolens and the M10 mutant was roughly comparable, whether the yeasts were grown in YNB glucose supplemented with 1 gL 1 of isoleucine or not. The same applies to the ADH activity measured on 2-methylpropanal, which showed no significant difference between the two strains.We noticed that aldehyde dehydrogenase (A1DH) activity was 5 times lower in the M10 mutant compared to the wild type and overall this enzyme activity was 10 times lower than that of ADH although the data on the measured VOC levels suggest a similar contribution of the oxidative and reductive parts of the Ehrlich pathway in the production of esters. This apparent discrepancy may be due to the fact that the in vitro enzyme measurement involves the activity of ADH which acts to produce ethanol from acetaldehyde and which may not be the one specific to the Ehrlich pathway.

[0123] Regarding the enzymes of the [3-oxidation] pathway, the most dramatic effect was observed on Enoyl-CoA hydratase, whose activity in the M10 mutant was within the detection limit of the assay method. We also found that the activity of acyl-CoA hydrolase (ACyH), which catalyzes a reaction upstream of enoyl-CoA hydratase, was 2 times higher than in the wild type. However, no conclusions could be given for acyl-CoA dehydrogenase (ACyD), since the activity of this enzyme was 2 times lower in M10 growing on YNB-Glc but 2 times higher than in the wild type when cultured in YNB-Glc-Ile. The activity of BCKAD, which catalyzes the initial step in the oxidation of branched-chain amino acids, was also measured.It was found that while this enzyme exhibited similar activity between the mutant and wild type, its activity was 4 to 10 times lower than that of the enzymes downstream of the . pathway. These data appear to be consistent with previous reports that support that this enzyme may be rate-limiting in the catabolism of branched-chain amino acids by the [3-oxidation] pathway as reported in mammalian cells. Moreover, this enzyme exhibits at least 4-fold lower activity than Ehrlich pathway enzymes, which may indicate that branched-chain amino acids would be preferentially catabolized by the Ehrlich pathway (EP), at least when yeast is grown with glucose as the carbon source. Finally, since we do not know what type of ester synthase is involved in ester formation in S. suave o lens, we decided to measure this activity as an esterase (EST), assuming that an acyl-CoA:ethanol O-acetyltransferase, like the one identified in S. cerevisiae, which has both ester synthase and esterase activity, also exists in S. suave o lens.As previously reported, the overall activity of this esterase was similar in both strains but was 2-fold higher in synthetic medium containing glucose and isoleucine, suggesting a positive effect of branched-chain amino acids on the expression of genes encoding this esterase. In summary, these enzymatic analyses support the idea that the increased production of VOCs in the S. suaveolens M10 mutant can be explained by the almost complete absence of enoyl-CoA hydratase (ECH) activity. This loss of ECH activity in turn leads to increased availability of acyl-CoA and enoyl-CoA precursors for ester synthesis, and may explain the inability of this mutant to grow on branched-chain amino acids. Conclusions

[0124] UV mutagenesis was employed to generate S. suaveolens mutants that harbor higher VOC production from branched-chain amino acids. Our selection was based on the inability to grow with these amino acids as the sole carbon source. It resulted in the isolation of 9 mutants, one of which (mutant M10) was found to exhibit 8 times more VOC than the wild-type S. suaveolens strain. By determining the specific activities of key enzymes involved in isoleucine catabolism, we provided evidence that the higher VOC production observed in this S. suaveolens mutant is primarily due to the loss of enoyl-CoA hydratase activity and an increase in acyl-CoA hydrolase activity.Sequencing of wild-type and mutant strains will confirm whether the loss of ECH activity is due to a mutation in the gene encoding this enzyme or in a regulatory gene of this metabolic system. Nevertheless, this work has shown the feasibility of using the UV mutagenesis strategy and selection on specific media containing branched-chain amino acids as the sole carbon source to significantly improve the ability of the yeast S. suaveolens to produce valuable VOCs.

Claims

Claims

1. Strain of microorganism capable of producing at least 400 mg / L of α-unsaturated esters characterized in that: a. said strain is genetically modified b. the initial strain before genetic modification produces a basal level of α-unsaturated esters c. The genetic modifications consist of a reduction in the activity of the enoyl-CoA hydratase enzyme and an increase in the acyl-coA hydrolase activity.

2. Strain according to claim 1, characterized in that the activity of the enzyme enoyl-CoA hydratase is reduced and the activity of the enzyme acyl-CoA hydrolase is increased by at least 50%.

3. Strain according to one of claims 1 or 2 chosen from yeasts, fungi or bacteria.

4. A process for preparing a genetically modified microorganism strain capable of producing at least 400 mg / L of α-unsaturated esters, comprising: a. Providing a strain producing a basal level of α-unsaturated esters b. Reducing enoyl-CoA hydratase activity c. Increasing acyl-CoA hydrolase activity, wherein steps b. and c. can be carried out by mutagenesis of the gene(s) coding for these enzymes.

Images