The present invention relates to a novel method of production of 2,4-dihydroxybutyric acid from malate and/or glyoxylate and/or succinyl-CoA by the implementation of a synthetic pathway that converts malate and/or glyoxylate and/or succinyl-CoA into malylCoA, malylCoA into malate-4- semialdehyde, and then converting said malate-4-semialdehyde into 2,4- dihydroxybutyric acid (2,4-DHB). The carboxylic acids cited within the present application are equally named under their salt (e.g. 2,4-dihydroxyburyrate) or acid forms (e.g. 2,4- dihydroxybutyric acid).

2,4-dihydroxybutyric acid (equally 2,4-DHB or DHB) is a compound of considerable economic interest. DHB can be readily converted into a- hydroxy-y-butyrolactone in aqueous media by adjusting the appropriate pH. o hydroxy-y-butyrolactone is a prominent precursor for the production of the methionine substitute, 2-hydroxy-4-(methylthio)-butyrate (HMTB), (US 2009/318715) which has a large market in animal nutrition. At present, a- hydroxy-y-butyrolactone is derived from γ-butyrolactone by a multi-stage process that implies halogenation of the γ-butyrolactone in position a, and subsequent substitution of the halogen atom by a hydroxyl group in alkaline medium (US 2009/318715).

From growing oil prices, the need for the production of DHB from renewable resources arises. Microorganisms are capable of transforming biomass-derived raw material, e.g. sugars or organic acids, into a large variety of different chemical compounds (Werpy & Petersen, 2004). With the growing body of biochemical and genomic information, it is possible to modify microorganisms such that they overproduce naturally occurring metabolic intermediates with high yield and productivity (Bailey, 1991 ). Optimization of production microorganisms often requires rational engineering of metabolic networks which ensures, among others, overexpression of enzymes required for the biosynthesis of the metabolite of interest, and alleviation of product feedback inhibition. Another possibility is the implementation of novel enzymatic systems that catalyze the production of a metabolite of interest.

Metabolic engineering approaches and enzymatic catalyses require detailed knowledge of the biochemistry and regulation of the metabolic pathway leading to the metabolite of interest. In the case of 2,4-DHB production, this information is not available. Only few studies report the occurrence of 2,4-DHB in patients with succinic semialdehyde dehydrogenase deficiency (Shinka et al., 2002) without, however, identifying enzymatic reactions implicated in DHB production. The zymotic or enzymatic production of 2,4-DHB, therefore, requires (i) the identification of a thermodynamically feasible pathway which transforms an accessible precursor into 2,4-DHB, (ii) the identification or construction of enzymes that are capable to catalyze individual reaction steps in the pathway and (iii) the functional expression of the pathway enzymes in an appropriate production organism.

The present invention has as an objective to satisfy these needs.

Accordingly, one object of the present invention is a method of producing 2,4-DHB comprising a first step of converting malate and/or glyoxylate and/or succinyl-CoA in malyl-CoA, a second step of converting malyl-CoA in malate-4-semialdehyde and, a third step of converting malate-4- semialdehyde in 2,4-DHB. Accordingly, one object of the present invention is a method of producing 2,4-DHB which comprises a first reaction (see Figure 1 ), wherein malate is converted into malyl-CoA by the action of an enzyme which possesses malyl-CoA synthetase activity [1.1], and/or wherein succinyl-CoA is converted into malyl-CoA by the action of an enzyme having a succinyl- CoA:(L)-malate CoA transferase activity [1.2], and/or wherein glyoxylate is converted into malyl-CoA by the action of an enzyme which possesses malyl- CoA lyase activity [1.3]. In the second reaction [2], malyl-CoA is converted into malate-4-semialdehyde by the action of an enzyme which possesses malyl- CoA reductase activity. In the third reaction [3], malate-4-semialdehyde is converted into DHB by the action of an enzyme which possesses DHB dehydrogenase activity. More precisely, reaction [3] is catalysed by an enzyme bearing malate-4-semialdehyde reductase activity in the biosynthetic sense of the pathway. Within another aspect of the invention, the first step of the method of producing 2,4-DHB involves an enzyme having malyl-CoA synthetase (equally named malate thiokinase or malate-coenzyme A ligase (ADP forming), EC 6.2.1.9), succinyl-CoA:(L)-malate CoA transferase (EC 2.8.3.-), or malyl- CoA lyase (EC 4.1.3.24) activity that transforms malate, succinyl-CoA, or glyoxylate, respectively, into malyl-CoA.

Said enzymes have been identified in methylotrophic bacteria that employ the serine cycle for fixation of formaldehyde (Chistoserdova et al., 2009; Smejkalova er a/., 2010; Vuilleumier et al., 2009), in bacteria that rely on acetate assimilation pathways that are independent from the glyoxylate cycle and isocitrate lyase activity (Meister et al., 2005), and in bacteria that employ the 3-hydroxypropionate C02-fixation cycle for autotrophic growth (Zarzycki et al., 2009).

Proteins sharing homology with the above enzymes are also another aspect of the invention such as functional variants or functional fragments.

Malyl-CoA synthetase consists of two subunits, MtkA and MtkB. Therefore, according to the invention, proteins comprising a malyl-CoA synthetase activity designate all polypeptides having at least 30 % of identity with the protein sequences of the malyl-CoA synthetase subunits MtkA and MtkB of M. petroleiphilum (YP 001022444 and YP 001022445) Methylobacter extorquens (YP002962851 and YP 002962852) or two subunits SucC and SucD of M. capsulatus (YP 114179 and YP 114180), preferentially at least 50 % and more preferentially 70 % of identity.

Malyl-CoA lyase is a homohexamer and found in bacteria that do not employ the glyoxylate cycle for acetate assimilation (Meister et al., 2005). Therefore, according to the invention, proteins comprising a malyl-CoA lyase activity designate all polypeptides having at least 30 % of identity with the protein sequences of the malyl-CoA lyase, Mcl, of Methylobacter extorquens, Rhodobacter capsulatus, or Streptomyces coelicolor, preferentially at least 50 % and more preferentially 70 % of identity.

In a further aspect of the invention, the malyl-CoA lyase of the invention is represented by SEQ ID No. 1 or by any variant thereof. Succinyl-CoA:(L)-malate CoA transferase consists of two subunits, SmtA and SmtB (Zarzycki et al., 2009)(Friedmann et al., 2006). Therefore, according to the invention, proteins having a succinyl-CoA:(L)-malate CoA transferase activity designate all polypeptides having at least 30 % of identity with the protein sequences of the succinyl-CoA:(L)-malate CoA transferase subunits SmtA and SmtB of Chloroflexus aurantiacus (represented by SEQ ID No. 191 and SEQ ID No. 193 or encoded by SEQ ID No. 192 and SEQ ID No. 194), preferentially at least 50 % and more preferentially 70 % of identity. More generally, within the meaning of the invention the identity between two protein sequences can be determined by methods well known by the skilled man in the art. Examples of such methods are the use of the CLUSTALW (Larkin et al., 2007) software

(http://www.ebi.ac.uk/Tools/msa/clustalw2/ with the default parameters indicated on the website) or the BLAST alignment program (http://blast.ncbi.nlm.nih.gov/Blast.cgi with the default parameters indicated on the website).

The term functional variant encompasses enzymes that may present substantial sequence modifications when compared to the sequences specifically described within the present application but that still retain the original enzymatic activity.

The term functional fragment, according to the invention, means that the sequence of the enzyme may comprise less amino acids than the original one but said truncated enzyme still retains the original enzymatic activity.

Improvement of said enzymes can be obtained by at least one mutation, said mutation(s) improving the activity and/or substrate affinity of the mutated enzyme for malate, succinyl-CoA, or glyoxylate respectively. Within the present invention, the expression "improve the activity and/or substrate affinity" means that the enzyme before mutation was either

- unable to use the substrate, and/or

- synthesized the product of the reaction at a maximum specific rate at least three times lower, and/or - had an affinity for malate, succinyl-CoA or glyoxylate, malyl-CoA or malate-4-semialdehyde that was at least two more preferably three times lower. The malyl-CoA synthetase and the malyl-CoA lyase activities can be measured by the enzymatic tests described by (Smejkalova et al., 2010) or (Meister et al., 2005), respectively. The succinyl-CoA:(L)-malate CoA transferase activity can be measured as described by (Friedmann et al., 2006) Within a still further aspect, the second step of the method of producing 2,4-DHB according to the invention involves an enzyme having malyl-CoA reductase activity characterized in that it transforms malyl-CoA into malate-4-semialdehyde.

Said enzyme can be identified among enzymes having malonyl- CoA reductase, a succinyl-CoA reductase or reported 3-hydroxy-3- methylglutaryl-CoA (HMG-CoA) reductase, cinnamoyl-CoA reductase, or acetaldehyde dehydrogenase activity, or they can be obtained by modification of said enzymes.

Malonyl-CoA reductase (EC 1.2.1.75) and succinyl-CoA reductase (EC 1.2.1.76) were found in bacteria that possess a modified 3- hydroxypropionate cycle for carbon dioxide fixation (Alber et al., 2006; Kockelkorn & Fuchs, 2009), and in bacteria that employ an anaerobic succinate degradation pathway (Seedorf et al., 2008; Sohling & Gottschalk, 1993). HMG- CoA reductase (EC 1.1.1.38, EC 1.1.1.88) is part of the biosynthetic pathway of isoprenoids in eukaryotes and some bacteria. Cinnamoyl-CoA reductase (EC 1.2.1.44), is an enzyme implicated in lignin biosynthesis (Kawasaki et al., 2006). Acetaldehyde dehydrogenase (EC 1.2.1.10) is found in a large variety of bacteria and catalyses the entry into the ethanol producing pathway or the detoxification of acetaldehyde.

Within a further aspect of the invention, the malyl-CoA reductase is represented by ID No. 7, or SEQ ID No.10 or by any functional variant thereof or any functional fragment thereof. Therefore, according to the invention, proteins having a malonyl- CoA reductase activity designate all polypeptides having at least 30 % of identity with the protein sequences of the Sulfolobus tokodaii malonyl-CoA reductase, Mcr (SEQ ID No. 7). Preferentially they have at least 50 % and more preferentially 70 % of identity.

The malonyl-CoA reductase of Chloroflexus auranthiacus (SEQ ID No.189 encoded by SEQ ID No. 190) constitutes another aspect of the invention. Polypeptides having at least 30 % of identity with the protein sequences of Chloroflexus auranthiacus are also part of the invention. Preferentially they have at least 50 % and more preferentially 70 % of identity.

Therefore, according to the invention, proteins having a succinyl- CoA reductase activity designate all polypeptides having at least 30 % of identity with the protein sequences of the Porphyromonas gingivalis succinyl- CoA reductase, SucD (SEQ ID No. 10), or with the bifunctional S. tokodaii malonyl-CoA and succinyl-CoA reductase, Mcr (SEQ ID No. 7). Preferentially they have at least 50 % and more preferentially 70 % of identity.

The malyl-CoA reductase activity can be measured by the enzymatic test described in Example 2 (see "Enzymatic assay").

This enzyme activity can be improved by at least one mutation of an enzyme, said mutation(s) improving the activity and/or substrate affinity of the mutated enzyme for malyl-CoA or decreasing its activity on the natural substrate.

The present invention also encompasses modified malyl-CoA reductase having improved activities. The malyl-CoA reductase according to the invention corresponds in a specific aspect to Sulfolobus tokodaii malonyl-CoA reductase comprising at least one mutation when compared to the wild type enzyme in at least one of the positions P111 , L152, T154, L202, G203, D204, Y206, D207, K209, T210, T238.T239, D295, R318, wherein the naturally occurring amino acid in said positions is replaced by anyone of the other 19 naturally existing proteinogenic amino acids, that is by either alanine, arginine, asparagine, aspartic acid, cysteine, glutamic acid, glutamine, glycine, histidine, isoleucine, leucine, lysine, methionine, phenylalanine, proline, serine, threonine, tryptophan, tyrosine, or valine. Within another aspect, the third step of the method of producing

2,4-DHB according to the invention involves a DHB dehydrogenase characterized in that it transforms malate-4-semialdehyde into 2,4-DHB, said enzyme bearing malate-4-semialdehyde reductase activity in the biosynthetic sense of the pathway.

Candidate DHB dehydrogenase enzymes that potentially already possess DHB dehydrogenase activity can be chosen from the class of beta- hydroxyacid dehydrogenases that act on C3, C4, or C5 compounds. According to a still further aspect of the invention, said DHB dehydrogenase enzymes can be structurally and mechanistically related to β- hydroxyacid dehydrogenases such as tartronate semialdehyde reductases, succinate semialdehyde reductases, 4-hydroxybutyrate dehydrogenases, malonate semialdehyde reductases, methylbutyraldehyde reductases, zinc- type alcohol dehydrogenases, L-threonine-3-dehydrogenases, cinnamyl alcohol dehydrogenases, alcohol dehydrogenases, or homoserine dehydrogenases.

The present invention also deals with the use of a methylbutyraldehyde reductase, or of a succinic semialdehyde reductase (equally named as 4-hydroxybutyrate dehydrogenase), or of an alcohol dehydrogenase, to transform malate-4-semialdehyde in 2,4-DHB.

In another specific aspect of the invention, the DHB dehydrogenase corresponds to methylbutyraldehyde reductase (Ypr1 ) of S. cerevisiae, the succinic semialdehyde reductase of M. sedula, the 4- hydroxybutyrate dehydrogenase (4hbd) of P. gingivalis, or to the alcohol dehydrogenase (YqhD) of Escherichia coli.

In specific embodiments, said methylbutyraldehyde reductase is represented by SEQ ID No. 14, said succinic semialdehyde reductase is represented by SEQ ID No. 16, said 4-hydroxybutyrate dehydrogenase is represented by SEQ ID No. 187, said alcohol dehydrogenase is represented by SEQ ID No. 185. The DHB dehydrogenase activity can be measured by the enzymatic test described in Example 3 (see "Enzymatic assay"). The affinity of DHB dehydrogenase for malate-4-semi aldehyde can be increased by at least one mutation of an enzyme, said mutation(s) increasing the activity and/or substrate affinity of the mutated enzyme for malate-4-semialdehyde, and/or reducing the activity or affinity for its natural substrate by at least factor 2.

The DHB dehydrogenase according to the invention corresponds in a specific aspect to M. sedula succinic semialdehyde reductase (SEQ ID No. 16) comprising at least one mutation when compared to the wild type enzyme in at least one of the positions S40, N43, H39 T49, F85, Q108, L281 and N305 wherein the naturally occurring amino acid in said positions is replaced by anyone of the other 19 naturally existing proteinogenic amino acids, that is by either alanine, arginine, asparagine, aspartic acid, cysteine, glutamic acid, glutamine, glycine, histidine, isoleucine, leucine, lysine, methionine, phenylalanine, proline, serine, threonine, tryptophan, tyrosine, or valine.

As demonstrated in a non-exclusive example, the affinity of M. sedula succinic semialdehyde reductase for (L)-malate-4-semialdehyde was increased by introducing the double mutation H39R N43H by site-directed mutagenesis, as represented by SEQ ID No. 36. Simple mutants H39R (SEQ ID No. 32) and N43H (SEQ ID No. 34) are also encompassed by the present invention (Example 5).

. DHB dehydrogenase can be used to transform malate-4- semialdehyde into 2,4-DHB, which constitutes a further aspect of the invention.

The nucleic acid sequence of genes can be adapted to the codon usage of the host organism thereby increasing the production of the heterologously expressed proteins. This constitutes a further aspect of the invention.

The synthesis of a synthetic gene coding for M. sedula succinic semialdehyde reductase H39R N43H whose nucleotide sequence was optimized for the expression of said enzyme in E. coli as represented by SEQ ID No. 38 is a further aspect of the invention.

In a still further aspect, the present invention also deals with nucleic acids, and more particularly with isolated nucleic acid sequences encoding malyl-CoA synthetase.

In a still further aspect, the present invention deals with isolated nucleic acid sequences encoding malyl-CoA lyase and more specifically by SEQ ID No.2.

In a still further aspect, the present invention deals with isolated nucleic acid sequences encoding malyl-CoA reductase and more specifically by SEQ ID No. 8, SEQ ID No.11 , or SEQ ID No.190.

In a still further aspect, the present invention also deals with isolated nucleic acid sequences encoding a DHB dehydrogenase as described above.

In another aspect, said nucleic acid is represented by SEQ ID No. 15, SEQ ID No. 17, SEQ ID No. 37,SEQ ID No. 186 or SEQ ID No. 188.

In accordance with this invention, a "nucleic acid sequence" refers to a DNA or RNA molecule in single or double stranded form, preferably a DNA molecule. An "isolated DNA", as used herein, refers to a DNA which is not naturally-occurring or no longer in the natural environment wherein it was originally present, e.g., a DNA coding sequence associated with other regulatory elements in a chimeric gene, a DNA transferred into another host cell, or an artificial, synthetically-made DNA sequence having a different nucleotide sequence compared to any naturally-occurring DNA sequence.

The present invention also relates to a chimeric gene comprising, functionally linked to one another, at least one promoter which is functional in a host organism, a polynucleotide encoding anyone of the malyl-CoA synthetase or malyl-CoA lyase, malyl-CoA reductase, malonyl-CoA reductase, succinyl- CoA reductase or DHB dehydrogenase activities as defined according to the invention, and a terminator element that is functional in the same host organism. The various elements which a chimeric gene may contain are, firstly, elements regulating transcription, translation and maturation of proteins, such as a promoter, a sequence encoding a signal peptide or a transit peptide, or a terminator element constituting a polyadenylation signal and, secondly, a polynucleotide encoding a protein. The expression "functionally linked to one another" means that said elements of the chimeric gene are linked to one another in such a way that the function of one of these elements is affected by that of another. By way of example, a promoter is functionally linked to a coding sequence when it is capable of affecting the expression of said coding sequence. The construction of the chimeric gene according to the invention and the assembly of its various elements can be carried out using techniques well known to those skilled in the art. The choice of the regulatory elements constituting the chimeric gene depends essentially on the host organism in which they must function, and those skilled in the art are capable of selecting regulatory elements which are functional in a given host organism. The term "functional" is intended to mean capable of functioning in a given host organism.

The promoters which the chimeric gene according to the invention may contain are either constitutive or inducible. By way of example, the promoters used for expression in bacteria may be chosen from the promoters mentioned below. For expression in Escherichia coli mention may be made of the lac, trp, Ipp, phoA, recA, araBAD, prou, cst-l, tetA, cadA, nar, tac, trc, Ipp- lac, Psyn, cspA, PL, PL-9G-50, PR-PL, T7, [lambda]PL-PT7, T3-lac, T5-lac, T4 gene 32, nprM-lac, VHb and the protein A promoters or else the Ptrp promoter (WO 99/64607). For expression in Gram-positive bacteria such as Corynebacteria or Streptomyces, mention may be made of the PtipA or PS1 and PS2 (FR91/09870) promoters or those described in application EP0629699A2. For expression in yeasts and fungi, mention may be made of the K. lactis PLAC4 promoters or the K. lactis Ppgk promoter (patent application FR 91/05294), the Trichoderma tefl or cbhl promoter (WO 94/04673), the Penicillium his, csl or apf promoter (WO 00/68401 ) and the Aspergillus gla promoter.

According to the invention, the chimeric gene may also comprise other regulatory sequences, which are located between the promoter and the coding sequence, such as transcription activators (enhancers).

As such, the chimeric gene of the invention comprises, in a specific embodiment at least, in the direction of transcription, functionally linked, a promoter regulatory sequence which is functional in a host organism, a nucleic acid sequence encoding the malyl-CoA synthetase, and/or the succinyl- CoA:(L)- malate-CoA transferase, and/or the malyl-CoA lyase, the malyl-CoA reductase and the DHB dehydrogenase of the invention and a terminator regulatory sequence which is functional in said host organism.

The present invention also relates to a cloning and/or expression vector comprising a chimeric gene according to the invention or a nucleic acid sequence of the invention. The vector according to the invention is of use for transforming a host organism and expressing in this organism anyone of the malyl-CoA synthetase, and/or the succinyl-CoA:(L)- malate CoA transferase, and/or the malyl-CoA lyase, the malyl-CoA reductase and/or DHB dehydrogenase. This vector may be a plasmid, a cosmid, a bacteriophage or a virus. Preferentially, the transformation vector according to the invention is a plasmid. Generally, the main qualities of this vector should be an ability to maintain itself and to self-replicate in the cells of the host organism, in particular by virtue of the presence of an origin of replication, and to express anyone of the malyl-CoA synthetase, and/or the succinyl-CoA:(L)- malate CoA transferase and/or the malyl-CoA lyase, the malyl-CoA reductase and/or DHB dehydrogenase therein. For the purpose of stable transformation of a host organism, the vector may also integrate into the genome. The choice of such a vector, and also the techniques of insertion of the chimeric gene according to the invention into this vector and are part of the general knowledge of those skilled in the art. Advantageously, the vector used in the present invention also contains, in addition to the chimeric gene according to the invention, a chimeric gene encoding a selectable marker. This selectable marker makes it possible to select the host organisms which are effectively transformed, i.e. those which incorporated the vector. According to a particular embodiment of the invention, the host organism to be transformed is a bacterium, a yeast, a fungus. Among the selectable markers which can be used, mention may be made of markers containing genes for resistance to antibiotics, such as, for example, the hygromycin phosphotransferase gene. Other markers may be genes to complement an auxotrophy, such as the pyrA, pyrB, pyrG, pyr4, arg4, argB and trpC genes, the molybdopterin synthase gene or that of acetamidase. Mention may also be made of genes encoding readily identifiable enzymes such as the GUS enzyme, or genes encoding pigments or enzymes regulating the production of pigments in the transformed cells. Such selectable marker genes are in particular described in patent applications WO 91/02071 , WO 95/06128, WO 96/38567 and WO 97/04103.

The present invention also relates to transformed host organisms containing at least one chimeric gene according to the invention, either integrated into their genome or carried on an extra-chromosomal genetic element, for example a plasmid. In a more specific aspect of the invention, the transformed host organism comprises a nucleic acid of the invention encoding a polypeptide having malyl-CoA synthetase activity, and/or succinyl-CoA:(L)- malate-CoA transferase and/or malyl-CoA lyase activity or a chimeric gene comprising a nucleic acid encoding a polypeptide having malyl-CoA synthetase activity, and/or succinyl-CoA:(L)- malate CoA transferase and/or a malyl-CoA lyase activity or an expression vector comprising a nucleic acid encoding a polypeptide having malyl-CoA synthetase activity, and/or succinyl-CoA:(L)- malate CoA transferase and/or a malyl-CoA lyase activity, and/or a nucleic acid encoding a polypeptide having malyl-CoA reductase activity, or a chimeric gene comprising a nucleic acid encoding a polypeptide having malyl-CoA reductase activity or an expression vector comprising a nucleic acid encoding a polypeptide having malyl-CoA reductase activity, and/or a nucleic acid encoding a polypeptide having DHB dehydrogenase activity, a chimeric gene comprising a nucleic acid encoding a polypeptide having DHB dehydrogenase activity or an expression vector comprising a nucleic acid encoding a polypeptide having DHB dehydrogenase activity. The activity of heterologously enzymes in the host organism is often limited by their poor solubility and the formation of inclusion bodies. Therefore, the present invention also relates to chimeric proteins in that a functional enzyme is physically fused to another protein or peptide (equally named fusion protein) in order to increase the activity of said enzyme upon expression in the host organism. Such fusion proteins are known in the art and are commonly selected among the following non-exclusive examples: maltose binding protein, Mbp, thioredoxin, ThrX, glutathione-S-transferase, Gst, transcription termination factor, NusA. The term "host organism" is intended to mean any lower monocellular organism into which the chimeric gene(s), nucleic acid(s) or vector(s) according to the invention may be introduced in order to produce 2,4- DHB. Preferably, the host organism is a microorganism, in particular a fungus, for example of the Penicillium, Aspergillus and more particularly Aspergillus flavus, Chrysosporium or Trichoderma genus, a yeast, in particular of the Saccharomycetaceae, Pichiaceae or Schizosaccharomycetaceae, most preferentially Saccharomyces cerevisiae, Schizosaccharomycespombe, Kluyveromyceslactis, Kluyveromyces marxianus, or Pichia adinii, Pichia stipitis or Pichia pastoris, a bacterium, preferentially selected among Enterobacteriaceae, Clostridiaceae, Bacillaceae, Streptomycetaceae, Streptococcaceae, Methylobacteriacae, and Corynebacteriaceae, most preferentially Escherichia coli, Bacillus subtilis, Corynebacterium glutamicum, Clostridium acetobutylicum, Methylobacterium extorquens, or Lactococcus lactis.

The host organism can be a host organism that naturally overproduces malate or succinate from sugars such as glucose or a host organism that was engineered to overproduce malate or succinate from sugars such as glucose and in which all potential membrane transporters that facilitate export of organic acids, such as malate, pyruvate, succinate, and fumarate have been deleted. The host organism can be an organism that was engineered to overproduce DHB and in which membrane transporters that facilitate export of organic acids such as malate, pyruvate, succinate, and fumarate have been deleted. Examples of permeases that facilitate export of malate and other organic acids are Mae1 from Schizosaccharomyces pombe (Camarasa et al., 2001 )(Grobler et al., 1995), DctA from Bacillus subtilis (Groeneveld et al., 2010), Dct 1 -4 from coli, Jen1 from S. cerevisiae (Akita et al., 2000). For an expert, it will be possible to identify candidate permeases in E. coli based on sequence identity. These constructions will serve to keep malate and other organic acids inside the cell to make them available for DHB production.

The expression "transformed host organism" is intended to mean a host organism which has incorporated into its genome, or in an extra chromosomal genetic element, for example a plasmid, at least one chimeric gene according to the invention, and consequently produces any one of malyl- CoA synthetase, malyl-CoA lyase, malyl-CoA reductase and/or DHB dehydrogenase in cell interior or in a culture medium. To obtain the host organisms according to the invention, those skilled in the art may use one of the many known transformation methods.

One of these methods consists in bringing the cells of the host organisms to be transformed into contact with polyethylene glycol (PEG) and with the vectors according to the invention. Electroporation is another method, which consists in subjecting the cells to be transformed and the vectors of the invention to an electric field. Another method consists in directly injecting the vectors into the cells or the tissues by microinjection. The "biolistic" method may be used. It consists in bombarding cells or tissues with particles onto which the vectors of the invention are adsorbed (U.S. Pat. No. 4,945,050).

Several methods for transforming bacteria are described in the literature for Escherichia coli and other Gram-negative bacteria. Conjugation may also be used. For Gram-positive bacteria, electroporation may be used, and also protoplast transformation, in particular for bacteria of the Streptomyces genus. Several methods for transforming fungi are also described in the literature. Protoplast transformation with PEG is described for Aspergillus in EP 0260762, and an adaptation of this method to the species Penicillium funiculosum is described in WO 00/36120. Transformation by restriction enzyme mediated integration, or REMI, is also known, as is protoplast transformation using bacteria of the Agrobacterium genus. Techniques for transforming yeasts are also described in the literature,

In a further aspect, the invention deals with a process of production of 2,4- DHB comprising the step of cultivating a transformed microorganism of the invention.

For the production of DHB various carbohydrates could be utilized individually or as a mixture such as glucose, fructose, sucrose, molasses, maltose, blackstrap molasses, starch hydrolysate (glucose, oligosaccharides), lactose, maltose, starch and starch hydrolysates, cellulose, cellulose hydrolysate, glycerol, acetate and certain hydrocarbons, oils and fats such as soy bean oil, sunflower oil, groundnut oil and coconut oil as well as fatty acids such as e.g. palmitic acid, stearic acid and linoleic acid. Those substances may be used individually or as mixtures.

Various sources of nitrogen could be utilized individually or as mixtures for the commercial and pilot scale production, including inorganic compounds such as gaseous and aqueous ammonia, ammonium salts of inorganic or organic acids such as ammonium sulphate, ammonium nitrate, ammonium phosphate, ammonium chloride, ammonium acetate and ammonium carbonate. Alternatively, natural nitrogen containing organic materials like soybean-hydrolysate, soy protein HCI-hydrolysate (total nitrogen of about 7%), soy bean meal, soybean cake hydrolysate, corn steep liquor, casein hydrolysate, yeast extract, meat extract, malt extract, urea, peptones and amino acids may also be utilized

The production process can be carried out under aerobic, anaerobic, and oxygen limited conditions. It can be carried out as a fed-batch process or a batch process.

Said production of 2,4-DHB can be made by cultivating the host organism in media where malate (or another organic acid such as pyruvate, succinate, or fumarate) was added alone or together with another carbon source that ensures growth. Malate (and other organic acids) can be added either directly, or by designing a two-stage fermentation process where malate (or other organic acids) is produced in a first process stage by a malate- overproducing microorganism, and 2,4-DHB production is realised in the following stage by a host organism according to the invention.

Product separation and purification is very important factor enormously affecting overall process efficiency and product costs. Methods for product recovery commonly comprise the steps cell separation, as well as product purification, concentration and drying, respectively.

Cell separation

Ultra filtration and centrifugation can be used to separate cells from the fermentation medium. Cell separation from fermentation media is often complicated by high medium viscosity. Therefore, we can add additives such as mineral acids or alkali salts, or heating of the culture broth to optimize cell separation. Product recovery

A variety of ion-exchange chromatographic methods can be applied for the separation of DHB either before or after biomass removal. They include the use of primary cation exchange resins that facilitate separation of products according to their isoelectric point. Typically, the resin is charged with the solution, and retained product is eluted separately following increase of pH (e.g. by adding ammonium hydroxide) in the eluent. Another possibility is the use of ion-exchange chromatography using fixed or simulated moving bed resins. Different chromatographic steps may have to be combined in order to attain adequate product purity. Those purification methods are more economical compared with a costly crystallization step, also providing additional advantages and flexibility regarding the form of final product.

Product concentration and drying

The purification process can also comprises a drying step which may involve any suitable drying means such as a spray granulator, spray dryer, drum dryer, rotary dryer, and tunnel dryer. Concentrated DHB solutions can be obtained by heating fermentation broths under reduced pressure by steam at 130°C using a multipurpose concentrator or thin film evaporator.

Efficient production of DHB can be ensured by optimizing carbon flux repartitioning in the metabolic network of the host organism and by ensuring sufficient NADPH and ATP supply for the three enzymes of the DHB pathway. Channeling of carbon flux into a desired metabolic pathway and supply of NAD(P)H cofactor is commonly facilitated by deleting or alleviating competing natural fermentative pathways. Nonexclusive examples are

the optimization of malate production in S. cerevisiae by impeding the formation of ethanol (by the deletion of pyruvate decarboxylases (Zelle er a/., 2008)(Zelle er a/., 2010). - the optimization of succinate or malate production in E. coli by impeding the formation of lactate (e.g. deletion of IdhA), the formation of acetate (e.g. deletion of pta, ackA), the formation of ethanol (e.g. deletion of adhE), the formation of formate (e.g. deletion of pfIB, focA), the oxidation of pyruvate (e.g. deletion of poxB), the degradation of malate (deletion of maeB and scfA), the formation of succinate (e.g. deletion of frdBC), the formation of methylglyoxal (deletion of mgsA) (Jantama et al., 2008a)(Jantama et al., 2008b)(Lin et al., 2005)(Zhang et al., 201 1 )(Sanchez et al., 2005).

the deletion of phosphoglucose isomerase, pgi, to channel carbon flux across the pentose phosphate pathway thereby increasing

NADPH availability for biosynthetic reactions (Auriol et al., 2011 ).

Another possibility to increase carbon flux and ATP supply for the production of organic acids is the engineering of the phosphoenolpyruvate (PEP)/pyruvate/oxaloacetate branch node (reviewed in (Sauer & Eikmanns, 2005)). Nonexclusive examples for metabolic engineering strategies that ensure the increase of carbon flux from phosphoenolpyruvate to oxaloacetate are:

the optimization of malate production in S. cerevisiae by impeding the function of pyruvate kinase and increasing the activity of PEP carboxykinase (Zelle et al., 2010).

the optimization of succinate production in E. coli by increasing the activity of natural or heterologously expressed PEP carboxylase, PEP carboxykinase, or pyruvate carboxylase (Millard et al., 1996)(Sanchez et al., 2005)(Zhang et al., 2011 ).

Another possibility to increase carbon flux and ATP supply for the production of organic acids in E. coli and other bacteria employing the PEP- consuming phosphotransferase system (PTS) for the initial phosphorylation step of glucose is the deletion of essential components of the PTS system (for example ptsl or ptsG) (Lin et al., 2005)(Zhang et al., 2009). To ensure further glucose uptake in mutants carrying deleterious mutations of the PTS system, the activity of alternative glucose uptake systems (e.g. GalP) has to be ensured. Another possibility to increase carbon flux into the desired pathways for the production of organic acids is the engineering of the citric acid and glyoxylate cycle. Non-exclusive examples are

the optimization of succinic acid production in E. coli by increasing the activity of isocitrate lyase (deletion of transcriptional repressor iclR) (Lin et al., 2005)(Sanchez et al., 2005)(Lin et al., 2005;

Sanchez et al., 2005a). the optimization of succinic acid production by the deletion of isocitrate dehydrogenase, and/or succinate dehydrogenase (Lin et al., 2005).

Another possibility to increase the availability of malate, glyoxylate and acetyl-CoA, which are the substrates of the entry reactions into the DHB- producing pathways, is the attenuation of aspartate transaminase (aspC, tyrB), fumarase (fumABC) , fumarate reductase (frdBC), malate synthase (aceS) and glyoxylate reductase (ghrAB) enzymes.

In another metabolic setting it is possible to produce the 2,4-DHB precursor malate exclusively via the Krebs cycle and the glyoxylate shunt. This setting requires deletion of the cytosolic and membrane bound malate dehydrogenases, mdh and mqo, respectively. The approach largely avoids potential leakage of carbon flux into aspartate and its derivatives.

Another possibility to increase carbon flux into the desired pathways for the production of 2,4-DHB is the expression of appropriate pyruvate dehydrogenases and citrate synthases in the production organism. Natural pyruvate dehydrogenase and citrate synthase of E. coli are inhibited by high intracellular NADH concentrations rendering these enzymes less active under anaerobic conditions. In E. coli, the expression of a pyruvate dehydrogenase mutant that is insensitive to NADH resulted in the overproduction of acetyl-CoA under anaerobic conditions and modified carbon flux repartitioning between the fermentative end-products (acetate, lactate, ethanol, formate, and pyruvate) (Wang et al., 2010). The heterologous expression of the Bacillus subtilis citrate synthase which is insensitive to NADH increased succinic acid production in engineered E. coli strains (Sanchez et al., 2005). In combination with the above described mutations, the use of the appropriate pyruvate dehydrogenases and citrate synthases (NADH sensitive or insensitive) enables the tuning of carbon flux repartitioning between glyoxylate and citric acid cycle reactions and fermentative pathways under anaerobic and aerobic conditions.

Another possibility to increase carbon flux through the DHB pathway is the deletion of enzymatic reactions that may degrade the pathway intermediates malyl-CoA, or 4-malate semialdehyde. Candidate enzymes that may degrade malate semialdehyde are- succinic semialdehyde dehydrogenases (sad, gabD), and other dehydrogenases that are able to oxidize short and medium carbon chain molecules with terminal aldehyde groups. Furthermore, it is known that malyl-CoA may be degraded by citrate synthase.

Another possibility to increase 2,4-DHB productivity of the host organism is the deletion of metabolic reactions that degrade 2,4-DHB. 2,4-DHB is a competitive inhibitor of malic enzyme, thus, having comparatively high affinity for the active site of this enzyme (Rognstad & Katz, 1979). Therefore, 2,4-DHB may be recognized by other enzymes and potentially degraded. These enzymes can be identified and deleted from the host organism.

When 2,4-DHB production is based on addition of malate or other organic acids, the 2,4-DHB-producing microorganisms should functionally express a membrane transport protein that facilitates uptake of malate (or other organic acids such as pyruvate, succinate, etc).

The transformed host organisms of the invention may further contain an additional pathway of synthesizing 2,4-DHB, said host organism comprises at least one chimeric gene, either integrated into their genome or carried on an extra-chromosomal genetic element, for example a plasmid encoding a malate kinase or a chimeric gene comprising a nucleic acid encoding a malate kinase or an expression vector comprising a nucleic acid encoding a malate kinase, and/or a nucleic acid encoding a malate semialdehyde dehydrogenase, or a chimeric gene comprising a nucleic acid encoding a malate semialdehyde dehydrogenase or an expression vector comprising a nucleic acid encoding a malate semialdehyde dehydrogenase, and/or a nucleic acid encoding a DHB dehydrogenase, a chimeric gene comprising a nucleic acid encoding a DHB dehydrogenase or an expression vector comprising a nucleic acid encoding a DHB dehydrogenase. Said enzymes are described in the International patent application WO 2012/056318.

The following examples illustrate the invention. These examples are for purposes of illustration only and are not to be construed as limiting the scope of the invention in any manner. Brief description of the figures

Figure 1 : (i) Reaction scheme that describes the conversion of (L)-malate, succinyl-CoA, or glyoxylate into (L)-2,4-dihydroxybutyrate (2,4-DHB).

Figure 2: Figure shows (top graph) activity on malate semialdehyde, (middle graph) activity on succinic semialdehyde, (lower graph) changes of enzyme specificity compared to the wild-type enzyme expressed as the logarithm of the ratio of mutant activity on malate semialdehyde and succinic semialdehyde over the ratio of wild type activity on malate semialdehyde and succinic semialdehyde. (positive values indicate changes of specificity in favour of malate semialdehyde). Figure 3: Chromatograms showing the presence of 2,4-DHB after incubation of 2 mM acetyl-CoA, 2 mM glyoxylate, and 2 mM NADPH, with different combinations of DHB pathway enzymes (Reaction 1 : malyl-CoA lyase (150 pg/mL Me-Mcl), malyl-CoA reductase (100 pg/mL St-Mcr), and malate semialdehyde reductase (100 pg/mL Ms-SSAred H39R N43H); Reaction 2:same as reaction 1 but using 100 pg/mL Pg-SucD as malyl-CoA reductase; Control 1 : same as reaction 1 but without malyl-CoA reductase; Control 2: same as reaction 1 but without malate semialdehyde reductase.)

Examples

Example 1: Demonstration of malyl-CoA lyase activity

Construction of plasmids containing wild-type genes coding for malyl- CoA lyase: The DNA sequences of the mcl genes coding for malyl-CoA lyase in M. extorquens (Arps et a/., 1993) and Rhodobacter capsulatus (Meister et a/., 2005)were optimized for the expression in Escherichia coli using the GENEius software (Eurofins). The optimized sequences were synthesized by Eurofins MWG OPERON® adding Nhe\ and Ecofil restriction sites upstream of the start codon and downstream of the stop codon of mcl, respectively, which allowed direct cloning of the synthesized DNA fragments into the pET28a+ vector (Novagen) using T4 DNA ligase (Biolabs). Ligation products were transformed into E. coli DH5a cells, amplified, and the plasmids pET28-Mex-mcl (expressing the malyl-CoA lyase from M. extorquens) and pET28-Rca-mcl (expressing the malyl-CoA lyase from R. capsulatus) were isolated using standard genetic protocols (Sambrook ef a/., 1989). NCBI and Integrated Genomics references of the utilized mcl protein sequences, and the references for the corresponding natural and synthetic DNA sequences are listed in Table 1 .

Table 1 : References to proteins from different organisms having annotated malyl-CoA lyase activity, and references to natural and optimized DNA sequences.

Expression of enzymes: E. coli BL21 (DE3) cells were transformed with the appropriate plasmids using standard genetic protocols (Sambrook et al., 1989). Enzymes with an N-terminal hexa-His tag were expressed in 250 ml_ LB cultures that were inoculated from an overnight culture at OD ₆ oo of 0.1 and grown to OD ₆ oo of 0.6 before protein expression was induced by addition of 1 mM isopropyl β-D-l-thiogalactopyranoside (IPTG) to the culture medium. After 3 h of protein expression, cells were harvested by centrifugation at 13000 g for 10 min and the supernatant is discarded. Cell pellets were stored at - 20°C until further analysis. Growth and protein expression were carried out at 37 °C. Culture media contained 50pg/mL kanamycin.

Purification of enzymes: Frozen cell pellets of expression cultures were resuspended in 0.5 mL of breakage buffer (50 mM Hepes, 300mM NaCI, pH 7,5) and broken open by four successive rounds of sonication (sonication interval: 20 sec, power output: 30 %, sonicator: Bioblock Scientific, VibraCell™ 72437). Cell debris were removed by centrifuging the crude extracts for 15 min at 4°C at 13000g and retaining the clear supernatant. RNA and DNA were removed from the extracts by adding 15 mg/mL streptomycin (Sigma), centrifuging the samples at 13000 g for 10 min at 4 °C and retaining the supernatant. Clear protein extract was incubated for 20 min at room temperature (1 h at 4 °C) with 0.3 (0.75 mL) (bed volume) of Talon™ Cobalt affinity resin (Clontech). The suspension was centrifuged at 700 g in a table top centrifuge and supernatant was removed. The resin was washed with 10 bed volumes of wash buffer (50 mM Hepes, 300 mM NaCI, 15 mM Imidazole, pH 7,5) before proteins were eluted with 0.5 mL of elution buffer (50 mM Hepes, 300 mM NaCI, 200 mM Imidazole, pH 7,5). Purity of eluted enzymes was verified by SDS-PAGE analysis. Protein concentrations were estimated with the method of Bradford. Enzymatic assays: Malyl-CoA lyase activity was assayed using a method adapted from (Meister et a/., 2005). Malyl-CoA synthesis by malyl-CoA lyase was coupled to the citrate synthase-catalyzed release of coenzyme A which was monitored by its spontaneous reaction with DTNB. Reaction scheme

Malyl-CoA lyase:

acetyl-CoA + glyoxylate ->(L)-malyl-CoA Citrate synthase:

(L)-malyl-CoA ->(L)-malate + Coenzyme A

Spontaneous :

coenzyme A + DTNB -> CoA-DTNB disulfide

The reaction mixture according to Assay 1 contained 50 mM MOPS/KOH (pH 7.5), 0.25 mM DTNB, 5 mM MgCI ₂ , 1 mM acetyl-CoA, 20 U/mL citrate synthetase (all products from Sigma), and appropriate amounts of purified malyl-CoA lyase or cell extract. Reactions were started by adding 10 mM glyoxylate. Enzymatic assays were carried out at 37 °C in 96-well flat bottomed microtiter plates in a final volume of 250 pL. The reactions were followed by the characteristic absorption of DNTB at 412 nm (EDNTB+COA = 13.6 m ^'1 cm ^"1 ) in a microplate reader (BioRad 680XR).

Purified malyl-CoA lyase from M. extorquens characterized had a Vmax of 36 pmol/(min mg prot), and a Km on glyoxylate of 0.5 mM.

Example 2: Demonstration of malyl-CoA reductase activity

Construction of plasmids containing wild-type genes coding for malonyl- CoA reductase and succinyl-CoA reductase: The DNA sequence of the mcr gene coding for malyl-CoA reductase in Sulfolobus tokodaii str 7 (Alber et al., 2006) was optimized for the expression in Escherichia coli using the GENEius software (Eurofins). The optimized mcr sequence, and the natural DNA sequence of the sucD gene coding for succinyl-CoA reductase in Porphyromonas gingivalis W83 were synthesized by Eurofins MWG OPERON® adding Nhe\ and EcoR\ restriction sites upstream of the start codon and downstream of the stop codon of mcr, respectively, which allowed direct cloning of the synthesized DNA fragments into the pET28a+ vector (Novagen) using T4 DNA ligase (Biolabs). Ligation products were transformed into E. coli DH5a cells, amplified, and the plasmids, pET28-St-mcr (expressing the malonyl-CoA reductase from S. tokodaii), and pET28-Pgi-sucD (expressing the succinyl-CoA reductase from P. gingivalis), were isolated using standard genetic protocols (Sambrook et al., 1989). NCBI references of the utilized mcr and sucD protein sequences, and the references for the corresponding natural and synthetic DNA sequences are listed in Table 2. Organism Protein NCBI/lntegrated Natural DNA Optimized DNA

Genomics accession sequence sequence number

S. tokodaii St-Mcr NP 378167 SEQ ID No. 8 SEQ ID No. 9

SEQ ID No. 7

P. gingivalis Pg-SucD AAQ65862 SEQ ID No. 11

SEQ ID No. 10

Table 2: References to proteins from different organisms having annotated malyl-CoA reductase or succinyl-CoA reductase activity, and references to natural and optimized DNA sequences.

Expression and purification of Pg-SucD was carried out as described in Example 1 using plasmid pET28-Pgi-sucD.

The St-mcr gene was amplified from plasmid pET28-St-mcr using primers 5'- TATAATGAGCTCGTTTAACTTTAAGAAGGAGATATACCATGATTCTGATGC GCCGT-3'(SEQ ID No. 12) and 5'-

TATAATGGATCCCTCGAATTCTTACTTCTC-3' (SEQ ID No. 13) which added a Sacl and a BamHI restriction site upstream of the start codon and downstream of the stop codon, respectively. The PCR fragment was ligated into the pACT3 expression vector using the Sacl and BamHI restriction sites. The resulting plasmid pACT3-St-Mcr was transformed into strain E. coli MG1655. The resulting expression strain was cultivated on mineral medium at 37 °C. One liter mineral medium contained 20 g glucose, 18 g Na ₂ HPO ₄ * 12 H ₂ O, 3 g KH ₂ PO ₄ , 0.5 gNaCI, 2 g NH ₄ CI, 0.5 g MgSO ₄ * 7 H ₂ O, 0.015 CaCI ₂ ^* 2 H ₂ O, 1 mL of 0.06 mol/L FeCI ₃ stock solution prepared in 100 times diluted concentrated HCI, 2 mL of 10 mM thiamine HCI stock solution, 20 g MOPS, 50 pg kanamycin sulphate (and 25 pg chloramphenicol when necessary), and 1 mL of trace element solution (containing per liter: 0.04 g Na ₂ EDTA ^* 2H ₂ O, 0.18 g CoCI ₂ ^* 6 H ₂ O, ZnSO4 ^* 7 H ₂ O, 0.04 g Na ₂ MoO4 ^* 2 H ₂ O, 0.01 g H ₃ BO ₃ , 0.12 g MnSO ₄ ^* H ₂ O, 0.12 g CuCI ₂ ^* H2O.). Medium pH was adjusted to 7 and medium was filter-sterilized.

When the exponentially growing culture reached an OD(600nm) of 0.6, 1 mM IPTG was added and cultures were incubated at 20 °C during 14 h before harvesting the cells by centrifugation (13000 x g, 10 min). After discarding the supernatant cell pellets were stored at -20 °C.

To purify St-Mcr, frozen cell pellets of expression cultures were resuspended in 0.5 ml_ of breakage buffer (50 mM Hepes, 300mM NaCI, pH 7,5) and broken open by four successive rounds of sonication (sonication interval: 20 sec, power output: 30 %, sonicator: Bioblock Scientific, VibraCell™ 72437). Cell debris were removed by centrifuging the crude extracts for 15 min at 4°C at 13000 x g and retaining the clear supernatant. Native proteins of E. coli were removed by heat precipitation at 85 °c during 30 min followed by centrifugation at 13000 x g. Purity of the protein preparations was analysed by SDS-page analysis which showed only one band corresponding to the expected size of the St-Mcr protein.

Enzymatic assays: Malyl-CoA reductase activity was assayed in the reductive and in the oxidative sense of the reaction employing Assay 1 or Assay 2, respectively.

Assay 1 (reaction scheme):

Malyl-CoA lyase: glyoxylate + acetyl-CoA -> malyl-CoA + acetate Malyl-CoA reductase:

(L)-Malyl-CoA + NADPH -> (L)-Malate semialdehyde + Coenzyme A +

NADP

Assay 2 (reaction scheme):

(L)-Malate semialdehyde + Coenzyme A + NADP

NADPH The reaction mixture according to Assay 1 contained 50 mM MOPS/KOH (pH 7.5), 10 mM glyoxylate, 4 mM acetyl-CoA, 5 mM MgCI ₂ , 0.25 mM NADPH (all products from Sigma), 5 U/mL of malyl-CoA lyase, and appropriate amounts of purified malyl-CoA reductase or cell extract. Reactions were started by adding glyoxylate. Enzymatic assays were carried out at 37 °C in 96-well flat bottomed microtiter plates in a final volume of 250 μΙ_. The reactions were followed by the characteristic absorption of NADPH at 340 nm (ENADPH = 6.22 mM ^"1 cm ^"1 ) in a microplate reader (BioRad 680XR).

The reaction mixture according to assay 2 contained 200 mM HEPES (pH 9), 5 mM MgCI ₂ , 1 mM NADP, 0.5 mM coenzyme A (all products from Sigma), and appropriate amounts of purified malyl-CoA reductase. Reactions were started by adding 5 mM (L)-malate semialdehyde. Enzymatic assays were carried out at 37°C in 96-well flat bottomed microtiter plates in a final volume of 250 μΙ_. The reactions were followed by the characteristic absorption of NADPH at 340 nm (8NADPH = 6.22 mM ^"1 cm ^"1 ) in a microplate reader (BioRad 680XR). Unstable malate semialdehyde was produced freshly prior to the enzymatic tests by the deprotection of the stable malate semialdehyde derivative 2-[(4S)- 2,2-dimethyl-5-oxo-1 ,3-dioxolan-4-yl]acetaldehyde(DMODA) (provided by Activation®). Malate semialdehyde was obtained by dissolving appropriate amounts of DMODA in 2 M hydrochloric acid, short heating of the suspension to boiling temperature, and leaving the hot suspension for 15 min at room temperature. The released acetone was evaporated at 35 °C and 50 mbar in a rotary evaporator. The pH of the malate semialdehyde solution was fixed at 3.5 using sodium bicarbonate.

Results listed in Tables 3 and 4 demonstrate malyl-CoA reductase activity for malonyl-CoA reductase, Mcr, of S. tokodaii and succinyl-CoA reductase, SucD, of P. gingivalis. Substrate Malonyl-CoA Succinyl-CoA Malyl-CoA

Enzyme Vmax Km Vmax Km Vmax Km

[ _j timol/(mi [mM] [μΐηο1/(ιηί [mM] [j½mol/(mi [mM] n mg)] n mg)] n mg)]

St-Mcr 0.67 + nd 0.98 0.2 0.24 nd

0.15 +0.17 +0.045

Pg-SucD nd nd 1 1 0.025 nd

Table 3: Kinetic parameters for the reductive sense of reaction (malonyl-CoA reductase and succinyl-CoA reductase activities were estimated by directly adding the substrates malonyl-CoA or succinyl-CoA to the reaction mixture).

Table 4: Kinetic parameters for the oxidative sense of reaction

Example 3: Demonstration of DHB dehydrogenase activity

To identify a suitable 2,4 DHB dehydrogenase, beta-hydroxyacid dehydrogenases from different biological sources were tested for their ability to reduce malate semialdehyde. Among the tested enzymes were the methylbutyraldehyde reductase from Saccharomyces cerevisiae, Ypr1 (Ford & Ellis, 2002)(SEQ ID No.14),the 4-hydroxybutyrate dehydrogenase, 4hbdh, of P. gingivalis (SEQ ID No.187), the alcohol dehydrogenase, YqhD, of E; coli (SEQ ID no. 185), and the succinic semialdehyde reductase, Ms-Ssr, from Metallosphaera sedula (Kockelkorn& Fuchs, 2009) (SEQ ID No. 16). The genes YPR1, 4hbdh, yqhD, and Ms-SSR were amplified using primers listed in Table 5 and cloned into vector pET28 (restriction enzymes see Table 5) yielding plasmids pET28-Sce-YPR1 , pET28-Pgi-4hbdh, pET28-Eco-yqhd and pET28-Mse-SSR, respectively. The proteins were expressed and purified as described in Example 1 .

Table 5 Primers and restriction enzymes used to clone candidate beta- hydroxyacid dehydrogenases

Test for malate semialdehyde reductase activity:

Reaction scheme:

(L)-Malate semialdehyde + NAD(P)H -> (L)-2,4-dihydroxybutyric acid + NAD(P)

The assay mixture contained 200 mM Hepes (pH 7.5), 50 mM KCI, 5 mM MgCI ₂ , 0,24 mM NADH or NADPH, and appropriate amounts of purified enzyme or cell extract. Reactions were started by adding 10 mM (L)-malate semialdehyde (malate semialdehyde was prepared freshly for each test, see Example 3). Enzymatic assays were carried out at 30 °C in 96-well flat bottomed microtiter plates in a final volume of 250 μΙ_. The reactions were followed by the characteristic absorption of NAD(P)H at 340 nm (£ _N ADPH = 6.22 mM ^"1 cm ^"1 ) in a microplate reader (BioRad 680XR). Results are listed in Table

6.

Enzyme Origin Reported function Activity on malate Activity on malate semialdehyde semialdehyde (cofactor NADH) (cofactor NADPH) [^mol/(min*mg_prot) [^mol/(min*mg_prot)

Ms-SSR M. Succinic 4.9 4.9

(SEQ sedula semialdehyde

ID reductase

No.16)

YqhD E. coli Alcohol nd 1.2

(SEQ dehydrogenase

ID No

185)

4hbdh P. 4-hydroxybutyrate 33 nd

(SEQ gingivalis dehydrogenase

ID No

187)

YPR1 S. Methylbutyraldehyd nd 0.19

(SEQ cerevisiae e reductase

ID No.

14)

Table 6: Reducing activity of selected beta-hydroxyacid dehydrogenases on malate semialdehyde (Results represent the average of at least two independent experiments). The succinic semialdehyde dehydrogenase from M. sedula and the methylbutyraldehyde reductase from S. cerevisiae have malate semialdehyde reductase activity. The Km of Ms-SSR for malate semialdehyde was 4 mM. Example 4: Rational construction of an improved malyl-CoA reductase enzyme

Site-directed mutagenesis was carried out using the oligonucleotide pairs listed in Table 7 and the pET28-Sto-mcr plasmid as the template. Point mutations to change the amino acid sequences were introduced by PCR (Phusion 1 U, HF buffer 20 % (v/v), dNTPs 2.5 mM, direct and reverse primers 1 μΜ each, template plasmid 200 ng, water). When possible, plasmids created by PCR contained new restriction sites (introduced using silent mutations) in addition to the functional mutation to facilitate identification of mutated clones. The PCR products were digested by Dpnl at 37 °C for 2 x 2h to remove template DNA, and transformed into NEB DH5-a competent E. coli cells (NEB). The mutated plasmids were identified by restriction site analysis and verified to carry the desired mutations by DNA sequencing.

Table 7: Primer pairs used to mutate the mcr gene of S. tokodaii.

The impact of the genetic modifications of St-Mcr was tested in the oxidative sense of the reaction using Assay 3 described in Example 2. Figure 3 shows that replacing the natural Tyr206 by other amino acids decreases the activity on the natural substrate, succinic semialdehyde, causing at the same time an increased or at least constant activity on malate semialdehyde. Thus, replacing Tyr206 by appropriate amino acid residues provides a selective advantage regarding the specificity of Mcr for the DHB pathway intermediate.

Preferred amino acid residues in position 206 are therefore phenylalanine, histidine, isoleucine, lysine, methionine, glycine, asparagine, proline, arginine, glutamine, leucine, serine, tryptophane, and threonine.

The protein wherein the Tyrosine 206 is replaced by a Proline residu is represented by SEQ ID No. 202.

Example 5: Rational construction of an improved DHB dehydrogenase

Site-directed mutagenesis was carried out using the oligonucleotide pairs listed in Table 6 and the pET28-Mse-SSR plasmid as the template. Point mutations to change the amino acid sequences were introduced by PCR (Phusion 11), HF buffer 20 % (v/v), dNTPs 2.5 mM, direct and reverse primers 1 μΜ each, template plasmid 200 ng, water). When possible, plasmids created by PCR contained new restriction sites (introduced using silent mutations) in addition to the functional mutation to facilitate identification of mutated clones. The PCR products were digested by Dpnl at 37 °C for 2 x 2h to remove template DNA, and transformed into NEB DH5-a competent E. coli cells (NEB). The mutated plasmids were identified by restriction site analysis and verified to carry the desired mutations by DNA sequencing. Table 8 summarizes kinetic parameters of the mutants. The results demonstrate that the double mutant Ms- SSR H39R N43H (SEQ ID No.38) has improved affinity for malate semialdehyde when compared to the wild type enzyme.

Table 8: Primer pairs used to mutate M. sedula succinic semialdehyde reductase (Ms-SSR)

Table 9: Summary of kinetic parameters of M. sedula succinic semialdehyde reductase (Ms-SSR) mutants (Results represent the average of at least two independent experiments).

The corresponding nucleic sequences are represented by SEQ ID No. 17, SEQ ID No. 33, SEQ ID No. 35 and SEQ ID No. 37.

The coding sequence of M. sedula succinic semialdehyde reductase including the mutations H39R and N43H was optimized for maximum expression in E. coli, using the GeneOptimizer® software. The synthetic gene was produced by GeneArt® Gene Synthesis (Invitrogen Life Technologie). Nhe\ and EcoR\ restriction sites were introduced upstream of the start codon and downstream of the stop codon, respectively, allowing direct cloning into pET28a+ (Novagen).

The resulting pET28-Mse-DHB-Dh-H39R_N43H-opt plasmid was isolated and shown by DNA sequencing to contain the full-length M. sedula SSR H39R N43H gene having the correct sequence (SEQ ID No.38).

Example 6: Demonstration of in vitro production of DHB by the synthetic malyl-CoA pathway

The enzymes malyl-CoA lyase (Me-Mcl), malyl-CoA reductase (St-Mcr or Pg-SucD), and DHB dehydrogenase (Ms-SSA-red H39N N43H) were expressed and purified as described in Examples 1 , 2, and 3.

Production of DHB by the pathway comprising malyl-CoA lyase, malyl-

CoA reductase, and DHB dehydrogenase was demonstrated in vitro by adding 2 mM glyoxylate to a reaction mixture that contained 50 mM Hepes (pH 7.5), 2 mM acetyl-CoA, 2 mM NADPH, 100 pg/mL DHB dehydrogenase, 150 pg/mL malyl-CoA lyase, and 100 pg/rnL malyl-CoA reductase (which was either St-Mcr (reaction 1 ), or Pg-SucD (reaction 2)).

Control reactions contained all components but were lacking either DHB dehydrogenase (Control 1 ) or malyl-CoA reductase (Control 2). After 120 min of incubation at 37 °C the DHB content in the reaction mixture was analysed by gas chromatography [GCMS-QP2010 Ultra Shimadzu; equipped with a FID detector (FID-2010 Plus Shimadzu); autosampler AOC20s (Shimadzu) ; splitless injector AOC20i (Shimadzu) (230 °C); column: Zebron ZB-FFAP, 30 m x 0.25 mm, df 0.25 pm; and liner: Tapered focus Liner5 x 95 x 3.4 mm (SGE). Carrier gas was hydrogen at a total flow rate of 25 mL/min. Flame ionization was carried out using an air-hydrogen mixture (flow rates were 300 mL/min and 30 mL/min, respectively). Detector temperature was 240 °C. Injected sample volume was 1 pL. Temperature program is provided in Table 10.

Chromatograms showing presence of DHB in the reactions containing all pathway enzymes and absence of DHB in samples containing only two out of three pathway enzymes are shown in Figure 2.

Table 10: Temperature program for GC analysis of reaction mixtures

Example 7: Construction of optimized DHB producer strains

Construction of a plasmid for simultaneous expression of malyl-CoA synthetase, malyl-CoA reductase, and DHB-dehydrogenase: The coding sequence of the malyl-CoA lyase from M. extorquens, Me- mcl, was amplified from plasmid pET28-Mex-mcl using the high fidelity polymerase Phusion (Fermentas) and the forward and reverse primers 5'- TCACACAGGAAACAGAATTCGAGCTCGGTAATGTCGTTTACCCTGATTCAG CAAGCGACT-3' (SEQ ID No. 39) and 5'- GGTATATCTCCTTCTTAAAGTTAAACTTATTTGCCGCCCATTGCATCCGCTT TCTG-3' (SEQ ID No. 40) which contained restriction sites for Sacl upstream of the start codon (underlined). The coding sequence of the malonyl-CoA reductase from S. tokodaii, St-mcr, was amplified from plasmid pET28-Sto-mcr using the forward and reverse primers 5'- GTTTAACTTTAAGAAGGAGATATACCATGATTCTGATGCGCCGTACCCTGA AAGCG-3' (SEQ ID No. 41 ) and 5'-

GGTATATCTCCTTCTTAAAGTTAAACTTACTTCTCGATGTAGCCTTTCTCCA CGAG-3' (SEQ ID No. 42) which contained restriction sites for BamHI downstream of the stop codon. The plasmid pET28-Mse-DHB-Dh- H39R_N43H-opt (Example 5) was used as the template to amplify the optimized coding sequence of the succinic semialdehyde reductase H39R N43H from M. sedula using the forward and reverse primers 5'- GTTTAACTTTAAGAAGGAGATATACCATGAAAGCAGCAGTTCTGCATACCT ATAAAGAACCGCTGAGCAT-3' (SEQ ID No. 43) and 5'- ATGCCTGCAGGTCGACTCTAGAGGATCCTTACGGAATAATCAGGCTACGA ATTGCTTC-3' (SEQ ID No. 44) that introduced a BamHI restriction site downstream of the stop codon (underlined).

The forward primers for St-mcr and the succinic semialdehyde reductase H39R N43H from M. sedula contained a rbs motif. The three genes were simultaneously cloned into the pACT3 expression vector by homologous recombination using the In-Fusion cloning kit (Clontech).

The resulting and pACT3-MCL-DHB (SEQ ID No. 45) plasmid was isolated and shown by DNA sequencing to have the correct sequence. Construction of a plasmid for simultaneous expression of malyl-CoA synthetase, malyl-CoA reductase, and DHB-dehydrogenase:

The DNA sequences coding for the two protein subunits of malyl-CoA synthetase, mtkA (YP_00296285) and mtkB (YP_002962852), from Methylobacterium extorquens AM1 were optimized for the expression in Escherichia coli using the GENEius software (Eurofins). The optimized DNA sequences of the subunit were physically linked by the DNA sequence naturally occurring between the mtkA and mtkB genes in M. Extorquens genome (CGAACGGGGGAGGAATCACGCC, SEQ ID No. 46). The resulting DNA fragment, 'mtkA gene - linker DNA - mtkB gene', was synthesized by Eurofins MWG OPERON® and subcloned into pET28b expression vector using Nhel and EcoRI restriction enzymes. The resulting DNA plasmid pET28-Mex-mtkAB (SEQ ID No. 47) was used to simultaneously amplify the two codons optimized genes encoding malyl-CoA synthetase from M. extorquens, Me-mtkA and Me- mtkB using the high fidelity polymerase Phusion (Fermentas) and the forward and reverse primers 5'-

CAGGAAACAGAATTCGAGCTCGGTAATGGATGTGCACGAATATCAGGCGA AAGAACTGCT-3' (SEQ ID No. 48) and 5'- TACGGCGCATCAGAATCATtacgccgcacgtgctaacacatcggcaac-3' (SEQ ID No. 49) which contained restriction sites for Sacl upstream of the start codon (underlined). The coding sequence of the malonyl-CoA reductase from S. tokodaii, St-mcr, was amplified from plasmid pET28-Sto-mcr using the forward and reverse primers 5'-

GGCGTAATGATTCTGATGCGCCGTACCCTGAAAGCG-3' (SEQ ID No. 50) and 5'-CTGCTGCTTTCATTACTTCTCGATGTAGCCTTTCTCCACGAG-3' (SEQ ID No. 51 ) which contained restriction sites for BamHI downstream of the stop codon. The plasmid pET28-Mse-DHB-Dh-H39R_N43H-opt (Example 5) was used as the template to amplify the optimized coding sequence of the succinic semialdehyde reductase H39R N43H from M. sedula using the forward and reverse primers 5'-

TACATCGAGAAGTAATGAAAGCAGCAGTTCTGCATACCTATAAAGAAC-3' (SEQ ID No. 52) and 5'- CCTGCAGGTCGACTCTAGAGGATCCTTACGGAATAATCAGGCTACGAATT GCTTCAC-3' (SEQ ID No. 53) that introduced a BamHI restriction site downstream of the stop codon (underlined).

The three genes were simultaneously cloned into the pEXT20 expression vector by homologous recombination using the In-Fusion cloning kit (Clonetch).

The resulting pEXT20-MCS-DHB (SEQ ID No.54) plasmid was isolated and shown by DNA sequencing to have the correct sequence.

Construction of plasmids for overexpression of phosphoenolpyruvate (PEP) carboxy kinase, PEP carboxylase, pyruvate kinase, pyruvate carboxylase, isocitrate lyase enzymes and the galactose symporter permease:

The plasmid pACT3-pck harbouring the PEP carboxykinase encoding pck gene of E. coli was constructed by amplifying the pck coding sequence using genomic DNA from E. coli MG1655 as the template and the forward and reverse primers, respectively, ⁵ TATAATCCCGGGATGCGCGTTAACAATGGTTTGACC ³ (SEQ ID No. 56 and ⁵ TATAATTCTAGATTACAGTTTCGGACCAGCCG ³ (SEQ ID No. 57). The DNA fragment was digested with Xmal and Xba\, ligated into the corresponding sites of the pACT3 expression vector (Dykxhoorn et al., 1996) using T4 DNA ligase (Biolabs), and transformed into E. coli DH5a cells. The transformants were selected on solid LB medium containing chloramphenicol (25 pg/mL). The resulting plasmid was isolated and correct insertion of the pck gene was verified by sequencing. Plasmids pACT3-aceA, pACT3-ppc, pACT3-galP, pACT3-pck and pACT3-pyc harbouring, respectively, aceA, ppc, galP, or pykA (all E. coli) or pck from Lactococcus lactis were constructed analogously using the primers listed in Table 11.

Gene Primer Linker Sequence

Ec _pck Ec _pck_clonJor Xmal tataatcccgggatgcgcgttaacaatggtttgacc

(SEQ ID No.57) Ec _pckj:lon rev Xbal tataattctagattacagtttcggaccagccg

(SEQ ID No.58)

Ec _ppc E _ppc_clonJor Xmal tataatcccgggatgaacgaacaatattcc

(SEQ ID No.59)

Ec jpc clon rev Xbal tataattctagattagccggtattacgcat

(SEQ ID No.60)

Ec _pykA Ec jpykA clonjor Xmal tataatcccgggatgtccagaaggcttcgcagaaca (SEQ ID

No.61)

Ec _pykA_clon_rev Xbal tataattctagattactctaccgttaaaatac

(SEQ ID No.62)

Ec aceA Ec aceAjclon or Xmal tataatcccgggatgaaaacccgtacacaacaaatt (SEQ ID

No63)

Ec_aceA_clon_rev Xbal tataattctagattagaactgcgattcttcag

(SEQ ID No.64)

LI _pycA LI _pycA_clonJor Xmal tataatcccgggatgaaaaaactactcgtcgccaat (SEQ ID

No.65)

LI _pycA_clon_rev Xbal tataattctagattaattaatttcgattaaca

(SEQ ID No.66)

Ec galP Ec galP ' clonjor Xmal tataatcccgggatgcctgacgctaaaaaacaggggcggt (SEQ ID

No.67)

Ec_galP_clon_rev Xbal tataattctagattaatc gt gage gectattte

(SEQ ID No.68)

Table 11 : Primers used for construction of plasmids for gene overexpression. Restriction sites used for cloning into pACT3 are underlined It is understood that removal of the lacl gene from the backbone of the above described plasmids along with the genomic deletion of lacl in the host strain may render protein expression from above described plasmids constitutive. Construction of strains with optimized carbon flux repartitioning for DHB production

Several genes were disrupted in E. coli strain MG1655 in order to optimise carbon flux repartitioning and cofactor supply for DHB production. Gene deletions were carried out using either the lambda red recombinase method according to Datsenko et a/.(Datsenko & Wanner, 2000), or the phage transduction method adapted from Miller(Miller, 1992).

Protocol for introduction of gene deletions using the lambda red recombinase method: the deletion cassettes were prepared by PCR using high fidelity polymerase Phusion™ (Finnzymes), and the FRT-flanked kanamycin resistance gene (kan) of plasmid pKD4 as the template (Datsenko & Wanner, 2000). Sense primers contained sequences corresponding to the 5'-end of each targeted gene (underlined) followed by 20 bp corresponding to the FRT- kan-FRT cassette of pKD4. Anti-sense primers contained sequences corresponding to the 3'-end region of each targeted gene (underlined) followed by 20 bp corresponding to the cassette. The primers are described in Table 12. PCR products were digested with Dpn\ and purified prior to transformation.

E. coli MG1655 strain was rendered electro-competent by growing the cells to an OD ₆ oo of 0.6 in LB liquid medium at 37 °C, concentrating the cells 100-fold, and washing them twice with ice-cold 10 % glycerol. The cells were transformed with plasmid pKD46 (Datsenko & Wanner, 2000) by electroporation (2.5 kV, 200 Ω, 25 pF, in 2 mm gap cuvettes). Transformants were selected at 30 °C on ampicillin (100 pg/mL) LB solid medium.

Disruption cassettes were transformed into electro-competent E. coli strains harbouring the lambda Red recombinase-expressing plasmid pKD46. The cells were grown at 30 °C in liquid SOB medium containing ampicillin (100 pg/mL). The lambda red recombinase system was induced by adding 10 mM arabinose when OD ₆ oo of the cultures reached 0.1. Cells were further grown to an OD ₆ oo of 0.6 before they were harvested by centrifugation, washed twice with ice-cold 10 % glycerol, and transformed with the disruption cassette by electroporation. After an overnight phenotypic expression at 30 °C in LB liquid medium, cells were plated on solid LB medium containing 25 pg/mL kanamycin. Transformants were selected after cultivation at 30 °C.

The gene replacement was verified by colony PCR using Crimson Taq polymerase (NEB). A first reaction was carried out with the flanking locus- specific primers (see Table 12) to verify simultaneous loss of the parental fragment and gain of the new mutant specific fragment. Two additional reactions were done by using one focus-specific primer together with one of the corresponding primers kl rev, or k2for (see Table 12) that align within the FRT- kanamycin resistance cassette (sense locus primer/kl rev and k2for/reverse locus primer).

The resistance gene (FRT-kan-FRT) was subsequently excised from the chromosome using the FLP recombinase-harbouring plasmid pCP20 (Cherepanov & Wackernagel, 1995) leaving a scar region containing one FRT site. pCP20 is an ampicillin and CmR plasmid that shows temperature-sensitive replication and thermal induction of FLP recombinase synthesis. Kanamycin resistant mutants were transformed with pCP20, and ampicillin-resistant transformants were selected at 30 °C. Transformants were then grown on solid LB medium at 37 °C and tested for loss of all antibiotic resistances. Excision of the FRT-kanamycin cassette was analysed by colony PCR using crimson Taq polymerase and the flanking locus-specific primers (Table 12). Multiple deletions were obtained by repeating the above described steps.

Gene Primer Sequence

IdhA AjdhAJ gaaggttgcgcctacactaagcatagttgttgatgagtgtaggctggagctgcttc

or (SEQ ID No.69)

AJdhAjr ttaaaccagttcgttcgggcaggtttcgcctttttcatgggaattagccatggtcc

ev SEQ ID No.70)

adhE A adhE atggctgttactaatgtcgctgaacttaacgcactcgtagagcgtgtgtaggctggagct gcttc

for (SEQ ID No.71)

A_adhE_ ttaagcggattttttcgcttttttctcagctttagccggagcagccatatgaatatcctc cttag rev (SEQ ID No.72)

ackA A ackAJ atgtcgagtaagttagtactggttctgaactgcggtagttcttcagtgtaggctggagct gcttc

or (SEQ ID No. 3)

A_ackA_ tcaggcagtcaggcggctcgcgtcttgcgcgataaccagttcttccatatgaatatcctc cttag rev (SEQ ID No.74)

focA- Δ focA- ttactccgtatttgcataaaaaccatgcgagttacgggcctataagtgtaggctggagct gcttc

pflB pflBJor (SEQ ID No.75)

AjocA- atagattgagtgaaggtacgagtaataacgtcctgctgctgttctcatatgaatatcctc cttag pflBjrev (SEQ ID No.76)

pta Ajptajo gtgtcccgtattattatgctgatccctaccggaaccagcgtcggtgtgtaggctggagct gcttc

r (SEQ ID No.77)

Ajptajre ttactgctgctgtgcagactgaatcgcagtcagcgcgatggtgtacatatgaatatcctc cttag

V (SEQ ID No.78)

poxB A poxB atgaaacaaacggttgcagcttatatcgccaaaacactcgaatcggtgtaggctggagct gcttc

for (SEQ ID No.79)

A _poxB_ ttaccttagccagtttgttttcgccagttcgatcacttcatcacccatatgaatatcctc cttag rev (SEQ ID No.80) sad A__sad_fo atgaccattactccggcaactcatgcaatttcgataaatcctgccgtgtaggctggagct gcttc

r (SEQ ID No.81)

A_sad_re tcagatccggtctttccacaccgtctggatattacagaattcgtgcatatgaatatcctc cttag

V (SEQ ID No.82)

gabD A gabD atgaaacttaacgacagtaacttattccgccagcaggcgttgattgtgtaggctggagct gcttc

for (SEQ ID No.83)

A_gabD_ ttaaagaccgatgcacatatatttgatttctaagtaatcttcgatcatatgaatatcctc cttag rev (SEQ ID No.847)

gadA A gadA atggaccagaagctgttaacggatttccgctcagaactactcgatgtgtaggctggagct gcttc

for (SEQ ID No.85)

A_gadA_ tcaggtgtgtttaaagctgttctgctgggcaataccctgcagtttcatatgaatatcctc cttag rev (SEQ ID No.86) gadB A gadB atggataagaagcaagtaacggatttaaggtcggaactactcgatgtgtaggctggagct gcttc for (SEQ ID No.87)

A_gadB_ tcaggtatgtttaaagctgttctgttgggcaataccctgcagtttcatatgaatatcctc cttag rev (SEQ ID No.88)

gadC A gadC atggctacatcagtacagacaggtaaagctaagcagctcacattagtgtaggctggagct gcttc

for (SEQ ID No.89)

A_gadC_ ttagtgtttcttgtcattcatcacaatatagtgtggtgaacgtgccatatgaatatcctc cttag rev (SEQ ID No.90) sfcA A_sfcAJ atggaaccaaaaacaaaaaaacagcgttcgctttatatcccttacgtgtaggctggagct gcttc

or (SEQ ID No.91)

A_sfcA_r ttagatggaggtacggcggtagtcgcggtattcggcttgccagaacatatgaatatcctc cttag ev (SEQ ID No.92)

maeB A maeB atggatgaccagttaaaacaaagtgcacttgatttccatgaatttgtgtaggctggagct gcttc

or (SEQ ID No.93)

A_maeB ttacagcggttgggtttgcgcttctaccacggccagcgccaccatcatatgaatatcctc cttag

_rev (SEQ ID No.94)

ppc Δ _PPC_fo atgaacgaacaatattccgcattgcgtagtaatgtcagtatgctcgtgtaggctggagct gcttc

r (SEQ ID No.95)

Δ _ppc_re ttagccggtattacgcatacctgccgcaatcccggcaatagtgaccatatgaatatcctc cttag

V (SEQ ID No.96)

pykA A_pykA_f atgtccagaaggcttcgcagaacaaaaatcgttaccacgttaggcgtgtaggctggagct gcttc

or (SEQ ID No.97)

A_pykA_ ttactctaccgttaaaatacgcgtggtattagtagaacccacggtcatatgaatatcctc cttag rev (SEQ ID No.98)

pykF AjpykFJ atgaaaaagaccaaaattgtttgcaccatcggaccgaaaaccgaagtgtaggctggagct gcttc

or (SEQ ID No.99)

A _pykF_r ttacaggacgtgaacagatgcggtgttagtagtgccgctcggtaccatatgaatatcctc cttag ev (SEQ ID No.100) mgsA A mgsA atggaactgacgactcgcactttacctgcgcggaaacatattgcggtgtaggctggagct gcttc

Jor (SEQ ID No. 101)

A_mgsA ttacttcagacggtccgcgagataacgctgataatcggggatcagcatatgaatatcctc cttag

_rev (SEQ ID No. 102) icIR AJclRJo atggtcgcacccattcccgcgaaacgcggcagaaaacccgccgttgtgtaggctggagct gcttc

r (SEQ ID No. 103)

A_iclR_r tcagcgcattccaccgtacgccagcgtcacttccttcgccgctttcatatgaatatcctc cttag ev (SEQ ID No. 104)

icd A_icd_for atggaaagtaaagtagttgttccggcacaaggcaagaagatcaccgtgtaggctggagct gcttc

(SEQ ID No. 105)

A_icd_re ttacatgttttcgatgatcgcgtcaccaaactctgaacatttcagcatatgaatatcctc cttag

V (SEQ ID No.106)

sucA A_sucA_f atgcagaacagcgctttgaaagcctggttggactcttcttacctcgtgtaggctggagct gcttc

or (SEQ ID No. 107)

A_sucA_ ttattcgacgttcagcgcgtcattaaccagatcttgttgctgtttcatatgaatatcctc cttag rev (SEQ ID No. 108)

Table 12: Primers used for gene disruptions. Sequences homologous to target genes are underlined

Deleted gene Sequence (5' - 3') I Forward primer Reverse primer

K2for I kl rev cggtgccctgaatgaactgc cagtcatagccgaatagcct

(SEQ ID No.129) (SEQ ID No.130)

IdhA atacgtgtcccgagcggtag tacacatcccgccatcagca

(SEQ ID No.131) (SEQ ID No.132) adhE gaagtaaacgggaaaatcaa agaagtggcataagaaaacg

(SEQ ID No.133) (SEQ ID No.134) ackA ccattggctgaaaattacgc gttccattgcacggatcacg

(SEQ ID No. 135) (SEQ ID No.136) focA _pflB atgccgtagaagccgccagt tgttggtgcgcagctcgaag

(SEQ ID No.137) (SEQ ID No.138) pta gcaaatctggtttcatcaac tcccttgcacaaaacaaagt

(SEQ ID No. 139) (SEQ ID No.140) poxB ggatttggttctcgcataat agcattaacggtagggtcgt

(SEQ ID No. 141) (SEQ ID No.142) sad gctgattctcgcgaataaac aaaaacgttcttgcgcgtct

(SEQ ID No.143) (SEQ ID No. 144) gabD tctgtttgtcaccaccccgc aagccagcacctggaagcag

(SEQ ID No.145) (SEQ ID No.146) gadA aagagctgccgcaggaggat gccgccctcttaagtcaaat

(SEQ ID No. 147) (SEQ ID No. 148) gadB ggattttagcaatattcgct cctaatagcaggaagaagac

(SEQ ID No. 149) (SEQ ID No.150) gadC gctgaactgttgctggaaga ggcgtgcttttacaactaca

(SEQ ID No. 151) (SEQ ID No.152) sfcA tagtaaataacccaaccggc tcagtgagcgcagtgtttta

(SEQ ID No. 153) (SEQ ID No.154) maeB attaatggtgagagtttgga tgcttttttttattattcgc

(SEQ ID No.155) (SEQ ID No.156) ppc gctttataaaagacgacgaa gtaacgacaattccttaagg

(SEQ ID No.157) (SEQ ID No.158) pykA tttatatgcccatggtttct atctgttagaggcggatgat

(SEQ ID No. 159) (SEQ ID No.160) pykF ctggaacgttaaatctttga ccagtttagtagctttcatt

(SEQ ID No.161) (SEQ ID No.162) iclR gatttgttcaacattaactcatcgg tgcgattaacagacaccctt

(SEQ ID No.163) (SEQ ID No.164) mgsA tctcaggtgctcacagaaca tatggaagaggcgctactgc

(SEQ ID No. 165) (SEQ ID No.166) icd cgacctgctgcataaacacc tgaacgctaaggtgattgca

(SEQ ID No. 167) (SEQ ID No.168) sucA acgtagacaagagctcgcaa catcacgtacgactgcgtcg

(SEQ ID No.169) (SEQ ID No.170) sucB tgcaactttgtgctgagcaa tatcgcttccgggcattgtc

(SEQ ID No.171) (SEQ ID No. 172) frdA aaatcgatctcgtcaaatttcagac aggaaccacaaatcgccata

(SEQ ID No. 173) (SEQ ID No.174) frdB gacgtgaagattactacgct agttcaatgctgaaccacac

(SEQ ID No.175) (SEQ ID No.176) frdC tagccgcgaccacggtaagaaggag cagcgcatcacccggaaaca

(SEQ ID No. 177) SEQ ID No. 178) frdD atcgtgatcattaacctgat ttaccctgataaattaccgc

(SEQ ID No.179) (SEQ ID No.180) lacl gaatctggtgtatatggcga tcttcgctattacgccagct

(SEQ ID No.181) (SEQ ID No.182)

UdD cgtcagcggatgtatctggt gcggaatttctggttcgtaa

(SEQ ID No. 183) (SEQ ID No. 184) ttgtcaacgatggggtcatg aaaaatgccgacataacgtc

Pgi (SEQ ID No.195) (SEQ ID No. 196) ptsG ccatccgttgaatgagtttt tggtgttaactggcaaaatc

(SEQ ID No. 197) (SEQ ID No. 198) ptsl gtgacttccaacggcaaaag- - eegttggtttgatagcaata

(SEQ ID No. 199) (SEQ ID No.200) Table 13: Primer pairs used for verification of gene disruptions

Protocol for introduction of gene deletions using the phage transduction method: strains carrying the desired single deletions were obtained from the Keio collection (Baba et a/., 2006). Phage lysates of single deletion mutants were prepared by inoculating 10 mL of LB medium containing 50 pg/mL kanamycin, 2 g/L glucose, and 5 mM CaCI ₂ with 100 pL of overnight precultures. Following an incubation of 1 h at 37 °C, 200 pL of phage lysate prepared from the wild-type MG1655 strain were added, and cultures were incubated for another 2-3 h until cell lysis had completed. After addition of 200 pL chloroform, cell preparations were first vigorously vortexed and then centrifuged for 10 min at 4500 x g. The clear lysate was recovered and stored at 4 °C.

The receptor strain was prepared for phage transduction by an overnight cultivation at 37 °C in LB medium. A volume of 1.5 mL of the preculture was centrifuged at 1500 x g for 10 min. The supernatant was discarded and the cell pellet was resuspended in 600 pL of a solution containing 10 mM MgS0 ₄ and 5 mM CaCI ₂ . The transduction was carried out by mixing 100 pL of the solution containing the receptor strain with 100 pL of lysate and incubating this mixture at 30 °C for 30 min. Thereafter, 100 pL of a 1 M sodium citrate solution were added followed by vigorous vortexing. After addition of 1 mL LB medium, the cell suspension was incubated at 37 °C for 1 h before spreading the cells on LB agar dishes containing 50 pg/mL kanamycin. Clones able to grow in presence of the antibiotic were confirmed by colony PCR to contain the desired deletion using the primers listed in Table 13. After the introduction of each gene deletion, the antibiotic marker was removed as described above following the method of (Cherepanov & Wackernagel, 1995) The plasmids co-expressing malyl-CoA synthetase, malyl-CoA reductase, and DHB dehydrogenase (pEXT20-MCS-DHB or pACT3-MCS- DHB); or plasmids co-expressing malyl-CoA lyase, malyl-CoA reductase, and DHB dehydrogenase(pEXT20-MCL-DHB or pACT3-MCL-DHB);or the empty control plasmids (pEXT20 or pACT3) were transformed alone or together with one of the plasmids pACT3-aceA, pACT3-ppc, pACT3-galP, pACT3-pck or pACT3-pyc into the optimized host strains. Transformants containing both a plasmid expressing the DHB-pathway enzymes, and a plasmid expressing an anaplerotic enzyme were selected on solid LB medium containing chloramphenicol (25 pg/mL) and kanamycin (50 pg/mL). Non-exclusive examples of constructed strains are listed in Table 14.

Table 14: Examples of strains constructed for DHB production

Example 8: Demonstration of zymotic production of DHB by the synthetic malyl-CoA pathway

Strains and cultivation conditions: Experiments were carried out using strain ECE69 which expressed the DHB pathway from plasmid pACT3- MCL-DHB represented by SEQ ID No. 203 (the wild-type Mcr enzyme was replaced by the Mcr Tyr206Pro mutant in this experiment) and the isogenic control strain ECE70 containing the empty plasmid pACT3. All cultivations were carried out at 37 °C on an Infors rotary shaker running at 170 rpm. Overnight cultures (3 mL medium in test tube) were inoculated from glycerol stocks and used to adjust an initial Οϋβοο of 0.05 in 100 mL growth cultures cultivated in 500 mL shake flasks. IPTG was added at a concentration of 1 mmol/L when OD ₆ oo of the growth cultures reached 1. The composition of the growth mineral medium is provided in Example 2. Estimation of DHB concentration by LC-MS/MS analyses.DHB was quantified using LC-MS: Liquid anion exchange chromatography was performed on an ICS-3000 system from Dionex (Sunnyvale, USA) equipped with an automatic eluent (KOH) generator system (RFIC, Dionex), and an autosampler (AS50, Dionex) holding the samples at 4 °C. Analytes were separated on an lonPac AS11 HC (250 x 2 mm, Dionex) column protected by an AG11 HC (50 x 2 mm, Dionex) pre-column. Column temperature was held at 25 °C, flow rate was fixed at 0.25 mL/min, and analytes were eluted applying the KOH gradient described earlier (Groussac E, Ortiz M & Francois J (2000) Improved protocols for quantitative determination of metabolites from biological samples using high performance ionic-exchange chromatography with conductimetric and pulsed amperometric detection. Enzyme. Microb. Technol.26, 715-723). Injected sample volume was 15 μΙ_. For background reduction, an ASRS ultra II (2 mm, external water mode, 75 mA) anion suppressor was used. Analytes were quantified a mass-sensitive detector (MSQ Plus, Thermo) running in ESI mode (split was 1/3, nitrogen pressure was 90 psi, capillary voltage was 3.5 kV, probe temperature was 450 °C). Results: After 24 h of cultivation the supernatant of strains ECE69 and

ECE70 contained 0.05 mM DHB and 0 mM DHB, respectively, demonstrating DHB production via the synthetic pathway.

References Akita, O., Nishimori, C, Shimamoto, T., Fujii, T. & Iefuji, H. (2000).

Transport of pyruvate in Saccharomyces cerevisiae and cloning of the gene encoded pyruvate permease. Biosci Biotechnol Biochem 64, 980-984.

Alber, B., Olinger, M., Rieder, A., Kockelkorn, D., Jobst, B., Hiigler, M. & Fuchs, G. (2006). Malonyl-coenzyme A reductase in the modified 3- hydroxypropionate cycle for autotrophic carbon fixation in archaeal

Metallosphaera and Sulfolobus spp. J Bacteriol 188, 8551-8559.

Arps, P. J., Fulton, G. F., Minnich, E. C. & Lidstrom, M. E. (1993). Genetics of serine pathway enzymes in Methylobacterium extorquens AMI: phosphoenolpyruvate carboxylase and malyl coenzyme A lyase. J Bacteriol 175, 3776-3783.

Auriol, C., Bestel-Corre, G., Claude, J.-B., Soucaille, P. & Meynial-Salles, I.

(2011). Stress-induced evolution of Escherichia coli points to original concepts in respiratory cofactor selectivity. Proc Natl Acad Sci U S A 108, 1278-1283.

Baba, T., Ara, T., Hasega a, M., Takai, Y., Okumura, Y., Baba, M., Datsenko, K. A., Tomita, M., Wanner, B. L. & Mori, H. (2006).

Construction of Escherichia coli K-12 in- frame, single-gene knockout mutants: the Keio collection. Mol Syst Biol 2, 2006.0008.

Bailey, J. E. (1991). Toward a science of metabolic engineering. Science 252, 1668- 1675.

Camarasa, C., Bidard, F., Bony, M., Barre, P. & Dequin, S. (2001).

Characterization of Schizosaccharomyces pombe malate permease by expression in Saccharomyces cerevisiae. Appl Environ Microbiol 67, 4144- 4151.

Cherepanov, P. P. & Wackemagel, W. (1995). Gene disruption in Escherichia coli: TcR and KmR cassettes with the option of Flp-catalyzed excision of the antibiotic-resistance determinant. Gene 158, 9-14.

Chistoserdova, L., Kalyuzhnaya, M. G. & Lidstrom, M. E. (2009). The expanding world of methylotrophic metabolism. Annu Rev Microbiol 63, 477- 499.

Datsenko, K. A. & Wanner, B. L. (2000). One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci U S A

97, 6640-6645.

Ford, G. & Ellis, E. M. (2002). Characterization of Yprlp from Saccharomyces cerevisiae as a 2-methylbutyraldehyde reductase. Yeast Chichester Engl 19, 1087-1096.

Friedmann, S., Steindorf, A., Alber, B. E. & Fuchs, G. (2006). Properties of succinyl-coenzyme A:L-malate coenzyme A transferase and its role in the autotrophic 3-hydroxypropionate cycle of Chloroflexus aurantiacus. J Bacteriol 188, 2646-2655.

Grobler, J., Bauer, F., Subden, R. E. & Van Vuuren, H. J. (1995). The mael gene of Schizosaccharomyces pombe encodes a permease for malate and other

C4 dicarboxylic acids. Yeast Chichester Engl 11, 1485-1491.

Groeneveld, M., Weme, R. G. J. D. O., Duurkens, R. H. & Slotboom, D. J.

(2010). Biochemical characterization of the C4-dicarboxylate transporter DctA from Bacillus subtilis. J Bacteriol 192, 2900-2907.

Jantama, K., Zhang, X., Moore, J. C, Shanmugam, K. T., Svoronos, S. A. & Ingram, L. O. (2008a). Eliminating side products and increasing succinate yields in engineered strains of Escherichia coli C. Biotechnol Bioeng 101, 881-893.

Jantama, K., Haupt, M. J., Svoronos, S. A., Zhang, X., Moore, J. C, Shanmugam, K. T. & Ingram, L. O. (2008b). Combining metabolic engineering and metabolic evolution to develop nonrecombinant strains of Escherichia coli C that produce succinate and malate. Biotechnol Bioeng 99, 1140-1153.

Kawasaki, T., Koita, H., Nakatsubo, T., Hasegawa, K., Wakabayashi, K., Takahashi, H., Umemura, K., Umezawa, T. & Shimamoto, K. (2006).

Cinnamoyl-CoA reductase, a key enzyme in lignin biosynthesis, is an effector of small GTPase Rac in defense signaling in rice. Proc Natl Acad Sci U S A 103, 230-235.

Kockelkorn, D. & Fuchs, G. (2009). Malonic semialdehyde reductase, succinic semialdehyde reductase, and succinyl-coenzyme A reductase from Metallosphaera sedula: enzymes of the autotrophic 3-hydroxypropionate/4- hydroxybutyrate cycle in Sulfolobales . J Bacteriol 191, 6352-6362.

Larkin, M. A., Blackshields, G., Brown, N. P., Chenna, R., McGettigan, P. A., McWilliam, H., Valentin, F., Wallace, I. M., Wilm, A. & other authors. (2007). Clustal W and Clustal X version 2.0. Bioinforma Oxf Engl 23, 2947- 2948.

Lin, H., Bennett, G. N. & San, K.-Y. (2005). Metabolic engineering of aerobic succinate production systems in Escherichia coli to improve process productivity and achieve the maximum theoretical succinate yield. Metab Eng 7, 116-127.

Meister, M., Saum, S., Alber, B. E. & Fuchs, G. (2005). L-malyl-coenzyme A/beta-methylmalyl-coenzyme A lyase is involved in acetate assimilation of the isocitrate lyase-negative bacterium Rhodobacter capsulatus. J Bacteriol 187, 1415-1425.

Millard, C. S., Chao, Y. P., Liao, J. C. & Donnelly, M. I. (1996). Enhanced production of succinic acid by overexpression of phosphoenolpyruvate carboxylase in Escherichia coli. Appl Environ Microbiol 62, 1808-1810.

Miller, J. H. (1992). A Short Course in Bacterial Genetics: A Laboratory Manual andHandbook for Escherichica coli and Related Bacteria. Plainview (NY):

Cold Spring Harbor Laboratory Press.

Rognstad, R. & Katz, J. (1979). Effects of 2,4-dihydroxybutyrate on lipogenesis in rat hepatocytes. J Biol Chem 254, 11969-11972.

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: A

Laboratory Manual, 2nd edn. Cold Spring Harbor: Cold Spring Harbor

Laboratory Press.

Sanchez, A. M., Bennett, G. N. & San, K.-Y. (2005). Novel pathway engineering design of the anaerobic central metabolic pathway in Escherichia coli to increase succinate yield and productivity. Metab Eng 7, 229-239.

Sauer, U. & Eikmanns, B. J. (2005). The PEP-pyruvate-oxaloacetate node as the switch point for carbon flux distribution in bacteria. Ferns Microbiol Rev 29, 765-794.

Seedorf, H., Fricke, W. F., Veith, B., Briiggemann, H., Liesegang, H., Strittmatter, A., Miethke, M., Buckel, W., Hinderberger, J. & other authors. (2008). The genome of Clostridium kluyveri, a strict anaerobe with unique metabolic features. Proc Natl Acad Sci U S A 105, 2128-2133.

Shinka, T., Inoue, Y., Ohse, M., Ito, A., Ohfu, M., Hirose, S. & Kuhara, T.

(2002). Rapid and sensitive detection of urinary 4-hydroxybutyric acid and its related compounds by gas chromatography-mass spectrometry in a patient with succinic semialdehyde dehydrogenase deficiency. J Chromatogr B Anal

Technol Biomed Life Sci 776, 57-63.

Smejkalova, H., Erb, T. J. & Fuchs, G. (2010). Methanol assimilation in

Methylobactenum extorquens AMI : demonstration of all enzymes and their regulation. Plos One 5.

Sohling, B. & Gottschalk, G. (1993). Purification and characterization of a coenzyme-A-dependent succinate-semialdehyde dehydrogenase from

Clostridium kluyveri. Eur J Biochem Febs 212, 121-127.

Vuilleumier, S., Chistoserdova, L., Lee, M.-C, Bringel, F., Lajus, A., Zhou, Y., Gourion, B., Barbe, V., Chang, J. & other authors. (2009).

Methylobactenum genome sequences: a reference blueprint to investigate microbial metabolism of CI compounds from natural and industrial sources.

Plos One 4, e5584.

Wang, Q., Ou, M. S., Kim, Y., Ingram, L. O. & Shanmugam, K. T. (2010).

Metabolic flux control at the pyruvate node in an anaerobic Escherichia coli strain with an active pyruvate dehydrogenase. Appl Environ Microbiol 76, 2107-2114.

Werpy, T. & Petersen, G. (2004). Top value added chemicals from biomass.

Results of screening for potential candidates from sugars and synthesis gas. Washington DC.

Zarzycki, J., Brecht, V., Miiller, M. & Fuchs, G. (2009). Identifying the missing steps of the autotrophic 3-hydroxypropionate C02 fixation cycle in Chloroflexus aurantiacus. Proc Natl Acad Sci U S A 106, 21317-21322.

Zelle, R. M., de Hulster, E., van Winden, W. A., de Waard, P., Dijkema, C, Winkler, A. A., Geertman, J.-M. A., van Dijken, J. P., Pronk, J. T. & van Maris, A. J. A. (2008). Malic acid production by Saccharomyces cerevisiae engineering of pyruvate carboxylation, oxaloacetate reduction, and malate export. Appl Environ Microbiol 74, 2766-2777.

Zelle, R. M., de Hulster, E., Kloezen, W., Pronk, J. T. & van Maris, A. J. A.

(2010). Key process conditions for production of C(4) dicarboxylic acids in bioreactor batch cultures of an engineered Saccharomyces cerevisiae strain.

Appl Environ Microbiol 76, 744-750.

Zhang, X., Jantama, K., Shanmugam, K. T. & Ingram, L. O. (2009).

Reengineering Escherichia coli for Succinate Production in Mineral Salts

Medium. Appl Environ Microbiol 75, 7807-7813.

Zhang, X., Wang, X., Shanmugam, K. T. & Ingram, L. O. (2011). L-malate production by metabolically engineered Escherichia coli. Appl Environ

Microbiol 77, 427-434.