REFERENCE TO SEQUENCE LISTING

The present application contains a sequence listing in computer readable form, which is incorporated by reference.

FIELD OF THE INVENTION

The present invention relates to the production of monoterpenoids, cannabinoids, iri- doids, monoterpene indole alkaloids, and prenylated aromatic compounds in eukaryotic cells, such as yeast cells. The invention further relates to engineered yeast cells, partic ularly adapted to such production.

BACKGROUND OF THE INVENTION

Terpenes, terpenoids, derivatives thereof and other prenylated aromatic compounds are widely used e.g. as pharmaceuticals, cosmetics, nutraceuticals, flavors, fragrances and pesticides. Methods for increasing the production of these compounds in natural or engineered cells are abundant in the art.

Using engineered microorganisms for producing valuable molecules from renewable feedstock is a desirable alternative from conventional means of production. However, achieving economically viable yield, titers and productivity is a major roadblock towards industrialization. Obstacles often encountered arise from the standoff between the en gineered pathway and the native metabolism that are pulling in opposite directions. Me tabolism has evolved towards meeting the needs for growth and rerouting it can be challenging due to multiple layers of control, such as gene regulation, negative feed back loops at the enzyme level by downstream products, and efficient competing path ways.

Monoterpenes and other geranyl diphosphate (GPP)-derived compounds, which are widely used as flavors, fragrances, pesticides and could find applications as drop in jet fuel or biopolymers, are a prime example of these issues. On one hand, extraction from plant natural sources can hardly meet the increasing demands and represents an envi ronmental challenge, whereas, on the other hand, production by microbial host leads to low yield and is hindered by native metabolism constraints. Monoterpene production by engineered microbes relies on either the MEP pathway (mainly prokaryotes), the MVA pathway, or the alternative MVA pathway, all three lead ing to the formation of DMAPP and IPP, which are, in turn, condensed to form GPP. GPP is converted either into a wide array of monoterpenes by monoterpene synthases (MTSs) that rearrange the 10-carbons backbone of GPP into various monoterpenes or precursors thereof, or it is further elongated into FPP or GGPP by successive addition of IPP molecules to form sesquiterpenes and diterpenes respectively. GPP also serves as the precursor for the synthesis of a number of compounds that contain a terpene moiety, such as cannabinoids, iridoids, monoterpene indole alkaloids, prenylated aro matic compounds, and other meroterpenoids.

Yeast is considered a good host for terpene production because of its ease to be engi neered, its native mevalonate pathway, and a good capacity to harbor functional cyto chromes P450 in its endoplasmic reticulum (ER) membrane for terpene scaffold deco ration. It has shown great capacity at producing sesquiterpenes, such as artemisinin and farnesene, at industrial scale. However, the production of monoterpenes has so far been far less successful.

This can be mainly explained by the rope-pulling game that is played at the GPP branch-point between native sterol biosynthesis and the heterologous pathway leading to monoterpenes, and which is largely in favor of the native metabolism. In wild-type yeast, there are no GPP-based compounds produced and the only purpose of GPP is to serve as an intermediate that is further elongated into FPP for the production of squalene in the sterol pathway. Because of this, no dedicated GPP synthase is present in yeast, and GPP is produced by a bi-functional GPP-FPP synthase, Erg20p, that has been shown to convert very efficiently GPP into FPP as soon as it is formed and chan nel it into sterol synthesis. Various strategies have been employed to downregulate Erg20p, either by converting it into a strict GPP synthase, or by reducing its activity, but the intrinsic essentiality of sterol synthesis have rendered those attempts only moder ate fruitful, while decreasing cells viability by posing a burden on sterol synthesis. Compartmentalization is a strategy used by eukaryotic cells to solve similar issues within their own metabolism. Organelles, such as mitochondria, peroxisomes, and the endoplasmic reticulum (ER), are designed to protect the rest of the cells from toxic compounds, isolate intermediates from competing pathways, shield enzymes from in hibitors, and, overall, provide a more suitable environment for a reaction to occur away from the main bulk of the metabolism. An example of such a strategy has been reported recently, where geraniol production has been improved 11.5-fold by compartmentalizing an extra copy of the entire MVA pathway into the mitochondria together with a geraniol synthase, in comparison with the same modification in the cytosol.

While this proved to be a successful strategy, hijacking the mitochondria appeared to pose a metabolic burden to the strain with lower cells viability and growth. This can be attributed to the essential nature of the mitochondria as the powerhouse of the cells, which may hinder further engineering to reach the significantly higher titer needed for industrial application. These findings also showed that mitochondria might come with limitations on how far they can be engineered without compromising the integrity of the metabolism.

US 20150010978 discloses methods for producing terpenoids in a vast number of cells by transforming the cells with genes encoding enzymes involved in the biosynthesis of the terpenoids. The genes may be introduced into the genomes of chloroplasts for cells having chloroplasts. The exemplification discloses production of di-terpenes. KR20190079575A discloses a recombinant yeast wherein the number of peroxisomes is increased, leading to increased terpenoid production. Also disclosed is insertion of a heterologous geranylgeranyl pyrophosphate synthase.

US20130302861 A1 discloses terpenoid production in yeast by localizing a terpene syn thase to the mitochondria. The exemplification focuses on FPP-derived sesquiter penes.

Guo-Song Liu et al (J. Agric. Food Chem. 2020, 68, 7, 2132-2138) reported the produc tion of squalene, the FPP-based precursor of ergosterol, in yeast peroxisome demon strating the functionality of the MVA pathway in this organelle. However, the resulting strain did not outperform its cytosolic counterpart, most probably due to the fact that the original pathway is already well-tuned and designed to efficiently produce squalene in the cytosol.

SUMMARY OF THE INVENTION

In a first aspect the invention relates to a yeast cell comprising a peroxisomally-localized enzyme catalyzing the formation of the branch point compound, which branch point com pound can be converted in a prioritized pathway and in a non-prioritized pathway; and a peroxisomally-localized enzyme catalyzing the first step of the non-prioritized pathway. In a preferred embodiment, the invention relates to a yeast cell comprising a peroxiso mally-localized GPP synthase and a peroxisomally-localized monoterpene synthase. In a second aspect, the invention relates to a method for producing monoterpenoids, cannabinoids, iridoids, monoterpene indole alkaloids, and prenylated aromatic com pounds using a yeast cell of the invention.

DESCRIPTION OF THE DRAWINGS

Fig. 1 shows a chart disclosing the limonene production in yeast, where the limonene synthase (MTS) and/or the GPP synthase were localized to either the cytoplasm or the peroxisomes. For further details, see example 1.

Fig. 2 shows a chart disclosing the effect of localizing genes of the MVA pathway to gether with the limonene synthase (MTS) and the GPP synthase to the peroxisomes. For further details, see example 2.

Fig. 3 shows a chart disclosing improved production of six monoterpenoids, camphene, sabinene, (S)-(-)-limonene, alpha-pinene, ( )-(+)-limonene and ( )-(+)-linalool, by pe roxisomal localization of the respective synthases. For further details, see example 3. Fig. 4 shows graphs of the titer development in a fermentation of yeast according to the invention. Figure 4A shows the production of ( )-(+)-limonene and Figure 4B shows the production of geraniol. For further details, see example 4.

Fig. 5 shows a chart disclosing the effect of localizing a bi-functional GPP synthase/ter- pene synthase enzyme (GPP synthase - terpene synthase fusion) to the peroxisome on terpene production. For further details, see example 5.

Fig. 6 shows a graph disclosing the production of trans- isopiperitenol and 8-hydroxy- geraniol. Figure 6A shows (-)-limonene and trans-isopiperitenol production in strains CYTLim06, PERLim29 and PERLim30. Figure 6B shows geraniol and 8-hydroxy-gera- niol production in strains PERMGer03 and PERGer04.

Fig. 7 shows a graph disclosing the production of cannabinoid precursors. Figure 7 A shows CBGA production in strains PERMvaOI and PERCanOI in culture supplemented with 0.5 mM OA. Figure 7B shows CBGA production at different concentrations of OA added in the culture. Figure 7C shows improved peroxisomal CBGA production by tar geting CsPT4 to the peroxisome using an N-terminal targeting signal in strain PER- Can02.

OVERVIEW OF SEQUENCE LISTING

SEQ ID NO: 1 is the amino acid sequence of the engineered geranyl diphosphate syn thase derived from the Saccharomyces cerevisiae Erg20p protein and indicated as Erg20p ^N127W . SEQ ID NO: 2: is the amino acid sequence of the geranyl diphosphate synthase de rived from the Saccharomyces cerevisiae Erg20p protein and indicated as Erg20p ^N127W , and provided with the SKL peroxisomal localization signal.

SEQ ID NO: 3: is the amino acid sequence of the (+)-limonene synthase derived from Citrus limon and encoded by the C/LimS gene.

SEQ ID NO: 4: is the amino acid sequence of the (+)-limonene synthase derived from Citrus limon and encoded by the C/LimS gene and provided with the SKL peroxisomal localization signal.

SEQ ID NO: 5: is the amino acid sequence of the camphene synthase derived from So- lanum elaeagnifolium and encoded by the SeCamS gene.

SEQ ID NO: 6: is the amino acid sequence of the camphene synthase derived from So- lanum elaeagnifolium and encoded by the SeCamS gene and provided with the SKL peroxisomal localization signal.

SEQ ID NO: 7: is the amino acid sequence of the a-pinene synthase derived from Pi- nus taeda and encoded by the PfPinS gene.

SEQ ID NO: 8: is the amino acid sequence of the a-pinene synthase derived from Pi- nus taeda and encoded by the PfPinS gene and provided with the SKL peroxisomal lo calization signal.

SEQ ID NO: 9: is the amino acid sequence of the sabinene synthase derived from Sal via pomifera and encoded by the SpSabS gene.

SEQ ID NO: 10: is the amino acid sequence of the sabinene synthase derived from Salvia pomifera and encoded by the SpSabS gene and provided with an SKL peroxiso mal localization signal.

SEQ ID NO: 11: is the amino acid sequence of the geraniol synthase derived from Oci- mum basilicum and encoded by the tObGES gene.

SEQ ID NO: 12: is the amino acid sequence of the geraniol synthase derived from Oci- mum basilicum and encoded by the tObGES gene, and provided with the SKL peroxi somal localization signal.

SEQ ID NO: 13: is the amino acid sequence of the geranyldiphosphate:olivetolate geranyltransferase derived from Cannabis sativa and encoded by the CsPT4 gene.

SEQ ID NO: 14: is the amino acid sequence of the geranyldiphosphate:olivetolate geranyltransferase derived from Cannabis sativa and encoded by the CsPT4 gene, and provided with the SKL peroxisomal localization signal.

SEQ ID NO: 15: is the amino acid sequence of the fusion protein between having a GPP synthase domain (Erg20p ^N127W ) and a terpene synthase domain (C/LimS) linked by a 5xGS polypeptide and targeted to the peroxisome by a C-terminal PTS1 SEQ ID NO: 16: is the amino acid sequence of the fusion protein between having a ter- pene synthase domain (C/LimS) and a GPP synthase domain (Erg20p ^N127W ) linked by a 5xGS polypeptide and targeted to the peroxisome by a C-terminal PTS1 SEQ ID NO: 17: is the amino acid sequence of the geraniol 8-hydroxylase derived from Catharanthous roseus and encoded by the C/G80H gene.

SEQ ID NO: 18: is the amino acid sequence of the cytochrome P450 reductase derived from Catharanthous roseus and encoded by the Ci OPR gene.

SEQ ID NO: 19 is the amino acid sequence of the aromatic prenyltransferase AtaPT derived from Aspergillus terreus.

SEQ ID NO: 20 is the amino acid sequence of the 7-dimethylallyltryptophan synthase (7-DMATS) derived from Neosartorya fumigatus.

SEQ ID NO: 21 is the amino acid sequence of the phenylpropane-specific prenyltrans ferase AcPT1 derived from Artemisia capillaris.

SEQ ID NO: 22: is the amino acid sequence of the (f?)-(+)-linalool synthase derived from Mentha citrata and encoded by the McL\S gene.

SEQ ID NO: 23: is the amino acid sequence of the (f?)-(+)-linalool synthase derived from Mentha citrata and encoded by the McL\S gene and provided with the SKL peroxi somal localization signal.

SEQ ID NO: 24: is the amino acid sequence of the (S)-(-)-limonene synthase derived from Mentha spicata and encoded by the /WsLimS gene.

SEQ ID NO: 25: is the amino acid sequence of the (S)-(-)-limonene synthase derived from Mentha spicata and encoded by the /WsLimS gene and provided with the SKL pe roxisomal localization signal.

SEQ ID NO: 26: is the amino acid sequence of the beta-myrcene synthase derived from Ocimum basilicum and encoded by the ObMyrS gene and provided with the SKL peroxisomal localization signal.

SEQ ID NO: 27: is the amino acid sequence of the limonene-3-hydroxylase derived from Mentha spicata and encoded by the /WsLim3H gene.

SEQ ID NO: 28: is the amino acid sequence of the cytochrome P450 reductase derived from Taxus cuspidata and encoded by the fcCPR gene.

SEQ ID NO: 29: is the amino acid sequence of the geranyldiphosphate:olivetolate geranyltransferase derived from Cannabis sativa and encoded by the CsPT4 gene, and provided with the N-terminal peroxisomal localization signal. DEFINITIONS AND ABBREVIATIONS

Branch point molecule: A branch point molecule is according to the invention intended to mean a molecule in a biochemical pathway that can be converted into two or more different other molecules or pathways. An example is GPP that can be converted into FPP and thereby be directed into the synthesis of sesqui- and higher terpenes, or it can be converted into a monoterpene by a monoterpene synthase, into cannabinoids by a prenyltransferase enzyme, or into a prenylated aromatic compound by a corresponding prenyltransferase. For branch point molecules there will typical exist a favored or priori tized pathway, which in the natural yeast cells is favored due to e.g. biosynthetic need; and one or more other pathways that are non-prioritized.

DMAPP and IPP: Dimethylallyl pyrophosphate (or dimethylallyl diphosphate; DMAPP) and isopentenyl pyrophosphate (or isopentenyl diphosphate; IPP) are 5-carbon precur sors which are used to make isoprenoids

GPP: Geranyl diphopsphate (or geranyl pyrophosphate; GPP). GPP is formed by con- densation of a DMAPP and an IPP molecule. GPP is a branch point molecule in isopre- noid synthesis and it can, by addition of an IPP molecule, be converted into FPP, and thereby be directed into the biosynthesis of sesqui-, di- or tri-terpenes or sterol synthe sis, or it can, by the action of a monoterpene synthase, be directed into the synthesis of monoterpenoids, iridoids, and monoterpene indole alkaloids. Other prenyltransferases can also direct GPP towards the production of cannabinoids, prenylated aromatic com pounds, or meroterpenoids in general.

FPP: Farnesyl pyrophosphate (or farnesyl diphosphate; FPP) is formed by condensing GPP with an IPP molecule. FPP is the precursor for the synthesis of sesquiterpenes, diterpenes, triterpenes and sterols. GGPP: Geranylgeranyl pyrophopsphate (or geranylgeranyl diphosphate; GGPP).

GGPP is formed by condensing an FPP with an IPP molecule. GGPP is precursor for the synthesis of diterpenes.

Higher terpenes: are in this application intended to mean molecules comprising more than 10 carbon atoms of isoprenoid structure. Examples include sesquiterpenes, diter penes and triterpenes. Higher terpenes may include moieties not having the isoprenoid structure in addition to the terpene structure. Monoterpenes: Monoterpenes (or monoterpenoids) are molecules comprising a 10- carbon isoprenoid structure. Monoterpenoids may, in addition to the 10-carbon isopre- noid structure, comprise moieties not having isoprenoid structure. Frequently, the bio synthesis of monoterpenoids involves several additional steps following the initial con- version of GPP to the basic monoterpene skeleton. These additional steps may be oxi dations (e.g. catalyzed by a cytochrome P450 enzyme), reductions, isomerizations. acetylations, methylations, etc.

Iridoids: are a group of compounds found in plants and some animals, which are bio- synthetically derived from 8-oxogeraniol. Monoterpene indole alkaloids are a large and diverse group of plant chemical com pounds derived from a unit of tryptamine and a 10-carbon or 9-carbon unit of terpenoid origin that is, in turn, derived from 8-oxo-geraniol.

Cannabinoids: are a group of compounds members of which were initially isolated from the plant Cannabis sativa. Many cannabinoids are bio-synthesized by the addition of GPP to olivetolic acid.

MEP pathway: The methylerythritol 4-phosphate (MEP) pathway forming IPP and DMAPP. The pathway is found e.g. in most bacteria, in algae and is the plastids of higher plants.

MVA pathway: The mevalonate pathway (MVA pathway) is an essential metabolic pathway present in eukaryotes and in some bacteria forming IPP and DMAPP starting from acetyl-CoA.

Alternative MVA pathway: The alternative MVA pathway is found in archaea and pro vides IPP and DMAPP, starting from acetyl-CoA but utilizing isopentenyl phosphate as intermediate.

Monoterpene synthases. The term includes any enzyme that is able to catalyze the re arrangement of GPP into monoterpenoids. Monoterpene synthases typically synthesize multiple products, but the diversity of products varies among terpene synthases. Some terpene synthases have high product specificity, catalyzing the synthesis of a limited number of products, and other terpene synthases have low product specificity, catalyz ing the synthesis of a large variety of different terpenes. Examples of the products of monoterpene synthases include, but are not limited to, the following compounds: tricy- clene, a/p/7a-thujene, a/p a-pinene, a/pba-fenchene, camphene, sabinene, befa-pi- nene, myrcene, delta- 2-carene, alpha- phellandrene, 3-carene, 1 ,4-cineole, alpha- ter- pinene, befa-phellandrene, 1,8-cineole, limonene, (Z)-befa-ocimene, (E)-beta-ocimene, gamma-terpinene, terpinolene, linalool, perillene, allo-ocimene, c/s-beta-terpineol, cis- terpine-1-ol, isoborneol, de/fa-terpineol, borneol, chrysanthemol, lavandulol, alpha-ter- pineol, nerol, geraniol. In addition to GPP, certain terpene synthases (or terpene syn thase variants developed by protein engineering) have been reported to convert non- canonical prenyl diphosphate substrates, such as the 11 -carbon substrate 2-methyl- GPP, to terpenes with non-canonical prenyl scaffolds (Ignea et al. 2018). In the context of this disclosure, enzymes that are able to convert non-canonical prenyl-diphosphates with carbon lengths that differ from 10 into non-canonical terpenoids with 8, 9, 11, or 12 carbons are also included in the definition of monoterpene synthases. Prenyltransferases: Are enzymes that append a prenyl moiety to isoprenoid or non-iso- prenoid skeletons. Many prenyltransferases that append a prenyl moiety to other iso prenoid chains are involved in the synthesis of the prenyl diphosphate precursors, such as GPP (GPP synthases), FPP (FPP synthases), GGPP (GGPP synthases) or geranyl- farnesyl diphosphate synthases (GFPP synthases). These enzymes typically add IPP units to extend DMAPP to larger size prenyl-diphosphates in the trans- configuration. For this reason they are also called trans-polyprenyl synthases or trans-polyprenyl- transferases. Several prenyltransferase enzymes exist that catalyze the cis- condensa tion and elongation of DMAPP with IPP. These enzymes are termed cis-prenyltransfer- ase, or cis-polyprenyl diphosphate synthase, or cis-polyprenyltransferases, are respon sible for the synthesis of neryl diphosphate, cis.cis-farnesyl diphosphate, and nerylneryl diphosphate.

Furthermore, certain isoprenoid prenyltransferases have been reported to condense two DMAPP molecules to lavandulyl diphosphate or chrysanthemyl diphosphate. Prenyltransferases that append a prenyl moiety to non-isoprenoid scaffolds add DMAPP, GPP, FPP or GGPP to non-isoprenoid compounds, including flavonoids, amino acid residues and peptides, aromatic compounds, and other chemical com pounds in general. Such prenyltransferase enzymes are involved in the biosynthesis many different natural products including, but not limited to, cannabinoids, prenylated flavonoids, or other meroterpenoids. In the case of cannabinoid synthesis, this enzyme is a geranyldiphosphate:olivetolate geranyltransferase.

The prenyl ransferase may be part of separate polypeptides or fused into one polypeptide chain. The prenyltransferase may also be fused to a GPP synthase, a terpene synthase, or another non-terpene synthesizing protein. The prenyltransferase may also be fused to an enzyme that naturally localizes to the peroxisome matrix or its membrane in yeasts or in another organism, or that it is fused to a polypeptide chain that is itself fused to a peroxisomal targeting signal.

An aromatic prenyltransferase is selected among any enzyme with prenyltransferase ac tivity, identified from any organism or engineered, that is able to transfer an isoprenoid moiety to another isoprenoid or non-isoprenoid compound. The prenyl ransferase may be part of separate polypeptides or fused into one polypeptide chain. The prenyltrans ferase may also be fused to a GPP synthase, a terpene synthase, or another non-ter- pene synthesizing protein. The prenyltransferase may also be fused to an enzyme that naturally localizes to the peroxisome matrix or its menbrane in yeasts or in another or ganism, or that it is fused to a polypeptide chain that is itself fused to a peroxisomal targeting signal.

DETAILED DESCRIPTION OF THE INVENTION

The invention is based on the observation that in biochemical pathways branch points exist where branch point molecules can be diverted into different biochemical pathways and that the prioritized pathway leads to compounds that are mandatory for growth of the particular organism, whereas other pathways are not prioritized (designated non- prioritized pathway). It follows that special measures are required to make the cell pri oritize a non-prioritized pathway.

According to the invention, the production of a compound of a non-prioritized pathway is increased by peroxisomal localization of an enzyme catalyzing the formation of the branch point compound and an enzyme catalyzing the first step of the non-prioritized pathway. The enzyme catalyzing the formation of a branch point compound and the enzyme cat alyzing the first step of the non-prioritized pathway may be present as separate mole cules, they may be present as a single molecule comprising a domain catalyzing the formation of a branch point compound and another domain that catalyzes the first step of the non-prioritized pathway or they may even be present in form of a multidomain molecule that in addition to one or two of the two enzymatic activities comprise one or more additional domains with different function.

GPP is such a branch point molecule in terpene synthesis and it can be converted into monoterpenoid or compounds comprising a 10-carbon monoterpenoid structure at tached to a non-prenyl moiety; into sesqui-, di- or tri-terpenoids or compounds compris ing a 15-carbon sesqui-, 20-carbon di- or a 30-carbon tri-terpenoid structure; or into sterols.

In yeasts, sterols are essential for the growth and survival of the cells, therefore the conversion of GPP into FPP, and eventually into sterols, is prioritized.

Thus, in a first aspect, the invention relates to a yeast cell wherein an enzyme catalyz ing the formation of GPP and an enzyme catalyzing the first step in a pathway starting from GPP and forming a monoterpenoid, cannabinoid, iridoid, monoterpene indole al kaloid, or a prenylated aromatic compound are localized to the peroxisomes.

Enzymes catalyzing the formation of GPP, also called GPP synthases, are known for the skilled person. The invention is not limited to any particular GPP synthase, so, in principle, any GPP synthase may be peroxisomally localized and used according to the invention. The GPP synthase may be a homologous GPP synthase, i.e. an enzyme originating from the same species as the host cell, it may be a heterologous enzyme, i.e. an enzyme that originate from a different species than the host cell, or it may even be a synthetic enzyme, i.e. an enzyme that does not occur in nature but is artificially created using techniques known in the art of genetic engineering. The GPP synthase can be a single subunit, or multi-subunit enzyme that is composed from identical or non-identical subunits (several such examples exist in nature and are known to the skilled person, i.e. the combination of the large with the small subunit of snapdragon GGPP synthase (Orlova I. et al 2009)). The subunit(s) of the GPP synthase may be part of separate polypeptides or fused into one polypeptide chain. The subunit(s) may also be fused to a terpene synthase, a prenyltransferase, or another non-terpene syn thesizing protein. In particular, it may be fused to an enzyme that naturally localizes to the peroxisome in yeasts or in another organism, or that it is fused to a polypeptide chain that is, in turn, fused to a peroxisomal targeting signal. An example of a preferred GPP synthase according to the invention is an engineered GPP synthase Erg20p ^N127W (SEQ ID NO: 1) which is the native S. cerevisiae GPP syn thase containing the substitution N127W. The N127W substitution blocks the catalytic site of the enzyme to prevent further conversion of GPP into FPP by addition of an IPP molecule.

Other preferred GPP synthases includes polypeptides having, alone or in combination with other polypeptides, GPP synthase activity, said polypeptides originating from an organism that belongs to any of the kingdoms of life, i.e. Bacteria, Archaea, Protozoa, Chromista, Plantae, Fungi, or Animalia. Other preferred synthases include enzymes that have been engineered to have, alone or in combination with other polypeptides, GPP synthase activity, using protein engineering.

Enzyme catalyzing the first step in a pathway starting from GPP and forming a mono terpenoid, iridoid, cannabinoid, monoterpene indole alkaloid, a prenylated aromatic compound, or other meroterpenoids; are also known for the person skilled in the art. Non-limiting examples include monoterpene synthases, such as (+)-limonene syn thase, (-)-limonene synthase, a-pinene synthase, 1,8-cineole synthase, sabinene syn thase, camphene synthase, linalool synthase, myrcene synthase, or geraniol synthase, and prenyltransferases, such an geranyldiphosphate:olivetolate geranyltransferase, or a broad specificity aromatic prenyltransferase .

Examples of preferred monoterpene synthase according to the invention includes the (+)-limonene synthase derived from Citrus limon and having the amino acid sequence of SEQ ID NO: 3; the camphene synthase derived from Solarium elaeagnifolium and having the amino acid sequence of SEQ ID NO: 5; the (-)-limonene synthase derived from Mentha spicata and having the amino acid sequence of SEQ ID NO: 25; the (+)- linalool synthase derived from Mentha citrata and having the amino acid sequence of SEQ ID NO: 23; the myrcene synthase derived from Ocimum basilicum and having the amino acid sequence of SEQ ID NO: 26; the a-pinene synthase derived from Pinus taeda and having the amino acid sequence of SEQ ID NO: 7; the sabinene synthase derived from Salvia pomifera and having the amino acid sequence of SEQ ID NO: 9; and the geraniol synthase form Ocimum basilicum and having the amino acid se quence of SEQ ID NO: 11 Other preferred monoterpene synthases includes polypeptides having activity as beta- pinene synthase, (-)-limonene synthase, linalool synthase, myrcene synthase, bornyl diphosphate synthase, alpha-terpineol synthase, isoborneol synthase, tricyclene syn thase, a/p/7a-thujene synthase, a/p/7a-fenchene synthase, delta- 2-carene synthase, al- p/7a-phellandrene synthase, 3-carene synthase, 1 ,4-cineole synthase, a/p/7a-terpinene synthase, befa-phellandrene synthase, 1,8-cineole synthase, (Z)-befa-ocimene syn thase, (E)-beta-ocimene synthase, gamma-terpinene synthase, terpinolene synthase, allo-ocimene synthase, c/s-beta-terpineol synthase, c/s-terpine-1-ol synthase, delta- ter- pineol synthase, borneol synthase, alpha-terpineol synthase, nerol synthase, 2-methyl- isoborneol synthase, 2-methylenebornene synthase, 2-methyl-2-bornene synthase, or beta-phellandrene synthase.

Prenyltransferases able to attach the geranyl moiety to a non-isoprenoid scaffold in clude the geranyldiphosphate:olivetolic geranyltransferase CsPT4 derived from Canna bis sativa and having the amino acid sequence of SEQ ID NO: 13, or the aromatic prenyltransferase AtaPT from Aspergillus terreus and having the amino acid sequence of SEQ ID NO: 19.

Other preferred prenyltransferases able to add the prenyl group to isoprenoid scaffolds include neryl diphosphate synthase, chrysanthemyl diphosphate synthase, or la- vandulyl diphosphate synthase, while preferred prenyltransferases able to append the prenyl group to non-isoprenoid scaffolds include the 7-dimethylallyltryptophan synthase from Aspergillus fumigatus (7-DMATS) (SEQ ID NO: 20) and the phenylpropane-spe- cific prenyltransferase AcPT 1 from Artemisia capillaris (SEQ ID NO: 21).

In one preferred embodiment, a Saccharomyces cerevisiae cell is provided wherein a GPP synthase and a limonene synthase are localized to the peroxisomes. The inven tors have found that this alone is sufficient to induce a 32-fold improvement in the mon oterpene (limonene) production compared with the production obtained when these two enzymes are expressed in the cytosol, or only one of the two enzymes is present in the peroxisome and the other in the cytosol. Further, additional peroxisomal compartmen- talization of the complete MVA pathway, comprised of EfmvaS, EfmvaE, Erg12p, Erg9p and Idilp improved the monoterpene production by 14-fold, 17-fold, 17-fold, 20.5-fold, 22-fold, and 125 fold, respectively for camphene, pinene, (-)-limonene, (+)-linalool, sab- inene and (+)-limonene, compared with an identical yeast cell where the enzymes are localized in the cytosol. In another preferred embodiment, a S. cerevisiae cell is provided wherein a GPP syn thase and a geraniol synthase are localized to the peroxisomes. The yeast cell pro duces improved level of geraniol, a precursor for iridoids and monoterpene indole alka loids, compared with same cell wherein the enzymes are localized in the cytosol.

In a further preferred embodiment, a S. cerevisiae cell is provided, wherein a GPP syn thase and an olivetolic acid prenyltransferase are localized to the peroxisomes. The yeast cell is efficient in producing cannabigerolic acid, the precursor of several canna- binoid compounds.

In a further embodiment a S. cerevisiae cell is provided, wherein a GPP synthase and the aromatic prenyltransferase AtaPT, from Aspergillus terreus are localized to the pe roxisomes. When provided with umbeliferone, quercetin, isoquercetin, resveratrol, or naringenin, the yeast cell afforded efficient synthesis of osthrutin, geranylated querce tin, geranylated isoquercetin, geranylresveratrol, and geranyl-naringenin respectively.

In a further embodiment, a S. cerevisiae cell is provided, wherein an isopentenyl di phosphate isomerase (IDI), which is a DMAPP synthesizing enzyme, and a terpene synthase catalyzing the synthesis of isoprene (isoprene synthase; ISPS) are localized to the peroxisomes. The yeast cell produced afforded efficient synthesis of isoprene.

In a further embodiment, a S. cerevisiae cell is provided, wherein an isopentenyl di phosphate isomerase (IDI) and the lavandulyl diphosphate synthase from Lavandula x intermedia are localized to the peroxisomes. The yeast cell afforded efficient synthesis of lavandulol.

In a further embodiment, a S. cerevisiae cell is provided, wherein an isopentenyl di phosphate isomerase (IDI) and the chrysanthemyl diphosphate synthase from Tanace- tum cinerariifolium are localized to the peroxisomes. The yeast cell afforded efficient synthesis of chrysanthemol.

In a further embodiment, a S. cerevisiae cell is provided, wherein an isopentenyl di phosphate isomerase (IDI) and the 7-dimethylallyltryptophan synthase from Aspergillus fumigatus (7-DMATS) are localized to the peroxisomes. The yeast cell afforded effi cient synthesis of prenyl-tryptophan.

In a further embodiment, a S. cerevisiae cell is provided, wherein an isopentenyl di phosphate isomerase (IDI) and the phenylpropane-specific prenyltransferase AcPT1 from Artemisia capillaris are localized to the peroxisomes. When provided with p-cou- maric acid, the yeast cell afforded efficient synthesis of drupanin and artepillin C. In a further embodiment, a S. cerevisiae cell is provided, wherein an isopentenyl di phosphate isomerase (IDI) and the O-prenyltrasferase AcaPT from Antrodia campho- rata are localized to the peroxisomes. When provided with apigenin, kaempherol, dai- dzein, naringenin, genistein, isoliquiritigenin, equol, umbelliferone, curcumin, resvera- trol, or diethylstilbestrol, the yeast cell afforded efficiently synthesis of 4'-dimethylallyl- apigenin, 4'-dimethylallyl-naringenin, 4'-dimethylallyl-kaempferol, 4'-dimethylallyl-dai- dzein, 7-dimethylallyl-daidzein, 7,4'-di-(dimethylallyl)-daidzein, 4'-dimethylallyl- genistein, 7-dimethylallyl-genistein, 7,4'-di-(dimethylallyl)-genistein, 4-dimethylallyl- isoliquiritigenin, 4'-dimethylallyl-equol, 7-dimethylallyl-equol, 6-dimethylallyl-equol, 4'- dimethylallyl-daidzin, 7-dimethylallyl-umbelliferone, 8-dimethylallyl-curcumin, 8'-di- methylallyl-demethoxycurcumin, 8-dimethylallyl-demethoxycurcumin, 7-dimethylallyl-L- tryptophan, 4'-dimethylallyl-resveratrol, or 5-dimethylallyl-diethylstilbestrol.

Peroxisomal localization

According to the invention, the expression peroxisomal localization or grammatically equivalent terms in connection with biosynthetic enzymes for the terpene pathways, is intended to mean that the enzymes in question are translocated to the peroxisomes or the peroxisome membrane after synthesis and that the enzymes thereafter exerts their catalytic functions in the peroxisomes.

Peroxisomal localization can be effectuated by providing the gene encoding the en zyme to be peroxisomally localized with a peroxisomal localization signal.

Peroxisomal localization and peroxisomal localization signals are known in the art e.g. in W09424289A1 and KR101308971B1 (incorporated herein by reference); and such signals and methods known in the art are also useable according to the present inven tion.

A preferred peroxisomal localization signal is SKL (SerLysLeu) added to the C-termi- nus of the polypeptide to be peroxisomally localized or any C-terminal tripeptide with the canonical sequence (S/A/C)-(K/R/H)-(L/M).

For yeast, another preferred localization signal consists of the conserved peptide (R/K)- (L/V/I)-X ₅ -(H/Q)-(L/A/F) added to the N-terminus of the polypeptide to be peroxisomally localized. An additional method to effectuate peroxisomal localization of a protein is to fuse said protein with another protein that is naturally found in the peroxisomes of yeasts or other organisms. A further method to achieve peroxisomal localization of a protein is to con struct a protein fusion between said protein and another protein (or protein domain) that does not normally reside in the peroxisome but is engineered to localize to the pe roxisome by the addition of a localization signal as described above.

The peroxisomally localized enzymes may be homogeneous, meaning that the peroxi- somally localized enzyme is identically to enzyme naturally found in the cytoplasm of the host cell or it may be heterologous, meaning that it is different from the enzyme nat urally found in the cytoplasm of the host cell.

According to the invention, peroxisomal localization of an enzyme means that a gene encoding the enzyme in question, provided with an encoded peroxisomal localization signal is introduced into the host cell. If the enzyme in question is an enzyme that is naturally found in the host cell, it is believed that the peroxisomally localized enzyme provides for the improved synthesis of monoterpenoids, cannabinoids, iridoids mono- terpene indole alkaloids and other prenylated compounds according to the invention, whereas the naturally enzyme localized in the cytoplasm of the host cell provides for the normal biosynthesis of biomolecules necessary for survival and growth of the host cell. For example, if the enzyme in question is a GPP synthase, the peroxisomally lo calized GPP synthase will provide the improved synthesis of monoterpenoids, canna binoids, iridoids, monoterpene indole alkaloids, and other prenylated compounds ac cording to the invention and the natural GPP-synthesizing enzyme localized in the cyto plasm will secure that GPP is provided for the biosynthesis of required molecules e.g. sterols; necessary to secure survival and normal growth of the host cell.

If the selected host cell is a polyploid cell, e.g. a diploid or tetraploid cell; it may even be possible to provide for peroxisomal localization by gene editing techniques resulting that one or more allele of the gene encoding the enzyme in question is provided with a peroxisomal localization signal and leaving at least one allele unaltered. This will se cure that the edited allele(s) provide for the peroxisomal localized enzyme, whereas the non-edited allele(s) provides for the natural enzyme localized in the cytoplasm.

Host cells

The host cell is according to the invention a yeast cell, i.e. a eukaryotic single cellular organism; reviewed e.g. in: The yeasts. 5 ^th edition. A taxonomic study. Editors: Kurtz- man, Fell, Boekhout. Elsevier, 2011. Preferred host cells include cells belonging to the genera: Saccharomyces, Pichia, Candida, Yarrowia, Ogataea. More preferred, the host cells are selected among the species; Saccharomyces cerevisiae, Pichia pastoris, Yarrowia lipolytica, Ogataea poly- morpha, Candida albicans, Candida boidinii.

The invention is not limited to any particular methods for providing the peroxisomal lo calization for the enzymes according to the invention. Any method known in the art for providing suitable genes, optimizing codon usage, providing suitable regulatory ele ments such as promoters, terminators, adenylation sites, introns, exons, enhancer ele ments, ribosome binding sites, Kozak sequences, transforming yeasts, etc. may be used according to the invention.

Production

The invention is also related to the production of monoterpenoids, cannabinoids, iri- doids, monoterpene indole alkaloids and other prenylated compounds using a yeast cell of the invention.

According to the invention monoterpenoids, cannabinoids, iridoids, monoterpene indole alkaloids, and prenylated compounds may be produced using a method comprising the steps of: a. Providing a yeast cell according to the invention; b. Growing the yeast cell in a substrate supporting growth of the yeast cell; and c. When required, providing a co-substrate to be prenylated. d. Recovering the compound from the fermentation broth, or converting the compound to more complex products within the yeast cells by the action of additional native or heterologously expressed enzymes.

The compound provided by this method may even be further converted to more com plex products within the yeast cells by the action of additional native or heterologously expressed enzymes.

Growing the yeast can in principle be done in any known method for growing yeast, but in order to facilitate the recovery it is preferred to grow the yeast cells in liquid medium in a container such as in shake-flasks or fermenters.

It is particular preferred to grow the yeast cells in a fermenter and the fermentation pro cess may be performed as a batch fermentation, fed-batch fermentation, or a continu ous fermentation, as known in the art. The substrate supporting growth of the yeast cell may be any suitable medium com prising a carbon source, nitrogen source, minerals and nutritionals required by the par ticular yeast cell.

The substrate may be a complex substrate comprising ingredients that are not fully de- fined, or it may be a defined medium comprising only defined ingredients.

As examples of ingredients for a complex medium can be mentioned molasses, dex- trins, hydrolysates of starch and/or proteins.

As examples of ingredients for a defined medium can be mentioned: glucose, sucrose, ammonia, salts, minerals and vitamins. The fermentation process generates a fermentation broth comprising cells, water, prod uct, remaining nutritionals and minerals and waste products generated by the cells.

The recovery of the monoterpenoids, cannabinoids, iridoids, monoterpene indole alka loids, and other prenylated compounds from the fermentation broth is done using meth ods known in the art for recovering such compounds.

MATERIALS AND METHODS

Genes used:

Table 1 - List of genes used in the study. Accession numbers are for Uniprot.org. Where unavailable a reference is cited.

Gene Origin Function Accession number/reference

EfmvaE Enterococcus fae- Acetyitransferase + HMG- Q9FD70 calis CoA reductase

Delta-iso erase Farnesyl pyrophosphate P08524 synthase ERG20 ^N127W S. cerevisiae Geranyl pyrophosphate (Ignea et al 2014) synthase tcCPR/POR Taxus cuspidata cytochrome P450 reduc AAT76449.1 tase

Yeast strains

The yeast strains used in this application were based on the EGY48 Saccharomyces cerevisiae strain disclosed in (Ignea et al (2011), Thomas B.J. and R. Rothstein (1989) and (Ellerstrom M et al (1992)), and modified according to Table 2.

Table 2: Strains

Constructions of plasmids:

Plasmids were generated using standard methods used within genetic engineering and known in the art. Detailed protocols for methods for plasmid constructions can be found in general handbooks containing methods for molecular cloning.

Plasmids designed to provide for peroxisomal localization of enzymes, named pPER includes the peroxisomal localization signal (-SKL) fused C-terminally to the amino acid sequence of the enzymes or an N-terminal peroxisomal localization signal, whereas plasmids designed to provide cytoplasmatic localization of enzymes (pCYT) did not contain this signal.

Genes were amplified by PCR and placed under the control of the dual inducible pro moter PGAU and PGALIO- Coding genes sequences were then ligated using USER clon ing (Nour-Eldin et al (2010)) into the backbone of the pESC-URA, pESC-LEU, pESC- TRP, and pESC-HIS, vectors (Agilent Technologies) to construct the plasmids listed in Table 3.

Table 3 - List of plasmids used in the study

Designation of strains harboring plasmids used in this application

Plasmids (Table 3) were then used to transform yeast cells (Table 2) using the lithium acetate/PEG method. Transformants were selected by their respective auxotrophy on the corresponding minimal media.

Table 4- Strains harboring episomal vectors used in the study

Culture conditions

The yeast cells were first cultured on selective minimal media with glucose at 30°C over night. Complete minimal media consisted of 0.13% w/v dropout powder, 0.67% w/v yeast nitrogen base without amino acids with ammonium sulphate (YNB+AS), 2% w/v glucose. Dropout powder was purchased to lack leucine, histidine, uracil and tryptophan. When required, these four nutrients were added at 0.01-0.02% w/v. Cells were then harvested by centrifugation to remove medium and resuspended in selective minimal production media with an initial ODeoo _nm around 0.5. This media was used to induce galactose pro- moters, with additional raffinose as an alternative carbon source. Media composition: 0.13% w/v dropout powder, 0.64% w/v YNB+AS, 2% galactose, 1% w/v raffinose. When appropriate, the same four nutrients as above were added at 0.01-0.02% w/v. Isopropylmyristate (I PM) was added as an overlay corresponding to 10% of the culture volume. The cultures were grown at 30°C, 150 rpm, for the indicated time, the cells were then harvested by centrifugation, and the I PM phase recovered and analyzed using GC- FID and/or GC-MS.

EXAMPLES

Example 1 : Peroxisomal co-localization of a GPP synthase and a terpene synthase improves terpene production

Yeast strain construction

The Saccharomyces cerevisiae strains used were derived from the strain EGY48 ( Mat a, ura3, trp1, his3, 6xLexA operators: :LEU2). An engineered GPP synthase, ERG20 ^N127W from S. cerevisiae, and a monoterpene synthase (MTs) characterized as a limonene synthase, C/LimS from Citrus limon, were expressed under the control of the PGAU - PGALIO promoter.

First, the C/LimS was expressed in the cytosol of the strain EGY48 (strain CYTLimOI) As shown in Fig. 1 , this strain produced only 0.31 mg/L of limonene. A very similar result was observed when targeting C/LimS to the peroxisome by addition of the C-terminal PTS1 SKL (strain PERLimOI). Indeed, although GPP could be translocated from the cytosol into the peroxisome, the naturally very low cytosolic GPP pool is most likely to result in a very limited transport of this molecule into other compartments. Overexpressing the entire MVA pathway in the cytosol using the Enterococcus faecalis EfmvaE and EfmvaS genes (equivalent to Erg10p, Erg13p and HmgRp in yeast), Erg8p, Erg12p, Erg9p, Idilp and Erg20p ^N127W (as a GPP synthase) resulted in a 3.6-fold in crease in limonene production to 1.12 mg/L (Strain CYTLim02).

However, by only targeting the GPP synthase ERG20p ^N127W and C/LimS to the peroxi some (strain PERLim02) the limonene production was improved drastically by 32 fold, compared to CYTLim02, to reach 35 mg/L. This indicates that the precursors IPP and/or DMAPP can be transported into the peroxisome and can be converted to GPP in this organelle.

The significant increase in limonene production also shows that the peroxisome is effec tively able to act as a barrier and protect this newly formed GPP from the cytosolic ERG20p, thus allowing its uptake by C/LimS. The results are also shown in Figure 1.

Example 2: Localization of the mevalonate pathway into peroxisomes enhances production of terpenoids

To assess the possibility of harvesting peroxisomal acetyl-CoA for GPP production in this organelle, the eight MVA pathway enzymes were targeted to the peroxisome by ad dition of a C-terminal peroxisomal targeting signal of type 1 (PTS1) composed of the tri peptide SKL (see sequences part). Although the presence of the GPP synthase Erg20p ^N127W and C/LimS was sufficient to observe a notable jump in limonene production, stepwise localization of additional enzymes of the MVA pathway to the peroxisome im proved limonene production only slightly when the pathway was not complete in this organelle. However, when all eight enzymes were peroxisomally targeted, an additional 4-fold increase in limonene production to 141 mg/L was observed (Fig. 2). Moving the entire pathway from acetyl-CoA to limonene from the cytosol (strain CYTLim02) to the peroxisome (strain PERLim05) gives an overall 125-fold improvement in the production.

Example 3: Expanding the invention to other monoterpenoids. Construction of a yeast strain for improved camphene, pinene, (S)-(-)-limonene, (/?)-(+)-limonene, (/?)-(+)-linalool and sabinene production using the peroxisome.

In order to assess whether the peroxisome-linked improvements reported in examples 1 and 2 were specific to limonene production or were applicable to monoterpenes in gen eral, we targeted five additional MTPs either to the cytosol or to the peroxisome, along with Erg20p ^N127W , with overexpression of the rest of the MVA pathway. A camphene syn thase (SeCamS), a (S)-(-)-limonene synthase (/WsLimS), a (f?)-(+)-limonene synthase (C/LimS), a (f?)-(+)-linalool synthase (/WcLiS), an alpha-pinene synthase (PfPinS), and a sabinene synthase (SpSabS), were chosen and evaluated by determining the production titer of their major product. The same positive effect of peroxisome targeting of these five monoterpene synthase (MTSs) together with a GPP synthase (Erg20p ^N127W ) was ob served with an improvement of 14-fold (PERCam02 vs CYTCam02), 17-fold (PERPin02 vs CYTPin02), 22-fold (PERSab02 vs CYTSab02), 17-fold (PERLim27 vs CYTLim04), 125-fold (PERLim05 vs CYTLim02) and 20.5-fold (PERLinOI vs CYTLinOI), for cam phene, pinene, sabinene, (S)-(-)-limonene, ( )-(+)-linonene and ( )-(+)-linalool respec tively, compared to the corresponding cytosolic expression of the same enzymes (Figure 3).

Example 4: Improved monoterpene production using an optimized buffered synthetic minimal media.

A synthetic minimal defined media was used in order to assess production in an indus trially relevant media. This synthetic minimal defined media was composed of the follow ing: 5 g/L (NH ₄ ) ₂ S0 ₄ ; 3 g/L KH ₂ P0 ₄ ; 1 g/L MgS0 ₄ ^* 7 H ₂ 0; 0.0064 g/L D-biotin; 0.03 g/L nicotinic acid; 0.1 g/L thiamin HCL; 0.04 g/L D-panthothenic acid; 0.08 g/L myo-inositol; 0.02 g/L pyridoxine; 0.067 g/L tritriplex III; 0.067 g/L (NH ₄ ) ₂ Fe(S0 ₄ ) ₂ .6H ₂ 0; 0.0055 g/L CuS0 ₄ ; 0.02 g/L ZnS0 ₄ ; 0.02 g/L MnS0 ₄ ; 0.00125 g/L NiS0 ₄ ; 0.00125 g/L CoCL ₂ ; 0.00125 g/L boric acid; 0.00125 g/L Kl and 0.00115 Na ₂ Mo0 ₄ . The pH is hereby buffered with MES at a starting value of 6.3.

Strains PERLim05, PERGer02, PERPin02 and PERLinOI were used to determine mon oterpene production levels in this medium. Additionally, a strain named PERMyrOI, pro ducing beta-myrcene, was constructed by introducing the beta-myrcene synthase ObMyrS together with the GPP synthase ERG20N127 and the rest of the MVA pathway targeted to the peroxisome by fusion with the C-terminal tri-peptide SKL (SEQ ID NO: 26). All strain were cultivated in shake flask for 72h at 30 degrees with 10% isopropyl myristate overlay. The synthetic minimal defined medium described previously was used and supplemented with 4% galactose for growth and gene inductions. Strain PERLim05 produces 770 mg of limonene /L of culture, strain PERGer02 produces 1681 g of geraniol /L of culture, strain PERPin02 produces 250 mg of alpha-pinene /L of culture, strain PERLinOI produces 547 mg of linalool /L of culture and strain PER- Myr01 produces 251 mg of myrcene /L of culture. This represents a 5.4-fold, 5.2-fold, 5.1-fold and 2.9-fold improvement for strains PERLim05, PERGer02, PERPin02 and PERLinOI respectively, compared to the same strains cultivated in the un-buffered com plete media used in example 3. Example 5: High levels of (+)-limonene and geraniol production by a combined strategy of genomic integration and plasmid-based expression of the MVA pathway genes together with a GPPS and LimS or GES.

A single copy of each gene of the MVA pathway targeted to the peroxisome was inte- grated into the genome of strain EGY48 together with ERG20p ^N127W and C/LimS/tObGES giving strain PERLim06 and strain PERGerOI . Furthermore, an additional copy of each gene of the MVA pathway, Erg20p ^N127W and C/LimS ortObGES were introduced on plas mids in strains PERLim06 and strain PERGerOI giving strains PERLim07 and PERGer02, respectively. In order to determine the maximum limonene and geraniol titer achievable by compart mentalizing the pathway into the peroxisome, semi-continuous fed-batch experiment were carried out with strains PERLim07 and PERGer02. The cultures were fed every 48 h with 40 g/L galactose and 20 g/L raffinose and the pH was adjusted to 4.5. The I PM layer was also harvested every 48h to measure monoterpenes production. Fed batch flask culture with strains PERLim07 and PERGer02 resulted on a continuous accumulation of limonene and geraniol that was proportional, to a great extent, to the amount of biomass formed. After 700 h, titers of 2575 mg of limonene /L of culture (Figure 4A) and 5516 mg of geraniol /L of culture (Figure 4B) were determined. Example 6: Peroxisomal localization of a fusion protein comprising of a GPP synthase domain and a terpene synthase domain increases terpenes production. The possibility of using a single polypeptide having both a GPP synthase and a terpene synthase activity is investigated in this example of the present invention. Such a bifunc tional enzyme could be already found in nature or synthetically created. For exemplifi cation purpose, we created such a polypeptide by fusing a GPP synthase domain and a terpene synthase domain together and targeted it to the peroxisome.

To do so, the GPP synthase Erg20p ^N127W was fused to the terpene synthase C/LimS by a linker polypeptide comprised of five glycine-serine repeats (5xGS). The GPP syn thase domain can be at the N-terminal end of the protein and the terpene synthase do main at the C-terminal end of the protein. Alternatively, the terpene synthase domain can be at the N-terminal end of the protein and the GPP synthase domaine at the C- terminal end of the protein. In this example, both these configurations were tested. Construction of those two synthetic enzymes gave two new polypeptides described in SEQ ID NO 15 and SEQ ID NO 16. Both sequences were subsequently introduced into an expression vector for yeast to give plasmids pPER15 and pPER16 and the strain PERLim06 was transformed with either one of the two plasmids to give PERLimlO and PERLiml 1. After culturing PERLimlO and PERIim11 for 72 h at 30°C with an I PM over lay, limonene production was measured and compared to the one of strains PERLim08 and PERIim09. As seen in Figure 5, limonene production obtained by peroxisomally targeting one or the other bifuctional GPP-terpene synthase fusion (PERLimlO and PERLiml 1) is similar or better compared to the production observed when peroxiso mally targeting a GPP synthase and a terpene synthase as separate enzyme (PER- Lim09). Furthermore, a 16-fold and an 11-fold improvement in limonene production are observed when the bifuctional GPP-terpene synthase is targeted to the peroxisome compared with targeting only the GPP synthase activity to the peroxisome (PER- Lim08).

Example 7A: Efficient production of frans-isopiperitenol, precursor of menthol.

We further evaluated the contribution of the present invention in the production of trans- isopiperitenol, which is the precursor of the high-value compound menthol. We introduced the limonene-3-hydroxylase from Mentha spicata (/WsLim3H; Q6IV13.1) into strain PERLim27, together with the cytochrome P450 reductase (fcCPR/POR) from Taxus cuspidata to give strain PERLim30 or an empty vector (pESC-Leu) to give PER- Lim29. As a comparison we introduced the limonene-3-hydroxylase from Mentha spi cata (/WsLim3H; Q6IV13.1) into strain CYTLim04, together with the cytochrome P450 reductase (fcCPR/POR) from Taxus cuspidata to give strain CYTLim06. After 72h of growth in complete minimal media, production of trans- isopiperitenol was evaluated by GC-FID of the culture extracts. As shown in Figure 6A, both (-)-limonene and trans- iso piperitenol could be extracted from the strain PERLim30, while only limonene was de tected in strain PERLim29 lacking the limonene-3-hydroxylase /WsLim3H and the cyto chrome P450 reductase fcCPR (Figure 6A). Reduction of the (-)-limonene recovered from the PERLim30 culture compared to PERLim29 is in accordance to a 37% conver- sion into trans- isopiperitenol (19.24 mg/L) (Figure 6A). However when limonene pro duction was done in the cytosol in strain CYTLim06, only 0.28 mg/L trans-isopiperitenol was obtained, corresponding to a limonene conversion of only 14%. These results demonstrate that increased limonene production in the peroxisome allow for a signifi cant amount of it being channeled through the endoplasmic reticulum (ER), where it can be hydroxylated by /WsLim3H.

Example 7B: Efficient production of 8-hydroxygeraniol, precursor of iridoids and monoterpene indole alkaloids.

We further evaluated the contribution of the present invention in the production of 8-hy- droxygeraniol, which is the precursor of large groups of high-value compounds, includ ing the iridoids and the monoterpene indole alkaloids. We introduced the geraniol 8-hy- droxylase from Catharanthous roseus (CrG80H; CYP76B6) into strain PERGer02, to gether with the cytochrome P450 reductase (C/CPR/POR) from the same species to give strain PERGer04 or an empty vector (pESC-Leu) to give PERGer03. After 72h of growth in complete minimal media, production of 8-hydroxy-geraniol was evaluated by GC-FID of the culture extracts. As shown in Figure 6B, both geraniol and 8-hydrox- ygeraniol could be extracted from the strain PERGer04, while only geraniol was de tected in strain PERGer03 lacking the geraniol 8-hydroxylase CrG80H and the cyto chrome P450 reductase Ci CPR (Figure 6B). Reduction of the geraniol recovered from the PERGer04 culture compared to PERGer03 is in accordance to a partial conversion into 8-hydroxy-geraniol (Figure 6B). These results demonstrate that increased geraniol production in the peroxisome allow for a significant amount of it being channeled through the endoplasmic reticulum (ER), where it can be hydroxylated by C/G80H.

Example 8: Efficient production of cannabinoids by targeting a GPP synthase and a geranyldiphosphate:olivetolate geranyltransferase to the peroxisome.

The invention was evaluated for its applicability in the production of another group of GPP-derived high-value compounds, beyond monoterpenoids and monoterpene indole alkaloids, that of cannabinoids. In the cannabinoid biosynthetic pathway, olivetolic acid (OA) is prenylated by GPP to form cannabigerolic acid (CBGA) via the action of a dedi cated geranyltransferase. CBGA represents a key step in the pathway because it is the last common precursor to various types of cannabinoids, such as tetrahydrocannabinolic acid (THCA) and cannabidiolic acid (CBDA). In yeast, this prenylation step represents a major bottleneck in the process of producing high titers of cannabinoids because of the limited pool of GPP in the cytosol. In this example, the GPP synthase Erg20p ^N127W and the geranyldiphosphate:olivetolate geranyltransferase CsPT4, from C. sativa, were targeted to the yeast peroxisome, using the C-terminal targeting signal SKL, for CBGA production. Both genes were introduced into the strain PERMvaOI under the control of the inducible promoters PGALI and PGAUO, to give strain PERCanOl After 72h of growth in complete minimal media under galac tose-induced conditions and supplemented with various concentration of olivetolic acid (0.05 mM, 0.1 mM, 0.25 mM or 0.5 mM), production of CBGA was analyzed by LC-MS. Cells were disrupted and CBGA extracted from the cell fraction (pellet) using ethyl ace tate/formic acid (0.05 % v/v) in a 1:1 ratio and glass bead beating. The organic layer was separated by centrifugation and evaporated using a SpinVac. The remaining dry fraction was dissolved in methanol and filtered through a 0.22pm pore size PVDF filter. Samples were diluted 10 times prior to LC-MS analysis.

As shown in Figure 7A, both OA and CBGA could be extracted from the strain PER CanOl , while only OA was detected in strain PERMvaOI lacking the GPP synthase ERG20p ^N127W and the geranyldiphosphate:olivetolate geranyltransferase CsPT4. Reduc tion of the OA recovered inside the cells in PERCanOl compared to PERMvaOI is in accordance to a partial conversion into CBGA. These results demonstrate that 1) OA can be transported and/or diffuse into the peroxisome, 2) CsPT4 is active in the peroxi some and 3) the pool of GPP is sufficient in the peroxisome to allow efficient OA prenyl- ation.

Additionally, an alternative N-terminal peroxisomal targeting signal was examined for CsPT4 (SEQ ID NO: 29). PTS2-CsPT4 was introduced in strain PERMvaOI together with the GPP synthase Erg20p ^N127W -SKL giving both enzymes peroxisomal localization. This new strain called PERCan02 was assessed as previously for OA consumption and CBGA production. For comparison, CsPT4 and Erg20p ^N127W were introduced in strain CYTMvaOI to give strain CYTCanOI for cytosolic CBGA production. As shown in Figure 7B, strain PERCan02 produced 82.3 mg/L CBGA, 19.5-fold more than strain CYTCanOI with only 4.2 mg/L CBGA. References:

Ellerstrom M et al (1992): Cloning of a cDNA for rape chloroplast 3-isopropylmalate dehydrogenase by genetic complementation in yeast. Plant Mol Biol, 18, 557-560. Ignea C. et al (2011): Improving yeast strains using recyclable integration cassettes for the production of plant terpoenoids. Microb. Cell. Fact, 10, 4 (3)

Ignea C. et al (2014): Engineering monoterpene production in yeast using a synthetic dominant negative geranyl diphosphate synthase. ACS Synth. Biol. 3, 298-306.(53) Ignea C. et al (2018): Synthesis of 11 -carbon terpenoids in yeast using protein and met- abolic engineering. Nature Chemical Biology 14, 1090-1098.

Nour-Eldin H.H et al (2010): USER cloning and USER fusion: The Ideal cloning tech niques for small and big laboratories. In Fett-Neto A (eds.) Plant Secondaty Metabolism Engineering. Methods in Molecular biology (Methods and Protocols), vil 643. Humana Press, Totowa, NJ. Orlova I et al. (2009): The Small Subunit of Snapdragon Geranyl Diphosphate Synthase Modifies the Chain Length Specificity of Tobacco Geranylgeranyl Diphosphate Synthase in Planta. Plant Cell; 21, 4002-4017.

Philips M. A. et al (2003): cDNA isolation, funbctional expression , and characterization of (+)- a-pinene synthase and (-)-a-pinene synthase form lobolly pine (Pinus taeda): Stereocontrol in pinene biosynthesis. Arch. Biochem. Biophys. 411 , 267-276. (57)

Thomas B.J. and R. Rothstein (1989): Elevated recombination rates in transcriptionally active DNA. Cell 56, 619-630 (4)