Field of the invention This invention relates to a eukaryotic host and a method for producing glycolic acid using eukaryotic host cells, especially cells of a genetically modified fungal host. Further this invention relates to glycolic acid obtained using the method described here and the use of eukaryotic cells in production of glycolic acid. Description of related art

Glycolic acid is a widely used chemical. It has applications within cosmetic field and organic synthesis of polyglycolic acid and other biocompatible polymers. In textile industry it is used as a dyeing and tanning agent and in food processing as a flavouring agent and as a preservative. Traditionally glycolic acid is usually derived from fossil fuels but it is possible to derive glycolic acid from bio mass by using metabolically engineered E.coli.

Glycolic acid is the smallest a-hydroxy acid and it is not naturally produced by microorganisms at least in feasible quantities. This is different to lactic acid which is also a small a-hydroxy acid, but this acid is produced naturally by a large number of microorganisms. To produce glycolic acid from carbohydrates in a fermentative way using microorganisms requires genetically engineered microorganisms. Previously Escherichia coli strains were described that were engineered to produce glycolic acid from glucose (WO 2007/141316 A2, WO2010/108909 Al and WO2011/036213). In these strains the glyoxylate cycle was disrupted by deleting the malate synthase and overexpressing an endogenous glyoxylate reductase so that the glyoxylate produced by the isocitrate lyase could be converted to glycolic acid. Also other modifications were done to channel more carbon to the glyoxylate cycle or to prevent the utilization of glyoxylic acid. E. coli strains engineered in this way were producing glycolic acid from glucose.

A characteristic of E. coli and bacteria in general is that the inner space of the cell or the cytosol of the organism is not compartmentalized. This is different in eukaryotic organisms. Eukaryotic cells, including yeast and filamentous fungal cells, have different compartments that are separated by membranes, such as vacuoles, mitochondria or peroxisomes. In difference to bacteria the cellular reactions take place in different compartments. For example the reactions of the TCA cycle (Krebs cycle, citric acid cycle) are located in the mitochondria, whereas the different reactions of the glyoxylate cycle are in separate compartments. According to current understanding isocitrate lyase and malate synthase are peroxisomal in filamentous fungi. In yeasts such as S. cerevisiae isocitrate lyase and malate synthase are cytosolic. The citrate synthase is located in the peroxisomes and in the mitochondria, the aconitase can be in mitochondria and cytosol and the succinate dehydrogenase is exclusively in mitochondria. The distribution of the enzyme activities to different compartments and inability of the metabolites to freely travel between the compartments makes a fundamental difference to metabolic pathways and consequently the approaches to engineer these pathways by means of genetic engineering are essentially different. It is therefore not clear whether the teaching disclosed in the WO2007/141316 A2, WO2010/108909 Al and WO2011/036213 can be applied to eukaryotic microorganism.

Using E. coli for the production of glycolic acid has several drawbacks. E. coli is requiring a complex growth medium which is in general more expensive than growth media for yeast or filamentous fungi. Another drawback is that E. coli is operating close to neutral pH. This increases the contamination risk and in the case of acid production it requires base to neutralize the acid. Further, in neutral conditions the produced glycolic acid is in dissociated form, which is a disadvantage in downstream processing.

Objects and summary of the invention

It is an aim of the invention to provide a eukaryotic host organism and a method for producing glycolic acid using said host. Particularly, the aim is to provide a production host and a method that allows using mildly acidic or acidic conditions. These and other objects are achieved by the present invention as hereinafter described and claimed. The first aspect of the invention is a eukaryotic production host. Characteristic to said host is that it is genetically modified to express glyoxylate reductase gene and produce glycolic acid. The second aspect of the invention is a method for producing glycolic acid. According to the invention the method comprises the steps of culturing a eukaryotic host encoding glyoxylate reductase activity and optionally recovering the glycolic acid from said medium.

The third aspect of the invention is a product obtained by culturing the production host of the invention or by using the method of the invention.

The fourth aspect of the invention is the use of eukaryotic host in production of glycolic acid.

Some of the embodiments of the invention are disclosed in the dependent claims. Brief description of the drawings

Figure 1. is a diagrammatic representation of plasmid pUG6 containing G418 gene, KanMX, between loxP sites, which can be used to loop out the KanMX resistance gene from the genomic integration site after Cre recombinase activation. Figure 2. is a diagrammatic representation of plasmid pl431 which is similar as the plasmid in Figure 1 except that instead of KanMX marker plasmid has BLE marker for bleomycin (Zeocin) resistance.

Figure 3. is a diagrammatic representation of plasmid p2159 which is a constitutive expression plasmid for yeast. It has TPI promoter and URA3 marker gene.

Figure 4. is a diagrammatic representation of plasmid p3466 where glyoxylate reductase GLYR1 is expressed under TPI promoter. Figure 5. is a diagrammatic representation of plasmid p3532 where kanMX marker is between the HO-L and HO-R sites which can be used for targeted integration in S. cerevisiae. Figure 6. is a diagrammatic representation of plasmid pRS426 which has URA3 marker gene and can be used as a backbone vector for constructing expression and deletion cassettes by homologous recombination in S. cerevisiae. Figure 7. is a diagrammatic representation of S. cerevisiae isocitrate dehydrogenase (IDP2) deletion cassette which in addition to the deletion over expresses S. cerevisiae isocitrate lyase (ICL1) gene. The cassette has KanMX marker between the /ox -sites.

Figure 8. is a diagrammatic representation of plasmid p609 which is a constitutive expression plasmid for yeast. It has mADH promoter and URA3 marker gene.

Figure 9. is a diagrammatic representation of pMLV94 plasmid which has C. krusei PGK promoter and S. cerevisiae ADH1 terminator. Figure 10. is a diagrammatic representation of pMLV128 plasmid for C. krusei which has hygromycin resistance, hph, marker gene.

Figure 11. is a diagrammatic representation of pMV45-AnMLS-del-pyrG plasmid for A. niger. The plasmid has flanks for malate synthase deletion and pyrG marker.

Figure 12. is a diagrammatic representation of pyrG-MLS deletion cassette for A. niger. It has flanks for malate synthase deletion and pyrG marker.

Figure 13. is a diagrammatic representation of JKpl-hph-Glyr plasmid for A. niger. The plasmid is an overexpression vector for GLYR1 gene and it has hygromygin resistance, hph, marker gene in addition to URA3 gene.

Figure 14. shows the production of glycolic acid in modified K. lactis strain where malate synthase and isocitrate dehydrogenase are deleted and GLYR1 expressed in a plasmid (H3986). The fermentation is done in synthetic medium containing 2 % xylose and 2 % ethanol. Also biomass and xylose and ethanol amounts are presented in the figure.

Figure 15. shows the pH curve of the K. lactis strain H3986 fermentation presented in Figure 14 which was done in unbuffered media. Figure 16. shows the production of glycolic acid in modified K. lactis strain where malate synthase and isocitrate dehydrogenase are deleted and GLYR1 expressed in a plasmid (H3986). The fermentation is done in synthetic medium containing 2 % xylose and 2 % ethanol and 5 g Γ ¹ ethanol was added daily to the culture. Also biomass and xylose and ethanol amounts are presented in the figure.

Figure 17. shows the pH curve of the K. lactis strain H3986 fermentation presented in Figure 16 which was done in unbuffered media.

Figure 18. is the schematic representation of the metabolic route leading to glycolic acid production.

Figure 19. is a diagrammatic representation of pMLV97 plasmid for C. krusei which has EX marker gene.

Figure 20. is a diagrammatic representation of pMLV97 plasmid for C. krusei which has MEL marker gene and where MLS flanks and GLYR1 expression cassette have been ligated.

Description of the preferred embodiments of the invention

This invention relates to a eukaryotic host which is genetically modified to express glyoxylate reductase gene and produce glycolic acid. The metabolic route leading to glycolic acid production is illustrated as Figure 18.

The eukaryotic host may be any eukaryotic organism but most usually it is a cell, preferably a micro-organism. In one embodiment the cell is a fungal cell. When compared to e.g. E. coli fungal cells are generally more tolerant to changing culture conditions such as low pH, lignocellulosic hydrolysates, impurities, degradation products of hydrolytic enzymes and toxins. They also have lower nutritional requirements compared to bacteria most commonly used in industrial processes. Thus fungal production is an economic approach to microbial production of glycolic acid. In one embodiment the host is a yeast cell. Yeasts like Saccharomyces cerevisiae or filamentous fungi like Aspergillus niger are naturally operating at very acidic pH. Yeast produced in fermentations is often used as cattle feed which benefits the economics of the process. Bacteria are generally not used as cattle feed but deemed a waste.

Suitable yeasts are for example the genera Saccharomyces, Kluyveromyces, Candida, Scheffersomyces, Pachysolen and Hansenula. Yeast species of particular interest include S. cerevisiae, S. exiguus, K. marxianus, K. lactis, K. thermotolerans, C. sonorensis, C. krusei (also known as Issachenkia orientalis and Pichia kudriavzevii), C. shehatae, Pachysolen tannophilus and Scheffersomyces stipitis.

In one embodiment the host is a filamentous fungus. One particular advantage of using filamentous fungi to produce glycolic acid is that it can be done in a consolidated process, meaning that the fungus produces the enzymes for biomass hydrolysis and ferments the resulting sugars in the same process.

Suitable filamentous fungi hosts are for example of the genera Aspergillus, Trichoderma, Monascus, and Penicillium. Fungal species of particular interest include A. niger, A. ficuum, A. phoenicis, T. reesei, T. harzianum, M. ruber, and P. chrysogenum. Filamentous fungi allow using only partially hydrolysed biomass as a carbon source which is benefit if some lignocellulosic waste is used as a carbon source. A. niger is also a well known citric acid producer and by metabolic engineering it can be made an efficient host for glycolic acid production. In a preferred embodiment the host includes those of the species S. cerevisiae, S. exiguus, K. marxianus, K. lactis, C. krusei and A. niger.

The host organism of this invention is genetically modified and may contain also other genetic modifications than those specifically described herein. Methods for making modifications of these types are generally well known and are described in various practical manuals describing laboratory molecular techniques.

A phrase "genetically modified to express" as used herein covers the cells where a protein- encoding polynucleotide has been transformed in such a manner that the host is capable of producing an active protein or where a promoter region of a cell has been modified to allow or enhance the expression of a heterologous or homologous gene encoding glyoxylate reductase activity. In one embodiment of the invention the host is capable of producing glycolic acid at pH below 6, preferably below 5.5, below 5.0, below 4.5, below 4.0, below 3.5, below 3.0, below 2.5 and even below 2.0, until pH 1.5. Acidic or mildly acidic culture conditions reduce risk of contaminations and thus improve the process hygiene and safety. Each contaminated large scale fermentation results in direct loss of money. For the downstream processing it is of advantage when the glycolic acid is produced at acidic pH. Glycolic acid has a pKa of 3.83 meaning that below a pH 3.8 it is mainly in the acid form which facilitates the separation.

In one embodiment of the invention the host is capable of producing glycolic acid in non- buffered culturing conditions. In other words the strain is tolerant to decreasing pH during the cultivation process. This simplifies the culturing process and thereby reduces costs. However, this characteristic naturally does not exclude possibility to regulate the pH conditions using bases (or acids) or even buffering agents. In one embodiment the glyoxylate reductase gene is a heterologous gene to a host. Especially when the Km value of the homologous gene is high it is beneficial to replace the gene by heterologous gene having lower Km value. The heterologous gene is preferably obtained from a plant source, gene obtained from Arabidopsis thaliana (Hoover et al. 2007) being the most preferred. It is also possible to use endogenous or native genes for production and modify the cell to overproduce glyoxylate reductase.

In one embodiment the glyoxylate reductase enzyme is characterized by having an EC number EC 1.1.1.79 (NADP+) or EC 1.1.1.26 (NAD+). Enzymes of class EC 1.1.1.79 (glyoxylate :NADP+ reductases, CAS 37250-17-2) catalyze reduction of glyoxylate using NADPH into glycolic acid and NADP or reduces hydroxypyruvate to gly cerate. A. thaliana glyoxylate reductase designated here as "GLYR1" is one example of this class. Enzymes of class EC 1.1.1.26 (glyoxylate:NAD+ reductases, CAS 9028-32-4) reduces glyoxylate to glycolic acid or hydroxypyruvate to D-glycerate. In the European chemical legislation the enzymes are defined according to the type of reaction they catalyze. Each enzyme is given a systematic name and an IUBMB (International Union of Biochemistry and Molecular Biology) number such as EC 1.1.1.26 (\vww.chem.qmul.ac.uk¾bmb/enzyme/). It is known that different enzymes can have the same enzyme activity but have very different amino acid sequences. The Glyrl from Arabidopsis thaliana (GenBank accession number AY044183, EC 1.1.1.79) and the glyoxylate reductase from Thermus thermophilus HB27 (Gene identifier TT C0431, Protein ID AAS80779.1) Biotechnology Progress (2008) 24 (2) 321-325 are examples for this. Both enzymes show glyoxylate reductase activity when heterologously expressed but when the amino acid sequences are aligned in a CLUSTALW2sequence alignment the identities are only 12%. Comparison of the identity percentages of different glyoxylate reductases is presented in Table 1.

Table 1. Comparison of different glyoxylate reductase protein sequences done with ClustalW2 alignment (Cost matrix: BLOSUM, Gap open cost 10, Gap extend cost 0,1)

In one embodiment the host cell comprises the genes encoding proteins characterized by SEQ ID NO: 3, or SEQ ID NO: 71 or a sequence having at least 40%, 50%, 55%, 60% or identity, preferably at least 65%, 70%, 75% or 80% identity, 85% identity, more preferably at least 90% identity and most preferably at least 95% or even 98% identity to gene encoding the polypeptide having SEQ ID NO: 3 or SEQ ID NO: 71 , or an active fragment thereof. In preferred embodiment the encoded enzymes has characteristics of SEQ ID NO: 3.

As used in the present context the term "identity" refers to the global identity between two amino acid sequences compared to each other from the first amino acid encoded by the corresponding gene to the last amino acid. For the purposes of the present invention identity is preferably determined by means of known computer programs using standard algorithms. An example of such a program is NCBI BLAST; BLASTp (comparison of known protein sequences, amino acids), BLASTn (comparison of nucleic acid sequences), BLASTx (comparison of translated nucleic acid sequences against know protein sequences).

"An active fragment" means a fragment having all the parts needed for completing the function typical for the protein.

It is understood by the skilled reader that the gene must be operably linked to the sequences regulating the expression of the gene. Two DNA sequences are operably linked when the function of the promoter results in transcription. An operable linkage is a linkage in which a sequence is connected to a regulatory sequence (or sequences) in such a way as to place expression of the sequence under the influence or control of the regulatory sequence.

In one embodiment the gene is genetically optimized. It is understood by a skilled man that heterologous gene obtained by a different organism may need genetic optimization in order to properly function in the host cell. In one embodiment the heterologous gene is genetically optimized to fit the host systems.

Standard molecular biology methods can be used in the cloning of glyoxylate reductase or other overexpressed genes. The basic methods used like isolation and enzyme treatments of DNA, E. coli transformations made for plasmid constructions, the isolation of the vectors or fragments containing the said gene and amplification of fragments by PCR are described in the standard molecular biology handbooks e.g. Sambrook et al. (1989) and Sambrook and Russell (2001). Genetic modification of the host fungus is accomplished in one or more steps via the design and construction of appropriate vectors and transformation of the host fungus with those vectors. Electroporation, protoplast-PEG and/or chemical (such as calcium chloride- or lithium acetate-based) transformation methods can be used. The heterologous gene can be stably introduced into the genome of the host cell. Stable transformation is obtained when the expression cassette is integrated to the chromosomal DNA of the host. Targeted integration can be accomplished by designing a vector having regions that are homologous to the upstream (5'-) and downstream (3'-) flanks of the target gene. Either or both of these regions may include a portion of the coding region of the target gene. The GLYRl cassette (including suitable promoters and terminators if different from those of the target gene) and/or selection markers (with suitable promoters and terminators) will reside between the regions that are homologous to the upstream and downstream flanks of the target gene. Stable transformation is preferred as no selection pressure is needed during cultivation but alternatively also episomal plasmids and other non- integrated constructs are within this invention.

The use of native (homologous to the host cell) or non-native (heterologous to the host cell) promoters and terminators, together with respective upstream and downstream flanking regions, can permit the targeted integration of the GLYRl or any other gene mentioned above, or any other gene further described below, into specific loci of the host cell's genome, and for simultaneous integration of the said gene and deletion of a native gene, such as, for example, a malate synthase (e.g. MLS1 or DAL7) encoding gene.

The exogenous glyoxylate reductase gene may be maintained on a self-replicating plasmid, integrated randomly into the host cell's genome or inserted at one or more targeted locations. Examples of targeted locations include the locus of a gene that is desirably deleted or disrupted, such as the MLS1 gene in S. cerevisiae or the malate synthase gene in A. niger. In one embodiment the cell has been further modified to overexpress the gene encoding glyoxylate reductase.

The overexpressed genes - such as glyoxylate reductase (e.g. GLYRl gene from A. thaliana), isocitrate lyase, aconitase, citrate synthase, fumarate reductase, acetyl-coenzyme A synthetase or a gene regulating glyoxylate cycle is under the control of a promoter and a terminator, both of which are functional in the modified fungal cell. As used herein, the term "promoter" refers to a sequence located upstream (e.g., 5') to the translation start codon of a structural gene and which controls the start of transcription of the structural gene. Similarly, the term "terminator" refers to a sequence located downstream (e.g., 3') to the translation stop codon of a structural gene and which controls the termination of transcription of the structural gene. A promoter or terminator is "operatively linked" to a structural gene if its position in the genome relative to that of the structural gene is such that the promoter or terminator, as the case may be, performs its transcriptional control function.

The genetically modified fungus may contain a single copy or multiple copies of the glyoxylate reductase gene (e.g. GLYR1 gene) or any other gene mentioned above, or any other gene further described below. If multiple copies of the glyoxylate reductase gene are present, from 2 to 10 or more copies may be integrated into the genome, or > 100 copies may be present on self-replicating plasmids. If multiple copies of the glyoxylate reductase gene are integrated into the genome, they may be integrated at a single locus (so they are adjacent each other), or at several loci within the host's genome. It is possible for different glyoxylate reductase genes to be under the control of different types of promoters and/or terminators.

In one embodiment the production host has been further modified by increasing the flux towards the glyoxylate cycle. The flux towards the cycle can be increased e.g. by overexpressing isocitrate lyase, aconitase, citrate synthase or fumarate reductase. Heterologous fumarate reductase would convert succinate into fumarate, which can be then metabolized further and used again in the glyoxylate cycle. Increase of the glyoxylate flux may require also reducing activity of the enzymes consuming the intermediates of the cycle, e.g. attenuating expression of malate synthase(s), NAD and/or NADP dependent mitochondrial and/or cytosolic isocitrate dehydrogenase(s). Increase of flux towards the cycle results in more efficient production of gly colic acid.

In this connection term attenuating refers to partial or total deletion or knock-out of the said gene, modification of regulatory regions of the said gene in order to decrease its activity or silencing or decreasing the activity of the said gene by any methods without restricting to those mentioned above.

The yield of the glycolic acid can be increased by several ways. Increased yield saves the fermentation costs and enhances the down-stream processing.

In one embodiment the production host has been further modified by modifying the genes regulating the glyoxylate cycle i.e. genes involved in negative regulation of glucoserepressible genes. This can be done for example by activating genes such as CATS in S. cerevisiae or in K. lactis by mutation or overexpression or by attenuating genes such as REG1 in S. cerevisiae.

In one embodiment the production host has been further modified by improving NADPH availability. This can be done e.g. by overexpressing cytosolic aldehyde dehydrogenase such as ALD6 in S. cerevisiae or deleting phosphoglucose isomerase gene such as PGI1 in S. cerevisiae or K. lactis. Glyoxylate reductases are usually NADPH dependent and thus need NADPH for functioning. Glyoxylate reductase can be also NADH dependent and in these cases the improvement of NADPH availability is unnecessary. In one embodiment, the production host has been further modified to control ethanol production and utilization.

Thus, in one embodiment the cell has been further modified by attenuating genes involved in alcohol production in yeast or production of hydrolytic enzymes in filamentous fungi. Alcohol production competes with the glyoxylate cycle and it can be preferred to decrease it. This can be done e.g. by overexpressing one or more of pyruvate carboxylases such as PYC2, PYC1 in S. cerevisiae and acetyl-coenzyme A synthetase genes such as ACS1 in S. cerevisiae; or reducing expression of gene encoding alcohol dehydrogenase, such as ADH2 gene of S. cerevisiae.

In another embodiment, ethanol is first produced from sugar source as hexose or pentose and then ethanol is further utilized to glycolic acid. In this latter case the deletion of alcohol dehydrogenase is not desired but instead ethanol utilization should be directed towards the glyoxylate cycle and glycolic acid production. If the host organism chosen for glycolic acid production is capable of further utilizing glycolic acid it is necessary to also attenuate the genes responsible for these enzyme reactions. E.g. some microbes are known to have glycolate oxidase which oxidases glycolic acid to glyoxylate. In addition glycolic acid production can also be reduced if the host has enzymes other than glyoxylate reductase utilizing glyoxylate. Deletion of malate synthase was already described above as an example how glyoxylate cycle needs to be modified to produce glycolic acid. In addition to malate synthase some fungi are known to have e.g. glyoxylate oxidase, which would need to be deleted in order to produce glycolic acid efficiently.

In one embodiment the cell has been further modified by any combination of the above described modifications. One embodiment of this invention is also a method for producing glycolic acid. In the method a eukaryotic cell encoding glyoxylate reductase is cultured in conditions allowing the expression the glyoxylate reductase gene, producing and secreting the glycolic acid and optionally recovering the glycolic acid from said medium. In the production process of the invention, the fungus is cultivated in a growth and production medium that includes a carbon source and typical nutrients required by the particular host, including but not limited to a source of nitrogen (such as amino acids, proteins, inorganic nitrogen sources such as ammonia or ammonium salts), and various vitamins and minerals. Alternatively, more than one different carbon source can be used.

Carbon source contains a sugar which can be hexose or pentose. Hexose can be e.g. glucose, fructose, mannose, or galactose and oligomers of glucose such as maltose, maltotriose, isomaltotriose, starch or cellulose. Examples of pentoses are xylose, xylan or other oligomer of xylose, and preferably also other carbon containing compounds to provide for growth and energy. The medium may also contain ethanol, glycerol, acetate, or amino acids, or any mixture thereof, preferably ethanol or acetate, most suitably ethanol, which further components can also function as carbon sources. The carbon substrates may be provided as pure substrates or from complex sources. The xylose containing sugars are suitably hydrolysates of plant biomass e.g. hemicellulose- containing biomass, such as lignocellulose. In addition, the medium may consist of or contain complex, poorly defined elements, such as would be present in relatively inexpensive sources like black liquor, corn steep liquor or solids, or molasses. In case of oligomeric sugars, it may be necessary to add enzymes to the fermentation broth in order to digest these to the corresponding monomeric sugar. It is also possible to use production hosts, such as filamentous fungus hosts that secrete hydro lytic enzymes enhancing the production of fermentative sugars.

Other fermentation conditions, such as temperature, cell density, selection of nutrients, and the like are not considered to be critical to the invention and are generally selected to be suitable for the cell used and to provide an economical process. Temperatures during each of the growth phase and the production phase may range from above the freezing temperature of the medium to about 50°C, although the optimal temperature will depend somewhat on the particular micro-organism. A preferred temperature, particularly during the production phase, is from about 25 to 30°C.

The pH of the process may or may not be controlled to remain at a constant pH, but should be between 1.5 and 6.5, depending on the production organism. In one embodiment the culturing pH is below 6, preferably below 5.5, below 5.0, below 4.5, below 4.0, below 3.5, below 3.0, below 2.5 and even below 2.0, until pH 1.5. Depending on the production organism the lower limit of the pH may vary between 1.5 and 4. Preferred pH of the culture media is 1.5 to 5, more preferably 2 to 4 and most preferably 2 to 3. In one embodiment the culture medium contains no buffering agent.

According to an embodiment of the invention, the pH is controlled to a constant pH of 3.5 to 5.5. Suitable buffering agents for regulating or buffering pH are basic materials that neutralize glycolic acid as it is formed, and include, for example, calcium hydroxide, calcium carbonate, sodium hydroxide, potassium hydroxide, potassium carbonate, sodium carbonate, ammonium carbonate, ammonia, ammonium hydroxide and the like. In general, those buffering agents that have been used in conventional fermentation processes are also suitable here. It is within the scope of the invention, however, to allow the pH of the fermentation medium drop from a starting pH that is typically 6 or higher, to below the pKa of the acid fermentation product, such as in the range of about 3 to about 4.

The fermentation is conducted aerobically or microaerobically. If desired, specific oxygen uptake rate can be used as a process control. The process of the invention can be conducted continuously, batch- wise, or some combination thereof.

During the fermentation gly colic acid is excreted out from the cells into the growth medium from which it may be recovered without disrupting the cells.

In one embodiment of the invention a cell described above is cultured and glycolic acid is recovered. Glycolic acid can be recovered from the fermentation medium by e.g. ion exchange chromatography or reactive extraction or it can be polymerized in the fermentation medium and recovered thereafter.

One embodiment of the invention is the use of eukaryotic host cells or eukaryotic organisms in production of glycolic acid or as a starting organism for preparation a production host suitable for production of glycolic acid. In preferred embodiment the eukaryotic organism is a fungal cell, preferably a yeast or filamentous fungus. The eukaryotic organism may be modified as described here and is suitable for the method as described here.

The invention is illustrated by the following non-limiting examples. It should be understood, however, that the embodiments given in the description above and in the examples are for illustrative purposes only, and that various changes and modifications are possible within the scope of the invention.

Examples Example 1A. Construction of a S. cerevisiae strain able to produce glycolic acid (selfreplicating plasmids)

The Saccharomyces cerevisiae malate synthase gene MLS1 (GenBank accession number NM 001182955) and DAL7 (GenBank accession number NM 001179553) was knocked out from the S. cerevisiae strains CEN.PK2 and CEN.PK113-1 A. MLSl deletion cassette was constructed by PCR using primer pair MLSlfwd (SEQ ID NO: 5) and MLSlrev (SEQ ID NO: 6) to amplify loxP-KanMX-loxP fragment of the pUG6 plasmid (Figure 1) (Guldener et al, 1996). In addition to the 20 / 23 bp long body annealing to loxP-KanMX- loxP fragment of the pUG6 plasmid, primers contained 40 bp flanks homologous to the sequence outside the S. cerevisiae MLSl orf enabling the deletion of MLSl gene by homologous recombination. The deletion cassette for the other malate synthase gene, DAL7, was constructed in a similar way by PCR with primer pair ScDAL7f (SEQ ID NO: 7) and ScDAL7r (SEQ ID NO: 8) which consisted of 20 / 23 bp long body annealing to pUG6 plasmid and 40 bp flanks outside the DAL7 orf enabling the deletion of the S. cerevisiae DAL7 gene. The MLSl and DAL7 deletion cassettes were transformed to CEN.PK2 (H1346) strain and to CEN.PK113-1A (H3675) strain where the latter is capable to utilize xylose. Deletion fragments included the KanMX gene conferring the ability to grow on media containing G418 (Geneticin). The double deletion strain where MLSl and DAL7 were both knocked out was constructed in a similar manner except that pi 431 (Figure 2) plasmid instead of pUG6 was used as a template. pl431 is a modification of pUG6 where the KanMX resistance gene has been replaced by BLE marker gene. The MLSl and DAL7 deletion cassettes were introduced to S. cerevisiae strains CEN.PK2 and CEN.PK113-1A in separate transformation procedures. DAL7 deletion fragment included the BLE gene conferring the ability to grow on media containing Zeocin. In order to verify the MLSl and DAL7 deletions, yeast colony PCR and sequencing reactions were carried out with the primer pairs MLSlseqf (SEQ ID NO: 9), MLSlseqr (SEQ ID NO: 10) and MLS2seqf (SEQ ID NO: 11), MLSlseqr (SEQ ID NO: 12) respectively. All deletion cassettes contained loxP sites in front and after the marker genes. The loxP sites function as a specific target site for the Cre DNA recombinase enzyme (Guldener et al, 1996). Cre catalyzes DNA recombination leading up to the cleavage of loxP sites. Cre recombinase was expressed in S. cerevisiae by transforming the Cre expression plasmid pSH47 (Guldener et al, 1996) into the yeast cell. Plasmid pSH47 conferred the ability to grow in the absence of uracil. In order to activate Cre recombinase the MLSl A and DAL7A strains and the double knock-out MLSl A, DAL7A strain were cultured on a media containing galactose as a carbon source to activate the recombinase. Once the resistance marker had looped out and the ability to grow on media containing G418 and/or Zeocin the strains were cultivated on a rich media (YPD) without selection in order to lose the pSH47 plasmid. Once the strains had lost the ability to grow on media lacking uracil the strains were named as presented in Table 2.

The original Arabidopsis thaliana glyoxylate reductase, GR1 (SEQ ID NO: 1) which codes for the protein GR1 (SEQ ID NO: 3) was codon optimized for S. cerevisiae by GenScript (GenScript, USA) and the codon optimized form of the sequence was renamed GLYR1 (SEQ ID NO: 2), which codes for the protein GLYR1 (SEQ ID NO: 3). GLYR1 was ordered and received so that it had additional 5' BamHl and 3' EcoRI sites. GLYR1 was ligated into the EcoRI and BamHl sites between the TPI promoter and polyA terminator of the p2159 vector (Figure 3), generating plasmid p3644 (Figure 4). Plasmid p3644 was transformed to Saccharomyces cerevisiae CEN.PK2 derived strains H3772, H3773 and H3774 to generate strains H3783, H3784 and H3785 and to CEN.PK113-1A derived strains H3775, H3776 and H3777 to generate strains H3788, H3789 and H3790 (Table 2). Plasmid p3644 conferred the ability to grow in the absence of uracil. Strains were tested for glycolic acid production and as an example, the results of flask cultivations of strain H3790 are presented in the Table 3.

All yeast transformations were carried out with the Gietz method (Gietz and Woods (2002) Transformation of yeast by the LiAc/SS carrier DNA/PEG method. Methods in Enzymology 350: 87-96). The proofreading DNA polymerase Phusion (Thermo Fisher Scientific Inc., Finland) was used for all PCR reactions.

Example IB. Construction of a S. cerevisiae strain able to produce glycolic acid (integrated cassettes)

Plasmids such as those described in Example 1 A are linearized using restriction enzymes, in such a way that the 2 micron sequence (or autonomously replicating sequence, ARS, if appropriate) is deleted from the fragment containing the GLYR1 expression cassette and selectable marker (e.g. KanMX or URA3). These linearized fragments are then used to transform S. cerevisiae, with the result that the fragment is integrated into the genome. Integration may be random, or may also occur at a location for which there is homologous sequence present on both the fragment and in the genome. The plasmids may be modified to include specific sequences for targeted integration into the genome at the site of a specific gene, such as HO-locus. Integration of the GLYRl gene into the S. cerevisiae HO- locus was carried out by first constructing a cassette including GLYRl under TPI promoter and HO-segments. GLYRl was first amplified by PCR from p3644 (Figure 4) with the primer pair ScGLYRp3466- Sallf (SEQ ID NO: 13) and ScGLYRp3466-SalIr (SEQ ID NO: 14) including Sail sites. The product from the PCR reaction was ligated into the TOPO-TA plasmid (Invitrogen, USA) according to the manufacturer's protocol and transformed into E. coli. Purified TOPO-TA plasmids were digested with Sail and the TPI-GLYRl-poly-A fragment was ligated into the Sail site of the p3532 (Figure 5) which has HO-poly-kanMX4-HO fragment (Voth et al., 2001). To verify the correct composition of the plasmid, colony PCR was carried out. The integration cassette was prepared by PCR using modified p3532 plasmid with GLYRl as a template and HOr-HOlf (SEQ ID NO: 15) and HOr-HOlr (SEQ ID NO: 16) as a primer pair. Resulting deletion fragment can be transformed by homologous integration to any of the strains presented in Examples 1-6 and KanMX gene of the deletion fragment confers the ability to grow on media containing G418. The strain H3777 where the GLYRl fragment was integrated, generated strain H3913 and the strain H3847 with integrated GLYRl generated strain H3963. The successful transformation can be seen as an increase in glyoxylate reductase activity compared to the strains where GLYRl is not introduced.

Example 2. Construction of a S. cerevisiae strain able to produce glycolic acid with high yield

Strains producing glycolic acid such as those described in examples 1A and IB can be further modified in order to increase the glycolic acid production. This is possible by increasing flux towards glyoxylate cycle and by overexpressing the genes essential for the pathway.

The overexpression of endogenous isocitrate lyase, ICL1, and deletion of cytosolic NADP- specific isocitrate dehydrogenase, IDP2 (GenBank accession number NM 001182061.1), were carried out by a single integration cassette (Figure 5). The ICL1 gene was amplified by PCR using genomic DNA from H3675 strain as a template. The primer pair ICLlfwd (SEQ ID NO: 17) and ICLlrev (SEQ ID NO: 18) including flanks for homologous recombination with p2159 was used in the PCR reaction. Amplified ICL1 fragment together with EcoRI and BamHI digested p2159 plasmid were transformed to S. cerevisiae H1346 strain were the vector was constructed by homologous recombination. Transformed yeast cells were screened by growth in the absence of uracil. Recombined plasmids were rescued from yeast and transformed to E. coli for plasmid amplification. In order to find the correct plasmid constructs, colony PCR was carried out with the primer pair EcoRI- 102fwd (SEQ ID NO: 19) and BamHI+27rev (SEQ ID NO: 20). The TPI-ICLl-poly-A fragment was amplified from correct plasmid construct by PCR with the primer pair Flankl9 TPI-ICLl-PolyAf (SEQ ID NO: 21) and Flank-TPI-ICLl-PolyAr (SEQ ID NO: 22). The primer pair included flanks for homologous recombination with the loxP-KanMX- loxP fragment and pRS426 shuttle vector (Christianson et al, 1992) (renamed p2974), which was used as a backbone for cassette construction (Figure 6).

The loxP-KanMX-loxP fragment was amplified by PCR using primer pair Flank- KanMXfwd (SEQ ID NO: 23) and Flank-KanMXrev (SEQ ID NO: 24) and the pUG6 plasmid (Figure 1) as a template. The primer pair included flanks for homologous recombination with the TPI-ICLl-poly-A fragment and pRS426 plasmid. For constructing the cassette expressing ICLl and deleting IDP2, EcoRI and Xhol digested pRS426 plasmid and TPI-ICLl-poly-A and loxP-KanMX-loxP fragments were transformed and recombined in HI 346 strain. Constructed plasmid was then extracted and transformed into E. coli. The correct plasmid composition was confirmed by colony PCR and sequenced with the primer pair Check-IDP2-del-ICLl-f (SEQ ID NO: 25) and Check-IDP2-del-ICLl-r (SEQ ID NO: 26). The cassette expressing ICLl and deleting IDP2 (Figure 7) was amplified from the correct plasmid with the primer pair IDP2-ICLl-f (SEQ ID NO: 27) and IDP2-ICLl-r (SEQ ID NO: 28) including the flanks for homologous recombination into 7D 2-locus.

Constructed strain H3947 was used for glycolic acid production after the plasmid p3644 was transformed to the strain. The resulting strain was named H3846 (Table 2). The results of glycolic acid production of the flask cultivations of the strain H3846 are presented in the Table 3. Table 2: Modifications of yeast strains.

Modifications of the H1346 strain.

Modifications of the H1346 Strain S. cerevisiae strain number

HI 346 (mlsIA) H3772

HI 346 (dal7A) H3773

HI 346 (mlsIA, dal7A) H3774

H1346 (p3644, mlsIA) H3783

H1346 (p3644, dal7A) H3784

H1346 (p3644, mlsIA, dal7A) H3785

H1346 (m/siA, dal7A, idp2 A:\ICL1) H3911

H1346 (m/s , < α/7Δ, idp2A : :ICL 1 , pgi 1 Δ) H3965

H1346 (p3644, mlsIA, dal7A, idp2A : :ICL 1 , pgi 1 Δ) H3973

is of the H3675 strain.

Modifications of the xylose utilizing Strain H3675 S. cerevisiae strain number

H3675 (mlsIA) H3775

H3675 (dal7A) H3776

H3675 (mlsIA, dal7A) H3777

H3675 (p3644, m/s ) H3788

H3675 (p3644, < α/7Δ) H3789

H3675 (p3644, m/s , dal7A) H3790

H3675 (p3644, m/s , < α 74 idp2 A:\ICL 1) H3846

H3675 (m/s , < α 74 idp2 A:\ICL 1) H3847

H3675 (HO- cus::GLYRl, mlsIA, dal7A) H3913

H3675 (HO- cus::GLYRl, mlsIA, dal7A,

H3963 idp2 A:\ICL1)

H3675 (p3644, m/s , </α/7Δ, idp2 A:\ICL 1, regIA) H3994 Modifications of the K. lactis strain.

Table 3. Glycolic acid production of selected yeast strains.

Organism Strain Used carbon Glycolic acid production (g Γ ¹ ) number sources

Day 1 Day 2 Day 3 Day 6

S. cerevisiae H1346 2 % glucose 0 0 0 0

2 % ethanol + 0 0 0 0 2 % glycerol

S. cerevisiae H3675 2 % xylose 0 0 0 0

2 % ethanol + 0 0 0 0 2 % glycerol

S. cerevisiae H3790 2 % glucose 0 0 0 0

2 % xylose 0 0 0 0

2 % ethanol + 0 1,6 0,34 0,45 2 % glyserol

S. cerevisiae H3846 2 % glucose 0,09 0,28 0,28 0,28

2 % ethanol 0,71 0,78 0,78 -

S. cerevisiae H3973 1.9 % fructose 0 0 0,46 0,54

+ 0.1 %

glucose K. lactis H3986 2 % xylose + 0,45 0,83 1,77 1,88

2 % ethanol

2 % xylose + 0,46 0,93 1,58 2,94 2 % ethanol +

5 g Γ ¹ ethanol

added daily

As can be seen from the above results, flask cultivations already provide advantageous results. However, even higher yields could be obtained when cultivating in a fermentor instead of flasks (Example 14).

Example 3. Further construction of a S. cerevisiae strain able to produce glycolic acid with high yield

It is possible to further modify the host in order to increase glycolic acid production. This can be done by directing the flux towards glyoxylate cycle and by overexpressing the genes essential for the pathway. These overexpressed genes can be such as aconitase {ACOl) which converts citrate into isocitrate via cis-aconitate that is further converted into glyoxylate in the glyoxylate cycle. The ACOl gene was amplified by PCR using genomic DNA from S288c (H190) strain as a template. The primer pair ACOlHindlllf (SEQ ID NO: 29) and ACOlHindlllr (SEQ ID NO: 30) introducing Hindlll sites was used and the fragment was ligated into the Hindlll site between the mADH promoter and terminator of the p609 plasmid (Figure 8). The correct plasmid composition was confirmed by colony PCR and sequenced and it can be used for transformation as such or the fragment including the promoter and terminator sequence can be further used for deletion cassette or integration constructs which contain flanks for homologous recombination. This resulting fragment can be used for genomic integration of the gene which can be done similarly as presented in Example IB and 2 or overexpressed genes can also be introduced as plasmids as described in the Example 1A. The constructs were checked by PCR and the overexpression of the said gene was confirmed by enzyme assays. After the GLYR1 was transformed to the strains as described in Examples 1A and IB the strain was tested for glycolic acid production as described in the Example 9. Example 4. Construction of a S. cerevisiae glycolic acid producing strain with reduced ethanol production

Increase in flux towards glycolic acid can also be achieved by knocking out genes directing the flux in unwanted directions. ADH2 gene of the S. cerevisiae (GenBank accession number NM 001182812) was knocked out in order to decrease ethanol formation. The deletion construct was amplified by PCR using primers ADH2delflankKpnINotIf (SEQ ID NO: 31), ADH2delflankEcoRISalIr (SEQ ID NO: 32) for the first flanking fragment and ADH2delflank2EcoRIBamHIf (SEQ ID NO: 33), ADH2delflank2SpeIr (SEQ ID NO: 34) were used for the second flanking fragment. The first ADH2 fragment was then digested with Kpnl and Sail restriction enzymes and ligated with the pRS426 vector (Figure 6) which has been linearized with the same restriction enzymes Kpnl and Sail. The resulting plasmid was redigested with EcoRI and Spel and the second ADH2 flanking fragment, first digested with these same enzymes EcoRI and Spel, was ligated to the vector. The resulting plasmid with pRS426 backbone and both ADH2 flanks was then digested with Sail and EcoRI. The loxP-KanMX-loxP fragment was amplified by PCR by primer pair loxPkanMX-loxP Sallf (SEQ ID NO: 35) and loxP-kanMX-loxP EcoRIr (SEQ ID NO: 36) using plasmid pUG6 (Figure 1) as a template. The resulting fragment was digested with Sail and EcoRI and ligated between the Sail and EcoRI sites of the ADH2 flanking regions. The construct was checked by primers pRS426XhoIcutFseq (SEQ ID NO: 37), pRS426SpeIcutRseq (SEQ ID NO: 38) and loxPKanMXloxPseqF (SEQ ID NO: 39).

Correct deletion cassette was then cleaved with Notl and the resulting fragment was transformed to S. cerevisiae by homologous recombination. Overexpression of selected genes is also possible with the same deletion fragment as a fragment expressing the gene with promoter and terminator can be introduced between the ADH2 flank and loxP- KanMX-loxP fragment either by homologous recombination or by ligation after digestion with suitable restriction enzymes. The overexpressed gene(s) can be e.g. pyruvate carboxylase (PYC or PYC2) or any of the genes which overexpression is described in the Examples 2, 3, 5 or 6. The constructs can be then checked by PCR or sequencing and deletion of the said gene can be confirmed by enzyme assay where the activity of alcohol dehydrogenase is decreased and the activity of the overexpressed gene is increased. After the GLYR1 was transformed to the strains as described in Examples 1A and IB the strains were tested for glycolic acid production as described in Example 9.

Example 5. Construction of a S. cerevisiae glycolic acid producing strain with increased NADPH availability

S. cerevisiae PGIl (GenBank accession number NM 001178544) knock-out strain was constructed in order to balance the NADPH availability as NADPH is required by the glyoxylate reductase reducing glyoxylate into glycolic acid. PGIl flanks were amplified from genomic DNA of S. cerevisiae S288c (H190) strain. Primers PGIflankKpnINotIf (SEQ ID NO: 40) and PGIflankEcoRISallr (SEQ ID NO: 41) were used for the construction of the first PGIl flank fragment. The second PGIl flank fragment was amplified with primers PGIflank2EcoRIBamHIf (SEQ ID NO: 42) and PGIflank2SpeIr (SEQ ID NO: 43). In addition, the loxP-KanMX-loxP fragment was amplified from the pUG6 vector (Figure 1) with primer pair loxP-kanMX-loxP Sallf (SEQ ID NO: 35) and loxP-kanMX-loxP EcoRIr (SEQ ID NO: 36). The Kpnl and EcoRI digested PGIl flankl and the pRS426 vector (Figure 6) linearized with the same restriction enzymes, Kpnl and EcoRI, were first ligated together. As a next step, the resulting plasmid as well as the PGIl flank 2 fragment were digested with EcoRI and Spel and ligated together. The resulting plasmid with the pRS426 backbone and both PGIl flanks and the loxP-KanMX-loxP fragment were digested with Sail and EcoRI and ligated together. The final construct was checked by colony-PCR with primers loxP-kanMX-loxP EcoRIr (SEQ ID NO: 36) and pRS426XhoIcutFseq (SEQ ID NO: 37) and sequenced with primers pRS426XhoIcutFseq (SEQ ID NO: 37) and pRS426SpeIcutRseq (SEQ ID NO: 38) and the correct vector construct was named as p4115.

The PGIl deletion cassette was released by Notl digestion from p4115 plasmid and transformed into strain H3911 resulting strain H3965 (Table 2). The PGIl negative mutants were screened based on glucose toxicity. The loxP-KanMX-loxP marker of the PGIl deletion cassette was looped out from the PGIl deficient strain H3965 by using the Cre recombinase plasmid (p902). After the GLYR1 was transformed to the strain as described in Example 1 A the strain was named as H3973 and it was used for glycolic acid production (Table 3). PGII can be replaced by some other genes such as cytosolic aldehyde dehydrogenase (ALD6) and/or acetyl-coA synthase (ACS1) (Example 16) and/or any of the genes which overexpression is described in the Examples 2, 3, 4 or 6. Different combinations of overexpressed genes can be used under different promoters and terminators.

Example 6. Construction of S. cerevisiae strain where glyoxylate cycle regulation has been affected in order to get higher glycolic acid yields.

REG1 of S. cerevisiae is a regulatory subunit of type 1 protein phosphatase Glc7p, involved in negative regulation of glucose-repressible genes. The gene was deleted in order to decrease the glucose repression of the glyoxylate cycle. The REG1 (GenBank accession number NM 001180336) deletion was done by using the primer pair Regldelfwd (SEQ ID NO: 44) and Regldelrev (SEQ ID NO: 45) for PCR where pUG6 (Figure 1) plasmid was used as a template. The resulting loxP-KanMX-loxP fragment with flanks homologous to the sequence upstream and downstream of the REG1 gene was transformed to yeast and the transformants were selected based on the G418 resistance. The strain H3847 was named as H3994 after the REG1 deletion (Table 2). After the GLYRI was transformed to the strains as described in Examples 1A and IB the strains were used for glycolic acid production as described in Example 9.

Example 7. Construction of a K. lactis glycolic acid producing strain.

Vector for expression of glyoxylate reductase in K. lactis: The pJJH958r (FEMS Yeast Res (2010) 333-342) was digested with Kpnl to remove the CreR gene with GALI promoter and CYC1 terminator and religated. The circular vector, p4150, was then digested with Sail. The GLYRI gene with TPII promoter and terminator from Example 1A was then amplified by PCR to introduce Sail restriction sites and ligated to the Sail site of p4150. The resulting vector, p4185, is a multicopy expression vector for the GLYRI with URA3 selection.

Deletion of a malate synthase gene in K. lactis: The K. lactis gene with the GenBank accession number CR382126.1 was identified as the gene with highest homology to the S. cerevisiae MLS I and was deleted following a protocol described previously (FEMS Yeast Res 10 (2010) 333-342). The HIS3 gene of the pJJH955H (FEMS Yeast Res 10 (2010) 333-342) was amplified by PCR using primers K1MLS1 (SEQ ID NO: 46) and K1MLS2 (SEQ ID NO: 47) that had 50 bp flanking regions upstream and downstream the open reading frame of the MLS1 homologue. The PCR product was then transformed to the K. lactis strain Os276 (FEMS Yeast Res 10 (2010) 333-342) and transformants selected for the ability to grow in the absence of histidine. The deletion cassette was then looped out as described previously (FEMS Yeast Res 10 (2010) 333-342). The resulting strain, H3968, was not growing on ethanol medium in contrast to the parent strain.

Deletion of the IDP2 homologue in K. lactis and expression of glyoxylate reductase: The K. lactis gene with the Gene ID: 2894935 was identified as the closest homologue to the S. cerevisiae IDP2 and was deleted in the H3968 in a similar way as the MLS1 homologue described above using the primers K1IDP2F (SEQ ID NO: 48) and K1IDP2R (SEQ ID NO: 49). The resulting strain is H3976. This strain was transformed with the glyoxylate reductase expression vector, p4185, resulting in the strain H3986. The strain was tested for glycolic acid production and the results are shown in Table 3.

Modification of the K. lactis CATS gene by changing the serine at 661 to glutamine, S661E: The mutation S661E in the K. lactis CAT8 gene changes the transactivation activity in a way that the CATS is constitutively active (Mol. Cell Biol. 24(2004) 4083-4091). To make this point mutation a CATS fragment with the S661E mutation was made by PCR with the primers KlCat8S661EHind (SEQ ID NO: 50) and KlCat8revPst (SEQ ID NO: 51). This fragment was then digested with Hindlll and Pstl and ligated to the Hindlll and Pstl digested plasmid pJJH955L (FEMS Yeast Res 10 (2010) 333-342). This vector has the LEU2 marker between loxP sites. The CATS fragment with the mutation followed by the LEU2 marker was the amplified by PCR with the primers KlCat8S661Ef (SEQ ID NO: 52) and KlCat8S661Er (SEQ ID NO: 53) and the PCR product transformed to the H3976. The LEU2 was then looped out as described before and the expression vector for the GLYR1, p4185, transformed. The resulting strain is H3987 and it was used for glycolic acid production as described in Example 9. Example 8. Construction of a Candida krusei strain able to produce glycolic acid with high yield

Glyoxylate reductase GLYR1 was introduced into C. krusei strain ATCC32196 under PGKl promoter of C. krusei. The loPGKI promoter was amplified from the genomic DNA of C. krusei strain ATCC32196 essentially as described in US 2009/0226989 Al, Sep. lO, 2009. pMLV94 (Figure 9) is a bacterial plasmid with C. krusei PGK promoter and S.cerevisiae ADH1 terminator. GLYR1 was amplified by PCR by primer pair BglllGLYRfwd (SEQ ID NO: 54) and BglllGLYRrev (SEQ ID NO: 55) introducing Bglil sites to the fragment.

After Bglil digestion the GLYR1 fragment was ligated to the Bglil site of the pMLV94 vector and the resulting plasmid was digested with BamHl to release the 2.0 kb fragment including the promoter and terminator regions and the GLYR1 gene. pMLV128 plasmid (Figure 10) containing the loxP-hph-loxP fragment, where hph stands for a gene giving resistance for hygromycin, was digested with Pvull, and BamHl digested prom-GLYRl-tevm fragment was ligated to the site after its 5' overhangs had been filled-in with Klenow polymerase reaction. The resulting plasmid was digested with Not! in order to release the fragment.

C. krusei malate synthase gene (SEQ ID NO: 72), MLS, was identified based on the homology to the S. cerevisiae MLS1 gene. The flanks homologous to the genome sequence upstream and downstream of the C. krusei MLS1 gene were amplified using primer pair IoMLSflank5fpRS (SEQ ID NO: 56), IoMLSfiank5rIopPGK (SEQ ID NO: 57) and primer pair IoMLSfiank3fioxP (SEQ ID NO: 58), IoMLSflank3rpRS (SEQ ID NO: 59). In addition also primer pairs IoMLSflank5fpRS_2 (SEQ ID NO: 73), IoMLSfiank5rIopPGK_2 (SEQ ID NO: 74) and primer pair IoMLSfiank3fioxP_2 (SEQ ID NO: 75), IoMLSflank3rpRS (SEQ ID NO:76) can be used.

All fragments, including the Not! released fragment with promoter-GLYRl-terminator- loxP-hph-loxP, the Notl digested pMLV128 vector and the two MLS1 flanking fragments were combined by homologous recombination in S. cerevisiae to produce a cassette with GLYR1 overexpression fragment combined to the hph resistance marker and at least 400- 500 bp flanks which were similar to the sequence upstream and downstream of the malate synthase gene of the C. krusei. The resulting malate synthase deletion fragment overexpressing GLYR1 was confirmed by sequencing and the 5.0 kb deletion fragment, released by Xmal digestion, was transformed to C. krusei. Transformants were plated on YPD plates containing 400 μg ml ^"1 hygromycin for the selection. Correct transformants were analysed by PCR and by enzyme assays where glyoxylate reductase activity was found to be increased. The strain was named as H4154.

C. krusei is diploid and its second copy of malate synthase gene was deleted by using following primer pairs for MLS deletion cassette construction: 221MLS3'fSalI (SEQ ID NO: 77), 222MLS3'rSalI (SEQ ID NO: 78), and 223MLS5'fSpeI (SEQ ID NO: 79), 224MLS5'rSpeIBamHI (SEQ ID NO: 80). The first fragment was digested by Sail and the second fragment by Spel. pMVL97 plasmid (Figure 18) containing the loxP-MEL-loxP fragment, where MEL stands for a gene encoding for a-galactosidase gene which gives the cells carrying the gene a blue colour when plated on a plate containing x-a-gal, was digested with Sail. The Sail digested fragment flanking to the upstream of the C. krusei malate synthase gene was ligated to the vector and correct plasmid construct was then further digested with Spel. The Spel digested fragment flanking downstream of the MLS gene was ligated to the vector and correct plasmid construct was selected based on the colony-PCR verification. The plasmid construct with MLS flanks was digested with BamHl and the above described GLYR fragment was ligated to the vector. The resulted plasmid (Figure 19) construct was then sequenced. MLS deletion cassette was cleaved with Notl digestion and the 5.5 kb fragment was purified from the agarose gel.

The 5.5 kb deletion fragment was transformed to C. krusei strain where already one copy of the malate synthase gene had been deleted. Transformants were plated on YPD plates containing 300 μg ml ^"1 hygromycin and 40 μg ml ^"1 x-a-gal. The correct transformants were selected among the blue colonies.

Correct transformants were analysed by PCR and by enzyme assays where malate synthase activity was decreased and glyoxylate reductase activity increased. The resulting strain H4155 was not growing on ethanol medium in contrast to the parent strain and the strain C. krusei H4154 where only one malate synthase had been deleted. H4154 and H4155 strains expressing glyoxylate reductase were used for glycolic acid production as described in Example 9. The production of glycolic acid is presented in Table 4. The pH of C. krusei strain H4155 after two days cultivation was measured to be 2.8 showing that the production of glycolic acid is possible in non-buffered conditions still at low pH.

Table 4. Glycolic acid production of C. krusei.

Example 9. Production of glycolic acid by modified yeast strains

K. lactis H3986 was grown in non-buffered conditions in modified synthetic complete (SC) medium lacking uracil (Sherman F, Fink G, Hicks JB. (1983) Methods in Yeast Genetics. A Laboratory Manual. Cold Springs Harbor Laboratory, Cold Springs Harbor, N.Y.). Xylose (20 g Γ ¹ ) and ethanol (20 g Γ ¹ ) were provided as a carbon source. Yeast was grown in 50 ml SC-URA medium shake flask at 30°C, 200 rpm.

The other yeast strains described in Examples 1-8 were grown in similar conditions with different carbon sources. The integrant strains S. cerevisiae (Example IB) and C. krusei (Example 8) were grown on SC medium instead of SC medium lacing uracil. Instead of xylose (20 g Γ ¹ ) and ethanol (20 g Γ ¹ ) mix also other carbon sources were tested: glucose, xylose, ethanol, ethanol + glycerol and fructose + glucose in different concentrations. The results of glycolic acid production of some selected strains are presented in Table 3 and the production in modified C. krusei strain is presented in Table 4. The base strains HI 346 and H3675 without any additional modifications did not produce glycolic acid (Table 3) according to the HPLC results.

Extracellular metabolites in cell-free spent culture medium (glycolic acid, ethanol), and xylose were analysed by HPLC on a Fast Acid Analysis Column (100 mm x 7,8 mm, Bio- Rad, Hercules, CA) linked to an Aminex HPX-87H column (BioRad Labs) with 2.5 mM H ₂ SO ₄ as eluent and a flow rate of 0.5 ml min ^"1 . The column was maintained at 55°C. Peaks were detected using a Waters 410 differential refractometer and a Waters 2487 dual wavelength UV (210 nm) detector. As an more detailed example the formation of glycolic acid and biomass and the consumption of xylose and ethanol of the K. lactis strain H3986 are presented in Figure 14 and the pH curve is presented in the Figure 15. Example 10. Production of glycolic acid by modified yeast strains with additional ethanol

Since glycolic acid production occurred at the highest rates when ethanol was present as the carbon source, the production of glycolic acid with addition of extra ethanol was demonstrated. Similar experiment as described more closely in the Example 9 was done with the same K. lactis strain H3986 so that in addition to the xylose (20 g Γ ¹ ) and ethanol (20 g Γ ¹ ), 5 g Γ ¹ of ethanol was added to each flask daily. The analysis of glycolic acid, ethanol and xylose concentrations in the media was done by HPLC as presented in the Example 8. The formation of glycolic acid and biomass and the consumption of xylose and ethanol are presented in Figure 16 and the pH curve is presented in the Figure 17.

Example 11. Construction of niger glycolic acid producing strain

A plasmid pMV45-AnMLS-del-cas (Figure 10) contained 1633 bp (region -969.. +664 from the both sides of the ATG start codon) from the A. niger acuE (malate synthase gene) promoter and ORF, followed by the pyrG gene (including the native pyrG promoter and terminator regions, -500..-1 from ATG and +1..+497 from the stop codon, respectively) and 1671 bp (region -720.. +951 from the from the both sides of the TAA stop codon) from A. niger acuE ORF and terminator. The acuE 5' and 3' fragments were obtained by PCR of genomic DNA from niger ATCC1015 using primers AnMLS-5'-For (SEQ ID NO: 60) and AnMLS-5'-Rev (SEQ ID NO: 61) and AnMLS-3'-For (SEQ ID NO: 62) and AnMLS-3'-Rev (SEQ ID NO: 63). The pyrG gene fragment was obtained by PCR of genomic DNA from A. niger ATCC1015 using primers pyrG-del-F_n (SEQ ID NO: 64) and pyrG-del-R_n (SEQ ID NO: 65). The proofreading DNA polymerase Phusion (Thermo Fisher Scientific Inc., Finland) was used for all PCR reactions. Construction of the plasmid was performed in three steps. First, plasmid pMV45 was digested by the restriction endonuclease EcoRI (NEB) and amplified A. niger AcuE fragments were digested by the restriction endonuclease EcoRI and Xrnal (both NEB). Ligation of digested pMV45 plasmid and A. niger acuE fragments was performed using T4 DNA ligase and the intermediary construct pMV45-AnMLS was obtained. Second, pMV45-AnMLS and amplified pyrG fragment were digested with the restriction endonuc leases Xrnal (NEB). Ligation was performed using T4 DNA ligase and the intermediary construct pMV45- AnMLS-del-/?yrG was obtained. Third, pMV45-AnMLS-del-/?yrG was digested with the restriction endonuclease EcoRI (NEB) and the final construct pMV45-AnMLS-del- pyrGcas was obtained (Figure 11).

A uracil auxotroph A. niger strain Anl015-pyrGA was generated as described by Mojzita et al. 2010 (Mojzita D, Wiebe M, Hilditch S, Boer H, Penttila M, Richard P: Metabolic engineering of fungal strains for the conversion of D-galacturonate to meso-galactarate. Appl Environ Microbiol 2010, 76: 169-175.). Anl0l5-pyrGA was transformed with the deletion cassette pMV45-AnMLS-del-/?yrG-cas, to generate strain Anl0l5-acuEA::pyrG. Transformation was performed using the basic PEG-mediated protoplast transformation method and transformants were screened for integration of the deletion cassette at the acuE locus by growth in the absence of uracil. Deletion of the acuE gene was confirmed by PCR with primers AnMLS-screen-For (SEQ ID NO: 66) and AnMLS-screen-Rev (SEQ ID NO: 67). Strain Anl0l5-acuEA::pyrG was identified as a transformant in which the acuE gene had been deleted. The strain had no malate synthase activity which was confirmed by the malate synthase assay (the utilization of acetyl-CoA in the presence of glyoxylate) and it produced less biomass from ethanol than the parent strain on an agarose plate.

Plasmid JKpl-hph-Glyr (Figure 13) contained hygromycin B phosphotransferase gene (hph) and A. niger codon optimized (by GenScript) GLYR1 gene from A. thaliana (SEQ ID NO: 4) under the gpdA promoter and trpC terminator. The codon optimized GLYR1 (from GenScript) was received with the 5' Sad and 3' Xmal sites. The received GLYR1 gene and the plasmid JKpl-hph were digested with the restriction endonucleases Sacl and Xmal (both NEB). Ligation of digested JKpl-hph plasmid and the GLYR1 fragment was performed using T4 DNA ligase and the intermediary construct JKpl-hph-Glyr was obtained. JKpl-hph-Glyr was digested with the restriction endonuclease Spel (NEB) and digested cassette containing gpdA-GLYRl-trpC-hph was transformed into the Anl015- acuEA::pyrG strain using the standard fungal transformation methods and transformants were screened for integration of the gpdA-GLYRl-trpC-hph cassette by growth in the presence of hygromycin. Integration of the transformed fragment into the genome was confirmed by PCR with the primers AnGlyrScreenFor (SEQ ID NO: 68) and AnGlyrScreenRev (SEQ ID NO: 69). The introduced glyoxylate reductase activity was confirmed by the standard glyoxylate reductase enzyme assay (the utilization of NADPH in the presence of glyoxylate). In contrast to the parental strain (Anl0l5-acuEA::pyrG), Anl0l5-acuEA::pyrG-GLYRl had an increased glyoxylate reductase activity in crude extract.

Example 12. Production of glycolic acid by A. niger strains

A. niger ATCC1015 AacuE+GLYRl was grown in the production medium (The defined medium of Vogel described by Mojzita et al. 2010) with 20g glucose, xylose or ethanol Γ ¹ as a carbon source. Pre-cultures were grown in the medium containing 10 g yeast extract Γ 20 g peptone Γ ¹ and 30 g gelatine Γ ¹ (50 ml medium in 250 ml flasks). Mycelium from 50 ml cultures was collected by filtration, washed with sterile H ₂ 0 and re-suspended in 50 ml of the production medium in 250 ml flasks. Cultures were incubated at 30°C, 250 rpm. Culture supernatant was analysed by HPLC.

A. niger ATCCl0l5AacuE+GLYRl produced 0.5-100 g glycolic acid Γ ¹ extracellularly in 1-100 h. A. niger ATCC1015 (wild-type) and the parent strain (A. niger ATCC1015 AacuE) produced no glycolic acid when grown under the same conditions. Example 13. Other possible exogenous glyoxylate reductases used for glycolic acid production

Instead of the Arabidopsis thaliana GLYRl glyoxylate reductase it is possible to use any other glyoxylate reductase for glycolic acid production. According to our knowledge there are at the moment nine known glyoxylate reductases described in literature where the gene coding for the enzyme is known (Table 1). With BLAST search it is possible to look for genes with high homologies to these and get new possible candidates for glyoxylate reductases. The glyoxylate reductase gene of Thermus thermophilus (SEQ ID NO: 70) was ordered from GenScript with additional 5' BamHl and 3' EcoRI sites. The gene was introduced between the BamHl and EcoRI sites of the p2159 plasmid similarly as described in the case of A. thaliana GLYRl in the Example 1 A. The resulted vector was used in the strains described above and the strains were tested for glycolic acid production as described in the Example 9.

Example 14. Production of glycolic acid by strain H3986 at pH 5.0 with high yield.

K. lactis strain H3986 was pregrown in flasks in modified synthetic complete (SC) medium lacking uracil (Sherman F, Fink G, Hicks JB. (1983) Methods in Yeast Genetics. A Laboratory Manual. Cold Springs Harbor Laboratory, Cold Springs Harbor, N.Y.). Glucose (20 g Γ ¹ ) was provided as carbon source. Pregrown cells were used for yeast cultivation started at OD ₆ oo ⁼ l in a Biostat CT-DCU bioreactor (max. working volume 5000 ml, Sartorius, Gottingen, Germany) at pH 5.0, 30°C, 1 volume air [volume culture] ^"1 min ^"1 (wm) and 500 rpm agitation with rushton turbines. Xylose (20 g Γ ¹ ) and ethanol (20 g Γ ¹ ) were provided as carbon source at the beginning of the fermentation. The pH was maintained constant by addition of 2 M NaOH. Silicone antifoaming agent (BDH, 0.2 ml Γ ^l ) was added to prevent excess foaming. Xylose and ethanol were fed to the fermentor as separate feeds based on their consumption. Extracellular metabolites in cell-free spent culture medium were analysed by capillary electrophoresis (P/ACE MDQ, Beckmann-Coulter) with UV detection according to Turkia et al. 2010. H3986 produced extracellularly 17.2 g glycolic acid Γ ¹ after 9 days at pH 5.0 (Table 5). The yield of glycolic acid on xylose ethanol media was approximately 0.13 g glycolic acid [g ethanol + xylose] ^"1 (Table 6). Table 5. Glycolic acid production by strain H3986 at pH 5.0 in a fermentor.

Table 6. Glycolic acid production yields by strain H3986.

Example 15. Production of glycolic acid by strain H3986 using acetate as substrate.

K. lactis strain H3986 was pregrown in flasks in modified synthetic complete (SC) medium lacking uracil (Sherman F, Fink G, Hicks JB. (1983) Methods in Yeast Genetics. A Laboratory Manual. Cold Springs Harbor Laboratory, Cold Springs Harbor, NY.). Glucose (20 g Γ ¹ ) was provided as carbon source. Pregrown cells were used for yeast cultivation started at OD ₆ oo ⁼ l in a Biostat Q bioreactor (max. working volume 1000 ml, Satrorius, Gottingen, Germany) at pH 4.0, 30°C, 1 volume air [volume culture] ^"1 min ^"1 (wm) and 500 rpm agitation with Rushton turbines. The pH was maintained constant by addition of 2 M NaOH. Silicone antifoaming agent (BDH, 0.2 ml Γ ¹ ) was added to prevent excess foaming. 5 % xylose and 0.2 % acetate were used as carbon sources and acetate was later on fed to the fermentor based on its consumption.

Extracellular metabolites in cell-free spent culture medium were analysed by capillary electrophoresis (P/ACE MDQ, Beckmann-Coulter) with UV detection according to Turkia et al. 2010. H3986 produced extracellularly 0.9 g glycolic acid Γ ¹ after 6 days at pH 4.0.

Example 16. Construction of a K. lactis glycolic acid producing strain with increased NADPH availability and asetyl-CoA synthase (ACS!) overexpression.

Primer pair 234KlMLSfiankTPIF (SEQ ID NO: 81), 235KlMLSflankURAR (SEQ ID NO: 82) was used to amplify S. cerevisiae aldehyde dehydrogenase, ALD6, gene which had been cloned to the p2159 plasmid (Figure 3) under the TPI promoter. The 3.3 kb promoter- ALD6-terminator fragment and MLS flanks flanking upsteam and downstream of the malate synthase gene of K. lactis. The fragment was transformed to the K. lactis strain H3976 (Table 2) and trasformants were plated on SCD-URA plates.

Plasmid p4124 also known as pJJH955L (Heinisch J. et al. FEMS Yeast Res 10 (2010) 333-342) was used as a template for PCR with primer pair 236KlMLSURAflankLEUf (SEQ ID NO: 83), 237KlMLSflankLEUr (SEQ ID NO: 84). The 1.9 kb PCR product was then digested with Notl and ligated to the 2.3 kb fragment purified from the Notl digested p4124 plasmid. The constructed plasmid was digested by Sail.

S. cerevisiae acetyl-coenzyme A synthetase, ACS1, gene which had been cloned to the p609 plasmid (Figure 8) under the mADH promoter was amplified by PCR by primer pair 240ACS_SalIF, 241ACS_SalIR. The 3.3 kb fragment was then digested by Sail and ligated with the above constructed and Sail digested vector. The resulting vector is overexpressing the S. cerevisiae ACS1 gene and having LEU marker between the loxP sites and having the flanks allowing integration to the MLS site of the above constructed K. lactis strain with ALD6 overexpression. The flanks were designed so that they also delete the URA marker introduced to the above constructed K. lactis strain with the ALD6 overexpression. The ACS1 overexpression cassette was digested by Notl and the 5.2 kb fragment was then transformed to the K. lactis (mlslA:ALD6, idp2A) strain. Transformants were plated on SCD-LEU plates. Finally p4185 (described in Example 7) overexpressing GLYR1 was transformed to the strain and the transformants were plated on SCD-LEU-URA plates. The resulting strain is H4158. The strain was used for flask cultivations to compare glycolic acid production with the previously constructed strain H3986. Cultivations were done in 25 ml flasks on synthetic complete media (SC) lacking uracil (both strains) and leucine (only strain H4158) and having 2 % xylose and 2 % ethanol as carbon source. It was seen that in these conditions after three days of cultivation in + 30 °C glycolic acid production of the strain H4158 was increased by 2.1 fold compared to the production of the H3986 strain grown in similar conditions. Glycolic acid production was analysed by HPLC as described in Example 9.

References

Charbon G, Breunig KD, Wattiez R, Vandenhaute J, Noel-Georis I. (2004) Key role of Ser562/661 in Snfl -dependent regulation of Cat8p in Saccharomyces cerevisiae and Kluyveromyces lactis. Mol Cell Biol. 2004 May;24(10):4083-91. Christianson TW, Sikorski RS, Dante M, Shero JH, Hieter P. (1992) Multifunctional yeast high-copy-number shuttle vectors. Gene. 1992 Jan 2;110(1): 119-22.

Gietz and Woods (2002) Transformation of yeast by the LiAc/SS carrier DNA/PEG method. Methods in Enzymology 350: 87-96

Guldener U, Heck S, Fielder T, Beinhauer J, Hegemann JH. (1996) A new efficient gene disruption cassette for repeated use in budding yeast. Nucleic Acids Res. 1996 Jul 1 ; 24(13): 2519-24.

Heinisch JJ, Buchwald U, Gottschlich A, Heppeler N, Rodicio R. (2010) A tool kit for molecular genetics of Kluyveromyces lactis comprising a congenic strain series and a set of versatile vectors. FEMS Yeast Res. 2010 May;10(3):333-42. Hoover, G.J., Van Cauwenberghe, O.R., Breitkreuiz, K.E., Clark, S.M., Merrill, A.R., and Shelp, B.J. (2007). Characteristics of an Arabidopsis glyoxylate reductase: General biochemical properties and substrate specificity for the recombinant protein, and developmental expression and implications for glyoxylate and succinic semialdehyde metabolism in planta. Can. J. Bot. 85: 883-895.

Mojzita D, Wiebe M, Hilditch S, Boer H, Penttila M, Richard P. (2010) Metabolic engineering of fungal strains for the conversion of D-galacturonate to meso-galactarate. Appl Environ Microbiol 2010, 76: 169-175.

Ogino H, Nakayama H, China H, Kawata T, Doukyu N, Yasuda M. (2008) Characterization of recombinant glyoxylate reductase from thermophile Thermus thermophilus HB27. Biotechnol Prog. 2008 Mar-Apr;24(2):321-5. Epub 2008 Feb 27. Sambrook et al. (1989)

Sambrook and Russell (2001)

Sherman F, Fink G, Hicks JB. (1983) Methods in Yeast Genetics. A Laboratory Manual. Cold Springs Harbor Laboratory, Cold Springs Harbor, N.Y.

Turkia H, Siren H, Pitkanen JP, Wiebe M, Penttila M. (2010) Capillary electrophoresis for the monitoring of carboxylic acid production by Gluconobacter oxydans. J. Chromatogr. A., 1217 (2010) 1537-1542.

Voth WP, Richards JD, Shaw JM, Stillman DJ. (2001) Yeast vectors for integration at the HO locus. Nucleic Acids Res. 2001 Jun 15;29(12):E59-9.