on the basis of second generation feedstocks

Field of the invention

The present invention relates to the utilization of second generation feedstocks in the microbial production of advanced fuels and chemicals. The invention is based on metabolic engineering of and fermentation by yeasts such as Saccharomyces cerevisiae. In particular, the invention relates to S. cerevisiae strains that have been engineered to express in the absence of fatty acids several of the enzymes that are required in the degradation of such acids along with certain other enzymes. The expression of some of the native yeast enzymes has been prevented in order to avoid the formation of undesired by-products. By varying the expression of the various enzymes in yeast, it has been possible to obtain a range of metabolically-engineered yeasts that each produce a particular valuable compound from sugars such as hexoses and/or pentoses. The invention further relates to the processes wherein the engineered strains of the invention produce compounds from sugars.

Background of the invention

The production of ethanol from first generation feedstocks such as corn, wheat and sugarcane has been optimized and it currently is applied world-wide. The process, however, has generic drawbacks at two levels. Firstly, ethanol is not an optimal fuel compound in view of its chemical structure and the associated physical properties. Secondly, first generation feedstocks are under pressure both reasons of economics and of sustainability. These two drawbacks call for a radical change in the bioproduction of fuels. Novel ways should be explored to arrive at production processes for advanced fuels that are more suitable than ethanol. And such compounds should be produced from second rather from first generation feedstocks. In developing processes for advanced fuels it is possible to develop hand in hand production processes for chemicals.

Second generation feedstocks are lignocellulosic materials. They may be derived from stalks, cobs, etc. from plants that now are used as first generation feedstocks. In other words bagasse, wheat straw, corn stover, etc. Such materials also can be obtained by growing dedicated energy crops on marginal lands thus not competing directly with food crops. Waste streams may be used as ell such as solid municipal wastes.

Lignocellulosic materials are cheaper than first generation sugar streams, but they require specific treatments. Harvesting may be an issue and the feedstocks require costly pretreatment and enzymatic treatments for liberating sugars.

An important aspect is that lignocellulosics contain not only C6 sugars (glucose, fructose) but also C5 sugars (xylose, arabinose). For a profitable process, it consequently is required that also the C5 sugars are converted into the desired product along with the C6 sugars.

Another issue with lignocellulosics as feedstocks is that pretreatment results often in compounds that are toxic to microbes such as acetic acid, furfural and hy droxymethy lfurfural .

Zhang et al (Current Opinion in Biotechnology 2011, 22:775-783) disclose that production of biofuels from renewable resources provides a source of liquid transportation fuel to replace petroleum-based fuels. This endeavor requires the conversion of cellulosic biomass into simple sugars, and the transformation of simple sugars into biofuels. Recently, microorganisms have been engineered to convert simple sugars into several types of biofuels, such as alcohols, fatty acid alkyl esters, alkanes, and terpenes, with high titers and yields. Several engineered metabolic pathways for the production of advanced biofuels were reported.

Peralta-Yahya et al (2012, Nature, 488(7411):320-8) disclose that advanced biofuels produced by microorganisms have similar properties to petroleum-based fuels, and can 'drop in' to the existing transportation infrastructure. However, producing these biofuels in yields high enough to be useful requires the engineering of the microorganism's metabolism. Pathways for fatty-acid-derived fuels were quoted from published data.

Clomburg and Gonzalez (2010, Biotechnol Bioeng. 108(4):867-79) disclose biofuel production in E. coli with an emphasis on metabolic engineering and synthetic biology.

Dellomonaco et al. (2011, Nature, 476:355-361; WO 2012/109176) disclose the engineered reversal of the β-oxidation cycle for the synthesis of fuels and chemicals in E. coli. They produced long-chain products via elongation of short-chain metabolites. They applied a reversal of the β-oxidation cycle as based on CoA-thioester intermediates that directly uses acetyl-CoA for acyl-chain elongation.

Gulevich et al. (2012, Biotechnol Lett. 34(3):463-9) disclose metabolic engineering of E. coli for 1-butanol synthesis through the inverted aerobic fatty acid β- oxidation pathway.

Zheng et al. (2012, Microbial Cell Factories 11 :65) disclose that they have studied the bioproduction of C12/C14 and C16/18 alcohols in E. coli. They improved the fatty alcohol production by systematically optimizing the fatty alcohol biosynthesis pathway, mainly by targeting three key steps from fatty acyl-acyl carrier proteins (ACPs) to fatty alcohols, which are sequentially catalyzed by thioesterase, acyl- coenzyme A (CoA) synthase and fatty acyl-CoA reductase.

Bernard et al. (2012, Plant Cell. 24(7):3106-18) disclose that reconstitution of plant alkane biosynthesis is possible in yeast.

The specific approaches in producing bio fuels compounds as quoted above depend on the enzyme machinery for either fatty acid oxidation (e.g. Dellamonaco et al. 2011, supra) or on the biosynthesis of fatty acids (e.g. Zheng et al. 2012, supra).

In several instances, compounds can be produced that find application outside the fuel's world as chemicals. For instance, Kataoka et al. 2012, Journal of Bioenegineering, htt : 7dx.doi.org/ 10.1016/j .jbiosc.2012.11.025 produced (R)-l,3- butanediol as a valuable chemical in an approach similar to the reversal of the β- oxidation cycle.

A basically different approach is available in which use is made of bio synthetic routes that depend on polyketide synthetases which are complexes of enzymes. Their basic design and functioning has been described in great detail (Hopwood and Sherman, 1990, Annu. Rev. Genet. 24:37-66; Smith and Tsai, 2007, Nat Prod Rep. 24(5): 1041-1072). Unlike the enzymes involved in the degradation or production of fatty acids, the polyketide synthetases produce a diverse array of products. Weissman and Leadlay (2005, Nature Rev. Microbiol. 3:925-936) describe how many compounds have been obtained for applications in medicine such as antibiotics and anticancer agents. Polyketide synthetases also can be employed in obtaining compounds that are relevant for the production of compounds for the chemical and fuel industries. Fortman et al. (WO 2012/050931) disclose that polyketide synthetases can be used in the production of olefins and they indicate that a large number of compounds potentially might be produced on the basis of the action of these enzymes.

Mendez-Perez et al. (2011, Appl. Environ. Microbiol. 77:4264-4267) disclose the production of traces of 1-alkenes as based on a polyketide synthetase in a Synechococcus strain. They suggest that the unsaturated hydrocarbons were produced from fatty acids via an elongation decarboxylation mechanism involving the so-termed Ols multi-domain enzyme.

Mendez-Perez et al (2012, Metabolic Engineering, 14:298-305) expressed the ols gene heterologously in E. coli but no production of an alkene was observed.

Although several bacterial processes have been suggested, it is expected that ultimately the yeast S. cerevisiae will be used for the commercial production of bulk compounds. This organism is preferred in large-scale processes because of its high efficiency and its robustness in industrial environments. Relative to other microbes, it can cope better with toxic compounds in lignocellulosic hydro lysates such as furans and acetic acid. The organism has the ability to adapt to toxic environments.

S. cerevisiae is incapable of fermenting C5-sugars. This aspect has been overcome to a certain extent by constructing xylose- and/or arabinose-utilizing S. cerevisiae strains as summarized by Madhavan et al (2012, Critical Reviews in Biotechnology, 32:22-48). It is very important that organisms can deal with C5-sugars because of the relative abundance of xylose and arabinose in second generation feedstocks.

Work on obtaining xylose- and/or arabinose-utilizing S. cerevisiae strains has been directed almost exclusively for obtaining ethanol-producing strains. Brat and Boles (2013 FEMS Yeast Research DOI: 10.1111/1567-1364.12028) have been exceptional in this respect since they constructed a S. cerevisiae strain that was able to convert xylose into isobutanol.

The present inventors have now surprisingly been able to successfully express in C5 -utilizing S. cerevisiae metabolic routes that previously have been implemented in E. coli. Consequently, the production of a number of highly-relevant chemicals is now feasible on industrial scales on the basis of lignocellulosic feedstocks.

Matsuda et al. (2011 Microb Cell Fact. 10:70) disclose that S. cerevisiae is a preferred host over E. coli but that their central metabolism are distinct from each other calling for novel metabolic engineering strategies in S. cerevisiae as compared to E. coli.

It is an object of the present invention to provide for modified eukaryotic microbial (host) cells that are capable of producing, on the basis of second generation feedstocks, compounds such as fatty acids, 1 -alcohols (other than ethanol or methanol), β-ketoacids, β-ketoalcohols, β-hydroxy acids, 1,3-diols, trans-A ² -fatty acids, alkenes, alkanes and/or derivatives thereof on the basis of second generation feedstocks. It is also an object of the present invention to provide for processes wherein such cells are used to produce such compounds from second generation feedstocks.

Summary of the invention

In a first aspect the invention pertains to a modified eukaryotic microbial host cell. The host cell preferably is modified to comprise: a) cytosolic expression of the enzymes of the fatty acid β-oxidation cycle, in the absence of fatty acids and in the presence of a non- fatty acid carbon source; b) a metabolic route for producing under oxic conditions and preferably in the cytosol, acetyl-CoA from the non-fatty acid carbon source to feed into and drive the β-oxidation cycle in the bio synthetic direction; c) expression of a termination enzyme to convert reaction intermediates of the β- oxidation cycle into a fermentation product selected from the group consisting of a fatty acid, a 1 -alcohol, an alkene, an alkane and derivatives thereof; and, d) the ability to grow and/or metabolize a pentose such as xylose and/or arabinose, preferably by comprising at least one of: i) expression of an exogenous xylose isomerase gene, which gene confers to the cell the ability to isomerize xylose into xylulose; and, ii) expression of exogenous genes coding for a L-arabinose isomerase, a L-ribulokinase and a L- ribulose-5-phosphate 4-epimerase, which genes together confer to the cell the ability to convert L-arabinose into D-xylulose 5-phosphate.

In a second aspect the invention pertains to the use of a cell modified according to the invention, for the production of at least one fermentation product selected from the group consisting of a fatty acid, a 1 -alcohol, an alkene, an alkane and derivatives thereof.

In a third aspect, the invention relates to processes for producing at least one fermentation product selected from the group consisting of a fatty acid, a 1 -alcohol, an alkene, an alkane and derivatives thereof, whereby the process comprises the steps of: a) fermenting a medium with a modified cell according to the invention, whereby the medium contains or is fed with a non- fatty acid carbon source and whereby the yeast cell ferments the non-fatty acid carbon source to the fermentation product; and optionally, b) recovery of the fermentation product. Preferably, the non-fatty acid carbon source comprises at least one of xylose and arabinose. Optionally the carbon source may comprise a source of hexoses like e.g. glucose.

Description of the invention

Definitions

Sequence identity is herein defined as a relationship between two or more amino acid (polypeptide or protein) sequences or two or more nucleic acid (polynucleotide) sequences, as determined by comparing the sequences. In the art, "identity" also means the degree of sequence relatedness between amino acid or nucleic acid sequences, as the case may be, as determined by the match between strings of such sequences. "Similarity" between two amino acid sequences is determined by comparing the amino acid sequence and its conserved amino acid substitutes of one polypeptide to the sequence of a second polypeptide. "Identity" and "similarity" can be readily calculated by known methods. The terms "sequence identity" or "sequence similarity" means that two (poly)peptide or two nucleotide sequences, when optimally aligned, preferably over the entire length (of at least the shortest sequence in the comparison) and maximizing the number of matches and minimizes the number of gaps such as by the programs ClustalW (1.83), GAP or BESTFIT using default parameters, share at least a certain percentage of sequence identity as defined elsewhere herein. GAP uses the Needleman and Wunsch global alignment algorithm to align two sequences over their entire length, maximizing the number of matches and minimizes the number of gaps. Generally, the GAP default parameters are used, with a gap creation penalty = 50 (nucleotides) / 8 (proteins) and gap extension penalty = 3 (nucleotides) / 2 (proteins). For nucleotides the default scoring matrix used is nwsgapdna and for proteins the default scoring matrix is Blosum62 (Henikoff & Henikoff, 1992, PNAS 89, 915-919). A preferred multiple alignment program for aligning protein sequences of the invention is ClustalW (1.83) using a blosum matrix and default settings (Gap opening penalty: 10; Gap extension penalty: 0.05). It is clear than when RNA sequences are said to be essentially similar or have a certain degree of sequence identity with DNA sequences, thymine (T) in the DNA sequence is considered equal to uracil (U) in the R A sequence. Sequence alignments and scores for percentage sequence identity may be determined using computer programs, such as the GCG Wisconsin Package, Version 10.3, available from Accelrys Inc., 9685 Scranton Road, San Diego, CA 92121-3752 USA or the open-source software Emboss for Windows (current version 2.7.1- 07). Alternatively percent similarity or identity may be determined by searching against databases such as FASTA, BLAST, etc.

Preferred methods to determine identity are designed to give the largest match between the sequences tested. Methods to determine identity and similarity are codified in publicly available computer programs. Preferred computer program methods to determine identity and similarity between two sequences include e.g. the GCG program package (Devereux, J., et al, Nucleic Acids Research 12 (1):387 (1984)), BestFit, BLASTP, BLASTN, and FASTA (Altschul, S. F. et al, J. Mol. Biol. 215:403-410 (1990). The BLAST X program is publicly available from NCBI and other sources (BLAST Manual, Altschul, S., et al, NCBI NLM NIH Bethesda, MD 20894; Altschul, S., et al, J. Mol. Biol. 215:403-410 (1990). The well-known Smith Waterman algorithm may also be used to determine identity.

Preferred parameters for polypeptide sequence comparison include the following: Algorithm: Needleman and Wunsch, J. Mol. Biol. 48:443-453 (1970); Comparison matrix: BLOSSUM62 from Hentikoff and Hentikoff, Proc. Natl. Acad. Sci. USA. 89: 10915-10919 (1992); Gap Penalty: 12; and Gap Length Penalty: 4. A program useful with these parameters is publicly available as the "Ogap" program from Genetics Computer Group, located in Madison, WI. The aforementioned parameters are the default parameters for amino acid comparisons (along with no penalty for end gaps). Preferred parameters for nucleic acid comparison include the following: Algorithm: Needleman and Wunsch, J. Mol. Biol. 48:443-453 (1970); Comparison matrix: matches=+10, mismatch=0; Gap Penalty: 50; Gap Length Penalty: 3. Available as the Gap program from Genetics Computer Group, located in Madison, Wis. Given above are the default parameters for nucleic acid comparisons.

Optionally, in determining the degree of amino acid similarity, the skilled person may also take into account so-called "conservative" amino acid substitutions, as will be clear to the skilled person. Conservative amino acid substitutions refer to the interchangeability of residues having similar side chains. For example, a group of amino acids having aliphatic side chains is glycine, alanine, valine, leucine, and isoleucine; a group of amino acids having aliphatic-hydroxyl side chains is serine and threonine; a group of amino acids having amide- containing side chains is asparagine and glutamine; a group of amino acids having aromatic side chains is phenylalanine, tyrosine, and tryptophan; a group of amino acids having basic side chains is lysine, arginine, and histidine; and a group of amino acids having sulphur-containing side chains is cysteine and methionine. Preferred conservative amino acids substitution groups are: valine-leucine-isoleucine, phenylalanine-tyrosine, lysine-arginine, alanine- valine, and asparagine-glutamine. Substitutional variants of the amino acid sequence disclosed herein are those in which at least one residue in the disclosed sequences has been removed and a different residue inserted in its place. Preferably, the amino acid change is conservative. Preferred conservative substitutions for each of the naturally occurring amino acids are as follows: Ala to ser; Arg to lys; Asn to gin or his; Asp to glu; Cys to ser or ala; Gin to asn; Glu to asp; Gly to pro; His to asn or gin; He to leu or val; Leu to ile or val; Lys to arg; gin or glu; Met to leu or ile; Phe to met, leu or tyr; Ser to thr; Thr to ser; Trp to tyr; Tyr to trp or phe; and, Val to ile or leu.

Nucleotide sequences of the invention may also be defined by their capability to hybridize with parts of specific nucleotide sequences disclosed herein, respectively, under moderate, or preferably under stringent hybridization conditions. Stringent hybridization conditions are herein defined as conditions that allow a nucleic acid sequence of at least about 25, preferably about 50 nucleotides, 75 or 100 and most preferably of about 200 or more nucleotides, to hybridize at a temperature of about 65 °C in a solution comprising about 1 M salt, preferably 6 x SSC or any other solution having a comparable ionic strength, and washing at 65 °C in a solution comprising about 0.1 M salt, or less, preferably 0.2 x SSC or any other solution having a comparable ionic strength. Preferably, the hybridization is performed overnight, i.e. at least for 10 hours and preferably washing is performed for at least one hour with at least two changes of the washing solution. These conditions will usually allow the specific hybridization of sequences having about 90% or more sequence identity.

Moderate conditions are herein defined as conditions that allow a nucleic acid sequences of at least 50 nucleotides, preferably of about 200 or more nucleotides, to hybridize at a temperature of about 45 °C in a solution comprising about 1 M salt, preferably 6 x SSC or any other solution having a comparable ionic strength, and washing at room temperature in a solution comprising about 1 M salt, preferably 6 x SSC or any other solution having a comparable ionic strength. Preferably, the hybridization is performed overnight, i.e. at least for 10 hours, and preferably washing is performed for at least one hour with at least two changes of the washing solution. These conditions will usually allow the specific hybridization of sequences having up to 50% sequence identity. The person skilled in the art will be able to modify these hybridization conditions in order to specifically identify sequences varying in identity between 50% and 90%.

A "nucleic acid construct" or "nucleic acid vector" is herein understood to mean a man-made nucleic acid molecule resulting from the use of recombinant DNA technology. The term "nucleic acid construct" therefore does not include naturally occurring nucleic acid molecules although a nucleic acid construct may comprise (parts of) naturally occurring nucleic acid molecules. The terms "expression vector" or expression construct" refer to nucleotide sequences that are capable of affecting expression of a gene in host cells or host organisms compatible with such sequences. These expression vectors typically include at least suitable transcription regulatory sequences and optionally, 3' transcription termination signals. Additional factors necessary or helpful in effecting expression may also be present, such as expression enhancer elements. The expression vector will be introduced into a suitable host cell and be able to effect expression of the coding sequence in an in vitro cell culture of the host cell. The expression vector will be suitable for replication in the host cell or organism of the invention.

As used herein, the term "promoter" or "transcription regulatory sequence" refers to a nucleic acid fragment that functions to control the transcription of one or more coding sequences, and is located upstream with respect to the direction of transcription of the transcription initiation site of the coding sequence, and is structurally identified by the presence of a binding site for DNA-dependent RNA polymerase, transcription initiation sites and any other DNA sequences, including, but not limited to transcription factor binding sites, repressor and activator protein binding sites, and any other sequences of nucleotides known to one of skill in the art to act directly or indirectly to regulate the amount of transcription from the promoter. A "constitutive" promoter is a promoter that is active in most tissues under most physiological and developmental conditions. An "inducible" promoter is a promoter that is physiologically or developmentally regulated, e.g. by the application of a chemical inducer.

The term "selectable marker" is a term familiar to one of ordinary skill in the art and is used herein to describe any genetic entity which, when expressed, can be used to select for a cell or cells containing the selectable marker. The term "reporter" may be used interchangeably with marker, although it is mainly used to refer to visible markers, such as green fluorescent protein (GFP). Selectable markers may be dominant or recessive or bidirectional.

As used herein, the term "operably linked" refers to a linkage of polynucleotide elements in a functional relationship. A nucleic acid is "operably linked" when it is placed into a functional relationship with another nucleic acid sequence. For instance, a transcription regulatory sequence is operably linked to a coding sequence if it affects the transcription of the coding sequence. Operably linked means that the DNA sequences being linked are typically contiguous and, where necessary to join two protein encoding regions, contiguous and in reading frame.

The terms "protein" or "polypeptide" are used interchangeably and refer to molecules consisting of a chain of amino acids, without reference to a specific mode of action, size, 3-dimensional structure or origin.

"Fungi" (singular fungus) are herein understood as heterotrophic eukaryotic microorganisms that digest their food externally, absorbing nutrient molecules into their cells. Fungi are a separate kingdom of eukaryotic organisms and include yeasts, molds, and mushrooms. The terms fungi, fungus and fungal as used herein thus expressly includes yeasts as well as filamentous fungi.

The term "gene" means a DNA fragment comprising a region (transcribed region), which is transcribed into an RNA molecule (e.g. an mRNA) in a cell, operably linked to suitable regulatory regions (e.g. a promoter). A gene will usually comprise several operably linked fragments, such as a promoter, a 5' leader sequence, a coding region and a 3'-nontranslated sequence (3 'end) comprising a polyadenylation site. "Expression of a gene" refers to the process wherein a DNA region which is operably linked to appropriate regulatory regions, particularly a promoter, is transcribed into an RNA, which is biologically active, i.e. which is capable of being translated into a biologically active protein or peptide. The term "homologous" when used to indicate the relation between a given (recombinant) nucleic acid or polypeptide molecule and a given host organism or host cell, is understood to mean that in nature the nucleic acid or polypeptide molecule is produced by a host cell or organisms of the same species, preferably of the same variety or strain. If homologous to a host cell, a nucleic acid sequence encoding a polypeptide will typically (but not necessarily) be operably linked to another (heterologous) promoter sequence and, if applicable, another (heterologous) secretory signal sequence and/or terminator sequence than in its natural environment. It is understood that the regulatory sequences, signal sequences, terminator sequences, etc. may also be homologous to the host cell. In this context, the use of only "homologous" sequence elements allows the construction of "self-cloned" genetically modified organisms (GMO's) (self-cloning is defined herein as in European Directive 98/81/EC Annex II). When used to indicate the relatedness of two nucleic acid sequences the term "homologous" means that one single- stranded nucleic acid sequence may hybridize to a complementary single-stranded nucleic acid sequence. The degree of hybridization may depend on a number of factors including the amount of identity between the sequences and the hybridization conditions such as temperature and salt concentration as discussed later.

The terms "heterologous" and "exogenous" when used with respect to a nucleic acid (DNA or RNA) or protein refers to a nucleic acid or protein that does not occur naturally as part of the organism, cell, genome or DNA or RNA sequence in which it is present, or that is found in a cell or location or locations in the genome or DNA or RNA sequence that differ from that in which it is found in nature. Heterologous and exogenous nucleic acids or proteins are not endogenous to the cell into which it is introduced, but have been obtained from another cell or synthetically or recombinantly produced. Generally, though not necessarily, such nucleic acids encode proteins, i.e. exogenous proteins, that are not normally produced by the cell in which the DNA is transcribed or expressed. Similarly exogenous RNA encodes for proteins not normally expressed in the cell in which the exogenous RNA is present. Heterologous/exogenous nucleic acids and proteins may also be referred to as foreign nucleic acids or proteins. Any nucleic acid or protein that one of skill in the art would recognize as foreign to the cell in which it is expressed is herein encompassed by the term heterologous or exogenous nucleic acid or protein. The terms heterologous and exogenous also apply to non-natural combinations of nucleic acid or amino acid sequences, i.e. combinations where at least two of the combined sequences are foreign with respect to each other. The "specific activity" of an enzyme is herein understood to mean the amount of activity of a particular enzyme per amount of total host cell protein, usually expressed in units of enzyme activity per mg total host cell protein. In the context of the present invention, the specific activity of a particular enzyme may be increased or decreased as compared to the specific activity of that enzyme in an (otherwise identical) wild type host cell.

"Aerobic conditions" "Oxic conditions" or an aerobic or oxic fermentation process is herein defined as conditions or a fermentation process run in the presence of oxygen and in which oxygen is consumed, preferably at a rate of at least 0.5, 1, 2, 5, 10, 20 or 50 mmol/L/h, and wherein organic molecules serve as electron donor and oxygen serves as electron acceptor.

Any reference to nucleotide or amino acid sequences accessible in public sequence databases herein refers to the version of the sequence entry as available on the filing date of this document.

Detailed description of the invention

E. coli has been engineered by constitutively expressing enzymes of the aerobic fatty acid β-oxidation cycle (the "β-oxidation cycle") in combination with expressing other relevant enzymes and by also removing certain enzyme activities that would lead to unwanted side products. Such engineered bacteria have been proposed as production organisms in the manufacturing from sugar of several chemicals (see e.g. Clomburg and Gonzalez, 2010 Biotechnol Bioeng. 108(4): 867-79; Dellomonaco et al. supra; and Gulevich et al, Biotechnol Letters 2012, 34:463-469).

The present inventors have found, however, that under relevant industrial conditions the production processes envisaged will be far from optimal due to the properties of the E. coli host. They have now found that it is possible to (constitutively) express all relevant genes for producing compounds via the reverse aerobic fatty acid β- oxidation cycle in yeasts such as S. cerevisiae. In combination with overexpressing and/or deleting enzymes, the present inventors have demonstrated that the industrially- relevant S. cerevisiae can be successfully employed in producing desired compounds under industrial conditions.

In a first aspect, the invention relates to a eukaryotic microbial host cell, preferably a yeast host cell, that has the desired properties for industrial processes and that will be modified following standard molecular techniques. A host cell of the invention preferably comprises the following features, which features may already be present in the host cell or which may be newly introduced and/or modified to be improved:

I) The host cell preferably functionally expresses (enzymes of) the fatty acid β- oxidation cycle, preferably in the cytosol of the host cell. Preferably the β-oxidation cycle enzymes are expressed under conditions that include the presence of non-fatty acid carbon sources, i.e. carbohydrates (such as hexoses and disaccharides), including catabolite repressing carbon sources, and preferably also in the absence of oxygen (anoxic/anaerobic conditions). These features are described in more detail under 4) herein.

II) The host cell preferably provides acetyl-CoA to feed into and drive the β- oxidation cycle in the reverse, i.e. biosynthetic direction. The acetyl-CoA is preferably provided in the cytosol of the host cell via pyruvate that is obtained from a non-fatty acid carbon source, preferably through at least (the last) part of the glycolytic pathway. These features are described in more detail under 3) herein. In certain instances, as described in 4.5) herein, the driving force towards product formation can be enhanced by introducing a malonyl-CoA shunt (Lan and Liao 2012, PNAS, 109:6018-6023).

III) The host preferably directs as much as possible of the carbon flow as acetyl- Co A into the reversed β-oxidation cycle. Reactions producing unwanted by-products

(such as e.g. ethanol, acetate, glycerol and/or acetaldehyde) as well as reactions that would compete for cytosolic acetyl-CoA with reversed β-oxidation cycle, are therefore preferably absent or modified so as to reduce or eliminate their activity in the host cell. These features are described in more detail under 2) herein.

IV) The host cell preferably functionally expresses termination enzymes to convert reaction intermediates of the β-oxidation cycle into the desired end product, such as fatty acids, 1 -alcohols, β-ketoacids, β-ketoalcohols, β-hydroxyacids, 1,3-diols, trans-A ² -fatty acids, alkanes, alkenes and/or derivatives thereof, preferably, 1 -alcohols other than methanol and ethanol, such as e.g. butanol, decanol, dodecanol and higher 1- alcohols and/or derivatives thereof. These features are described in more detail under 5) herein. These compounds can find use as (bio)fuels) or they can find application outside the fuel's world as chemicals. V) The host cell further preferably contains the ability to grow on pentoses such as xylose and arabinose. These features are described in more detail under 1.3) herein below. 1.1 The parent host cell

The present invention concerns the genetic modification of a host cell so as to enable the host cell to produce desired compounds via the reversal of the fatty acid β- oxidation pathway. To this end a number of genetic modifications will be introduced in a parent host cell in accordance with the invention. These modifications include the introduction of expression of a number of heterologous genes, as well as, the modification of the expression of a number of endogenous genes already present in the parent host cell, by reducing or inactivating the expression of some endogenous genes and/or by increasing, i.e. overexpressing, other endogenous genes. These genetic modification are further set out below herein. A parent host cell is thus understood to be a host cell prior to that any of the genetic modifications in accordance with the invention have been introduced in the host cell.

A parent host cell of the invention preferably is a eukaryotic host cell, more preferably, the host cell is a eukaryotic microorganism such as e.g. a fungal host cell. A most preferred parent host cell to be modified in accordance with the invention is a yeast host cell.

Yeasts are herein defined as eukaryotic microorganisms and include all species of the subdivision Eumycotina (Yeasts: characteristics and identification, J.A. Barnett, R.W. Payne, D. Yarrow, 2000, 3rd ed., Cambridge University Press, Cambridge UK; and, The yeasts, a taxonomic study, CP. Kurtzman and J.W. Fell (eds) 1998, 4th ed., Elsevier Science Publ. B.V., Amsterdam, The Netherlands) that predominantly grow in unicellular form. Yeasts may either grow by budding of a unicellular thallus or may grow by fission of the organism. Preferred yeasts cells for use in the present invention belong to the genera Saccharomyces, Kluyveromyces, Candida, Pichia, Schizosaccharomyces, Hansenula, Kloeckera, Schwanniomyces, Yarrowia, Cryptococcus, Debaromyces, Saccharomycecopsis, Saccharomycodes, Wickerhamia, Debayomyces, Hanseniaspora, Ogataea, Kuraishia, Komagataella, Metschnikowia, Williopsis, Nakazawaea, Torulaspora, Bullera, Rhodotorula, and Sporobolomyces. Preferably the parental yeast host cell is naturally capable of anaerobic fermentation, more preferably alcoholic fermentation and most preferably anaerobic alcoholic fermentation. However, as will become apparent below herein, in one embodiment of the invention, the host cell is modified to avoid or reduce the synthesis of ethanol.

Particularly when compared to bacteria, yeasts, such as Saccharomyces species, have many attractive features for industrial processes, including e.g. their high tolerance to acids, ethanol and other harmful compounds, their high osmo-tolerance and their capability of anaerobic growth, and of course their high fermentative capacity. Preferred yeast species as parent host cells for the invention include e.g. Saccharomyces cerevisiae, S. exiguus, S. bayanus, S. delbriickii, S. italicus, S. ellipsoideus, S. fermentati, S. kluyveri, S. krusei, S. lactis, S. marxianus, S. microellipsoides, S. montanus, S. norbensis, S. oleaceus, S. paradoxus, S. pastorianus, S. pretoriensis, S. rosei, S. rouxii, S. uvarum, S. ludwigii, Kluyveromyces lactis, K. marxianus, K. marxianus var. marxianus, K. thermotolerans, Candida utilis, C. tropicalis, C. albicans, C. lipolytica, C. versatilis, Pichia stipidis, P. pastoris and P. sorbitophila, Hansenula polymorpha and Schizosaccharomyces pombe.

The parent host cell of the invention preferably contains active glycolysis.

The host cell further preferably has a high tolerance to ethanol, a high tolerance to low pH (i.e. capable of growth at a pH lower than 5, 4, or 3) and towards organic acids like lactic acid, acetic acid or formic acid and sugar degradation products such as furfural and hydroxy-methylfurfural, and a high tolerance to elevated temperatures. Any of these characteristics or activities of the host cell may be naturally present in the host cell or may be introduced or modified by genetic modification, preferably by self cloning or by the methods of the invention described below.

A suitable cell is a cultured cell, a cell that may be cultured in fermentation process e.g. in submerged or solid state fermentation.

1.2 Methods for modifying and constructing the host cells of the invention

For the genetic modification of the parent host cells of the invention, i.e. for the construction of the modified host cells of the invention, standard genetic and molecular biology techniques are used that are generally known in the art and have e.g. been described by Sambrook and Russell (2001, "Molecular cloning: a laboratory manual" (3rd edition), Cold Spring Harbor Laboratory, Cold Spring Harbor Laboratory Press) and Ausubel et al. (1987, eds., "Current protocols in molecular biology", Green Publishing and Wiley Interscience, New York). Furthermore, the construction of genetically modified (yeast) host strains may be carried out by genetic crosses, sporulation of the resulting diploids, tetrad dissection of the haploid spores containing the desired auxotrophic markers, and colony purification of such haploid host cells in the appropriate selection medium. All of these methods are standard fungal and yeast genetic methods known to those in the art. See, for example, Sherman et al, Methods Yeast Genetics, Cold Spring Harbor Laboratory, NY (1978) and Guthrie et al. (Eds.) Guide To Yeast Genetics and Molecular Biology Vol. 194, Academic Press, San Diego (1991).

In general, suitable promoters for the expression of the heterologous nucleotide sequence coding for desired enzyme activities and/or for overexpression of endogenous genes in the context of the invention, include promoters that are preferably insensitive to catabolite (glucose) repression, that are active under oxic (aerobic) and under anoxic (anaerobic) conditions and/or that preferably do not require specific carbon sources for induction. Promoters having these characteristics are widely available and known to the skilled person. Suitable examples of such promoters include e.g. promoters from glycolytic genes such as the phosphofructokinase (PPK), triose phosphate isomerase (ΤΡΓ), glyceraldehyde-3 -phosphate dehydrogenase (GPD, TDH3 or GAPDH), pyruvate kinase (PYK), phosphoglycerate kinase (PGK), glucose-6-phosphate isomerase promoter (PGIl) promoters from yeasts. More details about such promoters from yeast may be found in (WO 93/03159). Other useful promoters are ribosomal protein encoding gene promoters (TEFI), the lactase gene promoter (LAC4), alcohol dehydrogenase promoters (ADH1, preferably a modified (constitutive) version of the ADH1 promoter (SEQ ID NO: 33), ADH4, and the like), the enolase promoter (ENO) and the hexose(glucose) transporter promoter (HXT7). Another promoter is e.g. the S. cerevisiae ANB1 promoter (SEQ ID NO: 73), which is however not preferred for use under oxic conditions. Other promoters, both constitutive and inducible, and enhancers or upstream activating sequences will be known to those of skill in the art. Preferably the promoter that is operably linked to nucleotide sequence as defined above is homologous to the host cell. Suitable terminator sequences are e.g. obtainable from the cytochrome cl (CYC1) gene or an alcohol dehydrogenase gene (e.g. ADH1).

To increase the likelihood that the enzymes are expressed at sufficient levels and in active form in the transformed host cells of the invention, the nucleotide sequence encoding of enzymes of the invention, are preferably adapted to optimize their codon usage to that of the host cell in question. The adaptiveness of a nucleotide sequence encoding an enzyme to the codon usage of a host cell may be expressed as codon adaptation index (CAI). The codon adaptation index is herein defined as a measurement of the relative adaptiveness of the codon usage of a gene towards the codon usage of highly expressed genes in a particular host cell or organism. The relative adaptiveness (w) of each codon is the ratio of the usage of each codon, to that of the most abundant codon for the same amino acid. The CAI index is defined as the geometric mean of these relative adaptiveness values. Non-synonymous codons and termination codons (dependent on genetic code) are excluded. CAI values range from 0 to 1, with higher values indicating a higher proportion of the most abundant codons (see Sharp and Li , 1987, Nucleic Acids Research 15: 1281-1295; also see: Jansen et al, 2003, Nucleic Acids Res. 3J_(8):2242-51). An adapted nucleotide sequence preferably has a CAI of at least 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8 or 0.9. Most preferred are the sequences which have been codon optimized for expression in the fungal host cell in question such as e.g. S. cerevisiae cells.

1.3 Introducing the ability to use pentoses as carbon and energy source

The parent host cell of the invention, further preferably contains the ability to (anaerobically) grow on pentoses such as xylose and arabinose. As most wild type yeasts do not have the ability to anaerobically ferment pentoses such as xylose and arabinose, a preferred parent host cell of the invention is a cell that has been modified to this end. Such modifications will at least include the expression of an exogenous xylose isomerase activity (for xylose) and/or expression of exogenous arabinose isomerase (araA), ribulokinase (araB), and ribulose-5-P-4-epimerase (araD) activities (for arabinose). Yeast strains modified for the ability to directly isomerize xylose into xylulose have been described in e.g. WO 2003/062430, US 20060234364, Madhavan et al, 2008, DOI 10.1007/s00253-008-1794-6, WO 2006/009434, WO 2009/109633, Brat et al, 2009, Appl. Environ. Microbiol. 75: 2304-2311, WO 2010/070549, WO 2010/074577 and WO 2011/006136. Yeast strains modified for the ability to convert L- arabinose into D-xylulose 5-phosphate have been described in Wisselink et al. (2007, AEM Accepts, published online ahead of print on 1 June 2007; Appl. Environ. Microbiol. doi: 10.1128/AEM.00177-07), WO 2008/041840 and WO 2009/011591. Further preferred genetic modifications that may improve the host cell's ability to anaerobically ferment pentoses such as xylose and arabinose include:

i) an increase of xylulokinase activity (by overexpression of endogenous genes and/or introduction of heterologous genes; see e.g. WO 2003/062430),

ii) a genetic modification that increases the flux of the (non-oxidative part of the) pentose phosphate pathway as described in WO 06/009434, e.g. by overexpression of one or more of the ribulose-5-phosphate isomerase, ribulose-5 -phosphate 3-epimerase, transketolase and transaldolase genes, and,

iii) reduction or inactivation of expression of unspecific aldose reductase activity, as described in WO 06/009434.

In a preferred embodiemt these genetic modifications are:

i) overexpression of the S. cerevisiae XKS1 gene or an orthologue thereof;

ii) overexpression of one or more of the S. cerevisiae RPI1, RPE1, TKL1 and TALI genes or orthologues thereof; and,

iii) reduction or elimination of the expression of the S. cerevisiae GRE3 gene or an orthologue thereof.

The host is further preferably is a host capable of active or passive pentose (xylose and preferably also arabinose) transport into the cell. 2. Inactivation or reduction of unwanted side reactions

In some embodiments of the invention, the host cells of the invention are genetically modified so as to reduce or inactivate unwanted side reactions.

2.1. Reducing pyruvate decarboxylase activity

In one embodiment of the invention, the endogenous pyruvate decarboxylase activity in the host cell is reduced or eliminated as a means to prevent the formation of at least one acetaldehyde, acetate and ethanol.

Endogenous pyruvate decarboxylase activity in yeast converts pyruvate to acetaldehyde, which is then converted to ethanol, or to acetyl-CoA via acetate (see Figure 1). Yeasts may have one or more genes encoding pyruvate decarboxylase. For example, there is one gene encoding pyruvate decarboxylase in Kluyveromyces lactis, while there are three isozymes of pyruvate decarboxylase encoded by the PDCI,

PCD 5, and PDC6 (EC 4.1.1.1) genes in Saccharomyces cerevisiae, as well as a pyruvate decarboxylase regulatory gene PDC2. Expression of pyruvate decarboxylase from PDC6 is minimal. In the cells of the invention, the specific pyruvate decarboxylase activity is preferably reduced by at least a factor 0.8, 0.5, 0.3, 0.1, 0.05 or 0.01 as compared to cells of a strain which is genetically identical except for the genetic modification causing the reduction in activity, at least under anaerobic conditions.

In the modified host cells of the invention the pyruvate decarboxylase activity can be reduced by modifying at least one gene encoding a pyruvate decarboxylase, or by modifying a gene regulating the expression of pyruvate decarboxylase gene(s). For example, in S. cerevisiae the PDC1 and PDC5 genes, or all three genes, can be modified to reduce or eliminate their expression e.g. by disruption. Alternatively, pyruvate decarboxylase activity may be reduced by modification, e.g. disruption, of the PDC2 regulatory gene in S. cerevisiae.

Thus, preferred genes to be modified for reducing the specific pyruvate decarboxylase activity in the cell of the invention are one or more or all of the S. cerevisiae PDC1, PDC5, PDC6 and PDC2 genes, encoding the amino acid sequences of SEQ ID NO's: 1, 2, 3 and 4, respectively, or orthologues thereof in other species. Therefore genes to be modified for reducing the specific pyruvate decarboxylase activity in the cell of the invention, preferably are one or more or all of the genes encoding an amino acid sequence with at least 70, 75, 80, 85, 90, 95, 98 or 99% sequence identity to one or more of SEQ ID NO's: 1, 2, 3 and 4, respectively.

Examples of yeast strains with reduced pyruvate decarboxylase activity due to disruption of pyruvate decarboxylase encoding genes have been reported for Saccharomyces in Flikweert et al. (Yeast (1996) 12:247-257), and disruption of the regulatory gene in Hohmann, (Mol Gen Genet. (1993) 241 :657-666). Saccharomyces strains having no pyruvate decarboxylase activity are available from the ATCC with accession no.'s #200027 and #200028.

Genes encoding pyruvate decarboxylases and/or PDC regulatory genes may be modified so as to reduce pyruvate decarboxylase specific activity in host cells of the invention using a variety of methods for genetic modification. Many methods for modifying endogenous target genes in host cells so as to reduce or eliminate the activity of the encoded target proteins are known to one skilled in the art and may be used for modifying the host cells of the invention. Modifications that may be used to reduce or eliminate expression of a target protein are disruptions that include, but are not limited to, deletion of the entire gene or a portion of the gene encoding the target protein (e.g. a pyruvate decarboxylase), inserting a DNA fragment into the target gene (in either the promoter or coding region) so that the protein is not expressed or expressed at lower levels, introducing a mutation into the target coding region which adds a stop codon or frame shift such that a functional protein is not expressed, and introducing one or more mutations into a target coding region to alter amino acids so that a non-functional target protein, or a target protein with reduced enzymatic activity is expressed. In addition, expression of the target gene may be blocked by expression of an antisense RNA or an interfering RNA, and constructs may be introduced that result in co-suppression. Moreover, a target coding sequence may be synthesized whose expression will be low because rare codons are substituted for plentiful ones, when this suboptimal coding sequence is substituted for the corresponding endogenous target coding sequence. Preferably such a suboptimal coding sequence will have a codon adaptation index (see above) of less than 0.5, 0.4, 0.3 0.2, or 0.1. Such a suboptimal coding sequence will produce the same polypeptide but at a lower rate due to inefficient translation. In addition, the synthesis or stability of the transcript may be reduced by mutation. Similarly the efficiency by which a protein is translated from mRNA may be modulated by mutation, e.g. by using suboptimal translation initiation codons. All of these methods may be readily practiced by one skilled in the art making use of the known or identified sequences encoding target proteins such as the pyruvate decarboxylase proteins.

DNA sequences flanking a target coding sequence such as the pyruvate decarboxylase coding sequences are also useful in some modification procedures and are available for yeasts such as for Saccharomyces cerevisiae in the complete genome sequence coordinated by Genome Project ID9518 of Genome Projects coordinated by NCBI (National Center for Biotechnology Information) with identification GOPID #13838.

In particular, DNA sequences surrounding a target coding sequence are useful for modification methods using homologous recombination. For example, in this method sequences flanking the target gene are placed on either site of a selectable marker gene to mediate homologous recombination whereby the marker gene replaces the target gene. Also partial target gene sequences and target gene flanking sequences bounding a selectable marker gene may be used to mediate homologous recombination whereby the marker gene replaces a portion of the target gene, e.g. a pyruvate decarboxylase gene. In addition, the selectable marker may be flanked by site-specific recombination sites, so that following expression of the corresponding site-specific recombinase, the resistance gene is excised from the genomic locus where the target gene was present without reactivating the latter. The site-specific recombination leaves behind a recombination site which disrupts expression of the target protein. The homologous recombination vector may be constructed to also leave a deletion in the target gene following excision of the selectable marker, as is well known to one skilled in the art.

Deletions may be made using mitotic recombination as described in Wach et al.

((1994) Yeast 10: 1793-1808). This method involves preparing a DNA fragment that contains a selectable marker between genomic regions that may be as short as 20 bp, and which bound, i.e. flank the target DNA sequence. This DNA fragment can be prepared by PCR amplification of the selectable marker gene using as primers oligonucleotides that hybridize to the ends of the marker gene and that include the genomic regions that can recombine with the yeast genome. The linear DNA fragment can be efficiently transformed into yeast and recombined into the genome resulting in gene replacement including with deletion of the target DNA sequence (as described in Methods in Enzymology, v 194, pp 281-301 (1991)).

Moreover, promoter replacement methods may be used to exchange the endogenous transcriptional control elements allowing another means to modulate expression such as described in Mnaimneh et al. ((2004) Cell 118(1):31-44) and in the Examples herein.

In addition, the activity of target proteins in any yeast cell may be disrupted using random mutagenesis, which is followed by screening to identify strains with reduced activity of the target proteins such as e.g. a pyruvate decarboxylase. Using this type of method, the DNA sequence coding for the target proteins (e.g. the pyruvate decarboxylase encoding region), or any other region of the genome affecting expression of pyruvate decarboxylase activity, need not even be known.

Methods for creating genetic mutations are common and well known in the art and may be applied to the exercise of creating mutants. Commonly used random genetic modification methods (reviewed in Methods in Yeast Genetics, 2005, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.) include spontaneous mutagenesis, mutagenesis caused by mutator genes, chemical mutagenesis, irradiation with UV or X-rays, or transposon mutagenesis.

Chemical mutagenesis of yeast commonly involves treatment of yeast cells with one of the following DNA mutagens: ethyl methanesulfonate (EMS), nitrous acid, diethyl sulfate, or N-methyl-N'-nitro-N-nitroso-guanidine (MNNG). These methods of mutagenesis have been reviewed in Spencer et al (Mutagenesis in Yeast, 1996, Yeast Protocols: Methods in Cell and Molecular Biology. Humana Press, Totowa, N.J.). Chemical mutagenesis with EMS may be performed as described in Methods in Yeast Genetics, 2005, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. Irradiation with ultraviolet (UV) light or X-rays can also be used to produce random mutagenesis in yeast cells. The primary effect of mutagenesis by UV irradiation is the formation of pyrimidine dimers which disrupt the fidelity of DNA replication. Protocols for UV-mutagenesis of yeast can be found in Spencer et al (Mutagenesis in Yeast, 1996, Yeast Protocols: Methods in Cell and Molecular Biology. Humana Press, Totowa, N.J.). Introduction of a mutator phenotype can also be used to generate random chromosomal mutations in yeast. Common mutator phenotypes can be obtained through disruption of one or more of the following genes: PMS1, MAGI, RAD18 or RAD51. Restoration of the non-mutator phenotype can be easily obtained by insertion of the wild type allele. Collections of modified cells produced from any of these or other known random mutagenesis processes may be screened for reduced activity of the target protein (US20090305363).

In a preferred embodiment of the host cell wherein the pyruvate decarboxylase activity is reduced or eliminated, the host comprises a further genetic modification of at least one copy of a transcriptional regulator involved in glucose sensing. S. cerevisiae strains in which the structural genes encoding pyruvate decarboxylase (PDC1, PDC5 and PDC6) are deleted are unable to grow in the presence of high glucose concentrations or under glucose-limited conditions without additional C ₂ carbon sources such as ethanol or acetate - due to their inability to form cytosolic acetyl-CoA for e.g. fatty acid synthesis. The addition of C ₂ carbon sources is however preferably avoided for processes at industrial scale. This ability of PDC-minus strain to grow on glucose in the absence of additional C ₂ sources can be restored by increasing the steady state level of a transcriptional regulator involved in glucose sensing, as described by Oud et al. (2012, Microbial Cell Factories, 11 : 131). The transcriptional regulator is preferably encoded by the MTH1 gene or an orthologue thereof. The genetic modification preferably is a genetic modification that increases the steady state level of the THi-encoded protein, by at least one of overexpressing the protein and inactivation of a phosphorylation site required for degradation of the transcriptional regulator. The transcriptional regulator to be modified preferably has an amino acid sequence with at least 70, 75, 80, 85, 90, 95, 98 or 99% sequence identity to SEQ ID NO: 5. The genetic modification that inactivates the phosphorylation site, preferably is an internal in- frame deletion of a segment comprising amino acid positions 57 - 131 of SEQ ID NO: 5, or a corresponding segment in an orthologue of SEQ ID NO: 5.

2.2 Reducing cytosolic acetaldehyde dehydrogenase activity

In one embodiment of the invention, the endogenous cytosolic acetaldehyde dehydrogenase activity in the host cell is reduced or eliminated so as to avoid or reduce the synthesis of acetate from acetaldehyde.

In the yeast genome, there are five genes known to encode aldehyde dehydrogenases, as well as an additional gene with sequence similarity. Ald2p and Ald3p are cytosolic enzymes which use only NAD ⁺ as cofactor (EC 1.2.1.3). Both genes are induced in response to ethanol or stress and repressed by glucose. Ald4p and Ald5p are mitochondrial, use NAD and NADP as co factors, and are K ⁺ dependent. Ald4p, the major isoform, is glucose repressed and ald4 mutants do not grow on ethanol, while Ald5p, the minor isoform, is constitutively expressed. ALD6 encodes the Mg ²⁺ activated cytosolic enzyme, which uses NADP ⁺ as cofactor and is constitutively expressed (EC 1.2.1.4). However, the cytosolic ALD6 gene product is the major enzyme responsible for catalyzing the oxidation of acetaldehyde to acetate in yeast.

Thus, in a preferred cell according to the invention the gene to be modified for reducing or eliminating the specific cytosolic acetaldehyde dehydrogenase activity in the cell of the invention is at least the S. cerevisiae ALD6 gene, encoding the amino acid sequence of SEQ ID NO: 6, or an orthologue thereof in another species. Therefore a gene to be modified for reducing or eliminating the specific cytosolic acetaldehyde dehydrogenase activity in the cell of the invention, preferably is a gene encoding a amino acid sequence with at least 70, 75, 80, 85, 90, 95, 98 or 99% sequence identity to SEQ ID NO: 6. However, in other preferred cells the expression of one or more or all of the ALD1, ALD2, ALD3, ALD4, ALD5 and ALD6 genes or their corresponding orthologues is reduced or eliminated.

In the cells of the invention, the specific cytosolic acetaldehyde dehydrogenase activity is preferably reduced by at least a factor 0.8, 0.5, 0.3, 0.1, 0.05 or 0.01 as compared to cells of a strain which is genetically identical except for the genetic modification causing the reduction in activity, at least under anaerobic conditions.

Methods for reducing or eliminating the specific cytosolic acetaldehyde dehydrogenase activity in the cell of the invention are as described above in 2.1 for pyruvate decarboxylase as target protein.

2.3. Reducing alcohol dehydrogenase activity

In one embodiment of the invention, the endogenous alcohol dehydrogenase activity in the host cell is reduced or eliminated so as to avoid or reduce the synthesis of ethanol from acetaldehyde.

In S. cerevisiae, there are five genes that encode alcohol dehydrogenases involved in ethanol metabolism, ADH1 to ADH5. Four of these enzymes, Adhlp, Adh3p, Adh4p, and Adh5p, reduce acetaldehyde to ethanol during glucose fermentation, while Adh2p catalyzes the reverse reaction of oxidizing ethanol to acetaldehyde. However, the cytosolic ADH1 gene product is the major enzyme responsible for catalyzing the reduction of acetaldehyde to ethanol with the concomitant regeneration of NAD ⁺ (EC 1.1.1.1).

Thus, a preferred gene to be modified for reducing or eliminating the specific ADH1 -encoded alcohol dehydrogenase activity in the cell of the invention is the S. cerevisiae ADH1 gene, encoding the amino acid sequence of SEQ ID NO: 7, or an orthologue thereof in another species. Therefore a gene to be modified for reducing or eliminating the specific ADH1 -encoded alcohol dehydrogenase activity in the cell of the invention, preferably is a gene encoding an amino acid sequence with at least 70, 75, 80, 85, 90, 95, 98 or 99% sequence identity to SEQ ID NO: 7.

Alternatively, preferred genes to be modified for reducing or eliminating the specific alcohol dehydrogenase activity in the cell of the invention are one or more of the S. cerevisiae ADH3, ADH4 and ADH5 genes or orthologues thereof in another species. The S. cerevisiae ADH3, ADH4 and ADH5 genes or their orthologues may be modified as such, or in combination with the above modification of the ADH1 gene or its orthologue. Preferably, at least the S. cerevisiae ADH3 gene or its orthologue is modified for reducing or eliminating its specific activity in the cell of the invention, more preferably, in combination with the modification of at least one of the S. cerevisiae ADH1, ADH4 and ADH5 genes or their orthologues. Therefore further genes to be modified for reducing or eliminating the specific alcohol dehydrogenase activity in the cell of the invention, preferably are genes encoding an amino acid sequence with at least 70, 75, 80, 85, 90, 95, 98 or 99% sequence identity to the amino acid sequences of the S. cerevisiae ADH3, ADH4 and ADH5 genes (with respectively Genbank accession no.'s CAA89229.1, CAA64131.1 and CAA85103.1 and SEQ ID NO's 90 - 92).

In the cells of the invention, the specific (ADH1 -encoded or overall) alcohol dehydrogenase activity is preferably reduced by at least a factor 0.8, 0.5, 0.3, 0.1, 0.05 or 0.01 as compared to cells of a strain which is genetically identical except for the genetic modification causing the reduction in activity, under aerobic or anaerobic conditions.

In the cells of the invention, the specific (^DHJ-encoded) alcohol dehydrogenase activity is preferably reduced by at least a factor 0.8, 0.5, 0.3, 0.1, 0.05 or 0.01 as compared to cells of a strain which is genetically identical except for the genetic modification causing the reduction in activity, at least under anaerobic conditions.

In the cells of the invention, the specific ( _^ 4ZXH¥-encoded) alcohol dehydrogenase activity is preferably reduced by at least a factor 0.8, 0.5, 0.3, 0.1, 0.05 or 0.01 as compared to cells of a strain which is genetically identical except for the genetic modification causing the reduction in activity, at least under anaerobic conditions.

In the cells of the invention, the specific ( _^ 4DH5-encoded) alcohol dehydrogenase activity is preferably reduced by at least a factor 0.8, 0.5, 0.3, 0.1, 0.05 or 0.01 as compared to cells of a strain which is genetically identical except for the genetic modification causing the reduction in activity, at least under anaerobic conditions.

Methods for reducing or eliminating the endogenous alcohol dehydrogenase activity in the cell of the invention are as described above in 2.1 for pyruvate decarboxylase as target protein.

2.4 Reducing endogenous acetyl- Co A synthetase activity In one embodiment of the invention, the endogenous acetyl-CoA synthetase activity in the host cell is reduced or eliminated so as to avoid or reduce the synthesis of acetyl-CoA from acetate. In this embodiment, whether or not one or all of the endogenous acetyl-CoA synthetase genes are inactivated depends on the introduction of heterologous genes for the production of cytosolic acetyl-CoA; see 3.1, 3.2 and 3.3 below).

An acetyl-CoA synthetase or acetate-CoA ligase (EC 6.2.1.1) is herein understood as an enzyme that catalyzes the formation of a new chemical bond between acetate and coenzyme A (CoA). In the modified host cells of the invention the acetyl- CoA synthetase activity can be reduced by modifying at least one gene encoding a acetyl-CoA synthetase. For example, in S. cerevisiae one or both (all) of the ACS1 and ACS2 genes can be modified to reduce or eliminate their expression e.g. by disruption.

Thus, preferred genes to be modified for reducing or eliminating the specific acetyl-CoA synthetase activity in the cell of the invention are at least one of the S. cerevisiae ACS1 and ACS2 genes, encoding the amino acid sequence of SEQ ID NO's: 8 and 9, respectively, or orthologues thereof in another species.

Therefore genes to be modified for reducing the specific acetyl-CoA synthetase activity in the cell of the invention, preferably are one or both of the genes encoding an amino acid sequence with at least 70, 75, 80, 85, 90, 95, 98 or 99% sequence identity to one or more of SEQ ID NO's: 8 and 9, respectively.

In the cells of the invention, the specific acetyl-CoA synthetase activity is preferably reduced by at least a factor 0.8, 0.5, 0.3, 0.1, 0.05 or 0.01 as compared to cells of a strain which is genetically identical except for the genetic modification causing the reduction in activity, at least under anaerobic conditions.

Methods for reducing or eliminating the specific acetyl-CoA synthetase activity in the cell of the invention are as described above in 2.1 for pyruvate decarboxylase as target protein.

2.5 Inactivation or repression of A CC1

In one embodiment of the invention, the activity of the product of the ACC1 gene or an orthologue thereof is reduced in the host cell by genetically modifying the gene so as to avoid or reduce the formation of fatty acids, via the yeast native fatty acid forming pathway because this pathway would compete with the reversed β-oxidation pathway introduced in the host cell of the invention.

The S. cerevisiae ACCI gene encodes the acetyl-CoA carboxylase (EC 6.4.1.2) which catalyzes the carboxylation of cytosolic acetyl-CoA to form malonyl-CoA and regulates histone acetylation by regulating the availability of acetyl-CoA; required for de novo biosynthesis of long-chain fatty acids.

Thus, preferably in a host cell of the invention, the S. cerevisiae ACCI gene or an orthologue thereof in another species, is modified to reduce or eliminate activity of its gene product. The S. cerevisiae ACCI gene encodes the amino acid sequence of SEQ ID NO: 10. Therefore a gene to be modified for reducing or eliminating the activity of the ACCI gene product or that of its orthologue in the cell of the invention, preferably is a gene encoding a amino acid sequence with at least 70, 75, 80, 85, 90, 95, 98 or 99% sequence identity to SEQ ID NO: 10.

Alternatively, the activity of the ACCI gene product can be reduced by adding the phospholipid precursors inositol and choline as ACCI is repressed in the presence of the phospholipid precursors, or by reducing the expression of the transcription factors Ino2p and Ino4p or by upregulating the negative ACCI regulator Opilp.

In the cells of the invention, the activity of the ACCI gene product or that of its orthologue is preferably reduced by at least a factor 0.8, 0.5, 0.3, 0.1, 0.05 or 0.01 as compared to cells of a strain which is genetically identical except for the genetic modification causing the reduction in expression, at least under anaerobic conditions. Methods for reducing or eliminating acetyl-CoA carboxylase activity in the cell of the invention are as described above in 2.1 for pyruvate decarboxylase as target protein. 2.6 Reducing fatty acid biosynthesis

In one embodiment of the invention, the expression of at least one of FAS1 and FAS2 is reduced or eliminated in the host cell so as to avoid or reduce (the initial steps of) fatty acid biosynthesis in which acetyl-CoA is consumed.

FAS1 and FAS2 encode the β- and a-subunits, respectively, of the fatty acid synthetase, which catalyzes the synthesis of long-chain saturated fatty acids. The FAS1- encoded β-subunit contains acetyltransacylase, dehydratase, enoyl reductase, malonyl transacylase, and palmitoyl transacylase activities. The ^S^-encoded a-subunit contains the acyl-carrier protein domain and β-ketoacyl reductase, beta-ketoacyl synthase and self-pantetheinylation activities.

Thus, preferably in a host cell of the invention, at least one the S. cerevisiae FAS1 FAS2 genes or their orthologues in another species, are modified in order to avoid or reduce the initial steps of fatty acid biosynthesis in which acetyl-CoA is consumed. The S. cerevisiae FAS1 and FAS2 genes encode the amino acid sequences of SEQ ID NO's: 11 and 12, respectively. Therefore, a gene to be modified for reducing or eliminating the initial steps of fatty acid biosynthesis in which acetyl-CoA is consumed, preferably is a gene encoding a amino acid sequence with at least 70, 75, 80, 85, 90, 95, 98 or 99% sequence identity to at least one of SEQ ID NO's: 11 and 12.

In the cells of the invention, fatty acid synthetase activity preferably is reduced by at least a factor 0.8, 0.5, 0.3, 0.1, 0.05 or 0.01 as compared to cells of a strain which is genetically identical except for the genetic modification causing the reduction in expression, at least under anaerobic conditions.

Methods for reducing or eliminating the fatty acid synthetase activity in the cell of the invention are as described above in 2.1 for pyruvate decarboxylase as target protein.

2.7 Reducing the activity of the glyoxylate cycle

In one embodiment, the activity of the glyoxylate cycle is reduced or eliminated in the host cell. The activity of the glyoxylate cycle is preferably at least reduced in the host cell of the invention because this cycle is another pathway competing (at least under aerobic conditions) for cytosolic acetyl-CoA. Its activity can be reduced or abolished by directly addressing the corresponding genes (CIT2, ICL1, MLS1 and MDH3) or by down-regulating their expression via down-regulation of the genes encoding one or more of the transcription factors HAP2, HAP3, HAP4 and HAP5.

Thus, preferably in a host cell of the invention, at least one the S. cerevisiae CIT2, ICL1, MLS1, MDH3, HAP2, HAP3, HAP4 and HAP5 genes or their orthologues in another species, are modified to reduce or eliminate activity of the glyoxylate cycle in the host cell. The S. cerevisiae CIT2, ICL1, MLS1 and MDH3 genes respectively encode enzymes with citrate synthase, isocitrate lyase, malate synthase and (peroxisomal) NAD-dependent malate dehydrogenase activities, which enzymes have the amino acid sequences of SEQ ID NO's: 13, 14, 15 and 16, respectively. Therefore a gene to be modified for reducing or eliminating activity of the glyoxylate cycle in the host cell, preferably is a gene encoding a amino acid sequence with at least 70, 75, 80, 85, 90, 95, 98 or 99% sequence identity to at least one of SEQ ID NO's: 13, 14, 15, 16, 17, 18, 19 and 20.

In the cells of the invention, the activity of the glyoxylate cycle is preferably reduced by at least a factor 0.8, 0.5, 0.3, 0.1, 0.05 or 0.01 as compared to cells of a strain which is genetically identical except for the genetic modification causing the reduction in expression, at least under anaerobic conditions.

Methods for reducing or eliminating the activity of the glyoxylate cycle in the cell of the invention are as described above in 2.1 for pyruvate decarboxylase as target protein.

2.8 Reducing pyruvate dehydrogenase activity

In one embodiment, the specific pyruvate dehydrogenase activity is reduced or eliminated in the host cell. Endogenous pyruvate dehydrogenase activity is located in the mitochondria in yeast and catalyzes oxidative decarboxylation of pyruvate to form acetyl-CoA. Acetyl-CoA is used in the TCA cycle and in fatty acid biosynthesis.

The pyruvate dehydrogenase enzyme is one enzyme of a multi-enzyme pyruvate dehydrogenase complex. Pyruvate dehydrogenase (EC 1.2.4.1) itself has alpha and beta subunits: encoded by the PDAI and PDBI genes, respectively, forming the El component. The complex includes an E2 core which has dihydrolipoamide acetyltransferase activity (EC 2.3.1.12) and E3 which has dihydrolipoamide dehydrogenase activity (ECl .8.1.4). E2 may be encoded by the LATI and E3 by LPDI genes. An additional complex protein is encoded by the PDXl gene. Thus, the pyruvate dehydrogenase complex may include the enzymes or subunits encoded by the PDAI, PDBI, LATI, LPDI, and PDXl genes. Any of the genes encoding pyruvate dehydrogenase complex enzymes of yeast may be modified to reduce pyruvate dehydrogenase activity in a yeast cell to prepare a strain of one embodiment of the invention.

Thus, preferably in a host cell of the invention, at least one the S. cerevisiae

PDAI, PDBI, LATI, LPDI, and PDXl genes or their orthologues in another species, are modified to reduce or eliminate pyruvate dehydrogenase activity in the host cell. The S. cerevisiae PDAI, PDBI, LATI, LPDI, and PDXl genes have the amino acid sequences of SEQ ID NO's: 21 , 22, 23, 24 and 25, respectively. Therefore a gene to be modified for reducing or eliminating pyruvate dehydrogenase activity in the host cell, preferably is a gene encoding a amino acid sequence with at least 70, 75, 80, 85, 90, 95, 98 or 99% sequence identity to at least one of SEQ ID NO's: 21, 22, 23, 24 and 25.

In the cells of the invention, the pyruvate dehydrogenase activity is preferably reduced by at least a factor 0.8, 0.5, 0.3, 0.1, 0.05 or 0.01 as compared to cells of a strain which is genetically identical except for the genetic modification causing the reduction in expression, at least under anaerobic conditions.

Methods for reducing or eliminating the pyruvate dehydrogenase activity in the cell of the invention are as described above in 2.1 for pyruvate decarboxylase as target protein.

2.9 Reduced transport of pyruvate into mitochondria

In one embodiment, the activity of mitochondrial pyruvate carriers is reduced or eliminated in the host cell in order to reduce or to avoid loss of pyruvate in the citric acid cycle. Preferably therefore, one or more of the S. cerevisiae MPCl, YIA6 and YEA6 genes (or their orthologues in another species) is genetically modified, so as to reduce or eliminate the activity of mitochondrial pyruvate carriers in the cell.

MPCl encodes a highly conserved subunit of the mitochondrial pyruvate carrier; a mitochondrial inner membrane complex comprised of Fmp37p/Mpclp and either Mpc2p or Fmp43p/Mpc3p mediates mitochondrial pyruvate uptake.

YIA6 and YEA6 are members of the mitochondrial carrier subfamily and encode for a mitochondrial NAD ⁺ transporter, involved in the transport of NAD ⁺ and pyruvate into the mitochondria.

Thus, preferably in a host cell of the invention, at least one the S. cerevisiae

MPCl, YIA6 and YEA6 genes or their orthologues in another species, are modified to reduce or eliminate the activity of mitochondrial pyruvate carriers in the cell. The S. cerevisiae MPCl, YIA6 and YEA6 genes have the amino acid sequences of SEQ ID NO's: 26, 27 and 28, respectively. Therefore a gene to be modified for reducing or eliminating the activity of mitochondrial pyruvate carriers in the cell, preferably is a gene encoding a amino acid sequence with at least 70, 75, 80, 85, 90, 95, 98 or 99% sequence identity to at least one of SEQ ID NO's: 26, 27 and 28. In the cells of the invention, the activity of mitochondrial pyruvate carriers is preferably reduced by at least a factor 0.8, 0.5, 0.3, 0.1, 0.05 or 0.01 as compared to cells of a strain which is genetically identical except for the genetic modification causing the reduction in expression, at least under anaerobic conditions.

Methods for reducing or eliminating the activity of mitochondrial pyruvate carriers in the cell of the invention are as described above in 2.1 for pyruvate decarboxylase as target protein.

2.10 Reduction of transport of acetyl-CoA from cytosol into mitochondria

In one embodiment, the host cell of the invention is genetically modified so as to reduce or eliminate transport of acetyl-CoA from cytosol into mitochondria. This is preferably achieved by reducing or eliminating the activity of the carnitine shuttle by genetically modifying at least one of the S. cerevisiae YAT1, YAT2 and CRC1 genes or their orthologue in another species.

YAT1 encodes an outer mitochondrial carnitine acetyltransferase, minor ethanol- inducible enzyme involved in transport of activated acyl groups from the cytoplasm into the mitochondrial matrix.

YAT2 encodes a carnitine acetyltransferase; has similarity to Yatlp, which is a carnitine acetyltransferase associated with the mitochondrial outer membrane.

CRC1 encodes a mitochondrial inner membrane carnitine transporter, required for carnitine-dependent transport of acetyl-CoA from peroxisomes to mitochondria during fatty acid β-oxidation.

Thus, preferably in a host cell of the invention, at least one the S. cerevisiae

YAT1, YAT2 and CRC1 genes or their orthologues in another species, are modified to reduce or eliminate transport of acetyl-CoA from cytosol into mitochondria in the cell via the carnitine shuttle. The S. cerevisiae YAT1, YAT2 and CRC1 genes have the amino acid sequences of SEQ ID NO's: 29, 30 and 31, respectively. Therefore a gene to be modified for reducing or eliminating transport of acetyl-CoA from cytosol into mitochondria in the cell, preferably is a gene encoding a amino acid sequence with at least 70, 75, 80, 85, 90, 95, 98 or 99% sequence identity to at least one of SEQ ID

NO's: 29, 30 and 31.

In the cells of the invention, the transport of acetyl-CoA from cytosol into mitochondria in the cell is preferably reduced by at least a factor 0.8, 0.5, 0.3, 0.1, 0.05 or 0.01 as compared to cells of a strain which is genetically identical except for the genetic modification causing the reduction in expression, at least under anaerobic conditions.

Methods for reducing or eliminating transport of acetyl-CoA from cytosol into mitochondria in the cell of the invention are as described above in 2.1 for pyruvate decarboxylase as target protein.

2.11 Reduced glycerol transmembrane transporter activity

In another embodiment, the formation of glycerol as by-product is prevented or reduced by genetically modifying a plasma membrane channel involved in the efflux of glycerol from the host cell so as to reduce or eliminate its activity. Reduction of glycerol efflux from the cell leads to a decreased production of glycerol by feed-back regulation as glycerol accumulates within the cells, thereby reducing the carbon flux towards glycerol biosynthesis. In S. cerevisiae the FSP1 gene encodes a aquaglyceroporin involved in efflux of glycerol. The FSP1 gene encoded aquaglyceroporin is a plasma membrane (glycerol) channel and a member of major intrinsic protein (MIP) family.

Thus, preferably in a host cell of the invention, the S. cerevisiae FSP1 gene, or its orthologue in another species, is genetically modified to reduce or eliminate the efflux of glycerol from the cell. The S. cerevisiae FSP1 gene has the amino acid sequence of SEQ ID NO: 32, respectively. Therefore a gene to be modified for reducing or eliminating the activity of the plasma membrane channel involved in the efflux of glycerol from the cell, preferably is a gene encoding a amino acid sequence with at least 70, 75, 80, 85, 90, 95, 98 or 99% sequence identity to SEQ ID NO: 32.

In the cells of the invention, the efflux of glycerol from the cell is preferably reduced by at least a factor 0.8, 0.5, 0.3, 0.1, 0.05 or 0.01 as compared to cells of a strain which is genetically identical except for the genetic modification causing the reduction in expression, under aerobic or anaerobic conditions.

Methods for reducing or eliminating the efflux of glycerol from the cell of the invention are as described above in 2.1 for pyruvate decarboxylase as target protein.

2.12 Catabolite derepression In one embodiment the host cell is modified to de-repress catabolite (glucose) repressed genes by modification or inactivation of at least one of the endogenous SNF1 and MIG1 genes, or their orthologues. 3. Conversion of pyruvate into acetyl-CoA

The host cell preferably provides acetyl-CoA to feed into and drive the β- oxidation cycle in the reverse, i.e. biosynthetic direction. The acetyl-CoA is preferably provided in the cytosol of the host cell through metabolic conversion from pyruvate. In microbes, several routes occur for converting pyruvate into acetyl-CoA. The host cells of the invention preferably at least comprise one of the following routes for converting pyruvate into acetyl-CoA.

3.1 Direct conversion of pyruvate to acetyl-CoA by a pyruvate oxidoreductase/dehydrogenase

Yeast has an endogenous pyruvate dehydrogenase. However, this enzyme complex is very slow and provides mitochondrial acetyl-CoA. In the host cells of the invention it is therefore preferred to introduce a heterologous pyruvate oxidoreductase or dehydrogenase in order to provide for acetyl-CoA in the cytosol of the host cell.

A variety of heterologous pyruvate oxidoreductase or dehydrogenase is available for expression in the host cells of the invention, including enzymes which use NAD ⁺ , NADP ⁺ , ferredoxin or flavin as co-factors (EC 1.2.4.1, EC 1.2.1.51, EC 1.2.7.1).

Preferably, the heterologous pyruvate dehydrogenase expressed in the host cell uses the same co-factor as the co-factor that is used by the acyl-CoA dehydrogenase / transenoyl-CoA reductase (see 4.4 herein below).

Suitable heterologous pyruvate dehydrogenases for expression in the host cells of the invention may be obtained from E. coli (aceE, aceF, and lpdA), Zymomonas mobilis (pdhA[alpha], pdhA[beta], pdhB, and lpd), S. aureus (pdhA, pdhB, pdhC, and lpd), Bacillus subtilis, Corymb acterium glutamicum, or Pseudomonas aeruginosa US20100062505 (US20100248233: NADP+ dependent enzyme is oxygen sensitive and is large multimeric enzyme. PDH-complex of Enterococcus faecalis is active under anoxic conditions (US20100062505). The yeast native PDH pyruvate dehydrogenase complex which uses NADH as co factor and which includes the enzymes or subunits encoded by the PDA1, PDB1, LAT1, LPD I, and PDX1 genes could also be used, when overexpressed in the cytosol, by deletion or inactivation of the mitochondrial targeting signals of each of the subunits and using promoters described in 1.2 above.

Preferred pyruvate dehydrogenases are (acetyl-transferring) enzymes which use NADH as co-factor (EC 1.2.4.1). Nucleotide sequences encoding such enzymes can be derived e.g. from Enterococcus faecalis, Arabidopsis thaliana, Bos taurus, Corymb acterium glutamicum, Escherichia coli, Homo sapiens, Zymomonas mobilis, Pseudomonas aeruginosa, Saccharomyces cerevisiae and Bacillus subtilis.

3.2 Conversion of acetaldehyde to acetyl-CoA by an acetylating acetaldehyde dehydrogenase

Another route for converting pyruvate into acetyl-CoA that can be used in the host cells of the invention is the direct conversion of cytosolic acetaldehyde (produced from pyruvate by pyruvate decarboxylase activity in the cytosol) to acetyl-CoA by introduction of a heterologous acetylating (NAD ⁺ -dependent) acetaldehyde dehydrogenase (E.C. 1.2.1.10). An acetylating acetaldehyde dehydrogenases is also referred to as (CoA-acetylating) acetaldehyde :NAD -oxidoreductase or an acetyl-CoA reductase. The conversion of acetaldehyde into acetyl-CoA by the acetylating acetaldehyde dehydrogenase is reversible and runs in the direction of acetyl-CoA when acetaldehyde accumulates in the cytosol. Such an accumulation can e.g. be achieved by reducing or eliminating the activities of alcohol dehydrogenase (ADH1, see 2.3 above) and acetaldehyde dehydrogenase (ALD6, see 2.2 above) in the cytosol of the host cell (see e.g. US20100248233).

The heterologous acetaldehyde dehydrogenases for expression in the host cell of the invention preferably is a mono functional enzyme having only acetaldehyde dehydrogenase activity (i.e. an enzyme only having the ability to oxidize acetaldehyde into acetyl-CoA), such as e.g. the enzyme encoded by the E. coli mhpF gene, as opposed bifunctional enzymes with both acetaldehyde dehydrogenase and alcohol dehydrogenase activities, such as e.g. the bifunctional enzyme encoded by the E. coli adhE gene.

A suitable heterologous gene for expressing a mono functional enzyme having only acetaldehyde dehydrogenase activity is e.g. the E. coli mhpF gene. A suitable exogenous gene coding for a mono functional enzyme with acetaldehyde dehydrogenase activity therefore comprises a nucleotide sequence coding for an amino acid sequence with at least 55, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with SEQ ID NO: 37 (the sequence of the E. coli mhpF gene). Suitable examples of prokaryotes comprising mono functional enzymes with acetaldehyde dehydrogenase activity are provided in Table 1. The amino acid sequences of these mono functional enzymes are available in public databases and can be used by the skilled person to design codon-optimized nucleotide sequences coding for the corresponding mono functional enzyme such as e.g. coding sequence of the E. coli mhpF gene codon optimized for expression in S. cerevisiae as depicted in SEQ ID NO's: 38 or 74). Table 1 : Enzymes with acetaldehyde dehydrogenase activity related to E. coli mhpF

3.3 Direct conversion of pyruvate to acetyl-CoA by a pyruvate formate lyase A further route for converting pyruvate into acetyl-CoA that can be used in the host cells of the invention is the direct conversion of pyruvate into acetyl-CoA by introduction into the host cell of expression of a heterologous pyruvate formate lyase. A pyruvate formate lyase is an enzyme with the ability to convert pyruvate and coenzyme-A into formate and acetyl-CoA catalyses the reaction (EC 2.3.1.54):

pyruvate + coenzyme A (CoA)→ acetyl-CoA + formate Such an enzyme is herein understood as an enzyme having pyruvate formate lyase activity and is referred to as a pyruvate formate lyase (PFL) or formate C- acetyltransferase .

A suitable heterologous gene coding for an enzyme with pyruvate formate lyase activity is e.g. a prokaryotic pyruvate formate lyase, such as the pyruvate formate lyase from E. coli. The E. coli pyruvate formate lyase is a dimer of PflB (encoded by pflB), whose maturation requires the activating enzyme PflAE (encoded by pflA), radical S- adenosylmethionine, and a single electron donor, which in the case of E. coli is flavodoxin (Buis and Broderick, 2005, Arch. Biochem. Biophys. 433:288-296; Sawers and Watson, 1998, Mol. Microbiol. 29:945-954). However, Waks and Silver (supra) have shown that for activation of the pyruvate formate lyase in yeast, only co- expression of an activating enzyme is required but expression of flavodoxin is not necessary.

A suitable heterologous gene coding for an enzyme with pyruvate formate lyase activity preferably comprises a nucleotide sequence coding for an amino acid sequence with at least 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with SEQ ID NO: 39 (the amino acid sequence of the E. coli PflB). Suitable examples of organisms comprising an enzyme with pyruvate formate lyase activity are provided in Table 2. Further examples of such organisms are listed by Lehtio and Goldman (2004, Prot. Engin. Design & Selection, 17:545-552). The amino acid sequences of these enzymes are available in public databases and can be used by the skilled person to design codon-optimized nucleotide sequences coding for the corresponding enzyme with pyruvate formate lyase activity (see e.g. SEQ ID NO: 40 depicting the coding sequence of the E. coli pflB gene optimized for expression in S. cerevisiae).

Table 2: Enzymes with pyruvate formate lyase activity related to E.coli pflB Organism Amino acid

identity (%)

Escherichia coli str. K12 substr. MG1655 100%

Shigella boydii 100%

Escherichia albertii TW07627 99%

Salmonella enterica 97%

Citrobacter rodentium ICC 168 97%

Klebsiella pneumoniae NTUH-K2044 96%

Yersinia aldovae ATCC 35236 91%

Proteus mirabilis HI4320 87%

Haemophilus influenzae Rd KW20 86%

Actinobacillus succinogenes 130Z 83%

Piromyces sp. E2 57%

The host cell of the invention further preferably comprises a heterologous gene coding for the PflA activating enzyme for activation of the pyruvate formate lyase. The pyruvate formate lyase activating enzyme is herein understood as an enzyme that catalyses the reaction:

S-adenosyl-L-methionine + dihydroflavodoxin + [pyruvate formate lyase] -glycine

X

5'-deoxyadenosine + L-methionine + flavodoxin semiquinone +

[formate C-acetyltransferase]-glycin-2-yl radical The heterologous gene coding for the pyruvate formate lyase activating enzyme preferably comprises a nucleotide sequence coding for an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with SEQ ID NO: 41. Suitable examples of organisms comprising an enzyme with pyruvate formate lyase activity are provided in Table 3. The amino acid sequences of these enzymes are available in public databases and can be used by the skilled person to design codon-optimized nucleotide sequences coding for the corresponding enzyme with pyruvate formate lyase activity (see e.g. SEQ ID NO: 42, depicting the coding sequence of the E. coli pflA gene optimized for expression in S. cerevisiae).

Table 3: Pyruvate formate lyase activating enzymes related to E.coli pflA Organism Amino

acid identity

(%)

Escherichia coli str. K12 substr. MG1655 100%

Shigella boydii 100%

Escherichia albertii TW07627 99%

Salmonella enterica 98%

Citrobacter rodentium ICC 168 98%

Klebsiella pneumoniae NTUH-K2044 97%

Yersinia rohdei ATCC 43380 89%

Proteus penneri ATCC 35198 85%

Haemophilus parasuis 29755 70%

In a preferred host cell of the invention, the heterologous genes coding for the enzyme with pyruvate formate lyase activity and the pyruvate formate lyase activating enzyme are from the same donor organism, i.e. be homologous to each other. However, the exogenous genes coding for the enzyme with pyruvate formate lyase activity and the pyruvate formate lyase activating enzyme may also be from different donor organisms, i.e. be heterologous to each other.

A preferred host cell expressing pyruvate formate lyase activity further comprises a genetic modification that increases the specific NAD ⁺ -dependent formate dehydrogenase activity in the cell to dispose of the potentially harmful formate that is produced by the pyruvate formate lyase. A NAD ⁺ -dependent formate dehydrogenase is herein understood as an enzyme that catalyses the reaction: formate + NAD ⁺ → C0 ₂ + NADH + H ⁺ (EC 1.2.1.2).

A preferred gene encoding such a formate dehydrogenase whose activity is to be increased in the cell of the invention expressing heterologous pyruvate formate lyase activity is at least one of the endogenous S. cerevisiae FDH1 and FDH2 genes, or their orthologues in another species. The S. cerevisiae FDH1 is described by van den Berg and Steensma (1997, Yeast 13:551-559) and the S. cerevisiae FDH2 is described by Overkamp et al. (2002, Yeast 19:509-520). Therefore a gene encoding a formate dehydrogenase whose activity is to be increased in the cell of the invention preferably is a gene encoding a formate dehydrogenase having an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98 or 99% sequence identity to at least one of SEQ ID NO's: 43 and 44. Alternatively a heterologous formate dehydrogenase gene may be introduced into the cell in order to increase specific NAD -dependent formate dehydrogenase activity in the cell, such as e.g. the FDH1 gene from Candida boidinii.

In the cells of the invention, the specific formate dehydrogenase activity is preferably increased by at least a factor 1.1, 1.2, 1.5, 2, 5, 10 or 20 as compared to a strain which is genetically identical except for the genetic modification causing the reduction in specific activity, preferably under anaerobic conditions. Formate dehydrogenase activity may be determined as described by Overkamp et al. (2002, supra).

Alternatively, it may be advantageous to accumulate formate, in which case the formate dehydrogenase activity is decreased in the cell of the invention. Formate dehydrogenase activity is decreased in the cell by a genetic modification that reduces or eliminates expression of the above-defined formate dehydrogenase genes, using methods described in 2.1 above.

Another preferred host cell expressing pyruvate formate lyase activity comprises a further genetic modification that reduces or eliminates endogenous pyruvate decarboxylase activity as described in 2.1 above. 3.4 Conversion of pyruvate to acetyl-CoA via the "PDH-bypass" route

Conversion of pyruvate to acetyl-CoA via the "PDH-bypass" route by, optionally, overexpression of pyruvate decarboxylase (PDC1, PDC5 and/or PDC6) and acetaldehyde dehydrogenase (ALD6) and acetyl-CoA synthetase (ACSI or ACS2 or both) and optional reduction ethanol production by reduction or elimination of expression of an alcohol dehydrogenase encoded by at least one of ADH1, -3, -4 and -5 or their orthologues in other species.

In addition, pyruvate decarboxylase, acetyl-CoA synthetase and acetaldehyde dehydrogenase genes from other fungal and bacterial species can be expressed for the modulation of this pathway. A variety of organisms could serve as sources for (nucleotide sequences coding for) these enzymes, including, but not limited to, Saccharomyces sp., including S. cerevisiae mutants and S. uvarum, Kluyveromyces, including K. thermotolerans, K. lactis, and K. mandanus, Pichia, Hansenula, including H. polymorpha, Candidia, Trichosporon, Yamadazyma, including Y. stipitis, Torulaspora pretoriensis, Schizosaccharomyces pombe, Cryptococcus sp., Aspergillus sp., Neurospora sp. or Ustilago sp. Examples of useful pyruvate decarboxylase are those from Saccharomyces bay anus (1PYD), Candida glabrata, K. lactis (KIPDCl), or Aspergillus nidulans (PdcA), and acetyl-CoA synthetase from Candida albicans, Neurospora crassa, A. nidulans, or K. lactis (ACSl), and acetaldehyde dehydrogenase from Aspergillus niger (ALDDH), C albicans, Cryptococcus neoformans (alddh). Sources of prokaryotic enzymes that are useful include, but are not limited to, E. coli, Z. mobilis, Bacillus sp., Clostridium sp., Pseudomonas sp., Lactococcus sp., Enterobacter sp. and Salmonella sp. Further enhancement of this pathway can be obtained through engineering of these enzymes for enhanced activity by site-directed mutagenesis and other evolution methods (which include techniques known to those of skill in the art).

In a preferred embodiment of the invention, the acetyl-CoA synthetase that is expressed in the cell of the invention, is a (heterologous) ADP-forming acetyl-CoA synthetase (EC 6.2.1.13), i.e. an acetyl-CoA synthetase which catalyzes the reaction:

ATP + acetate + CoA <→ ADP + phosphate + acetyl-CoA

Such an ADP-forming acetyl-CoA synthetase only consumes one ATP molecule for each acetyl- CoA synthesized, instead of two ATP molecules as is done by the native yeast enzyme which hydrolyses ATP to AMP and pyrophosphate. The use of an ADP-forming acetyl-CoA synthetase thus avoids loss of energy by the formation of AMP + pyrophosphate. Examples of ADP-forming acetyl-CoA synthetases (EC 6.2.1.13) include those of Salmonella enterica, P. aeruginosa, Pyrococcus furiosus, Archaeoglobus fulgidus and Pyrobaculum islandicum. A preferred ADP-forming acetyl-CoA synthetase for expression in the cell of the invention is a variant of the Salmonella enterica acetyl-CoA synthetase (acssE), which acssE contains a point mutation (L641P) that prevents the enzyme from being inhibited by acetylation (Shiba et al. 2007; see SEQ ID NO. 45). Thus, a preferred ADP-forming acetyl-CoA synthetase that is expressed in the cell is an ADP-forming acetyl-CoA synthetase comprising an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98 or 99% sequence identity to SEQ ID NO: 45. Preferably, ADP-forming acetyl-CoA synthetase further has a mutation that prevents the enzyme from being inhibited by acetylation, such as a point mutation in a position corresponding to position 641 in SEQ ID NO: 45 in an alignment with SEQ ID NO: 45, preferably the point mutation is a mutation to a proline. Preferably, a co don-optimized sequence is used for expression of the ADP-forming acetyl-CoA synthetase, such as SEQ ID NO: 76, a sequence encoding SEQ ID NO: 45 optimized for expression in S. cerevisiae.

In another preferred embodiment of the invention, a heterologous acetaldehyde dehydrogenase is expressed in the cell which uses NAD ⁺ as co factor (instead of NADP ⁺ as the native yeast enzyme Ald6p does). Thus, preferably a NAD ⁺ -dependent aldehyde dehydrogenase is expressed (optionally in combination with the inactivation of the endogenous yeast NADP ⁺ -dependent ALD6). A suitable NAD ⁺ -dependent aldehyde dehydrogenase is e.g. an enzyme encoded by the aid gene of Pseudomonas aeruginosa, or an orthologue thereof. Thus, a preferred NAD ⁺ -dependent aldehyde dehydrogenase that is expressed in the cell is an aldehyde dehydrogenase comprising an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98 or 99% sequence identity to SEQ ID NO: 46. Preferably, a co don-optimized sequence is used for expression of a NAD ⁺ -dependent aldehyde dehydrogenase, such as SEQ ID NO: 75, a sequence encoding the P. aeruginosa aid optimized for expression in S. cerevisiae.

4. Enzymes of the fatty acid β-oxidation cycle

In one embodiment, the host cell of the invention expresses enzymes of the fatty acid β-oxidation cycle. In the context of the present invention, the term "enzymes of the fatty acid β-oxidation cycle" are understood to comprise one or more or all enzymes having at least the following activities:

a) an enzyme that catalyses the synthesis of acetoacetyl-CoA (either directly from two acetyl-CoA molecules or indirectly via a malonyl-CoA molecule and an acetyl- CoA molecule, resp. an acetoacetyl-CoA thiolases and an acetoacetyl-CoA synthase) and/or that catalyses the synthesis of β-ketoacyl-CoA from acetyl-CoA and a higher acyl-CoA (3-ketoacyl-CoA thiolases) as defined in 4.1 and 4.5;

b) an enzyme having hydroxyacyl-CoA dehydrogenase activity, as defined in 4.2; c) an enzyme having enoyl-CoA hydratase or crotonase activity, as defined in 4.3; and

d) an enzyme that catalyses the reduction of a trans-A2-enoyl-CoA into an acyl-

CoA.

In the host, the at least one or all of the enzymes of fatty acid β-oxidation cycle are preferably expressed in the absence of fatty acids and in the presence of a non- fatty acid carbon source. Thus, the at least one or all enzymes do not require the presence of fatty acids in the growth medium for expression and, preferably, the expression of the at least one or all enzymes is insensitive to catabolite (glucose) repression, i.e. the enzymes are expressed in the presence of source of glucose. Further, preferred, the at least one or all enzymes are expressed at least under anoxic (anaerobic) and optionally also under oxic (aerobic) conditions. Examples of suitable promoters for the expressing the at least one or all enzymes are given in 1.2 above herein.

Further preferred is that the enzymes of the fatty acid β-oxidation cycle are expressed in the cytosol of the eukaryotic host cell. In most eukaryotes the enzymes of the fatty acid β-oxidation cycle are present in peroxisomes and/or in mitochondria. In cells of the invention, however, the at least one or all of the enzymes of fatty acid β- oxidation cycle are preferably expressed in the cytosol of the host cell. Therefore, preferably, the β-oxidation enzymes are expressed in the cell without any targeting signal, e.g. for entry into peroxisomes or mitochondria such peroxisomal targeting signals (PTS-1 or -2) or mitochondrial targeting signals. Nucleotide sequences coding for β-oxidation enzymes originating from eukaryotes are therefore preferably modified so as to delete or inactivated the targeting signals on the enzyme. Alternatively prokaryotic β-oxidation enzymes can be used that do not have any (cryptic) targeting signals.

In one embodiment, expression of β-oxidation enzymes in the absence of inducing substrates is effected by de-regulation of transcription factors involved in β- oxidation (ADR1, OAF I, PIP2) thus causing overexpression of yeast's endogenous β- oxidation enzymes (POT1, FOX2, POX1), which preferably have been modified to delete or inactivate their peroxisomal targeting signals (PTS-1 or -2).

4.1 Thiolases

In one embodiment, the invention relates to enzymes that condense acetyl-CoA either with another acetyl-CoA or with a higher acyl-CoA into a β-ketoacyl-CoA. These thiolases are known as acetoacetyl-CoA thiolases (EC 2.3.1.9) or 3-ketoacyl- CoA thiolases (EC 2.3.1.16), respectively. Yeast own thiolase (POT1) has no isoenzymes, therefore the enzyme must be able to metabolize substrates of all chain- length (Mursula 2002). Therefore, the yeast own enzyme could be used or as an alternative a heterologous enzyme could be used in case the yeasts enzyme exhibits poor conversion rate towards β-ketoacyl-CoA (acetoacetyl-CoA in case of C ₄ ). For the synthesis of acetoacetyl-CoA endogenous cytosolic ERG10 could be used as thiolase (as acetoacetyl-CoA thiolase for short chain products). However, the yeast POT I should be expressed without a functional N-terminal peroxisomal targeting sequence (PTS-2, peroxisome-targeting signal sequence) to enable cytosolic expression. Thus, preferably an N-terminally truncated or mutated POT1 is expressed, as e.g. described in Glover et al. (1994, J. Biol. Chem. 269: 7558-7563) (SEQ ID NO: 47 depicts the POT1 encoded S. cerevisiae thiolase lacking the N-terminal 16 amino acid PTS-2).

In mammalian cells ketoacyl-thiolases occur within the mitochondria and within the peroxisomes. They differ with respect to their substrate and stereospecifity. The mitochondria contain two short- and medium chain specific 3-ketoacyl-CoA thiolases. One is specific for acetoacetyl-CoA and 2-methylacetoacetyl-CoA and the other for substrates ranging from C ₆ to C ₁₆ (Eaton et al, 1996, Biochem. J. 320:345-357). In peroxisomes multiple enzyme iso forms exist (some inducible by peroxisome proliferators).

Suitable nucleotide sequences coding for homologous or heterologous thiolases are available from a number of sources, for example, Escherichia coli (GenBank No.'s: NP-416728, NC-000913); Clostridium acetobutylicum (GenBank Nos: NP-349476.1; NC-003030; NP-149242; NC-001988), Bacillus subtilis (GenBank Nos: NP-390297; NC-000964), and Saccharomyces cerevisiae (GenBank Nos: ERG10: NP-015297; POT1 : EEU05467.1), C. acetobutylicum sp. (e.g., protein ID AAC26026.1), C. pasteurianum (e.g., protein ID ABA18857.1), C. beijerinckii sp. (e.g., protein ID EAP59904.1 or EAP59331.1), Clostridium perfringens sp. (e.g., protein ID ABG86544.1, ABG83108.1), Clostridium difficile sp. (e.g., protein ID CAJ67900.1 or ZP-01231975.1), Thermoanaerobacterium thermosaccharolyticum (e.g., protein ID CAB07500.1), Thermoanaerobacter tengcongensis (e.g., AAM23825.1), Carboxydothermus hydrogenoformans (e.g., protein ID ABB13995.1), Desulfotomaculum reducens MI-1 (e.g., protein ID EAR45123.1), Candida tropicalis (e.g., protein ID BAA02716.1 or BAA02715.1), Saccharomyces cerevisiae (e.g., protein ID AAA62378.1 or CAA30788.1), Bacillus sp., Megasphaera elsdenii, and Butryivibrio fibrisolvens. In addition, the endogenous S. cerevisiae thiolases could also be active in a heterologously expressed pathway (ScERGlO, POT1) (see e.g. US20100062505). Further suitable thiolase coding sequences are listed in US2008182308.

A suitable gene coding for an enzyme with short chain specific thiolase activity for expression in a cell according to the invention is preferably derived from S. cerevisiae (ERG 10), E. coli (atoB), Ralstonia eutrophus (PhaA) and orthologues thereof comprising a nucleotide sequence coding for an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with at least one of SEQ ID NO: 48, 49 and 50 respectively.

A suitable gene coding for an enzyme with broad chain-length specificity thiolase activity for expression in a cell according to the invention is preferably derive from S. cerevisiae (POT1) or E. coli (FadA) and comprises a nucleotide sequence coding for an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with SEQ ID NO: 47 or 51, respectively.

Further preferred thiolases are encoded by the PhaA gene from Ralstonia eutrophus (Zhang et al. 2011, Curr. Op. Biotechnol, 22:775-783; Bond- Watts et al, 2011, Nat Chem Biol, 7:222-227) or the E. coli yqeF gene, which has an affinity for short chain substrates (Dellomonaco et al, 2011, supra).

4.2 Hydroxyacyl-CoA dehydrogenases

In one embodiment, the invention relates to enzymes that reduce a β-ketoacyl-

CoA into a trans-P-hydroxyacyl-CoA. Yeast's own hydroxyacyl-CoA dehydrogenase (FOX2) is a multifunctional peroxisomal enzyme catalyzing the D-specific hydroxyacyl-CoA dehydrogenase and D-specific enoyl-CoA hydratase activity. FOX2 catalyzes the same D-specific reactions as the mammalian MFE-2.

The yeast FOX2 encoded hydroxyacyl-CoA dehydrogenase has no isoenzymes, therefore the enzyme must be able to metabolize substrates of a broad range of chain- lengths (Mursula, A., 2002, "A3-A2-Enoyl-CoA Isomerase from the yeast Saccharomyces cerevisiae. Molecular and structural characterization". Academic Dissertation, Faculty of Science, University of Oulu, FI,). Therefore, the yeast endogenous FOX2 encoded enzyme could be used or as an alternative a heterologous enzyme could be used in case the yeasts enzyme exhibits poor conversion rate towards trans-P-hydroxyacyl-CoA. However, preferably the expression of the yeast FOX2 encoded enzyme is relocated from the peroxisome to the cytosol by expression of the enzyme without a functional peroxisome-targeting signal (PTS-1), i.e. the C-terminal tripeptide having the sequence SKL.. Thus, preferably an C-terminally truncated or mutated FOX-2 encoded enzyme, e.g. comprising an amino acid sequence as depicted in SEQ ID NO: 52 (the S. cerevisiae FOX2 lacking the C-terminal 3 amino acid PTS-1), or an orthologue thereof, is expressed in a cell of the invention.

Heterologous hydroxyacyl-CoA dehydrogenases (Hdb) may be reduced NADH- dependent, with a substrate preference for (S)-3-hydroxyacyl-CoA or (R)-3- hydroxyacyyl-CoA and are classified as E.C. 1.1.1.35 and E.C. 1.1.1.30, respectively. Alternatively, 3-hydroxyacyyl-CoA dehydrogenases may be reduced NADPH- dependent, with a substrate preference for (S)-3-hydroxyacyl-CoA or (R)-3- hydroxyacyyl-CoA and are classified as E.C. 1.1.1.157 and E.C. 1.1.1.36, respectively.

Suitable (nucleotide sequences coding for) heterologous 3-hydroxybutyryl-CoA dehydrogenase enzymes are available from a number of sources, for example, Clostridium acetobutylicum (GenBank NOs: NP-349314; NC-003030), B. subtilis (GenBank NOs: AAB09614; U29084), Ralstonia eutropha (GenBank NOs: YP- 294481; NC-007347), and Alcaligenes eutrophus (GenBank NOs: AAA21973; J04987). Suitable genes encoding 3-hydroxybutyryl-CoA dehydrogenases are also listed in US2008182308. Further suitable (nucleotide sequences coding for) are enzymes homologous to the Clostridium acetobutylicum genes include, but are not limited to: Clostridium kluyveri, which expresses two distinct forms of this enzyme (Miller et al, J. Bacteriol. 138:99-104, 1979), and Butyrivibrio fibrisolvens, which contains a bhbd gene which is organized within the same locus of the rest of its butyrate pathway (Asanuma et al, Current Microbiology 51 :91-94, 2005; Asanuma at al, Current Microbiology 47:203-207, 2003). A gene encoding a short chain acyl-CoA dehydrogenase (SCAD) was cloned from Megasphaera elsdenii and expressed in E. coli. In vitro activity could be determined (Becker et al., Biochemistry 32: 10736- 10742, 1993). Other homologues were identified in other Clostridium strains such as C. kluyveri (Hillmer et al, FEBS Lett. 21 :351-354, 1972; Madan et al, Eur. J. Biochem. 32:51-56, 1973), Clostridium tetani E88 (NP-782952.1), Clostridium perfringens SM101 (YP-699558.1), Clostridium perfringens str. 13 (NP-563213.1), Clostridium saccharobutylicum (AAA23208.1), Clostridium beijerinckii NCIMB 8052 (ZP- 00910128.1), Clostridium beijerinckii (AF494018-5), and in Thermoanaerobacter tengcongensis MB4 (NP-622220.1), Thermoanaerobacterium thermosaccharolyticum (CAB04792.1), Alkaliphilus metalliredigenes QYMF (ZP-00802337.1) and E. coli (see e.g. US20100062505).

Thus, a suitable gene for expression in a cell according to the invention and coding for an enzyme with 3-hydroxybutyryl-CoA dehydrogenase activity with higher specificity for short chain substrates preferably comprises a nucleotide sequence coding for an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with at least one of SEQ ID NO: 53 (Hbd; C. acetobutylicum) and 54 (Hbd; C. beijerinckii).

Suitable gene for expression in a cell according to the invention and coding for an enzyme with 3-hydroxybutyryl-CoA dehydrogenase activity with a broader chain- length specificity preferably comprises a nucleotide sequence coding for an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with SEQ ID NO: 55 (fadB; E. coli) or 52 (FOX2; S. cerevisiae).

4.3. Enoyl-CoA hydratases / crotonase

In one embodiment, the invention relates to enzymes that dehydrate a trans-β- hydroxyacyl-CoA into a trans-A2-enoyl-CoA.

As explained above in 4.2, yeast's own enoyl-CoA hydratase (FOX2) is a multifunctional enzyme that has both D-specific hydroxyacyl-CoA dehydrogenase and D-specific enoyl-CoA hydratase activities and it has a broad chain-length specificity. Similarly, also the E.coli FadB gene encodes a bifunctional enzyme having both dehydrogenase and hydratase activities as well as a broad chain-length specificity. Therefore, for the production of longer chain products, i.e. longer than 4, 6, 8 or 10 carbon atoms, preferably, a enoyl-CoA hydratase is expressed in the cell of the invention that is part of a multifunctional enzyme that also has hydroxyacyl-CoA dehydrogenase activity, such as described above in 4.2. Thus a preferred nucleotide sequence coding for a multifunctional enzyme having hydroxyacyl-CoA dehydrogenase and enoyl-CoA hydratase activities and having a broader chain-length specificity is a nucleotide sequence coding for an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with SEQ ID NO: 55 (FadB; E. coli) or 52 (FOX2; S. cerevisiae), whereby the latter is relocated to the cytosol by deleting or inactivating its peroxisomal targeting signal. For the production of shorter chain-length products, preferably a separate heterologous enoyl- CoA hydratase or crotonase is expressed in the cell of the invention together with a separate 3-hydroxybutyryl-CoA dehydrogenase with short chain substrate specificity such as the Hbd enzymes from Clostridium having amino acid sequence identity with at least one of SEQ ID NO: 53 and 54, as defined above in 4.2.

Such heterologous enoyl-CoA hydratases may have a substrate preference for (S)-3-hydroxyacyl-CoA or (R)-3-hydroxyacyl-CoA and are classified as E.C. 4.2.1.17 and E.C. 4.2.1.55, respectively. Suitable (nucleotide sequences coding for) enoyl-CoA hydratases are available from a number of sources, for example, E. coli (GenBank NOs: NP-415911; NC-000913), C acetobutylicum (GenBank NOs: NP-349318; NC- 003030), B. subtilis (GenBank NOs: CAB13705; Z99113), and Aeromonas caviae (GenBank NOs: BAA21816; D88825). Further genes are listed in US2008182308. The crotonases or enoyl-CoA hydratases are enzymes that catalyze the reversible hydration of cis and trans enoyl-CoA substrates to the corresponding β-hydroxyacyl CoA derivatives. In C. acetobutylicum, this step of the butanoate metabolism is catalyzed by EC 4.2.1.55, encoded by the crt gene (GenBank protein accession AAA95967, Kanehisa, Novartis Found Symp. 247:91-101, 2002; discussion 01-3, 19-28, 244-52). Unlike the mammalian crotonases that have a broad substrate specificity, the bacterial enzyme hydrates only crotonyl-CoA and hexenoyl-CoA. The structures of many of the crotonase family of enzymes have been solved (Engel et al, J. Mol. Biol. 275:847-859, 1998). A ClustalW alignment of 20 closest orthologues of crotonase from bacteria shows that the sequence identity varies from 40-85%.

Therefore, a suitable gene coding for a separate heterologous enzyme with crotonase or enoyl-CoA hydratase activity preferably comprises a nucleotide sequence coding for an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with at least one of SEQ ID NO: 56 (C. acetobutylicum crt) and 57 (C. beijerinckii crt). Such genes include, but are not limited to, Clostridium tetani E88 (NP-782956.1), Clostridium perfringens SM101 (YP- 699562.1), Clostridium perfringens str. 13 (NP-563217.1), Clostridium beijerinckii NCIMB 8052 (ZP-00909698.1 or ZP-00910124.1), Syntrophomonas wolfei subsp. wolfei str. Goettingen (YP-754604.1), Desulfotomaculum reducens MI-1 (ZP- 01147473.1 or ZP-01149651.1), Thermoanaerobacterium thermosaccharolyticum (CAB07495.1), and Carboxydothermus hydrogenoformans Z-2901 (YP-360429.1).

Studies in Clostridia demonstrate that the crt gene that codes for crotonase is encoded as part of the larger BCS operon. However, studies on B. fibriosolvens, a butyrate producing bacterium from the rumen, show a slightly different arrangement. While Type I B. fibriosolvens have the thl, crt, hbd, bed, etfA and etfB genes clustered and arranged as part of an operon, Type II strains have a similar cluster but lack the crt gene (Asanuma et al, Curr. Microbiol. 51 :91-94, 2005; Asanuma et al, Curr. Microbiol. 47:203-207, 2003). Other suitable genes are homologous genes from Fusobacterium nucleatum subsp. Vincentii (Q7P3U9-Q7P3U9_FUSNV), Clostridium difficile (P45361-CRT CLODI), Clostridium pasteurianum (P81357-CRT CLOPA), and Brucella melitensis (Q8YDG2-Q8YDG2 BRUME) (see e.g. US20100062505).

4.4. Enzymes that reduce a trans-A2-enoyl-CoA into an acyl-CoA

In one embodiment, the invention relates to enzymes that reduce a trans-A2- enoyl-CoA into an acyl-CoA. These enzymes (either acyl-CoA dehydrogenases or transenoyl-CoA reductases) may be either NADH- or NADPH- or flavoprotein dependent and are classified as E.C. 1.3.1.44, E.C. 1.3.1.38 or EC 1.3.8.1, respectively. Thus, the terms "trans-2-enoyl-CoA reductase" or "TER" are understood herein to refer to enzymes that are capable of catalyzing the conversion of trans-2-enoyl-CoA to acyl- CoA.

Acyl-CoA dehydrogenases or transenoyl-CoA reductases are available from a number of sources, for example, C. acetobutylicum (GenBank NOs: NP-347102); NC- 003030), Euglena gracilis (GenBank NOs: -AAW66853), AY741582), Streptomyces collinus (GenBank NOs: AAA92890; U37135), and Streptomyces coelicolor (GenBank NOs: CAA22721; AL939127). Further suitable genes are listed in e.g. US2008182308.

Such transenoyl-CoA reductases include, but are not limited to, the enzymes from Clostridium tetani E88 (NP-782955.1 or NP-781376.1), Clostridium perfringens str. 13 (NP-563216.1), Clostridium beijerinckii (AF494018-2), Clostridium beijerinckii NCIMB 8052 (ZP-00910125.1 or ZP-00909697.1), and Thermoanaerobacterium thermosaccharolyticum (CAB07496.1), Thermoanaerobacter tengcongensis MB4 (NP- 622217.1). Suitable trans-2-enoyl-CoA reductase (TER) enzymes can also be identified by generally well known bio informatics methods, such as BLAST. Examples of TER proteins include, but are not limited to Euglena spp. including, but not limited to, E. gracilis, Aeromonas spp. including, but not limited, to A. hydrophila, Psychromonas spp. including, but not limited to, P. ingrahamii, Photobacterium spp. including, but not limited, to P. profundum, Vibrio spp. including, but not limited, to V. angustum, V. cholerae, V alginolyticus, V parahaemolyticus, V vulnificus, V fischeri, V splendidus, Shewanella spp. including, but not limited to, S. amazonensis, S. woodyi, S. frigidimarina, S. paeleana, S. baltica, S. denitrificans, Oceanospirillum spp., Xanthomonas spp. including, but not limited to, X oryzae, X campestris, Chromohalobacter spp. including, but not limited, to C. salexigens, Idiomarina spp. including, but not limited, to /. baltica, Pseudoalteromonas spp. including, but not limited to, P. atlantica, Alteromonas spp., Saccharophagus spp. including, but not limited to, S. degradans, S. marine gamma proteobacterium, S. alpha proteobacterium, Pseudomonas spp. including, but not limited to, P. aeruginosa, P. putida, P. fluorescens, Burkholderia spp. including, but not limited to, B. phytofirmans, B. cenocepacia, B. cepacia, B. ambifaria, B. vietnamensis, B. multivorans, B. dolosa, Methylbacillus spp. including, but not limited to, M. flageliatus, Stenotrophomonas spp. including, but not limited to, S. maltophilia, Congregibacter spp. including, but not limited to, C. litoralis, Serratia spp. including, but not limited to, S. proteamaculans, Marinomonas spp., Xytella spp. including, but not limited to, X fastidiosa, Reinekea spp., Colwejfia spp. including, but not limited to, C. psychrerythraea, Yersinia spp. including, but not limited to, Y. pestis, Y. pseudotuberculosis, Methylobacillus spp. including, but not limited to, M flagellatus, Cytophaga spp. including, but not limited to, C. hutchinsonii, Flavobacterium spp. including, but not limited to, F. johnsoniae, Microscilla spp. including, but not limited to, M marina, Polaribacter spp. including, but not limited to, P. irgensii, Clostridium spp. including, but not limited to, C. acetobutylicum, C. beijerinckii, C. cellulolyticum, Coxiella spp. including, but not limited to, C. burnetii. These and other suitable TER enzymes are e.g. described in US2007022497, Hoffmeister et al, J. Biol. Chem., 280:4329-4338, 2005 and US20100062505.

It is understood herein that with respect to redox-balance, preferably TER enzymes are chosen that use the same co-factor as the (heterologous) pyruvate dehydrogenase (see 3.1 above). Suitable NADH-dependent enzymes (EC 1.3.1.44) are e.g. the TER enzymes from Treponema denticola (NP_971211; SEQ ID NO: 58) and Euglena gracilis (AAW66853; SEQ ID NO: 59), of which the TER encode by the gene from T. denticola is most preferred (Zhang et al 201 1 supra, Hu et al, 2013 Biochem J. 449(l):79-89). A ferrodoxin-dependent acyl-CoA dehydrogenase (coupled with NADH oxidation) from Clostridium kluyveri was characterized by Li (Li et al 2008). Thus, a suitable gene coding for an enzyme with transenoyl-CoA reductase activity preferably comprises a nucleotide sequence coding for an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with at least one of SEQ ID NO: 58 ( T. denticola) and 59 (E. gracilis). 4.5 Acetoacetyl-CoA synthases

In one embodiment, instead of or in addition to using the direct condensation of acetyl-CoA by a thiolase (see 4.1. above), an alternative route is used for generation acetoacetyl-CoA through the ATP-driven malonyl-CoA synthesis as described by Lan and Liao (2012, Proc. Natl. Acad. Sci. 109:6018-6023). In this embodiment, along with the acetyl-CoA pool, ATP is used to drive the thermodynamically unfavorable condensation of two acetyl-coA molecules to acetoacetyl-CoA. Thus, in this embodiment, the host cell expresses enzymes that catalyze ATP-driven malonyl-CoA synthesis and decarboxylative carbon chain elongation to drive the carbon flux into the formation of acetoacetyl-CoA, which then further undergoes the reverse β-oxidation to synthesize butyryl-CoA and/or subsequent higher acyl-CoA's.

The host cell therefore expresses or overexpresses an acetyl-CoA carboxylase (Acc), which synthesizes malonyl-CoA from acetyl-CoA, HCO ₃ and ATP. A suitable acetyl-CoA carboxylase for expression in the cell of the invention is e.g. (overexpression of) the Saccharomyces cerevisiae ACC1 gene (DAA10557; SEQ ID NO 10) or the Homo sapiens ACC1 (Q13085; SEQ ID NO 60. Thus, a preferred gene coding for an enzyme with acetyl-CoA carboxylase activity for expression in a cell of the invention, preferably comprises a nucleotide sequence coding for an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with at least one of SEQ ID NO's: 10 and 60. It is therefore understood that in this embodiment, preferably, the endogenous acetyl-CoA carboxylase in the host cell is not reduced or eliminated (see 2.5 above).

In this embodiment the host further expresses an enzyme with acetoacetyl-CoA synthase activity, which catalyses the decarboxylative condensation of acetyl-CoA and malonyl-CoA to form acetoacetyl-CoA. A suitable acetoacetyl-CoA synthase is e.g. the Streptomyces sp. CL190 nphT7 gene encoded acetyl-CoA:malonyl-CoA acyltransferase (AB540131 ; SEQ ID NO 61). Thus a preferred gene coding for an enzyme with acetoacetyl-CoA synthase activity for expression in a cell of the invention, preferably comprises a nucleotide sequence coding for an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with SEQ ID NO: 61.

The skilled person will understand that for the production of C ₄ -products this alternative route through the ATP-driven malonyl-CoA synthesis can replace the short chain thiolases of 4.1 above. In fact, in order to prevent a futile cycle it is preferred in this embodiment that the cell does not express a (heterologous) short-chain specific thiolase in the cytosol and/or that the activity of the endogenous cytosolic S. cerevisiae ERG10 short-chain thiolase (or its orthologue in another species) is reduced or eliminated. However, for the production of products with a carbon chain-length longer than 4, preferably, in addition also a thiolase with broad chain-length specificity, more preferably a thiolase with long chain-length specificity (i.e. C ₆ and longer), is expressed in accordance with 4.1 above,.

5. Termination enzymes

In further embodiments, the invention relates to termination enzymes, which are expressed in the cells of the invention to branch products off from the reverse β- oxidation cycle for the production of specific products, such as fatty acids, 1 -alcohols, β-ketoacids, β-ketoalcohols, β-hydroxy acids, 1,3-diols, trans-A ² -fatty acids, alkenes, alkanes and derivatives thereof, whereby the chain-length specificity of the termination enzyme can determine the chain-length of the final product. Preferred termination enzymes are enzymes for the production of 1 -alcohols, preferably 1 -alcohols other than methanol and ethanol, such as e.g. butanol, decanol, dodecanol and higher 1 -alcohols.

5.1 Enzymes that convert a β-ketoacyl-CoA into a desired product

In further embodiments, the invention relates to enzymes that convert a β- ketoacyl-CoA into a desired product.

5.1.1 A β-ketoacyl-CoA into a β-ketoalcohol In one embodiment, the invention relates to enzymes that catalyzes the formation of β-ketoalcohols from β-ketoacyl-CoA such as e.g. an alcohol- forming CoA thioester reductases, an aldehyde- forming CoA thioester reductases or an alcohol dehydrogenases.

5.1.2 A β-ketoacyl-CoA into a β-ketoacid

In one embodiment, the invention relates to enzymes that catalyzes the formation of β-ketoacids from β-ketoacyl-CoA such as e.g. thioesterases or CoA thioester hydrolases. A suitable enzyme for this purpose is e.g. the E. coli acyl-CoA thioesterase II tesB, which forms β-ketobutyric acid from β-ketobutyl-CoA.

5.2 Enzymes that convert a trans- β-hydroxyacyl- CoA into a desired product

In one embodiment, the invention relates to enzymes that convert a trans- - hydroxyacyl-CoA into a desired product.

5.2.1 A trans^-hydroxyacyl-CoA into a 1 ,3-diol

In one embodiment, the invention relates to enzymes that catalyzes the formation of a 1,3-diol from trans^-hydroxyacyl-CoA, such as e.g. an alcohol- forming CoA thioester reductases, aldehyde- forming CoA thioester reductase or an alcohol dehydrogenases. Preferably, for the production of a 1,3-diol such as e.g. 1,3 butanediol, the cell expresses an encoding NAD(P)H-dependent acetoacetyl-CoA reductase and at least one of a butyraldehyde dehydrogenase and an alcohol/aldehyde dehydrogenase

(see e.g. Kataoka et al, 2012, J. Biosci. Bioeng. http://dx.doi.Org/10.1016/i.ibiosc.2012.l l .025).

A preferred NAD(P)H-dependent acetoacetyl-CoA reductase is the PhaB acetoacetyl-CoA reductase from Ralstonia eutropha or an orthologue thereof such as an acetoacetyl-CoA reductase having an amino acid sequence with at least 45, 50, 60, 65,

70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with SEQ ID NO: 77

(PhaB of Ralstonia eutropha).

A preferred butyraldehyde dehydrogenase is the bid butyraldehyde dehydrogenase from Clostridium saccharoperbutylacetonicum ATCC 27012 or an orthologue thereof such as a butyraldehyde dehydrogenase having an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with SEQ ID NO: 78 (bid of C. saccharoperbutylacetonicum). A preferred alcohol/aldehyde dehydrogenase is the adhE NADH-dependent aldehyde/alcohol dehydrogenase from Clostridium acetobutylicum or an orthologue thereof, such as a dehydrogenase having an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with SEQ ID NO: 63 (adhE2 of C. acetobutylicum).

5.2.2 A trqfts-P-hydroxyacyl-CoA into a β-hydroxyacid

In one embodiment, the invention relates to enzymes that catalyzes the formation of a β-hydroxyacid from trans-P-hydroxyacyl-CoA, such as e.g. an acyl-CoA thioesterase. A suitable enzyme for production longer chain products is are e.g. the acyl-CoA thioesterase (EC 3.1.2.27) derived from Mus musculus (PTE-2, NM_133240) and expressed without a functional PTS-1 tripeptide, or an orthologue thereof. This enzymes converts 3-hydroxyhexadecanoyl-CoA into CoA and 3- hydroxyhexadecanoate. A suitable enzyme for production of short chain products is e.g. the acyl-CoA thioesterase I (tesA) from E. coli or orthologues thereof, e.g. for the synthesis of β-hydroxybutyric acid.

Thus, a suitable gene coding for an enzyme with acyl-CoA thioesterase activity preferably comprises a nucleotide sequence coding for an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with SEQ ID NO: 79 (Mus musculus PTE-2 without C-terminal PTS-1) or SEQ ID NO: 80 (E. coli tesA).

5.3 Enzymes that convert an acyl-CoA into a desired product

In one embodiment, the invention relates to enzymes that convert an acyl-CoA into a desired product.

5.3.1 An acyl-CoA into an aldehyde and/or alcohol

In one embodiment, the invention relates to enzymes that catalyze the conversion of acyl-CoA to an aldehyde and/or an alcohol. Preferably the enzyme catalyses the conversion into aldehydes or alcohols with (linear) chains longer than C ₂ , preferably at least C ₄ .

A preferred enzyme that catalyzes the conversion of acyl-CoA to an aldehyde is an enzyme that uses NADH or NADPH as co factor (EC 1.2.1.10 or EC 1.2.1.57). It is understood herein that in order to maintain a proper redox-balance either one of an NADH- or NADPH-dependent aldehyde dehydrogenases may be chosen in the host cell of the invention, depending on the co-factor dependencies of other enzymes expressed in the cell. A suitable aldehyde dehydrogenases is e.g. encoded by the E. coli mhpF gene (GenBank No CAA70751; SEQ ID NO: 37 encoded by the codon optimized SEQ ID NO's: 38 or 74) and orthologues thereof as defined in 3.2. Further suitable aldehyde dehydrogenases may be derived from Clostridium beijerinckii (GenBank NOs: AAD31841), AF157306 and C acetobutylicum (GenBank NOs: NP- 149325), NC-001988). Further suitable genes are listed in US2008182308.

A preferred enzymes that catalyze the conversion of an aldehyde to a 1 -alcohol, uses either NADH or NADPH as co factor. Again, it is understood herein that in order to maintain a proper redox-balance either one of an NADH- or NADPH-dependent alcohol dehydrogenases may be chosen in the host cell of the invention, depending on the co-factor dependencies of other enzymes expressed in the cell. Alcohol dehydrogenases are available from, for example, E. coli (fucO) (GenBank No YP_002927733; SEQ ID NO. 62), C acetobutylicum (GenBank NOs: NP-149325, NC- 001988; which enzyme possesses both aldehyde and alcohol dehydrogenase activity); NP-349891, NC-003030; and NP-349892, NC-003030 and E. coli (GenBank NOs: NP- 417-484, NC-000913). Depending on the culture conditions, the AdhE2 of C. acetobutylicum may be less preferred because it is oxygen sensitive. Alternatively, at least one the endogenous yeast alcohol dehydrogenases ADH1 and/or ADH2 used, e.g. by overexpression of the gene. Further suitable genes encoding alcohol dehydrogenases are listed in US2008182308.

Thus, a suitable gene coding for an enzyme with NADH-dependent alcohol dehydrogenases activity preferably comprises a nucleotide sequence coding for an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with SEQ ID NO: 62 (E. coli fucO) or 63 (C. acetobutylicum adhE2).

Further homologues to the C acetobutylicum aldehyde/alcohol dehydrogenase include without limitation Clostridium tetani E88 (NP-781989.1), Clostridium perfringens str. 13 (NP-563447.1), Clostridium perfringens ATCC 13124 (YP- 697219.1), Clostridium perfringens SM101 (YP-699787.1), Clostridium beijerinckii NCIMB 8052 (ZP-00910108.1), Clostridium acetobutylicum ATCC 824 (NP- 149199.1), Clostridium difficile 630 (CAJ69859.1), Clostridium difficile QCD-32g58 (ZP-01229976.1), and Clostridium thermocellum ATCC 27405 (ZP-00504828.1).

Two suitable n-butanol dehydrogenases from C. acetobutylicum (BDH I, BDH II) have been purified, and their genes (bdbA, bdhB) cloned. The GenBank accession for BDH I is AAA23206.1. The GenBank accession for BDH II is AAA23207.1. BDH I utilizes NADPH as the cofactor, while BDH II utilizes NADH, which is more preferred. Even more preferred at least for the production of n-butanol is adhE2 gene of C. acetobutylicum (GenBank accession #AF321779; SEQ ID NO: 63). Thus, a preferred gene coding for an enzyme for the production of n-butanol comprises a nucleotide sequence coding for an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with SEQ ID NO: 63.

Such homologues include, but are not limited to, Clostridium perfringens SM101 (YP-699787.1), Clostridium perfringens str. 13 (NP-563447.1), Clostridium perfringens ATCC 13124 (YP-697219.1), Clostridium tetani E88 (NP-781989.1), Clostridium beijerinckii NCIMB 8052 (ZP-00910108.1), Clostridium difficile QCD- 32g58 (ZP-01229976.1), Clostridium difficile 630 (CAJ69859.1), Clostridium acetobutylicum ATCC 824 (NP- 149325.1), and Clostridium thermocellum ATCC 27405 (ZP-00504828.1) (see e.g. US20100062505).

Other useful termination enzymes for the production of e.g. butanol include NADH-dependent or NADPH-dependent butanol dehydrogenases (E.C. 1.1.1.1), which convert butylaldehyde to 1 -butanol, or bifunctional NADH-dependent or NADPH- dependent aldehyde/alcohol dehydrogenases (E.C. 1.1.1.1./1.2.1.10), which converts butyryl-CoA to 1 -butanol via butyraldehyde (see e.g. US20100248233). Further genes coding for aldehyde- forming acyl-CoA reductases and aldehyde/alcohol dehydrogenases and that are useful as termination enzymes for the production of a wide variety of aldehydes and/or alcohols are listed in Table 7 of WO2012/109176.

For the biosynthesis of longer chain alcohols fatty acyl-CoA reductases (FAR) can be use as termination enzyme. Fatty acyl-CoA reductases (FAR) can be divided into two classes that differ with respect to the end-product synthesized, i.e. the aldehyde- and the alcohol-forming enzymes. Aldehyde-generating FAR catalyze a two- electron reduction of activated fatty acids, so that fatty aldehydes are formed. Such enzymes have been described in pea leaves, green algae and bacteria. The fatty aldehydes can be further reduced to fatty alcohols by fatty aldehyde reductases or can be involved in the biosynthesis of hydrocarbons. On the other hand, alcohol- forming FAR catalyze the reduction of activated fatty acids to fatty alcohols. This four-electron reduction takes place in two steps. In the first step an aldehyde is formed, that is subsequently reduced to a fatty alcohol in the second step. Proteins have been purified and genes encoding alcohol- forming FAR were identified in plants, mammals, insects, birds and protozoa. They usually require NADPH as electron donor but in certain organisms like Euglena gracilis NADPH can be substituted by NADH.

The FAR in Euglena gracilis (EgFAR accession no. ADI60057; SEQ ID NO: 64) uses 14:0, 16:0 and 18:0 as substrates and requires NADH as cofactor. Functional analysis of EgFAR in yeast indicated that it could effectively convert 14:0 and 16:0 fatty acids to their corresponding alcohols. Compared with other biochemically characterized FARs, EgFAR possesses a narrower substrate range, only using saturated fatty acids with 14 and 16 carbon chains as substrates with the preferred fatty acid being 14:0 when expressed in yeast. (Teerawanichpan and Qiu, 2010, Lipids, 45(3):263-73). Thus, a preferred gene coding for a termination enzyme for the production of C ₁₄ and/or C ₁₆ alcohols comprises a nucleotide sequence coding for an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with SEQ ID NO: 64.

Further suitable FAR genes for expression in the cells of the invention for the production of longer chains alcohols and/or aldehydes are described by Hellenbrand et al. (201 1 , BMC Biochem. 12:64) and include e.g. the bird fatty acyl-CoA reductases: AdFARl JN638548, AmFARl NP 001 180219.1 , AtCER4 NP 567936.5, AtFARl NP 197642.1 , AtFAR4 NP 190040.3, AtFAR5 NP 190041.2, AtFAR6 AEE79553.1 , AtFAR8 NP l 90042.2, AtMS2 AEE75132.1 , BmFAR NP 001036967.1 , EgFAR ADI60057.1 , FAR1 NP 001026350.1 , FAR2 XP 417235.2, HsFARl NP l 15604.1 , HsFAR2 NP 060569.3, MmFARl NP 080419.2, MmFAR2 NP 848912.1 , OnFARa ACY07547.1 , OnFARb ACY07546.1 , OsFARVIII ACJ06520.1 , ScFAR AAD38039.1 , TaFARl JN638549, TaFAR2 JN638550, TaTAAla CAD30693.1 , YeFARI ADD62438.1 , YeFARII ADD62439.1 , YeFARIII ADD62440.1 , YpFARII, ADD62442.1 and YrFARII ADD62441.1.

A further suitable termination enzyme for producing longer chain alcohols, i.e. C ₆ , Cs, or Cio or longer, is an iron-dependent, NAD ⁺ -dependent (long chain) aldehyde/alcohol dehydrogenase such as e.g. encoded by the E. coli yiaY gene. Thus, a preferred gene coding for a termination enzyme for the production of longer chains alcohols comprises a nucleotide sequence coding for an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with SEQ ID NO: 65.

5.3.2 An acyl-CoA into a fatty acid

In one embodiment, the invention relates to the biosynthesis of fatty acids. Preferred enzymes to be expressed in the cells of the invention for the biosynthesis of fatty acids are enzymes that catalyze the deacylation (hydrolysis) of acyl-CoA thioesters to fatty acids and CoA (EC 3.1.2.2) e.g. an acyl-CoA thioesterase from Anas platyrhynchos, Arabidopsis thaliana, Bumilleriopsis filiformis, Candida rugosa, Cricetulus griseus, Eremosphaera viridis, Escherichia coli, Euglena gracilis, Homo sapiens, Mougeotia scalaris, Mus musculus, Mycobacterium phlei, Mycobacterium smegmatis, Oryctolagus cuniculus, Pseudomonas aeruginosa, Rattus norvegicus, Rhodotorula aurantiaca, Rhodobacter sphaeroides, Saccharomyces cerevisiae, Streptomyces coelicolor, Streptomyces venezuelae, and Sus scrofa. A preferred enzyme is an enzyme that catalyzes the deacylation of long chain acyl-CoA thioesters to long chain fatty acids, i.e. fatty acid with a chain-length of C ₆ , C ₈ , or C ₁₀ or longer. Thus, a preferred gene coding for a termination enzyme for the production of longer chains fatty acids comprises a nucleotide sequence coding for an acyl-CoA thioesterase comprising an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with at least one of SEQ ID NO: 81 (hBACH; D88894) and SEQ ID NO: 82 (E. coli fadM; NP_414977).

5.3.3 An acyl-CoA into a 1-alkene

In one embodiment, the invention relates to the biosynthesis of alkenes. A preferred enzyme to be expressed in the cells of the invention for the biosynthesis of alkenes from acyl-CoA is an olefin synthase. A suitable olefin synthase is e.g. the polyketide synthase from the cyanobacterium Synechococcus sp. strain PCC 7002 (SYNPCC7002 A1173), or an orthologue thereof, such as e.g. available from Cyanothece sp. PCC 7424, Cyanothece sp. PCC 7822, Prochloron didemni Pl-Palau, Pseudomonas entomophila L48, and Haliangium ochraceum DSM 14365 (see also WO2012/050931). Thus preferably for the biosynthesis of alkenes, the cell of the invention expresses a codon-optimized nucleotide sequence coding for an olefin/polyketide synthase comprising an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with SEQ ID NO: 83.

Alternatively, the polyketide synthase is expressed in the cell of the invention in the form of several (e.g. two, three or four) individual polypeptide chains, each comprising one or more (preferably 2 or 3) individual modules of the synthase. E.g., a first polypeptide may comprise the acyl-loading domain (LD) and the acyl-carrier protein domain (ACP-1), a second polypeptide may comprise the ketosynthase (KS), acyltransferase (AT) and ketoreductase (K ) domains and a third polypeptide may comprise the acyl-carrier protein domain (ACP-2), the sulfotransferase (ST) domain and the thioesterase (TE) domain (as e.g. described in Mendez-Perez et al., 2012, Metab Eng. 14(4):298-305). Preferably each of the polypeptides comprising the synthase modules comprises a (different) affinity-tag and the cell expresses a scaffold protein that can accommodate each of the affinity-tags so as to bring together the synthase modules in the correct orientation. Thus preferably, the cell of the invention expresses three co don-optimized nucleotide sequence coding for three polypeptides which together comprise the modules of an olefin/polyketide synthase, the first polypeptide comprising an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with amino acids 1 - 675 of SEQ ID NO: 84, the second polypeptide comprising an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with amino acids 1 - 1254 of SEQ ID NO: 85, and the third polypeptide comprising an amino acid sequence with at least 45, 50, 60, 65, 70, 75, 80, 85, 90, 95, 98, 99% amino acid sequence identity with amino acids 1 - 724 of SEQ ID NO: 86. Suitable affinity tags that may be used include e.g. GBD (Kim A.S. et al, 2000, Nature. 404, (6774): 151- 158), SH3 (Nguyen, J.T. et al, 1998, Science 282 (5396): 2088-2092) and PDZ (Harris, B.Z. et al, 2001, Biochemistry 40 (20):5921-5930; Dueber, J.E. et al, 2003, Science 301 (5641): 1904-1908) ligands, which may be used in combination with a scaffold protein carrying the GBD (Kim A.S. et al, 2000, supra), the SH3 (Wu, X. et al, 1995, Structure. 3, (2):215-226) and PDZ (Schultz, J., et al, 1998, Nat Struct Biol. 5 (1): 19-24) binding domains, and preferably, having the amino acid sequence of SEQ ID NO: 87.

5.3.4 An acyl-CoA into an alkane In one embodiment, the invention relates to the biosynthesis of alkanes. For the biosynthesis of alkanes a two-step pathway is preferably used, which involves the reduction of acyl-CoA to fatty aldehydes by the action of fatty aldehyde- forming acyl- CoA reductases followed by the decarbonylation of the resulting aldehyde to alkane by aldehyde decarbonylases. Suitable heterologous genes encoding fatty aldehyde- forming acyl-CoA reductases are available from e.g. Acinetobacter calcoaceticus (acrl) and Acinetobacter sp. strain M-l (acrM) and can be used for expression in the cells of the invention (Ishige et al, 2002, Appl. Environ. Microbiol, 68: 1192-1195). Both enzymes are active with a range of acyl-CoAs. For the second step of the pathway a cell of the invention preferably expresses a heterologous aldehyde decarbonylase such as e.g. the aldehyde decarbonylase from Synechococcus elongatus PCC7942 (PCC7942;_orfl593) and other orthologues recently reported by Schirmer (2010, Science, 329:559- 562).

A comparable pathway for the biosynthesis of alkanes has been described by Bernard et al. (2012, Plant Cell, 24:3106-18). In this article the alkane biosynthetic route of Arabidopsis thaliana is described. In this pathway a heterodimer of two enzymes, CERl and CER3, catalyzes a two-step reaction starting with the reduction of acyl-CoA to an aldehyde intermediate followed by a decarbonylation step, in which a CytB5 provides electrons for the decarbonylation. Thus, in a preferred embodiment the Arabidobsis CERl, CER3 CYTB5, or orthologues thereof, are expressed in a cell of the invention for the synthesis of (very long chain) alkanes from (very long chain) acyl- CoAs in accordance with Bernard et al. (2012, supra).

Another possibility of alkane synthesis is the introduction of enzymes derived from insects in which aldehydes are converted to alkanes by a decarbonylation mechanism which results in the release of a CO group catalyzed by a NADPH- cytochrome P ₄₅ o reductase (Qiu et al, 2012, Proc Natl Acad Sci USA,109: 14858-63).

6. Modified host cells of the invention

In a first aspect the invention thus pertains to a modified eukaryotic microbial host cell. The host cell preferably is modified to comprise: a) cytosolic expression of the enzymes of the fatty acid β-oxidation cycle, preferably, in the absence of fatty acids and in the presence of a non-fatty acid carbon source; b) a metabolic route for producing under oxic conditions and preferably in the cytosol, acetyl-CoA from the non-fatty acid carbon source to feed into and drive the β-oxidation cycle in the biosynthetic direction; c) expression of a termination enzyme to convert reaction intermediates of the β-oxidation cycle into at least one fermentation product selected from the group consisting of a fatty acid, a 1 -alcohol, a β-ketoacid, a β-ketoalcohol, a β-hydroxyacid, a 1,3-diol, a trans-Δ ² - fatty acid, an alkene, an alkane and derivatives thereof, more preferably the fermentation product is selected from the group consisting of a fatty acid, a 1 -alcohol, an alkene, an alkane and derivatives thereof; and, d) at least one of: i) expression of an exogenous xylose isomerase gene, which gene confers to the cell the ability to isomerize xylose into xylulose; and, ii) expression of exogenous genes coding for a L-arabinose isomerase, a L-ribulokinase and a L-ribulose-5- phosphate 4-epimerase, which genes together confer to the cell the ability to convert L- arabinose into D-xylulose 5-phosphate.

In a preferred modified cell according to the invention, the acetyl-CoA is produced in the cytosol from pyruvate by at least one of: the following routes A, B, C and D.

A) Conversion of pyruvate to acetyl-CoA by decarboxylation of pyruvate to acetaldehyde by expression of pyruvate decarboxylase activity, conversion of acetaldehyde to acetate by expression of acetaldehyde dehydrogenase activity, preferably, a heterologous NAD ⁺ -dependent aldehyde dehydrogenase, and conversion of acetate to acetyl-CoA by expression of acetyl-CoA synthetase activity, preferably the acetyl-CoA synthetase is an heterologous ADP-forming acetyl-CoA synthetase (EC 6.2.1.13), whereby the host comprises a genetic modification that reduces or eliminates endogenous cytosolic alcohol dehydrogenase activity. Preferably, in A) the cell expresses a heterologous NAD ⁺ -dependent aldehyde dehydrogenase comprising an amino acid sequence with at least 45% amino acid sequence identity to SEQ ID NO: 46, the cell expresses a heterologous ADP-forming acetyl-CoA synthetase comprising an amino acid sequence with at least 45% amino acid sequence identity to SEQ ID NO: 45, the cell, optionally, comprises a genetic modification that increases cytosolic pyruvate decarboxylase activity, and whereby the cell comprises a genetic modification that reduces or eliminates expression of endogenous ADH1 gene or an orthologue thereof.

B) Conversion of pyruvate to acetyl-CoA by decarboxylation of pyruvate to acetaldehyde by expression of pyruvate decarboxylase activity and by direct conversion of acetaldehyde to acetyl-CoA by expression of acetylating acetaldehyde dehydrogenase (E.C. 1.2.1.10) activity, whereby the host comprises a genetic modification that reduces or eliminates endogenous cytosolic acetaldehyde dehydrogenase activity and a genetic modification that reduces or eliminates endogenous cytosolic alcohol dehydrogenase activity. Preferably, in b) the cell expresses a heterologous acetylating NAD ⁺ -dependent acetaldehyde dehydrogenase comprising an amino acid sequence with at least 55% amino acid sequence identity to SEQ ID NO: 37, and whereby the cell comprises a genetic modification that reduces or eliminates expression of endogenous ADH1 gene or an orthologue thereof and a genetic modification that reduces or eliminates expression of endogenous ALD6 gene or an orthologue thereof.

C) Direct conversion of pyruvate to acetyl-CoA and formate by expression of pyruvate formate lyase (EC 2.3.1.54) activity, whereby the host comprises a genetic modification that reduces or eliminates endogenous pyruvate decarboxylase activity and whereby the cell comprises formate dehydrogenase activity. Preferably, in C) the cell expresses a heterologous pyruvate formate lyase comprising an amino acid sequence with at least 50%> amino acid sequence identity to SEQ ID NO: 39, the cell expresses a heterologous pyruvate formate lyase activating enzyme comprising an amino acid sequence with at least 45% amino acid sequence identity to SEQ ID NO: 41, and whereby the cell comprises a genetic modification that reduces or eliminates expression of at least the endogenous PDCl and PDC5 genes or orthologues thereof, and the cell, optionally, comprises a genetic modification that increases formate dehydrogenase activity.

D) Direct conversion of pyruvate to acetyl-CoA by expression of pyruvate dehydrogenase (EC 1.2.4.1) activity, whereby the host comprises a genetic modification that reduces or eliminates endogenous pyruvate decarboxylase activity. Preferably, in d) the cell expresses a heterologous NAD ⁺ -dependent acetyl-transferring pyruvate dehydrogenase and whereby the cell comprises a genetic modification that reduces or eliminates expression of at least the endogenous PDCl and PDC5 genes or orthologues thereof.

A modified cell according to the invention further preferably comprises one or more genetic modifications selected from the group consisting of: a) a genetic modification that reduces or eliminates fatty acid synthesis; b) a genetic modification that reduces or eliminates activity of the glyoxylate cycle; c) a genetic modification that reduces or eliminates activity of the tricarboxylic acid cycle; d) a genetic modification that reduces or eliminates transport of pyruvate into mitochondria; e) a genetic modification that reduces or eliminates transport of acetyl-CoA into mitochondria; f) a genetic modification that reduces or eliminates transport of glycerol; g) a genetic modification that increases specific xylulose kinase activity; h) a genetic modification that increases specific activity of one or more of ribulose-5 -phosphate isomerase, ribulose-5 -phosphate 3-epimerase, transketolase and transaldolase; and, i) a genetic modification that reduces or eliminates unspecific aldose reductase activity. Preferably, a) the genetic modification is a modification that reduces or eliminates the expression of one or more of the S. cerevisiae FAS1, FAS2, ACC1, IN02, IN04 genes or orthologues thereof; b) the genetic modification is a modification that reduces or eliminates the expression of one or more of CIT2, ICL1, MLS1, MDH3, HAP2, HAP 3, HAP4 and HAP5 genes or orthologues thereof; c) the genetic modification is a modification that reduces or eliminates the expression of one or more of the S. cerevisiae PDA1, PDB1, LAT1, LPD1, and PDX1 genes or orthologues thereof; d) the genetic modification is a modification that reduces or eliminates the expression of one or more of the S. cerevisiae MPC1, YIA6 and YEA6 genes or orthologues thereof; e) the genetic modification is a modification that reduces or eliminates the expression of one or more of the S. cerevisiae YAT1, YAT2 and CRC1 genes or orthologues thereof; f) the genetic modification is a modification that reduces or eliminates the expression of the S. cerevisiae FPS1 gene or an orthologues thereof; g) the genetic modification is overexpression of the S. cerevisiae XKS1 gene or an orthologue thereof; h) the genetic modification is overexpression of one or more of the S. cerevisiae RPI1, RPE1, TKL1 and TALI genes or orthologues thereof; and, i) the genetic modification is a modification that reduces or eliminates the expression of the S. cerevisiae GRE3 gene or an orthologue thereof.

In the modified cells according to the invention, the enzymes of the fatty acid β- oxidation cycle are expressed preferably in the cytosol, and preferably from promoters that are insensitive to catabolite repression. In one preferred embodiment, the modified cell is a cell for producing fermentation products with a short chain-length, such as e.g. no more than C ₈ , C ₆ or C ₄ . In this embodiment the enzymes of the fatty acid β- oxidation cycle include: a) a thiolase with a short chain-length specificity, whereby, preferably, the thiolase comprises an amino acid sequence with at least 45% amino acid sequence identity with at least one of SEQ ID NO's: 48, 49 and 50; b) an enzyme with 3-hydroxybutyryl-CoA dehydrogenase activity, whereby, preferably, the 3- hydroxybutyryl-CoA dehydrogenase comprises an amino acid sequence with at least 45% amino acid sequence identity with at least one of SEQ ID NO's: 53 and 54; c) a crotonase, whereby, preferably, the crotonase comprises an amino acid sequence with at least 45% amino acid sequence identity with at least one of SEQ ID NO's: 56 and 57; and, d) a NADH-dependent trans-2-enoyl-CoA reductase, whereby, preferably, the trans-2-enoyl-CoA reductase comprises an amino acid sequence with at least 45% amino acid sequence identity with at least 45% amino acid sequence identity with at least one of SEQ ID NO's: 58 an 59. In another preferred embodiment, the modified cell is a cell for producing fermentation products with a longer chain-length, such as e.g. at least C ₆ , C ₈ , or C ₁₀ . In this embodiment the enzymes of the fatty acid β-oxidation cycle include: a) at least one of a peroxisomal thiolase that lacks a functional peroxisomal targeting signal and a heterologous thiolase with a broad chain-length specificity, whereby, preferably, the thiolase comprises an amino acid sequence with at least 45% amino acid sequence identity with at least one of SEQ ID NO's: 47 and 51 ; b) at least one of a peroxisomal enzyme with hydroxyacyl-CoA dehydrogenase and enoyl-CoA hydratase activities that lacks a functional peroxisomal targeting signal and a heterologous enzyme with hydroxyacyl-CoA dehydrogenase and enoyl-CoA hydratase activities with a broad chain-length specificity, whereby, preferably, the enzyme with hydroxyacyl-CoA dehydrogenase and enoyl-CoA hydratase activities comprises an amino acid sequence with at least 45% amino acid sequence identity with at least one of SEQ ID NO's: 52 and 55; and, c) a NADH-dependent trans-2-enoyl- CoA reductase, whereby, preferably, the trans-2-enoyl-CoA reductase comprises an amino acid sequence with at least 45% amino acid sequence identity with SEQ ID NO: 58.

A modified cell according to the invention can further comprise a) expression of an enzyme with acetyl-CoA carboxylase activity, whereby preferably the acetyl-CoA carboxylase comprises an amino acid sequence with at least 45% amino acid sequence identity with at least one of SEQ ID NO's: 10 and 60; and, b) expression of an enzyme with acetoacetyl-CoA synthase activity, which catalyses the decarboxylative condensation of acetyl-CoA and malonyl-CoA to form acetoacetyl-CoA, whereby preferably the acetoacetyl-CoA synthase comprises an amino acid sequence with at least 45% amino acid sequence identity with SEQ ID NO: 61. The cell can comprise this alternative route for producing acetoacetyl-CoA, in addition to but preferably as an alternative for the thiolase with short chain-length specificity in the event that products are produced not having a longer chain-length than C ₄ . In case products are to be produced with a longer chain-length than C ₄ , the alternative route for producing acetoacetyl-CoA is expressed in addition to the peroxisomal thiolase or the heterologous thiolase with a broad chain-length specificity, which are then needed for further rounds of the reverse β-oxidation cycle.

A preferred modified cell of the invention is a cell for producing butanol, wherein the cell expresses termination enzymes that catalyze the conversion of butyryl- CoA via butaldehyde to butanol, whereby, preferably the enzymes are selected from: a) a NADH-dependent enzyme having both aldehyde and alcohol dehydrogenase, preferably comprising an amino acid sequence with at least 45% amino acid sequence identity with SEQ ID NO: 63; and, b) a NADH-dependent aldehyde dehydrogenase, preferably comprising an amino acid sequence with at least 55% amino acid sequence identity with SEQ ID NO: 37, expressed together with a NADH-dependent alcohol dehydrogenase preferably comprising an amino acid sequence with at least 45% amino acid sequence identity with SEQ ID NO: 62.

A further preferred modified cell of the invention is a cell for producing 1,3- butanediol, wherein cell expresses termination enzymes that comprise an acetoacetyl- CoA reductase and at least one of a butyraldehyde dehydrogenase and an alcohol/aldehyde dehydrogenase, whereby, preferably, the enzymes are selected from: a) an acetoacetyl-CoA reductase comprising an amino acid sequence with at least 45% amino acid sequence identity with SEQ ID NO: 77; and at least one of: b) a butyraldehyde dehydrogenase comprising an amino acid sequence with at least 45% amino acid sequence identity with SEQ ID NO: 78; and, c) a NADH-dependent enzyme having both aldehyde and alcohol dehydrogenase, preferably comprising an amino acid sequence with at least 45% amino acid sequence identity with SEQ ID NO: 63.

Another preferred modified cell of the invention is a cell for producing an alcohol with a chain-length greater than C ₆ and preferably including dodecanol, wherein the cell expresses termination enzymes having fatty acyl-CoA reductase activity, whereby, preferably the enzymes are selected from: a) a NADH-dependent alcohol forming fatty acyl-CoA reductase, preferably comprising an amino acid sequence with at least 45% amino acid sequence identity with SEQ ID NO: 64; and, b) an NADH-dependent alcohol dehydrogenase, preferably comprising an amino acid sequence with at least 45%) amino acid sequence identity with SEQ ID NO: 65.

Still another preferred modified cell of the invention is a cell for producing an alkene, wherein the cell expresses a termination enzyme having olefin synthase activity, whereby preferably, the enzyme is selected from: a) a polyketide synthase comprising an amino acid sequence with at least 50% amino acid sequence identity with SEQ ID NO: 83; and, b) a polyketide synthase as defined in b) wherein the synthase is expressed in the form of more than one individual polypeptide chains, each comprising individual modules of the synthase.

Again another preferred modified cell of the invention is a cell for producing a fatty acid, preferably including dodecanoic acid, wherein the cell expresses a termination enzyme that catalyze the deacylation of acyl-CoA thioesters to fatty acids and CoA, whereby, preferably, the enzyme is an acyl-CoA thioesterase comprising an amino acid sequence with at least 45% amino acid sequence identity with at least one of SEQ ID NO: 81 and SEQ ID NO: 82.

The modified eukaryotic microbial host cell according to the invention preferably is a fungal cell, more preferably a yeast cell, and most preferably the cell is a cell of a yeast genus selected from the group consisting of Saccharomyces, Kluyveromyces, Candida, Pichia, Schizosaccharomyces, Hansenula, Kloeckera, Schwanniomyces, Yarrowia, Cryptococcus, Debaromyces, Saccharomycecopsis, Saccharomycodes, Wickerhamia, Debayomyces, Hanseniaspora, Ogataea, Kuraishia, Komagataella, Metschnikowia, Williopsis, Nakazawaea, Torulaspora, Bullera, Rhodotorula, and Sporobolomyces. It is further preferred that the cell is a cell of a yeast species selected from the group consisting of Saccharomyces cerevisiae, S. exiguus, S. bay anus, S. delbriickii, S. italicus, S. ellipsoideus, S. fermentati, S. kluyveri, S. krusei, S. lactis, S. marxianus, S. microellipsoides, S. montanus, S. norbensis, S. oleaceus, S. paradoxus, S. pastorianus, S. pretoriensis, S. rosei, S. rouxii, S. uvarum, S. ludwigii, Kluyveromyces lactis, K. marxianus, K. marxianus var. marxianus, K. thermotolerans, Candida utilis, C tropicalis, C albicans, C lipolytica, C versatilis, Pichia stipidis, P. pastoris and P. sorbitophila, Hansenula polymorpha and Schizosaccharomyces pombe. A preferred modified cell of the invention has the ability to grow on at least one of a hexose, preferably glucose, and a pentose, preferably at least one of xylose and arabinose, at a rate of at least 0.01, 0.02, 0.05, 0.1, 0.2, 0.25 or 0.3 h ^"1 under at least one of aerobic and anaerobic conditions.

7. Processes wherein the host cells of the invention are used for the production various compounds

In a second aspect, the invention relates to the use of a cell according to the invention for the preparation of at least one fermentation product that is or is derived from an intermediate in the β-oxidation cycle with a carbon chain-length of at least C ₄ , whereby, preferably, the fermentation product is selected from the group consisting of a fatty acid, a 1 -alcohol, a β-ketoacid, a β-ketoalcohol, a β-hydroxyacid, a 1,3-diol, a trans-A ² -fatty acid, a trans-Δ ² - fatty acid, an alkene, an alkane and derivatives thereof, more preferably the fermentation product is a 1 -alcohol other than methanol and ethanol, such as e.g. butanol, decanol, dodecanol and higher 1 -alcohols or a derivative thereof. The cell according to the invention is preferably used to produce the fermentation product in a process as defined herein below.

In a third aspect, the invention relates to processes for the production of these compounds by the host cells of the invention. Thus, in this aspect the invention relates to a process for producing at least one fermentation product that is or is derived from an intermediate in the β-oxidation cycle with a carbon chain-length of at least C ₄ , or wherein, more preferably, the fermentation product is selected from the group consisting of a fatty acid, a 1 -alcohol, a β-ketoacid, a β-ketoalcohol, a β-hydroxyacid, a 1,3-diol, a trans-Δ ² - fatty acid, a trans-Δ ² - fatty acid, an alkene and derivatives thereof. Most preferably, the fermentation product is a 1 -alcohol other than methanol and ethanol, such as e.g. butanol, decanol, dodecanol and higher 1 -alcohols or a derivative thereof. The process preferably, comprises the step of: a) fermenting a medium with a cell according to the invention, preferably under oxic conditions, whereby the medium contains or is fed with a non-fatty acid carbon source and whereby the yeast cell ferments the non-fatty acid carbon source to the fermentation product. The yeast cell preferably is a (host) cell as herein defined above. The process preferably comprises a further step wherein the fermentation product is recovered. The process may be a batch process, a fed-batch process or a continuous process as are well known in the art. In the process of the invention, the non-fatty acid carbon source preferably comprises at least one of a source of hexoses and a source of pentoses. The source of hexoses preferably comprises hexoses or multimers of hexoses that are assimilable by the cell, e.g. a yeast cell. Such assimilable hexoses or multimers thereof include e.g. glucose, fructose, galactose, mannose, maltose, saccharose, lactose and maltodextrines. The source of hexose preferably comprises or consists of at least glucose. Preferably the source of pentoses comprises or consists of at least one of xylose and arabinose.

Preferably, the medium fermented by the cells of the invention comprises or is fed with (fractions of) hydrolyzed biomass comprising at least one of a hexose and a pentose, such as e.g. glucose, xylose and/or arabinose.

The term "biomass" is understood to mean the biodegradable fraction of products, waste and residues from biological origin from agriculture (including vegetal, such as crop residues, and animal substances), forestry (such as wood resources) and related industries including fisheries and aquaculture, as well as biodegradable fraction of industrial and municipal waste, such as municipal solid waste or wastepaper. In a preferred embodiment, the biomass is plant biomass, more preferably a fermentable hexose/glucose/sugar-rich biomass, such as e.g. sugarcane, a starch-containing biomass, for example, wheat grain, or corn straw. Even more preferably, plant biomass is cereal grains, such as corn, wheat, barley or mixtures thereof. Another example of hydrolyzed biomass to be fermented in the processes of the invention is e.g. hydrolyzed cereal biomass. Methods for hydrolysis of biomass such as cereal are known in the art per se and include e.g. vapor and enzymes such as glucoamylases. Another example of hydrolyzed biomass to be fermented in the processes of the invention is e.g. hydrolyzed sugarcane biomass. Methods for hydrolysis of biomass such as sugarcane are also known in the art per se and include e.g. vapor.

Another highly preferred type of biomass for use in the process of the invention is a so-called "second generation" lignocellulosic feedstock. Lignocellulosic feedstocks are preferred if large volumes of fuels or chemicals are to be produced in a more sustainable way. Either dedicated crops, or by-products of existing industries, can be used since they will put less pressure on land use. Also, second generation processes are more sustainable than their first generation counterparts (Sims et al 2010, Bioresource Technology 101 : 1570-1580). Lignocellulosic feedstocks can be obtained from dedicated energy crops, e.g. grown on marginal lands, thus not competing directly with food crops. Or lignocellulosic feedstocks can be obtained as by-products, e.g. municipal solid wastes, wastepaper, wood residues (including sawmill and paper mill discards) and crop residues can be considered. Examples of crop residues include bagasse from sugar cane and also several corn and wheat wastes. In the case of corn by-products, three wastes are fiber, cobs and stover. Furthermore, forestry biomass may be used as feedstock.

Lignocellulosic feedstocks are mainly composed of cellulose, hemicellulose and lignin. Cellulose is a crystalline linear polymer of P-(l,4)-linked glucose moieties. The hemicelluloses present in the cell walls of grasses and cereals largely comprise xylan, a polymer of linear chains of P-(l,4)-linked D-xylopyranosyl residues. On this backbone, depending on the feedstock, various substitutions can be present, like a-L- arabinofuranosyl, (4-0-methyl)-a-D-glucuronic acids and O-acetylesters. These substitutions can be organized as short chains. Mono- or dimeric ferulic acid ester- linked to the arabinose moieties, tend to link the different polymers within plant cell walls.

In order to convert second generation feedstocks into the fermentation products of the invention, the cellulose and hemicellulose need to be released as monosaccharides. Hereto, either thermochemical approaches (usually referred to as pretreatment), enzymatic approaches or a combination of the two methodologies are applied. A pretreatment can serve to either completely liberate the sugars, or to make the polymeric compounds more accessible to subsequent enzymatic attack. Different types of pretreatment include liquid hot water, steam explosion, acid pretreatment, alkali pretreatment, and ionic liquid pretreatments. However, pretreatments can lead to the formation of compounds that are inhibitory to the fermenting organism. Inhibitory compounds in lignocellulose hydrolysates comprise aliphatic acids and in particular acetic acid, furans, aromatic compounds and extractives. Furans, such as furfural and 5- hydroxymethylfurfural (HMF), and some aliphatic acids, such as formic and levulinic acid, can be formed during pretreatment as degradation products from carbohydrates. Other compounds, such as acetic acid from hemicellulose, and phenolics from the partial breakdown of lignin, cannot be avoided since they are intrinsic parts of the cell wall structure (Eylen et al 2011, Bioresource Technology 102:5995-6004). Table 4 below shows an example of the composition of a hydrolysate from spruce biomass as reported by Koppram et al (Biotechnology for Biofuels 2012, 5:32).

Table 4: Example of the composition of a lignocellulosic hydrolysate

Values reported in g/L were recalculated to concentrations in mM. C6 sugars is the sum of glucose, mannose and galactose; C5 sugars is the sum of xylose and arabinose. Other compounds produced at low concentrations were formic and levulinic acids.

The relative amounts of the various compounds will depend both on the feedstock used and the pretreatment employed.

In the process of the invention, the sources of hexoses and/or pentoses may be hexoses and/or pentoses as such (i.e. as monomeric sugars, e.g. xylose, glucose and/or arabinose) or they may be in the form of any carbohydrate oligo- or polymer comprising hexose or pentose units (xylose, glucose and/or arabinose), such as e.g. lignocellulose, arabinans, xylans, cellulose, starch and the like. For release of xylose, glucose and/or arabinose units from such carbohydrates, appropriate carbohydrases (such as arabinases, xylanases, glucanases, amylases, cellulases, glucanases and the like) may be added to the fermentation medium or may be produced by the modified host cell. In the latter case the modified host cell may be genetically engineered to produce and excrete such carbohydrases. An additional advantage of using oligo- or polymeric sources of glucose is that it enables to maintain a low(er) concentration of free glucose during the fermentation, e.g. by using rate- limiting amounts of the carbohydrases preferably during the fermentation. This, in turn, will prevent repression of systems required for metabolism and transport of non-glucose sugars such as xylose and arabinose. In a preferred process the modified host cell ferments both the glucose and at least one of xylose and arabinose, preferably simultaneously in which case preferably a modified host cell is used which is insensitive to glucose repression to prevent diauxic growth. In addition to a source of at least one of xylose and arabinose (and glucose) as carbon source, the fermentation medium will further comprise the appropriate ingredient required for growth of the modified host cell. Compositions of fermentation media for growth of eukaryotic microorganisms such as yeasts are well known in the art.

The fermentation process may be an oxic (aerobic) or an anoxic (anaerobic) fermentation process. Preferably however, the process of the invention is an oxic (aerobic) fermentation process as herein defined above. Preferably, in an aerobic process under oxygen- limited conditions, the rate of oxygen consumption is at least 5.5, more preferably at least 6 and even more preferably at least 7 mmol/L/h. In a preferred aerobic oxygen-limited fermentation process of the invention, the yeast cell of the invention consumes less than 30, 20, 18, 15, 12, 10, 8 or 5% of the amount of oxygen on a C-molar basis related to the carbon source consumed during the conversion of the carbon source into the fermentation product. The conversion coefficient of oxygen consumed over substrate utilized on a C-molar basis (C _os ) is herein understood to mean mol 0 ₂ used per C-mol carbon source consumed. Thus, a process of the invention can be carried out under strict anaerobic conditions (i.e. C _os = 0.0), or the process of the invention can be carried out under aerobic, preferably oxygen-limited, conditions wherein the C _os is preferably less than 0.3, 0.2, 0.18, 0.15, 0.12, 0.1, 0.08, or 0.05.

The fermentation process is preferably run at a temperature that is optimal for the modified cells of the invention. Thus, for most yeasts cells, the fermentation process is performed at a temperature which is less than 42°C, preferably less than 38°C. For yeast cells, the fermentation process is preferably performed at a temperature which is lower than 35, 33, 30 or 28°C and at a temperature which is higher than 20, 22, or 25°C.

In a preferred fermentation process according to the invention the volumetric productivity of the fermentation product preferably is at least 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 5.0 or 10.0 g fermentation product per liter per hour. The yield of the fermentation product on glucose in the process preferably is at least 50, 60, 70, 80, 90, 95 or 98%. The yield is herein defined as a percentage of the theoretical maximum yield on glucose, which is for the following products in gram product per gram glucose:

1-butanol: 0.41, dodecanol: 0.34, 1,3-butanediol: 0.50, dodecanoic acid: 0.37 and tridecene: 0.29, palmitic acid: 0.36 and heptadecene: 0.30. In this document and in its claims, the verb "to comprise" and its conjugations is used in its non-limiting sense to mean that items following the word are included, but items not specifically mentioned are not excluded. In addition, reference to an element by the indefinite article "a" or "an" does not exclude the possibility that more than one of the element is present, unless the context clearly requires that there be one and only one of the elements. The indefinite article "a" or "an" thus usually means "at least one".

All patent and literature references cited in the present specification are hereby incorporated by reference in their entirety.

The following examples are offered for illustrative purposes only, and are not intended to limit the scope of the present invention in any way.

Description of the Figures Figure 1. A general pathway overview of a engineered yeast cell (e.g. S. cerevisiae) according to the invention wherein glucose is metabolized via pyruvate to acetyl-CoA to feed into a reversed β-oxidation cycle that is expressed in the cytosol of the yeast cell. Figure 2. Physical map of plasmid pUG6 (Guldener et al, 1996, Nucleic Acids Res., 24:2519-2524)

Figure 3 : Physical map of plasmid pSH47 (Guldener et al, 1996, Nucleic Acids Res., 24:2519-2524)

Figure 4: Schematic depiction of expression cassette for expression of TALI, TKL1, and RPE1 used for genomic integration into the S. cerevisiae GRE3 locus.

Figure 5 : Schematic depiction of expression cassette for expression of RPI1, XKS1, and the Clostridium phytofermentans xylose isomerase (XI Clos) used for genomic integration into the S. cerevisiae YHRCdeltal4 locus. Figure 6: Schematic depiction of expression cassette for expression of aldp _A and ACS _SE used for genomic integration into the S. cerevisiae FPS1 locus.

Figure 7: Schematic depiction of expression cassette for expression of E. coli mhpF, used for genomic integration into the S. cerevisiae ALD6 locus

Figure 8: Schematic depiction of expression cassette for expression of aldp _A and ACS _SE used for genomic integration into the S. cerevisiae HO locus. Figure 9: Physical map of plasmid pMA-C12-l as constructed in Example 8.

Figure 10: Physical map of plasmid pMA-C4-l as constructed in Example 9.

Figure 11 : Physical map of plasmid pMA-Tl-C12 as constructed in Example 10.

Figure 12: Physical map of plasmid pMA-T7-C4 as constructed in Example 10.

Figure 13: Physical map of plasmid pRS JC 1 as constructed in Example 11. Figure 14: Physical map of plasmid pRS JC 1 as constructed in Example 11.

Figure 15: Fermentation of strain CenPK 2a4 (Aadhl Afpsl Aald6::mhpF URA3) transformed with plasmids pMA-C4-l and pMA-T6-C4 in a 5L Labfors fermenter. Concentrations of cell dry weight (g/L), butanol ^g/L), ethanol (mg/L), acetic acid (mg/L) and glycerol (mg/L) are indicated over time.

Examples

Enzyme activity assays

Cell extracts for activity assays are prepared from exponentially growing aerobic or anaerobic batch cultures and analyzed for protein content as described by Abbot et al, (2009, Appl. Environ. Microbiol. 75 : 2320-2325).

NAD ⁺ -dependent acetaldehyde dehydrogenase (EC 1.2.1.10) activity is measured at 30°C by monitoring the oxidation of NADH at 340 nm. The reaction mixture (total volume 1 ml) contains 50 mM potassium phosphate buffer (pH 7.5), 0.15 mM NADH and cell extract. The reaction is started by addition of 0.5 mM acetyl-Coenzyme A.

The activity of pyruvate formate-lyase is estimated spectrophotometrically by recording the change in absorption at 340 nm at 35°C. The assay mixture of pyruvate formate-lyase contains 20 mM sodium pyruvate, 0.08 mM coenzyme A, 1 mM NAD, 6 mM sodium DL-malate, 2 mM dithiothreitol, 1.1 U of citrate synthase (pig heart, EC 4.1.3.7) per ml, 22 U of malate dehydrogenase (pig heart, EC 1.1.1.37) per ml, and cell- free extract in 100 mM potassium phosphate buffer (pH 7.6) (Takahashi et al, 1982, J. Bacteriol. 149: 1034-1040).

Thiolase activity is measured by monitoring the disappearance of acetoacetyl- CoA, corresponding to the thiolysis direction of the enzymatic reaction. The disappearance of acetoacetyl-CoA is monitored by the decrease in absorbance at 303 nm, which is the characteristic absorption band of an enolatecomplex formed by acetoacetyl-CoA with Mg ²⁺ . The reaction mixture contains 100 mM Tris-HCl, pH 8.0, 10 mM MgS04, 200 μΜ acetoacetyl-CoA, 200 μΜ CoA, and cell extract prepared as described above. A standard curve is constructed by measuring the absorbance of acetoacetyl-CoA at different concentrations with 10 mM Mg ²⁺ (Shen et al, 201 1, Appl Environ Microbiol, 77:2905-15).

The hydroxyacyl-CoA dehydrogenase activity is measured by monitoring the decrease of absorption at 340 nm, corresponding to consumption of NADH. The reaction mixture contains 100 mM 3-(N-morpholino) propanesulfonic acid (MOPS), pH 7.0, 200 _M NADH, 200 μΜ acetoacetyl-CoA, and crude cell extract. The reaction is initiated by the addition of the cell extract. (Shen et al., 2011, supra).

The enoyl-CoA hydratase activity is measured by the decrease of absorption at 263 nm, corresponding to disruption of the α-β unsaturation of crotonyl-CoA. The assay mixture contains 100 mM Tris-HCl pH 7.6, 100 μΜ crotonyl-CoA, and the crude extract. The reaction is initiated by the addition of the cell extract. The standard curves for crotonyl-CoA and 3-hydroxybutyryl-CoA are constructed by measuring the absorbance of the two compounds at 263 nm at different concentrations. (Shen et al, 2011, supra).

The trans enoyl-CoA reductase activity for crotonyl-CoA is measured at 340 nm. The reaction mixture contains 100 mM potassium phosphate buffer, pH 6.2, 200 μΜ NADH, 200 μΜ crotonyl-CoA, and crude extract. The reaction is initiated by the addition of the extract. To detect the activity for butyryl-CoA, the reaction mixture contains 1 mM NAD ⁺ , 0.4 mM butyryl-CoA, and crude extract in 100 mM Tris HC1, pH 7.5. The absorbance is monitored at 340 nm at 30°C. The reaction is initiated by the addition of the extract.

The aldehyde and alcohol dehydrogenase activities of AdhE2 are measured by monitoring the decrease of absorbance at 340 nm corresponding to the consumption of NADH or NADPH. The reaction mixture contains 100 mM Tris-HCl, pH 7.5, 5 mM dithiothreitol (DTT), 300 μΜ NADH, and 1 mM butyryl-CoA for the butyraldehyde dehydrogenase (BYDH) reaction and 50 mM butyraldehyde for the butanol dehydrogenase (BDH) reaction. The reaction is initiated by the addition of the extract. (Shen et al, 2011, supra).

NADH assay: A fluorescent NAD/NADH detection kit purchased from Cell

Technology (Mountain View, CA) is used. Cells are harvested by centrifugation at 13,200 rpm at 4°C. The pellets are then resuspended with 0.2 ml of the NAD/NADH extraction buffer and 0.2 ml of the lysis buffer (provided). Lysis is allowed to proceed for 10 to 20 min at 60°C until the cell resuspension turns clear. The lysate is then centrifuged at 8,000 rpm for 5 min at 4°C. The supernatant is retrieved for subsequent NADH assays. For the measurement of intracellular NADH levels, the cell lysates are mixed with the enzyme and the fluorescent detection reagent provided in the kit. The reaction is allowed to proceed for 1 to 1.5 h at room temperature in the dark, and then readings are taken with excitation at 530 to 570 nm and emission at 590 to 600 nm.

Strain construction The construction of the strains featured in the following examples is based on strain CEN.PK2-1C (MATa; ura3-52; trpl-289; leu2-3,1 12; his3A 1 ; MAL2-8 ^C ; SUC2). For further details on this strain see:

http://web.uni-frankfurt.de/fbl5/mikro/euroscarf/data/cen.ht ml)

Table 5 : Overview of plasmids used for the strain construction (the plasmids are featured in the following Examples)

Ec, Escherichia coli; Td, Treponema denticola; Sc, Saccharomyces cerevisiae

Clostridium acetobutylicum; Eg, Euglena gracilis; Hs, Homo sapiens; Cb, Clostridium beijerinckii, 1 ,3-BDO, 1,3-butanediol.

Table 6: Overview of basis strains for the strain construction (the construction of the strains is described in Examples 1 to 7).

Name Genotype

CenPK MATa; ura3-52; trpl-289; leu2-3,1 12; his3A 1 ; MAL2-8C; SUC2

CenPK 1 CenPK Aere3 : :TALl TKLl RPE1 AYHRCdeltal4: : RPI1 XKS1 xvlA

CenPK 2al CenPK Aald6: :mhpF URA3

CenPK 2a2 CenPK Aadhl : :kanMX Aald6: :mhpF URA3

CenPK 2a3 CenPK Afpsl : :kanMX Aald6: :mhpF URA3

CenPK 2a4 CenPK Aadhl : :loxP Afpsl : :kanMX Aald6::mhpF URA3

CenPK 2a5 CenPK Aho: :aldp _A acs _S E URA3

CenPK 2a6 CenPK Aadhl : :kanMX Aho: :aldPA acs _SE URA3

CenPK 2a7 CenPK Afpsl : :kanMX Aho::aldPA acs _SE URA3 CenP 2a8 | CenP Aadhl : :kanMX Afps 1 : :ald _PA acs _SE URA3

Example 1 : Deletion of the gene GRE3 and constitutive chromosomal overexpression of the yeast genes TALI. TKL1 and RPEl

The yeast genes TALI, TKL1 and RPE1 (for sequences see: http://www.yeastgenome.org/cgi- bin ^/ getSeq?query=YPR074C&seqtype=Coding%20sequence%20% 28CDS%29&forma t=fasta, TKL1; http://www.yeastgenome.org/cgi- biri/getSeq?query=YLR354C&seqtype=Coding%20sequence%20%2 8CDS%29&forma t=fasta, TALI; http://www.yeastgenome.org/cgi- bin/getSeq?query=YJL 121 C&seqtype=Coding%20sequence%20%28CDS%29&format =fasta, RPEl) were synthesized (codon optimized for yeast) at GenScript USA Inc. (www.genscript.com) in one cluster (gene cassette). The three genes are clustered according to Figure 4 with the URA3 marker gene for selection in yeast (the URA3 coding region including 400 bp upstream and 200 bp downstream, for sequence see http://www.yeastgenome.org/cgi- bin/getSeq?query=YEL021 W&flankl= 1000&flankr= 1000&format=fasta) flanked by marker rescue sequences (MR; for recovery of the URA3 marker via counter selection on plates containing 5-FOA; SEQ ID NO: 70) and flanking regions, which are homologous to the native GRE3 locus (the first and last 40 bp of the GRE3 coding region, for sequence see: http://www.yeastgenome.org/cgi- bin/getSeq?query=YHR104W&seqtype=Coding%20sequence%20%28 CDS%29&form at=fasta) for the genomic integration of the gene cassette into the GRE3 locus . For the expression of the TALI gene the ENOl promoter and ENOl terminator are used (600 bp upstream and 300 bp downstream of the ENOl coding region, for sequences see: http : ww . yeas t gen o me . o rg c i- bin/getSeq?query=YGR254W&flankl= 1000&flankr= 1000&format=fasta), while for expression of the TKL1 gene the TDH3 promoter and TDH3 terminator are used (600 bp upstream and 300 bp downstream of the TDH3 coding region, for sequences see: http : www.yeastgenome.org cgi- bin/getSeq?query=YGR192C&flanki=1000&flankr=1000& ;format=fasta) and for the expression of the RPEl gene a modified (constitutive) version of the ADH1 promoter (SEQ ID NO: 33) and the TRP1 terminator are used (300 bp downstream of the TRP1 coding region, for sequences see: http://www.yeastgenome.org/cgi- bin/getSeq?query=YDR007 W&fiankl= 1000&fiankr= 1000&format=fasta).

Strain CEN.PK2-1C was used for transformation with this gene cassette (Figure 4). After transformation of the corresponding yeast strains with the gene cassette via the lithium acetate method described by Gietz et al. (1992, supra) the strains are cultivated on WMVIII (Lang and Looman, 1995, supra) agar plates lacking uracil for the selection of uracil prototrophic strains (transformants carrying the expression cassette instead of the GRE3 coding region in the genome).

Subsequently the resulting strain is cultivated in liquid WMVIII medium in presence of Uracil in order to allow growth of cells, which have lost the URA3 marker cassette (via homologous recombination by the two MR sequences, leaving one MR sequence in the genom). Subsequently the cells are plated on YE-medium agar plates supplemented with 1.5 g/L 5'-FOA (5-fluoroorotic acid) for the selection of uracil auxotrophic cells. The desired phenotype (Ura minus) is confirmed via replica plating.

Example 2: Constitutive chromosomal overexpression of the yeast genes RPI1 and XKS1 and the Clostridium phytofermentans xylose isomerase gene

The yeast genes RPI1 and XKS1 (for sequences see: http://www.yeastgenome.org/cgi- bin/getSeq?query=YIL 119C&seqtype=Coding%20sequence%20%28CDS%29&format =fasta, RPI1 and http://www.yeastgenome.org/cgi- bin/getSeq?query=YGR194C&seqtype=Coding%20sequence%20%28 CDS%29&form at=fasta, XKS1) and a Clostridium phytofermentans xylose isomerase coding sequence (codon-optimized for yeast: SEQ ID NO: 89, encoding the amino acid sequence of SEQ ID NO: 34) were synthesized at GenScript USA Inc. (www.genscript.com) in one cluster (gene cassette). The three genes are clustered according to Figure 5 with the URA3 marker gene for selection in yeast (the URA3 coding region including 400 bp upstream and 200 bp downstream, for sequence see http://www.yeastgenome.org/cgi- bin/ getSeq?query=YEL021 W&flankl= 1000&flankr= 1000&format=fasta) flanked by marker rescue sequences (MR; for recovery of the URA3 marker via counter selection on plates containing 5-FOA; SEQ ID NO: 70) and flanking regions, which are homologous to the native YHRCdeltal4 locus (the first and last 40 bp of the YHRCdeltal4 coding region, for sequence see: http://www.yeastgenome.org/cgi- bin/getSeq?query=YHRCdeltal4&seqtype=ORF%20Genomic%20DNA &format=fasta ) for the genomic integration of the gene cassette into the YHRCdeltal4 locus . For the expression of the RPI1 gene the ENOl promoter and ENOl terminator are used (600 bp upstream and 300 bp downstream of the ENOl coding region, for sequences see: http://www.yeastgenome.org/cgi- bin/getSeq?query=YGR254W&flanki= 1000&f!ankr= 1000&format=fasta), while for expression of the XKS1 gene the TDH3 promoter and TDH3 terminator are used (600 bp upstream and 300 bp downstream of the TDH3 coding region, for sequences see: http ://www.yeastgenome.org/cgi- bir^getSeq?query=YGR192C&flarJcl=1000&flankr=1000&am p;format=fasta) and for the expression of the Clostridium phytofermentans xylose isomerase a modified (constitutive) version of the ADH1 promoter (SEQ ID NO: 33) and the TRP1 terminator are used (300 bp downstream of the TRP1 coding region, for sequences see: http://www.yeastgenome.org/cgi- bin/getSeq?query=YDR007W&flankl= 1000&flankr= 1000&format=fasta) .

The strain resulting from Example 1 was used for transformation with this gene cassette (Figure 5).

After transformation of the corresponding yeast strains with the gene cassette via the lithium acetate method described by Gietz et al. (1992, supra) the strains are cultivated on WMVIII (Lang and Looman, 1995, supra) agar plates lacking uracil for the selection of uracil prototrophic strains (transformants carrying the expression cassette instead of the YHRCdeltal4 region in the genome).

Subsequently the resulting strain is cultivated in liquid WMVIII medium in presence of Uracil in order to allow growth of cells, which have lost the URA3 marker cassette (via homologous recombination by the two MR sequences, leaving one MR sequence in the genom). Subsequently the cells are plated on YE-medium agar plates supplemented with 1.5 g/L 5 '-FOA (5-fluoroorotic acid) for the selection of uracil auxotrophic cells. The desired phenotype (Ura minus) is confirmed via replica plating. The resulting strain is named CEN.PK 1 (see Table 6).

Example 3 : Deletion of gene ADH1 For the deletion of the ADHI gene a deletion cassette is amplified from plasmid pUG6 (Guldener et al, 1996, Nucleic Acids Research, 24:2519-2524; Figure 2) via PCR with the following oligonucleotides:

ADHlcrelox fw:

5-ATG TCT ATC CCA GAA ACT CAA AAA GGT GTT ATC TTC TAC GCC

AGC TGA AGC TTC GTA CGC-3' (SEQ ID NO: 66)

ADHlcrelox rev:

5'- TTA TTT AGA AGT GTC AAC AAC GTA TCT ACC AAC GAT TTG AGC ATA GGC CAC TAG TGG ATC TG-3' (SEQ ID NO: 67)

The resulting 1.4 kbp fragment containing the KanMX marker gene (which provides resistance to geniticin in yeast) flanked by loxP sites (for recovery of the marker gene) and sequences homologous to the native ADHI gene locus (homologous to the first and last 40 bp of the coding region of ADHI; for sequence see: htt : / www. ycastgenomc.org/cgi- bir^getSeq?querv=YOL086C&seqtvpe=QRF%20Genomic%20DNA& ;format=fasta is used to transform strain S. cerevisiae CEN.PK 1 (see Table 6; resulting from Example 2). After transformation with this construct via the lithium acetate method described by Gietz et al. (1992, Nucleic Acids Res., 20(6): 1425) the strain is cultivated on YE agar plates containing 200 μg/ml of geniticin for the selection of resistant cells (trans formants carrying the deletion cassette instead of the adhl coding region).

Example 4: Deletion of gene FPS1

For the deletion of the FPS1 gene a deletion cassette is amplified from plasmid pUG6 (Guldener et al, 1996, supra; Figure 2) via PCR with the following oligonucleotides:

FPSlcrelox fw:

5'-ATG AGT AAT CCT CAA AAA GCT CTA AAC GAC TTT CTG TCC AGC CAG CTG AAG CTT CGT ACG C-3' (SEQ ID NO: 68)

FPSlcrelox rev:

5'-TCA TGT TAC CTT CTT AGC ATT ACC ATA ATG CGA ATC TTC TGC ATA GGC CAC TAG TGG ATC TG-3' (SEQ ID NO: 69)

The resulting 1.4 kbp fragment containing the KanMX marker gene (which provides resistance to geniticin in yeast) flanked by loxP sites (for recovery of the marker gene) and sequences homologous to the native FPS1 gene locus (homologous to the first and last 40 bp of the coding region of FPS1; for sequence see http://www.yeastgenome.org/cgi- bin/getSeq?querv=YLL043W&seqtvpe=ORF%20Genomic%20DNA& ;format=fasta is used to transform strain S. cerevisiae CEN.PK 1 (see Table 6; resulting from Example 2). After transformation with this construct via the lithium acetate method described by Gietz et al. (1992, supra) the strain is cultivated on YE agar plates containing 200 μg/ml of geniticin for the selection of resistant cells (trans formants carrying the deletion cassette instead of the fpsl coding region).

The sequential deletion of gene adhl and fpsl in one strain is accomplished by recovering the kanMX marker in the strain resulting from Example 3 (adhl deletion strain) via the cre-recombinase procedure described by Guldener et al, (1996, supra) and transforming the resulting yeast strain with the fpsl deletion cassette (see above). The cre-recombinase procedure requires the transformation of the yeast strain with the plasmid pSH47 (see Figure 3), which carries the gene for the cre-recombinase. This enzyme is able to recombine the two loxP sites flanking the kanMX resistance gene. This recombination event leads to a loss of this marker gene leaving one loxP site at the target locus (in the present example locus adhl). To enable the loss of plasmid pSH47 the yeast strain is cultivated for at least 10 generations in WMVIII (Lang and Looman, 1995 Appl Microbiol Biotechnol, 44(1-2): 147-56) containing uracil. Colonies, which have lost the plasmid pSH47 are identified via counter selection on plates with 5- fluoroorotic acid and picked for further construction purposes.

Example 5: Deletion of the gene FPS1 and expression of the P. aeruginosa aldpA and S. enterica acssrjgenes

The P. aeruginosa aldp _A gene (codon optimized for yeast: SEQ ID NO: 75) and the S. enterica acss _E gene are synthesized (codon optimized for yeast: SEQ ID NO: 76) at GenScript USA Inc. (www. genscript.com) in one cluster (gene cassette). The two genes are clustered according to Figure 6 with the URA3 marker gene for selection in yeast (the URA 3 coding region including 400 bp upstream and 200 bp downstream, for sequence see http : "www.ycastgenome.org/cgi- bin/getSeq?querv=YEL021 W&flankH 1000&flankr= 1 OOO&format tasta) flanked by marker rescue sequences (MR; for recovery of the URA3 marker via counter selection on plates containing 5-FOA; SEQ ID NO: 70) and flanking regions, which are homologous to the native FPSI locus (the first and last 40 bp of the FPSI coding region, for sequence see: http://www.yeastgenome.org/cgi- bir^getSeq?query=YLL043W«&seqtype=ORF%20Genomic%20DNA&a mp;format=fasta) for the genomic integration of the gene cassette into the FPSI locus . For the expression of the aldp _A gene the ENOl promoter and ENOl terminator are used (600 bp upstream and 300 bp downstream of the ENOl coding region, for sequences see: http://www.yeastgenome.org/cgi- bin/getSeq?querv=YGR254W&flankl= 1000&flankr= 1 Q00&format=fasta , while for expression of the acss _E gene the TDH3 promoter and TDH3 terminator are used (600 bp upstream and 300 bp downstream of the TDH3 coding region, for sequences see: http://www.yeastgenome.org/cgi- bin/getSeq?query=YGRl 92C&flankl= 1000&flankr= 1000&format=fasta .

The strains resulting from Examples 2 and 3 (adhl deletion strain) are used for transformation with this gene cassette (Figure 6).

After transformation of the corresponding yeast strains with the gene cassette via the lithium acetate method described by Gietz et al. (1992, supra) the strains are cultivated on WMVIII (Lang and Looman, 1995, supra) agar plates lacking uracil for the selection of uracil prototrophic strains (transformants carrying the expression cassette instead of the fpsl coding region in the genome).

Example 6: Deletion of the gene ALD6 and expression of the E. coli mhpF gene

The E. coli mhpF gene is synthesized (codon optimized for yeast: SEQ ID NO: 74) at GenScript USA I nc. (www . enscnpt.com) in one cluster (gene cassette). The two genes are clustered according to Figure 7 with the URA3 marker gene for selection in yeast (the URA3 coding region including 400 bp upstream and 200 bp downstream, for sequence see: http://www.yeastgenome.org/cgi- bin/getSeq ?querv= YEL021 W&flankl= 1000&flankr= 1 QQO&format=fasta flanked by marker rescue sequences (MR; for recovery of the URA3 marker via counter selection on plates containing 5-FOA; SEQ ID NO: 70) and flanking regions which are homologous to the native ALD6 locus (the first and last 40 bp of the ALD6 coding region, for sequence see: http://www.yeastgenome.org/cgi- bin/ getSeq?query=YPL061 W&seqtype=ORF%20Genomic%20DNA&format=fasta) for the genomic integration of the gene cassette into the ALD6 locus. For the expression of the mhpF gene the ENOl promoter and ENOl terminator are used (600 bp upstream and 300 bp downstream of the ENOl coding region, for sequences see: http://www.yeastgenome.org/cgi-

The strains resulting from Examples 2, 3 and 4 (adhl and fpsl deletion strains) are used for transformation with this gene cassette (Figure 7).

After transformation of the corresponding yeast strains with the gene cassette via the lithium acetate method described by Gietz et al. (1992, supra) the strains are cultivated on WMVIII (Lang and Looman, 1995, supra) agar plates lacking uracil for the selection of uracil prototrophic strains (transformants carrying the expression cassette instead of the ALD6 coding region in the genome).

Example 7: Expression of the genes P. aeruginosa aldpA and S. enterica acssgjn the HO locus

The P. aeruginosa aldp _A gene (codon optimized for yeast: SEQ ID NO: 75) and the S. enterica acss _E gene are synthesized (codon optimized for yeast: SEQ ID NO: 76) at GenScript USA Inc. (www.genscript.com) in one cluster (gene cassette). The two genes are clustered according to Figure 8 with the URA3 marker gene for selection in yeast (the URA 3 coding region including 400 bp upstream and 200 bp downstream, for sequence see: htt : //www. ycastgcnomc.org/cgi- bm/getSeq?query=YELQ21 W&flankl= 1000&flankr= 1000&format=fasta) flanked by marker rescue sequences (MR; for recovery of the URA3 marker via counter selection on plates containing 5-FOA; SEQ ID NO: 70) and flanking regions which are homologous to the native HO locus (the first and last 40 bp of the HO coding region, for sequence see: http://www.yeastgenome.org/cgi- bin/getSeq?querv=YDL227C&seqtvpe=ORF%20Genomic%20DNA& ;format=fasta for the genomic integration of the gene cassette into the HO locus. For the expression of the aldp _A gene the ENOl promoter and ENOl terminator are used (600 bp upstream and 300 bp downstream of the ENOl coding region, for sequences see: http://www.yeastgenome.org/cgi- while for expression of the acss _E gene the TDH3 promoter and TDH3 terminator are used (600 bp upstream and 300 bp downstream of the TDH3 coding region, for sequences see: http://www.yeastgenome.org/cgi- bin/getSeq ?query=YGR 192C&flankl= 1000&flankr= 1000&format=fasta .

The strains resulting from Examples 2, 3 and 4 (adhl and fpsl deletion strains) are used for transformation with this gene cassette (Figure 8).

After transformation of the corresponding yeast strains with the gene cassette via the lithium acetate method described by Gietz et al. (1992, supra) the strains are cultivated on WMVIII (Lang and Looman, 1995, supra) agar plates lacking uracil for the selection of uracil prototrophic strains (transformants carrying the expression cassette instead of the HO coding region in the genome).

Example 8: Expression of genes for the production of long chain compounds (e.g. CI 2) (with plasmid pMA-C12-l and pMA-C12-2

The Ci2 gene cassette 1 is synthesized by GenScript USA Inc. (www. genscnpt.com) and cloned into plasmid pMA via the Kpnl restriction site resulting in plasmid pMA-C12-l; see Figure 9). The pMA vector comprises an ampicillin resistance gene and an origin of replication for replication in E. coli (for sequence of pMA see: SEQ ID NO: 71). The cassette contains in one cluster a CEN/ARS sequence (SEQ ID NO: 72) for autonomous replication and segregation in yeast, the LEU2 expression cassette for selection of positive transformants (the LEU2 coding region including 400 bp upstream and 200 bp downstream, for sequence see: http://www.yeastgenome.org/cgi- bin/ getSeq?query=YCLO 18W&flankl= 1 OOO&flankr 1 OOO&format =fasta ) and codon optimized variants (codon optimized for S. cerevisiae) of the genes FadA (E. coli), FadB (E. coli) and Ter ( T. denticola). For the expression of the FadA gene a modified (constitutive) version of the ADH1 promoter (SEQ ID NO: 33) and the TRP1 terminator are used (300 bp downstream of the TRP1 coding region, for sequences see: http://www.yeastgenome.org/cgi- bin/getSeq?query=YDR007 W&flankl= 1000&flankr= 1000&format=fasta) while fo expression of the FadB gene the ENOl promoter and ENOl terminator are used (600 bp upstream and 300 bp downstream of the ENOl coding region, for sequences see: http ://www.yeastgenome.org/cgi- bin/getSeq?querv=YGR254W&flankl= 1000&flankr= 1 OOO&format fasta). For the expression of the Ter gene the TDH3 promoter and the TDH3 terminator are used (600 bp upstream and 300 bp downstream of the TDH3 coding region, for sequences see: http://www.yeastgenome.org/cgi-

The strains resulting from Examples 5, 6 and 7 are used for transformation with plasmid pMA-C 12-1.

The corresponding yeast strains are transformed with pMA-C12-l via the lithium acetate method described by Gietz et al. (1992, supra) the strains are cultivated on WMVIII (Lang and Looman, 1995, supra) agar plates lacking leucine for the selection of leucine prototrophic strains (transformants carrying the plasmid pMA-C12-l).

Another variant of Example 8 is the construction and transformation of plasmid pMA-C12-2 in the corresponding yeast strains. This plasmid is analogous to plasmid pMA-C12-l except that POT1 (S. cerevisiae), FOX2 (S. cerevisiae) genes (both without PTS) are used instead of FadA (E. coli) and FadB (E. coli). The gene Ter (T denticola) is used on both plasmids.

Example 9: Expression of genes for the production of short chain compounds (e.g. C4) (with plasmids pMA-C4-l and pMA-C4-2)

The C4 gene cassette 1 is synthesized by GenScript USA Inc. (www.genscript.com) and cloned into plasmid pMA via the restriction site Kpnl resulting in plasmid pMA-C4-l; see Figure 10). The pMA vector comprises an ampicillin resistance gene and an origin of replication for replication in E. coli (for sequence of pMA please see SEQ ID NO: 71). The cassette contains in one cluster a CEN/ARS sequence (SEQ ID NO: 72) for autonomous replication and segregation in yeast, the LEU2 expression cassette for selection of positive transformants (the LEU2 coding region including 400 bp upstream and 200 bp downstream, for sequence see: http://www.yeastgenome.org/cgi- bin/getSeq?query=YCLO 18 W&flankl= 10Q0&flankr= 10Q0&format=fasta) and codon optimized variants (codon optimized for S. cerevisiae) of the genes ERG10 (S. cerevisiae), Hbd (C. acetobutylicum), crt (C. acetobutylicum) and Ter (T. denticola). For the expression of the ERG10 gene a modified (constitutive) version of the ADH1 promoter (SEQ ID NO: 33) and the TRPl terminator are used (300 bp downstream of the TRPl coding region, for sequences see: htt ://www.yeastgenome.org/cgi- bin/getSeq?querv=YDR007W&flankl= 10Q0&flankr= 10Q0&format=fasta , while for expression of the Hbd gene the ENOl promoter and ENOl terminator are used (600 bp upstream and 300 bp downstream of the ENOl coding region, for sequences see: http ://www.yeastgenome.org/cgi-

For the expression of the crt gene the TDH3 promoter and the TDH3 terminator are used (600 bp upstream and 300 bp downstream of the TDH3 coding region, for sequences see: htt : //www. yeastgenomc.org/cgi- bin/getSeq?query=YGRl 92C&flantd= 10Q0&flankr= 1 OOO&format fasta).

For expression of the Ter gene the ENOl promoter and ENOl terminator are used

(600 bp upstream and 300 bp downstream of the ENOl coding region, for sequences see http ://www.yeastgenome.org/cgi- bin/getSeq?querv=YGR254W&flankl= 10QQ&flankr= 1000&format=fasta ).

The strains resulting from Examples 5, 6 and 7 are used for transformation with plasmid pMA-C4- 1.

The corresponding yeast strains are transformed with pMA-C4-l via the lithium acetate method described by Gietz et al. (1992, supra) the strains are cultivated on WMVIII (Lang and Looman, 1995, supra) agar plates lacking leucine for the selection of leucine prototrophic strains (transformants carrying the plasmid pMA-C4-l).

Another variant of Example 9 concerns the construction and transformation of plasmid pMA-C4-2 in the corresponding yeast strains. This plasmid is analogous to plasmid pMA-C4-l except that atoB (E. coli), Hbd (C. beijerinckii), crt (C. beijerinckii)) and Ter (E. gracilis) genes are used instead of ERG10 (S. cerevisiae), Hbd (C. acetobutylicum), crt (C. acetobutylicum) and Ter ( T. denticola), respectively.

Yet another variant of Example 9 concerns the construction and transformation of plasmid pMA-C4-3 in the corresponding yeast strains. This plasmid is analogous to plasmid pMA-C4-l except that the plasmid only contains the genes ERG 10 (S. cerevisiae) and Hbd (C. acetobutylicum) and the corresponding promoters and terminators.

Example 10: Expression of termination enzymes

Expression of genes encoding for enzymes converting intermediates of the reversed β-oxidation cycle into desired final compounds (termination enzymes) is performed using plasmids pMA-Tl-C12, pMA-T2-C12, pMA-T4-C12, pMA-T5-C12, pMA-T6-C4 and pMA-T7-C4 (see Table 4 for plasmids).

Plasmids pMA-Tl-C12, pMA-T2-C12, pMA-T4-C12, pMA-T5-C12 are used in combination with plasmids pMA-C12-l and pMA-C12-l (see Example 8) for the production of long chain compounds (in particular C ₁₂ ) and plasmids pMA-T6-C4 and pMA-T7-C4 were used in combination with plasmids pMA-C4-l and pMA-C4-l (see Example 9) for production of short chain compounds (in particular C ₄ ).

Constructs for the expression of the genes E. coli fadM or H. sapiens hBACH or M. musculus PTE-2 or E. gracilis FAR or E. coli yiaY or C. acetobutylicum adhE2 or E. coli mhpF+FucO are synthesized by GenScript USA Inc. (www . genscript.com) and cloned into plasmid pMA via the Kpnl restriction site resulting in plasmids pMA-Tl- C12, pMA-T2-C12, pMA-T4-C12, pMA-T5-C12, pMA-T6-C4 and pMA-T7-C4 (see Figures 9 and 10 for pMA-Tl-C12 and pMA-T7-C4, respectively). The pMA vector comprises an ampicillin resistance gene and an origin of replication for replication in E. coli (for sequence of pMA see SEQ ID NO: 71). The cassettes contain in one cluster a CEN/ARS sequence (SEQ ID NO: 72) for autonomous replication and segregation in yeast, the HIS3 expression cassette for selection of positive transformants after transformation (the HIS3 coding region including 500 bp upstream and 200 bp downstream, for sequence see: http://www.yeastgenome.org/cgi- and codon optimized variants (codon optimized for S. cerevisiae) of the genes E. coli fadM or H. sapiens hBACH or E. gracilis FAR or E. coli yiaY or C. acetobutylicum adhE2 or E. coli mhpF+FucO. For the expression of each of the gene the ENOl promoter and ENOl terminator are used (600 bp upstream and 300 bp downstream of the ENOl coding region, for sequences see: http://www.yeastgenome.org/cgi- bin/getSeq?querv=YGR254W&flankl= 1000&flankr= 1000&format=fasta . For the simultaneous expression of the E. coli mhpF and FucO gene the ENOl promoter and ENOl terminator are used for mhpF while for the FucO gene the TDH3 promoter and the TDH3 terminator are used (600 bp upstream and 300 bp downstream of the TDH3 coding region, for sequences see: http ://www.yeastgenome.org/cgi- bin/getSeq?query=YGRl 92C&flankl= 1000&flankr= 1000&format=fasta . The strains resulting from Examples 8 and 9 are used for transformation with plasmids pMA-Tl-C12, pMA-T2-C12, pMA-T4-C12, pMA-T5-C12, pMA-T6-C4 and pMA-T7-C4.

The strains were transformed via the lithium acetate method described by Gietz et al. (1992, supra) and were cultivated on WMVIII (Lang and Looman, 1995, supra) agar plates lacking histidine for the selection of histidine prototrophic strains (trans formants carrying plasmids pMA-Tl-C12 or pMA-T2-C12 or pMA-T4-C12 or pMA-T5-C12 or pMA-T6-C4 or pMA-T7-C4). Example 11 : Expression constructs for biosynthesis of alkenes

Two centromeric expression vectors pRSJCl and pRSJC2 are constructed for expression in S. cerevisiae CEN.PK strains of an OLS gene from Synechococcus sp strain PCC7002, which encodes an alfa-olefme synthase (SYNPCC7002 A1173). The OLS open reading frame encodes a multimodular Ols protein with eight distinguishable structural modules in a single polypeptide chain and each module has a different enzymatic activity (Mendez-Perez et al, 2012, supra). The pRSJCl vector is constructed to express the full-length Ols protein and the pRSJC2 vector is constructed for expression of three independent polypeptide chains (mOLs 1, 2 and 3) that together comprises all the different modules of the Ols protein. The latter will be expressed in combination with a molecular scaffold.

The complete, full length OLS gene is cloned into a centromeric pRSJCl vector (Figure 13) that is constructed by ligating a 4.442 bp fragment from the shuttle pRS316 vector (Sikorski and Hieter, 1989, Genetics, 122: 19) generated by Pvull digestion -to delete a 445bp fragment containing its multiple cloning site (MCS, 103bp)- to a 1357 bp Pvull fragment of the pESC vector (Stratagene), which contains the S. cerevisiae GAL1 and GAL 10 promoters in opposite orientation, two MCS sequences MCS1 and MCS2, and a transcription termination sequence downstream of each promoter. The OLS sequence will be cloned between the Spel and Sad sites of MCS 1 in pRSJCl to produce a plasmid designated as pOLS-JAMl .

Three DNA sequences named OLS1, OLS2 and OLS3, encoding three independent Ols-protein submodules - mOlsl, m01s2 and m01s3 (with amino acid sequences as depicted in SEQ ID NO's: 84, 85 and 86, respectively) - are synthesized, cloned and expressed from the two pRSJC vectors, pRSJCl and -2 (Figure 13 and 14), that are constructed in this work.

The OLS1 sequence is cloned in the MCS2 of the above described pRSJCl vector (URA3 marker) under the control of the GALI promoter; in addition, a DNA sequence encoding a synthetic protein scaffold (SCA; having the amino acid sequence depicted in SEQ ID NO: 87) is synthesized and cloned in the MCS1 of pRSJCl under the control of the GAL10 promoter to produce plasmid pOLSl/SCA-JAM2. The synthetic SCA sequence encodes three modular protein-protein interaction domains: the GTPase binding domain (GBD) from the rat actin polymerization switch N-WASP (Kim A.S. et al., 2000, supra), the SH3 domain from mouse c-Crk (Wu, X. et al., 1995, supra), and the PSD95/DlgA/Zo-l (PDZ) domain from the mouse adaptor protein syntrophin (Schultz, J. et al. 1998, supra). The three protein-protein interaction domains will be bound by sequences encoding two flexible nine-residue glycine-serine linkers (see SEQ ID NO: 87). According to domain encoding sequences used by Dueber et al. (2009, supra), a 771 bp Spel-Sacl DNA fragment is synthesized in vitro.

The OLS2 and OLS3 sequences are both cloned in pRSJC2: OLS2 at the MCS1 under the control of the GAL10 promoter and OLS3 at MCS2 under the control of the GALI promoter, to produce pOLS2/pOLS3-JAM3. pRSJC2 (Figure 14) is almost identical in features to pRSJCl, differing only in the auxotrophic marker (TRPI). For construction of pRSJC2, the above indicated 1357 bp Pvull fragment from the pESC vector containing the GAL1-GAL10 yeast promoter is subcloned into a 4338 bp derivative of pRS314 (TRPI; (Sikorski and Hieter, 1989, supra) obtained by deletion of the MCS with Pvull.

Amino acid sequences of mammalian protein-protein interaction domains are added at the C terminus of each Ols 1, 01s2 and 01s3 submodules, in order to co- localize at the correct domain into the scaffold device. The GBD ligand (interaction partner WASP GBP, amino acid sequence:

LVGALMHVMQRSRAIHSSDEGEDQAGDEDED; SEQ ID NO: 88; Kim A.S. et al., 2000, supra) is fused onto the C terminus of Olsl . The SH3 ligand (interaction partner Crk SH3, amino acid sequence: PPPALPPKRRR; SEQ ID NO: 35; Nguyen, J.T. et al., 1998, supra) is fused onto the C terminus of 01s2. The PDZ ligand (Syn PDZ interaction partner, amino acid sequence: GVKESLV; SEQ ID NO: 36; Harris, B.Z. et al., 2001, Dueber, J.E. et al., 2003) is fused onto the C-terminus of 01s3. The strains resulting from Example 8 are used for transformation with plasmid pOLS-JAMl, or with plasmids pOLSl/SCA-JAM2 and pOLS2/pOLS3-JAM3. In case that strains are used which carry the URA3 marker, the marker is recycled before transformation with plasmids pOLS-JAMl, or with plasmids pOLSl/SCA-JAM2 and pOLS2/pOLS3-JAM3. Therefore the strains are cultivated in WMVIII medium supplemented with 100 mg/L of uracil for 48 h to ensure growth of cells, which have lost the URA3 marker by pop-out via homologous recombination. Afterwards the cells are plated on YE -medium agar plates supplemented with 1.5 g/L 5'-FOA (5- fluoroorotic acid) for the selection of uracil auxotrophic cells.

The strains were transformed via the lithium acetate method described by Gietz et al. (1992, supra) and were cultivated on WMVIII (Lang and Looman, 1995, supra) agar plates lacking uracil for the selection of uracil prototrophic strains (pOLS-JAMl transformants) or agar plates lacking uracil and tryptophan for the selection of uracil/ tryptophan prototrophic strains (transformants of pOLSl/SCA-JAM2 and pOLS2/pOLS3-JAM3).

Example 12: Cultivation procedure for strain evaluation (growth and productivity)

The standard cultivation procedure for the strains of the yeast Saccharomyces cerevisiae was:

1) Preculture: 20 ml of WMVIII medium in a 100 ml shaking flask were inoculated with 20 μΐ of the corresponding glycerol stock (strain) and cultivated for 48 h at 30°C and 150 rpm on a rotary shaker..

2) Main culture: 50 ml of WMVIII medium (with 50 g/L xylose as sole carbon source) in a 250 ml shaking flask were inoculated with 1 % (v/v) of the preculture and cultivated for 72 h at 30°C and 150 rpm on a rotary shaker.

Strains with the genetic background of CEN.PK2-1C are auxotrophic for leucine, histidine, uracil and tryptophan. Therefore the medium was supplemented with leucine (400 mg/L), histidine (100 mg/L), uracil (100 mg/L) and tryptophan (100 mg/L) where necessary (to establish selection pressure on plasmids, plasmid containing strains were cultivated in medium lacking the corresponding supplement).

Composition of WMVIII medium (for 1L); according to Lang and Looman, 1995 (except for the substitution of sucrose for xylose): 50 g xylose, 250 mg NH ₄ H ₂ PO ₄ , 2,8 g NH ₄ CI, 250 mg MgCl ₂ x 6H ₂ 0, 100 mg CaCl ₂ x 2H ₂ 0, 2 g KH ₂ P0 ₄ , 550 mg MgS0 ₄ x 7H ₂ 0, 75 mg meso-Inositol and 10 g Na-glutamate.

Weigh out substances above and fill up with 1 L of distilled water. After autoclaving add filter sterilized (0,22 μιη) trace elements, vitamins and amino acids: 1 ml 1000 x trace elements, 4 ml 250 x vitamin solution of stock solutions.

Trace elements: lOOOx concentrated: 1,75 g ZnS0 ₄ x 7H ₂ 0; 0,5 g Fe ₂ S0 ₄ x 7 H ₂ 0; 0,1 g CuS0 ₄ x 5 H ₂ 0; 0,1 g MnCl ₂ x 4 H ₂ 0; 0,1 g NaMo0 ₄ x 2 H ₂ 0, (for one liter).

Vitamin solution: 250x concentrated: 2,5 g Nicotinic acid; 6,25 g Pyridoxine; 2,5 g Thiamine; 0,625 g Biotine; 12,5 g Ca-Pantothenate, (for one liter).

Filter sterilize (0,22 μιη) trace elements and vitamin solution.

Supplementation with leucine (400 mg/L), histidine (100 mg/L), uracil (100 mg/L) and tryptophan (100 mg/L) is carried out, if necessary (note: amino acid are prepared in stock solutions with a concentration of 20 mg/ml and filter sterilized).

Example 13: Metabolite analysis (identification and quantification of n-alcohols and fatty acids)

Identification of n-alcohols was conducted through gas chromatography- mass spectroscopy (GC-MS) following a modification of the method reported by Atsumi (2008). The analysis was performed on an AGILENT(TM) 6890 GC/5973 MS (AGILENT TECHNOLOGIES,(TM) Palo Alto, CA) instrument with a HP-5ms capillary column (30 m x 0.25 mm x 0.25 μιη). For the determination of intra- and extracellular metabolites 2 ml of culture broth (supernatant and cells) were mixed with 0.8 g of glass beads for cell disruption. After vortexing for 10 min and subsequent centrifugation at 14.000 x g for 10 min, 1ml of supernatant was extracted with 500 μΐ of GC standard grade hexane (Fluka). 0.5 μΐ of the extracted sample was injected using a 20: 1 split at 250° C. The oven temperature was initially held at 75° C for 2 min and then raised with a gradient of 5° C/min to 280° C and held for 2 min. Helium (MATHESON TRI-GAS,(TM) Longmont, CO) was used as the carrier gas with a 14- lb/in2 inlet pressure. The injector and detector were maintained at 255° C.

Identification of fatty acids was performed on a SHIMADZU(TM) Auto-System GC 2010 (SHIMADZU,(TM) Japan) equipped with a DB-5MS capillary column (30 m x 0.25 mm x 0.25 μιη) and directly connected to MS. The following method was used: an initial temperature of 50° C was held for 2 min and then ramped to 220° C at 4° C per min and held for 10 min. Extraction and derivatization procedures are described in section Metabolite Quantification.

The quantification of ethanol and butanol was conducted by high-performance liquid chromatography (HPLC). Samples (culture broth after cell disruption with glass beads) were analyzed with ion-exclusion HPLC using a SHIMADZU(TM) Prominence SIL 20 system (SHIMADZU SCIENTIFIC INSTRUMENTS, INC.,(TM) Columbia, MD) equipped with an HPX-87H organic acid column (BIO-RAD,(TM) Hercules, CA) with operating conditions to optimize peak separation (0.3 mL/min flow rate, 30 mM H2S04 mobile phase, column temperature 42° C).

Quantification of longer chain (C > 4) n-alcohols was conducted through gas chromatography (GC) in a VARIAN(TM) CP-3800 gas chromatograph (VARIAN ASSOCIATES, INC.,(TM) Palo Alto, CA) equipped with a flame ionization detector (GC- FID). Sample extraction procedure was as described above in section Metabolite Identification. The separation of alcohol compounds was carried out using a VF-5ht column (15 m, 0.32 mm internal diameter, 0.10 μιη film thickness; VARIAN ASSOCIATES, INC. (TM) Palo Alto, CA). The oven temperature was initially held at 40° C for 1 min and then raised with a gradient of 30° C/min to 130° C and held for 4 min. The temperature was then raised with a gradient of 15° C/min to 230° C and held for 4 min. Helium (1 ml min-1, MATHESON TRI-GAS,(TM) Longmont, CO) was used as the carrier gas. The injector and detector were maintained at 250° C. A 0.5-μ1 sample was injected in splitless injection mode.

Quantification of fatty acids was carried out in a VARIAN(TM) CP-3800 gas chromatograph (VARIAN ASSOCIATES, INC.,(TM) Palo Alto, CA) after hexane- methyl tertiary butyl ether (MTBE) extraction (Lalman 2004) and FA transesterification with a mixture of cholophorm:methanol:hydrochloric acid [10: 1 : 1, vol/vol/vol] as previously reported (Dellomonaco 2010). The resulting fatty acids methyl esters were quantified according to the following method: 50° C held for 1 min, 30° C/min to 160° C, 15° C/min to 200° C, 200° C held for 1.5 min, 10° C/min to 225° C, and 225° C held for 15 min.

Quantification of glycerol: Cell free supernatants of sampled cell suspensions of 1 ml are analyzed for extracellular glycerol by centrifugation at 3500 x g for 5 min and subsequent enzymatic determination of the glycerol content of the supernatant using commercial glycerol analysis kits (Free Glycerol Determination Kit, Sigma- Aldrich).

Determination of cell dry mass: cell dry weight is determined by harvesting 3 x 6 ml of culture after the desired cultivation time. The culture is centrifuged at 3500 x g for 5 min. Afterwards, the cells are washed once with H ₂ 0 and the pellet is weighted by a scale after vacuum-drying for 12 h at 80 °C in a cabinet desiccator and cooling to room temperature in a dehydrator.

Detection of olefin formation is carried out as follows:

Lipid and hydrocarbon extraction: Cells were harvested by centrifugation, resuspended in 3 mL of H ₂ 0 containing 50mg of hexadecane as an internal standard. (Mendez-Perez et al, 2011, Appl Environ Microbiol. 77(12):4264-7) Lipid and hydrocarbons were purified and analyzed by GC-MS as described in Lennen et al. (2010 Biotechnol Bioeng. 106(2): 193-202).:

To 2.5-mL samples of cell culture (three replicates for each culture at each sampling time), 5 mL of lOmg mL-1 heptadecanoic acid (Fluka, Buchs, Switzerland) dissolved in ethanol and 50 mL of 10 mg mL-1 pentadecanoic acid (Acros Organics, Geel, Belgium) dissolved in ethanol were added as internal standards. Next, 100 mL of glacial acetic acid and 5.0mL of a 1 : 1 (v/v) chloroform/methanol mixture were added. The samples were inverted several times, vortexed vigorously, and centrifuged. The aqueous layer and cell debris were removed by aspiration and the chloroform layer was stored at -80°C until further processing (extraction with chloroform/methanol 1 : 1, v/v).

Alternatively, lipids are extracted in a five-step procedure as described by Tronchoni et al. (2012, Int J Food Microbiol. 155(3): 191-8).

Determination of the double bond positions:

DMDS derivation: All chemicals were purchased from Sigma-Aldrich unless otherwise noted. To the dried hydrocarbon extract, ΙΟΟμί of dimethyl disulfide (Acros Organics) and 100 of carbon disulfide containing 6 mg/mL of iodine were added (according to Vicenti et al. 1987, Anal Chem. 59:694). The reaction mixture was kept at 60°C and after 40 hr, the reaction was quenched with aqueous Na ₂ S ₂ 0 ₃ (3xlO ^"4 M, Acros Organics) and the organic phase was then evaporated to dryness under a nitrogen stream and dissolved in hexane to be analyzed by GC-MS (Mendez-Perez et al. 2011, supra). Alternatively, the procedure described by Rude et al. (2011, Appl. Environ. Microbiol. 7: 1718-1727) is used. Permanganate/periodate oxidation: Isolation of the hydrocarbon for permanganate/periodate oxidation was performed by flash chromatography using a Pasteur pipette packed with silica gel (Silia flash P60, 40-63 μιη. 60A, Silicycle). The sample was applied after solvating the silica gel column with 8-10 column volumes of hexane. Five fractions were collected and analyzed by thin layer chromatography using commercial glass plates coated with silica gel. Spots were visualized using a permanganate solution. After identifying fractions containing the desired compound (verified by GC-MS), permanganate/periodate oxidation was used to identify the position of the double bond. Permanganate/periodate oxidation was done as described before (according to Goodloe et al. 1982, Biochimica et Biophysica Acta 710: 485- 492): the sample was dried under nitrogen and 0.3 ml of t-butanol, 0.1 ml of Na ₂ CC"3 (0.02M) and 0.12 ml of NaI0 ₄ /KMn0 ₄ (2.1mg/ml and 1.6 ml/ml respectively) were added to the dried extract. Reaction was maintained at 28°C with shaking for 6 hr. The sample was then decolorized with a few crystals of NaHS0 ₄ (Flinn Scientific) and acidified with 0.025ml of H ₂ S0 ₄ (0.5N, Fisher). 0.5 ml of diethyl ether (Alfa Aesar) was added and vortexed. The top layer was dried under nitrogen stream and methylation was done essentially as described by Lennen et al. (2010, supra) and GC/MS was used to identify the compounds (Mendez-Perez et al, 2011, supra).

Determination of branched point locations of hydrocarbons: Branched point locations of hydrocarbons were determines by the presence of characteristic secondary CnH2n-l fragment ions as described by Koster et al. (1999, Organic Geochemistry, 30 (11): 1367-1379).

Example 14: Fermentations

Table 7 (A and B) presents the results of shake flask fermentations as obtained with the various background strains as indicated, and transformed with plasmids as indicated. In particular Table 7 presents for each of the transformed strains the yields of biomass (cell dry weight) and product of interest (as indicated), as well as the yields of the by-products ethanol and glycerol, after 24 hours of cultivation. The strains were cultivated under the conditions described in Example 12.

Example 15: Fermentation in Labfors fermenter Strain CenPK 2a4 (Aadhl Afpsl Aald6::mhpF URA3) transformed with plasmids pMA-C4-l and pMA-T6-C4 was cultured in a 5L Labfors fermenter. The strain was precultured in 20 ml of WMVIII medium supplemented with 100 mg/L tryptophan in a 100 ml shaking flask were inoculated with 20 μΐ of the corresponding glycerol stock (strain) and cultivated for 48 h at 30°C and 150 rpm on a rotary shaker.

For the main culture 2L of WMVIII medium supplemented with 100 mg/L tryptophan in a 5L Labfors fermenter was inoculated with 40 mL (2 % v/v) of preculture and cultivated for 216 h. For agitation the stirrer speed was kept at 250 rpm and for aeration 0.1 wm were applied.

Metabolites were detected as described in Example 13. The results are presented in Figure 15.

Example 16: Effects of additional deletions of ADH genes

The ADH3 gene, both the ADH3 and ADH5 genes, and both the ADH3 and ADH4 genes, were inactivated by deleted in strain CenPK 2a4 (Aadhl Afpsl Aald6::mhpF URA3). The deleted strains were then transformed with transformed with plasmids pMA-C4-l and pMA-T6-C4 and tested in oxic shake flask fermentation as described in Examples 12. Metabolites were detected as described in Example 13. The results are presented in Table 8.

Deletion of the ADH3 gene leads to an increased yield of product (butanol) and to a decrease in the formation of ethanol by-product as compared to the strain that has only the ADH1 gene deletion. The additional deletion of the ADH4 gene or the ADH5 gene, in a strain that already has ADH1- and ^DHJ-deletions, results in a further increase in the yield of product (butanol). However, the ethanol by-products formation is not influenced by these additional deletions.

Table 8

a see Table 6 for genotypes of the strains b see Table 5 for details of the plasmids

Table 7A

a see Table 6 for genotypes of the strains b see Table 5 for details of the plasmids

Table 7B

a see Table 6 for genotypes of the strains b see Table 5 for details of the plasmids