FIELD OF THE INVENTION

The invention relates to a genetically modified bacterium for production of thiamine; where the bacterium is characterized by a transgene encoding a thiamine monophosphate phosphatase (TMP phosphatase having EC 3.1.3.-) as well as transgenes encoding polypeptides that catalyze steps in the thiamine pathway. The genetically modified bacterium is characterized by enhanced synthesis and release of thiamine into the extracellular environment. The invention further relates to a method for producing thiamine using the genetically modified bacterium of the invention; as well as the use of the genetically modified bacterium for thiamine production. BACKGROUND OF THE INVENTION

Thiamine, also known as vitamin B 1, is a member of the water-soluble B- complex of vitamins and is a nutritional requirement for mammals. In nature, thiamine exists in multiple phosphorylation states: thiamine, thiamine monophosphate (TMP) and thiamine diphosphate (TPP) (there are also traces of thiamine triphosphate found in cells). All living organisms use thiamine, but thiamine and its active form thiamine pyrophosphate (TPP) are only synthesized in bacteria, fungi, and plants. Animals depend on their diet for a source of thiamin, and thus, for humans, it is an essential nutrient. TPP acts in vivo as the coenzyme of enzymes executing several vital metabolic processes such as pentose phosphate pathway and the TCA cycle. Thiamine is on the World Health Organization's List of Essential Medicines that lists the most important medications needed in a basic health system and thiamine deficiency is a widespread health problem. The commercially relevant form of thiamine is un-phosphorylated, since this is the most stable form that can be assimilated and phosphorylated by humans and animals to produce the biologically active cofactor TPP.

The thiamine biosynthetic pathways characterized in bacteria, some protozoans, plants, and fungi, share some common features (Figure 1). The thiazole and pyrimidine moieties are biosynthesized separately. The pyrimidine moiety, 4-amino-5-hydroxymethyl-2-methylpyrimidine phosphate (HMP- P), is derived from 5-aminoimidazole ribotide (AIR), an intermediate in the de novo purine biosynthetic pathway. In Gram-negative bacteria, conversion of AIR to HMP-P is catalyzed by the thiC gene product. HMP-P is then phosphorylated to HMP-PP by ThiD kinase prior to coupling with the thiazole unit. The thiazole moiety, 5- (2-hydroxyethyl)-4-methylthiazole phosphate (HET-P), is derived from L-tyrosine and 1-deoxy-D-xylulose phosphate (DXP) and cysteine; where the sulfur atom likely derives from L- cysteine. This latter reaction requires expression of at least five genes thiF, this, thiG, thiH and thil.

The pyrimidine and thiazole moieties are then combined to form TMP by the action of thiamine-phosphate synthase (EC 2.5.1.3) encoded by thiE. Thus TMP is the first product of all known thiamine biosynthetic pathways. In E. coli and other Enterobacteriaceae, TMP may be phosphorylated to the cofactor TPP by a thiamine-phosphate kinase (EC 2.7.4.16) encoded by thiL in the presence of ATP. Some bacteria and eukaryotes, rely on the salvage pathway from HMP and HET which requires ThiD, ThiE and ThiM. E. coli uses a salvage enzyme, thiamine kinase, encoded by thiK to convert exogenous thiamine, taken up by the cell, into TMP. In view of their capacity to synthesize TMP and TPP, micro-organisms can be used as cell factories for the recombinant production of this vitamin. Recombinant production of essential medicines such as the vitamin thiamine requires the use of a suitable host that is capable of producing and exporting thiamine. Un-phosphorylated thiamine (THI) is the desired target for biological production, not only for commercial reasons, but also because thiamine easily crosses cell membranes, in contrast to its phosphorylated forms, and therefore accumulates outside of the production host during fermentation, where it is the most stable form.

The advantages of recombinant Escherichia coli as a cell factory for production of bioproducts are widely recognized due to the fact that: (i) it has unparalleled fast growth kinetics; with a doubling time of about 20 minutes when cultivated in glucose-salts media and under optimal environmental conditions, (ii) it easily achieves a high cell density; where the theoretical density limit of an E. coli liquid culture is estimated to be about 200 g dry cell weight/I or roughly 1 x 10 ¹³ viable bacteria/ml. Additionally, there are many molecular tools and protocols at hand for the high-level production of bioproducts in E. coli; heterologous proteins can easily be expressed in E. coli and there are many specialty strains available for the production of specialist end-products.

In most bacteria, TPP is produced by the direct conversion of TMP to TPP, where cell growth requires an intracellular or extracellular supply of TPP. Thus the use of E. coli as a cell factory for thiamine production requires a genetically modified strain that is both viable and at the same time is capable of producing and releasing thiamine into the extracellular environment.

Summary of the invention

The invention provides a genetically modified bacterium for production of thiamine; wherein said bacterium is characterized by having seven transgenes encoding, respectively: a. a polypeptide having thiamine mono-phosphate phosphatase activity (E.C. 3.1.3.-) wherein the amino acid sequence of said polypeptide has at least 80% sequence identity to a sequence selected from the group consisting of SEQ ID No: 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30, 32, 34, 36, 38, 40, 42, 44, 46, 48, 50, 52, 54, 56, 58, 60, 62, 64, 66, 68, 70, 72 and

74; b. a polypeptide having 4-amino-5-hydroxymethyl-2- methylpyrimidine phosphate (HMP-P) synthase activity (E.C. 4.1.99.17); c. a polypeptide having thiamine phosphate synthase activity (E.C.

2.5.1.3); d. a polypeptide having ThiS adenylyltransferase activity (E.C.

2.7.7.73); a polypeptide having ThiS sulfur-carrier activity; e. a polypeptide having thiazole synthase activity (E.C. 2.8.1.10); f. a polypeptide having 2-iminoacetate synthase activity (E.C. 4.1.99.19), or having glycine oxidase activity (EC 1.4.3.19); and g. a polypeptide having phosphohydroxymethylpyrimidine kinase activity (E.C. 2.7.4.7), and wherein the genus of bacterium is selected from the group consisting of Acetobacter, Azotobacter, Brevibacterium, Burkholderia, Campylobacter, Corynebacterium, Escherichia, Propionibacterium, and Streptomyces.

According to a further embodiment, the genetically modified bacterium of the invention is further characterized by a genetically modified endogenous thiL gene capable of expressing reduced thiamine-phosphate kinase activity (EC 2.7.4.16) as compared to the parent endogenous thiL gene.

According to a further embodiment, the genetically modified bacterium of the invention is further characterized by inactivation or deletion of one or more genes encoding a protein selected from the group: thiamine ABC transporter periplasmic binding protein; thiamine ABC transporter permease; and thiamine ABC transporter ATPase.

The invention further provides a method for producing thiamine comprising the steps of: a. introducing a genetically modified bacterium according to any

embodiment of the invention into a growth medium to produce a culture; b. cultivating the culture; and c. recovering thiamine produced by said culture, and optionally purifying the recovered thiamine.

The invention further provides for the use of a transgene encoding a polypeptide having thiamine mono-phosphate phosphatase activity (E.C 3.1.3.-) to enhance thiamine export in a bacterium.

According to one embodiment for the use of the transgene encoding a polypeptide having thiamine mono-phosphate phosphatase activity (E.C 3.1.3.-), the amino acid sequence of said polypeptide has at least 80% sequence identity to a sequence selected from among : a. the group consisting of SEQ ID No: 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28 and 30; b. the group consisting of SEQ ID No : 32, 34, 36, 38, 40, 42, 44, 46, 48,

50, 52, 54, 56, 58, 60, 62, 64 and 66; and c. the group consisting of SEQ ID No: 68, 70, 72 and 74.

The invention further provides for the use of a genetically modified bacterium according to any one embodiment of the invention for the production of thiamine.

Brief description of the drawings

Figure 1 Cartoon showing intermediates of the thiamine pathway in microorganisms and the respective enzymatic steps leading to synthesis of thiamine (THI); thiamine monophosphate (TMP) and thiamine diphosphate (TPP). Abbreviation of intermediates: 5-aminoimidazole ribonucleotide (AIR), 4-amino-2-methyl-5-(phosphooxymethyl)pyrimidine (HMP-P), 4-amino-2- methyl-5-(diphosphomethyl) pyrimidine (HMP-PP), 1-deoxy-D-xylulose 5- phosphate (DXP), dehydroglycine (DHG), 4-methyl-5-(2- phosphooxyethyl)thiazole (THZ-P), adenosine triphosphate (ATP), adenosine monophosphate (AMP), S-adenosyl-L-methionine (SAM), reduced nicotinamide adenine dinucleotide phosphate (NADPH), nicotinamide adenine dinucleotide phosphate (NADP+), reduced ferredoxin (Fdx red), oxidized ferredoxin (Fdx ox). Figure 2 Bar diagram showing the extracellular and intracellular levels of THI, TMP and TPP detected in cultures of genetically modified E. coli strains transformed with a vector expressing one of 5 different phosphatase genes: pBS93 with the A. thaliana gene (At5g32470); pBS94 with the S. aureus Newmann phosphatase rgsA gene; pBS95 with the Pseudomonas phosphatase gene; pBS96 with the S. cerevisiae pho3 gene (YBR092C); and pBS97 with the an E. coli nudJ gene, as compared to a strain comprising pBS92, a control vector. Additionally each strain comprises the vector pBS140 comprising genes thiMD and thiCEFSGH constitutively expressing enzymes in the thiamine biosynthesis pathway; to be compared with the control strain transformed with empty vector (pGEN49) lacking the genes encoding the pathway enzymes. The intracellular concentrations are calculated using the entire culture volume, i.e. the indicated level is the effective concentration if the intracellular metabolite had been released into the culture volume.

Figure 3 Neighbor joining tree showing distribution of TMP phosphatases in 3 clades based on alignment of amino acid sequences. Distance scale bar indicates changes / amino acid residue. Clade 1 comprises the A. thaliana gene (At5g32470); Clade 2 comprises an Anaerotruncus colihominis gene (WP_006874980) (Hasnain et al.2016); Clade 3 comprises a Syntrophomonas wolfei gene (WP_011640074.1) (Hasnain et al.2016).

Figure 4 Phylogenetic tree of TMP phosphatases belonging to (A) clade 1 ; (B) clade 2 and (C) clade 3 based on alignment of amino acid sequences. Each member of the three clades comprises a haloacid domain (HAD) known to catalyze dephosphorylation of thiamine monophosphate. Clade 1 consists of homologues of At5G32470 and spans the Viridiplantae and features a TenA/Teni-4 family domain (PFam : PF03070) at the N-terminus and a HAD in the C-terminal half (see their alignment in figure 5). Clade 2 represent bacterial genes often found as fusion with other thiamine genes and include the gene from Anaerotruncus colihominis (WP_006874980). Clade 3 represents a clade of bacterial genes that includes Syntrophomonas wolfei (WP_011640074.1), which are often found with other thiamine genes. The outgroup leaves are indicated in italics and the representatives of each clade that have been characterized as TMP phosphatases are indicated in bold. Arrows identify genes tested in example 4.4.

Figure 5 This figure shows an alignment of the protein sequences of 11 TMP phosphatases belonging to clade 1 (Arabidopsis thaliana AT5G32470.1; Brassica napus BnaC08g46820D_ emb| CDY62623.1 | ; Citrus sinensis C17G9.12c-like isoform Xl_ref| XP_006484613.1 | ; Glycine max LOC100788250 isoform Xl_ref| XP_003536133.1 | ; Jatropha curcas CGZ_23571 [gb| KDP23738.1 | ; Nicotiana tomentosiformis LOC104108252_ref| XP_009615535.1 | ; Oryza sativa

Os08g0566000_ref| NP_001062539.1 | ; Pyrus x bretschneideri LOC103968121 isoform Xl_ref| XP_009379735.1 | ; Prunus persica

PRUPE_ppa003431mg_>ref| XP_007199656.1 | ; Populus trichocarpa POPTR_0019s03960g_ref| XP_002325785.2 | ; Zea mays

LOC103653246_ref| XP_008678418.1 | ). The protein sequences were aligned using the 'geneious alignment' function (standard settings) of the geneious 6.0.6 software (Biomatters Ltd.). The TenA domain is underlined with a dashed line and the HAD domain is underlined with a dotted line. TenA active site residues (based on the crystal structure of representative a representative TenA protein - TenA from B. subtilis, PDB ID Number: 1YAF - are marked with * and residues shown to form H-bonds with HMP are marked with +. Conserved residues of a HAD domain are marked with # and the conserved DXD and DXXXD motifs are underlined with a solid line. Figure 6 Bar diagram showing the extracellular and intracellular levels of THI, TMP and TPP detected in cultures of E. coli strains with the following genetic modifications: cells deleted for thiK gene (thiK ) in the central bars; cells carrying a mutated (partially inactivated) thiL gene (thiL*) in the right hand bars; and cells with wild-type thiK and thiL genes in the left hand bars. The E. coli strains in the lower panel comprise the pBS93 vector encoding the A. thaliana TMP phosphatase (At5g32470); and the strains in the upper panel comprise the control vector pBS92. Additionally, all strains comprise the vector pBS140 comprising genes thiMD and thiCEFSGH constitutively expressing enzymes in the thiamine biosynthesis pathway. Figure 7 . Bar diagram showing the relative extracellular levels of THI detected in cultures of E. coli strains carrying a mutated (partially inactivated) thiL gene (thiL*) and comprising either the vector pBS140 comprising genes thiMD and thiCEFSGH constitutively expressing enzymes in the thiamine biosynthesis pathway or the empty control vector pBSlOO. Additionally all strains comprise a vector encoding a TMP phosphatase or the corresponding empty control vector pBS92. The phosphatases tested from clade 1 are pBS93 : A. thaliana TMP phosphatase At5g32470 [SEQ ID No. : 2], pBS791 : Jatropha curcas TMP phosphatase KDP23738.1 [SEQ ID No. : 14], and pBS792: Picea sitchensis TMP phosphatase ABR16455 [SEQ ID No. : 26] . Those tested from clade 2 are pBS793 : Eubacterium ventriosum TMP phosphatase WP_005362972 [SEQ ID No. : 34] and pBS794: Anaerotruncus colihominis TMP phosphatase WP_006874980 [SEQ ID No. : 32] . And those tested from clade 3 are pBS797 : Desulfitobacterium hafniense TMP phosphatase WP_018212876 [SEQ ID No. : 70] and pBS798: Syntrophomonas wolfei TMP phosphatase WP_011640074 [SEQ ID No. : 68] . Data shown is the average of triplicates; error bars indicate standard deviation.

Abbreviations and terms:

gi number: (genlnfo identifier) is a unique integer which identifies a particular sequence, independent of the database source, which is assigned by NCBI to all sequences processed into Entrez, including nucleotide sequences from DDBJ/EMBL/GenBank, protein sequences from SWISS-PROT, PIR and many others.

Amino acid sequence identity: The term "sequence identity" as used herein, indicates a quantitative measure of the degree of homology between two amino acid sequences of substantially equal length. The two sequences to be compared must be aligned to give a best possible fit, by means of the insertion of gaps or alternatively, truncation at the ends of the protein sequences. The sequence identity can be calculated as ((Nref- Ndif) 100)/(Nref), wherein Ndif is the total number of non-identical residues in the two sequences when aligned and wherein Nref is the number of residues in one of the sequences. Sequence identity calculations are preferably automated using the BLAST program e.g. the BLASTP program (Pearson W.R and DJ. Lipman (1988)) (www.ncbi.nlm.nih.gov/cgi-bin/BLAST). Multiple sequence alignment is performed with the sequence alignment method ClustalW with default parameters as described by Thompson J., et al 1994, available at http://www2.ebi.ac.uk/clustalw/.

Preferably, the numbers of substitutions, insertions, additions or deletions of one or more amino acid residues in the polypeptide as compared to its comparator polypeptide is limited, i.e. no more than 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 substitutions, no more than 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 insertions, no more than 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 additions, and no more than 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 deletions. Preferably the substitutions are conservative amino acid substitutions: limited to exchanges within members of group 1 : Glycine, Alanine, Valine, Leucine, Isoleucine; group 2: Serine, Cysteine, Selenocysteine, Threonine, Methionine; group 3 : proline; group 4: Phenylalanine, Tyrosine, Tryptophan; Group 5: Aspartate, Glutamate, Asparagine, Glutamine.

Genetically modified bacterium for production of thiamine, is derived from a member of a genus of bacterium wherein TPP biosynthesis is mediated via the thiL pathway and not the thiN pathway (Figure 1).

Mutant gene: a mutant gene in the genome of a bacterial cell may exhibit reduced function of the gene and hence where the mutant gene encodes a polypeptide the mutation may results in a loss/reduction of expression of the encoded polypeptide. Alternatively the mutant gene may encode a mutant polypeptide, and where the polypeptide is an enzyme, the mutation may result in a loss of detectable enzymatic activity in the bacterial cell.

Native gene: endogenous gene in a bacterial cell genome, homologous to host micro-organism.

Detailed description of the invention

Adaptation of E. coli for use as a cell factory for thiamine production requires the provision and/or manipulation of enzymes capable of controlling the intracellular phosphorylation state of thiamine. Two approaches can be taken to shift the intracellular equilibrium away from TPP and TMP and towards thiamine.

The first approach involves modification (e.g. down-regulation) of one or more of the kinases that phosphorylate thiamine (e.g. salvage kinase, ThiK in E. coli); the kinases that pyro-phosphorylate thiamine (ThiN in Bacillus subtilis; commonly found in eukaroytes, but not found in E. coli); and the kinases that phosphorylate TMP (thiL in E. coli). This approach has been used in B. subtilis (WO2004106557).

The second approach that has not previously been described, involves the expression of recombinant enzymes in the cell to dephosphorylate TMP, and optionally TPP in combination with the upregulation of the thiamine biosynthesis pathway. This approach may be further combined with one or more modifications of the first approach. TMP phosphatases and TPP phosphatases suitable for this approach are limited to those that selectively act on TMP and TPP. However, the best-studied phosphatases are promiscuous nucleotide phosphatases (e.g. E. coli nudJ encoded phosphatase) that act on both nucleoside di- and triphosphate and the structurally similar TMP and TPP. Such phosphatases are unsuitable, since their non-specific activity will destabilize cellular metabolism.

In some organisms (mostly eukaryotes) TMP is dephosphorylated before being pyrophosphorylated to TPP (see figure 1); whereby THI is a key intermediate in TDP and TPP biosynthesis. However, the identity of the responsible phosphatase has long been the subject of speculation, and its encoding gene was described as a missing gene (Goyer et al., 2013). Thiamine metabolism, in vivo, not only requires phosphorylases for the dephosphorylation of TDP and TMP, but also for the disposal of their corrupt forms, oxy- and oxo-thiamin. In a search for TMP phosphatases capable of selectively disposing of toxic forms of thiamine in plants, Goyer et al., 2013 suggested that an Arabidopsis protein sequence (At5g32470) was a plausible candidate, because annotation of its sequence predicted a haloacid dehalogenase (HAD) domain fused to a thiaminase 2 (TenA) family protein, which in other organisms TenA proteins were known to serve this selective function for disposal of oxy- and oxo-thiamin. However, the only phosphatase activity detected in extracts of Arabidopsis by Komeda Y et al., (1988) corresponded to a protein of lOkDa, significantly smaller than the 69kDa protein encoded by the AT5G32470.1 gene.

The present invention relates to the provision of a genetically modified bacterium that is capable of producing and releasing enhanced amounts of thiamine into the extracellular environment as compared to the parent from which was derived.

I A genetically modified bacterium for production and export of thiamine

The present invention provides a genetically modified bacterial cell capable of producing and exporting enhanced levels of thiamine. The bacterial cell of the invention comprises a transgene encoding a thiamine monophosphate phosphatase (TMP phosphatase having EC 3.1.3.-) as well as transgenes encoding polypeptides that catalyze steps in the thiamine pathway. The activity of the polypeptides that catalyze steps in the thiamine pathway enhances the synthesis of both intermediates in the thiamine pathway and products of the thiamine pathway (e.g. TMP and TPP) in the bacterial cell.

A polypeptide having thiamine mono-phosphate phosphatase activity (E.C 3.1.3.-) according to the invention is a member of a family of TMP phosphatase enzymes having a haloacid domain that catalyzes the dephosphorylation of thiamine monophosphate. The members of this family are encoded by genes belonging to three clades (Example 3). The amino acid sequence of the polypeptide having TMP phosphatase activity has at least 70, 75, 80, 85, 90, 95, 96, 98, 100% amino acid sequence identity to a sequence selected from any one of: SEQ ID No. : 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28 and 30 encoded by genes of clade 1; SEQ ID No. : 32, 34, 36, 38, 40, 42, 44, 46, 48, 50, 52, 54, 56, 58, 60, 62, 64 and 66 encoded by genes of clade 2; and SEQ ID No. : 68, 70, 72 and 74 encoded by genes of clade 3. The clade 1 gene from A. thaliana (AT5G32470.1) and its orthologues listed in Figure 4A; the Clade 2 genes from Anaerotruncus colihominis and Dorea longicatena and their orthologues listed in Figure 4B; and the clade 3 gene from Syntrophomonas wolfei and its orthologues listed in Figure 4C, all encode polypeptides having TMP phosphatase activity (E.C 3.1.3.-).

In one embodiment the amino acid sequence of the polypeptide having TMP phosphatase activity has at least 60, 65, 70, 75, 80, 85, 90, 95, 96, 98, 100% amino acid sequence identity to SEQ ID No. : 2; where the polypeptide is characterized by comprising TenA-like domain, likely important for recognition of thiamine, and a HAD-domain for hydrolysis of the phosphate bond (Example 2). Proteins with this signature structure can be overexpressed in a bacterial cell of the invention in order to shift the thiamine phosphorylation profile towards thiamine.

In one embodiment the amino acid sequence of the polypeptide having TMP phosphatase activity has at least 80, 85, 90, 95, 96, 98, 100% amino acid sequence identity to a sequence selected from any one of: SEQ ID No. : 2, 14, 26, 32, 34, 40, 68 and 70. In one embodiment the amino acid sequence of the polypeptide having TMP phosphatase activity has at least 80, 85, 90, 95, 96, 98, 100% amino acid sequence identity to a sequence selected from any one of: SEQ ID No. : 2, 14, 26, 32, 34, 68, and 70.

The polypeptides that are encoded by the transgenes in the genetically modified bacterium, and whose activity serves to enhance the synthesis of both intermediates and products of the thiamine pathway, are as follows: b) a polypeptide having 4-amino-5-hydroxymethyl-2-methylpyrimidine phosphate (HMP-P) synthase activity (E.C. 4.1.99.17); such as a polypeptide with an amino acid sequence having 80, 85, 90, 95 or 100% sequence identity to SEQ ID No. : 76, 78, 80 and 82;

c) a polypeptide having thiamine phosphate synthase activity (2.5.1.3), such as a polypeptide with an amino acid sequence having 80, 85, 90, 95 or 100% sequence identity to SEQ ID No. : 84;

d) a polypeptide having ThiS adenylyltransferase activity (2.7.7.73), such as a polypeptide with an amino acid sequence having 80, 85, 90, 95 or 100% sequence identity to SEQ ID No. : 86;

e) a polypeptide having ThiS sulfur-carrier activity (immediate sulfur donor in thiazole formation), such as a polypeptide with an amino acid sequence having 80, 85, 90, 95 or 100% sequence identity to SEQ ID No. : 88.

f) a polypeptide having thiazole synthase activity (2.8.1.10), such as a polypeptide with an amino acid sequence having 80, 85, 90, 95 or 100% sequence identity to SEQ ID No. :90;

g) a polypeptide having 2-iminoacetate synthase activity (4.1.99.19; also called L-tyrosine 4-methylphenol-lyase) such as a polypeptide with an amino acid sequence having 80, 85, 90, 95 or 100% sequence identity to SEQ ID No. :92; or a polypeptide having Glycine oxidase activity (EC 1.4.3.19) such as a polypeptide with an amino acid sequence having 80, 85, 90, 95 or 100% sequence identity to SEQ ID No. :94, 96 and 98; and

h) a polypeptide having phosphohydroxymethylpyrimidine kinase activity (2.7.4.7), such as a polypeptide with an amino acid sequence having 80, 85, 90, 95 or 100% sequence identity to SEQ ID No. : 100. The genetically modified bacterium may additionally be characterized by a genetic modification to reduce the expression of one or more of the three native endogenous thiBPQ genes and thereby reduce thiamine export from the bacterium. The three genes thiB, thiP and thiQ encode a thiamine ABC transporter periplasmic binding protein; a thiamine ABC transporter permease; and a thiamine ABC transporter ATPase, respectively. For example, thiamine export can be reduced by inactivating or deleting (by gene knockout) one or more of the three genes thiB (or thiP or thiQ) in the genome of the bacterium. Preferably, the amino acid sequence of the polypeptide encoded by the inactivated/deleted thiB gene has at least 80% 85, 90, 95 or 100% sequence identity to SEQ ID No. : 102; the polypeptide encoded by the inactivated/deleted thiP gene has at least 80% 85, 90, 95 or 100% sequence identity to SEQ ID No. : 104; and the polypeptide encoded by the inactivated/deleted thiQ gene has at least 80% 85, 90, 95 or 100% sequence identity to SEQ ID No. : 106.

The genetically modified bacterium may additionally be characterized by a further transgene encoding a polypeptide having hydroxyethylthiazole kinase activity (2.7.1.50), such as a polypeptide with an amino acid sequence having 80, 85, 90, 95 or 100% sequence identity to SEQ ID No. : 108.

The expression of a transgene encoding a thiamine monophosphate phosphatase (TMP phosphatase having EC 3.1.3.-) in combination with transgenes encoding polypeptides that catalyze steps in the thiamine pathway in the genetically modified bacterium of the invention, not only enhances the production of thiamine compounds, but additionally, it drives the pathway towards the production of un-phosphorylated thiamine (Example 1). As shown in Example 4, levels of extracellular thiamine were increased 6 fold, while total production of thiamine compounds was enhanced by 20%, when compared to cells lacking the transgene encoding TMP phosphatase. The surprisingly high levels of thiamine produced and released into the extracellular medium by the genetically modified bacterium of the invention meet the key requirements for its use as a cell factory for thiamine production. In a further embodiment, the genetically modified bacterium of the invention is further characterized by a genetically modified endogenous thiL gene that expresses reduced thiamine-phosphate kinase activity (EC 2.7.4.16) as compared to the parent endogenous thiL gene. The genetically modified thiL gene may be mutated to express lower amounts of the encoded polypeptide; or it can be mutated to encode a polypeptide having reduced thiamine- phosphate kinase activity (EC 2.7.4.16). For example, when the amino acid sequence of the encoded polypeptide has at least 80% sequence identity to SEQ ID No. : 110 and with the proviso that the sequence has an amino acid residue substitution G133D (or a G→D substitution in an equivalent position on the basis of sequence alignment); then it will have a reduced thiamine- phosphate kinase activity (EC 2.7.4.16). As shown in Example 4, when genetically modified bacteria comprising transgenes for the thiamine pathway alone, were compared with bacteria further comprising a transgene encoding a TMP phosphatase and a mutant thiL gene (reduced THIL activity), the levels of extracellular thiamine production were increased 7 fold, while total production of thiamine compounds was enhanced by 24%.

The genetically modified bacterium of the invention has a transgene that comprises a promoter operably linked to a coding sequence encoding a thiamine monophosphate phosphatase (TMP phosphatase having EC 3.1.3.-); as well as transgenes encoding polypeptides that catalyze steps in the thiamine pathway. A promoter may be operably linked to each thiamine pathway polypeptide coding sequence; or the respective coding sequences may form an operon that is operably linked to a single promoter.

Suitable promoters include both constitutive promoters (e.g. apFAB46 [SEQ ID No. : 147] apFAB70 [SEQ ID No. : 148], apFAB71 [SEQ ID No. : 149]), as well as inducible promoters (e.g pBAD ara promoter [SEQ ID No. : 150] when co-expressed with araC gene for regulation (Guzman et al. 1995) or pLac promoter with lacO operator site [SEQ ID No. : 151] (Norrander et al. 1983)).

The genetically modified bacterium according to the invention, for the production and export of thiamine, is a member of a genus of bacterium wherein TPP biosynthesis is mediated via the thiL pathway (Figure 1) and not the thiN pathway. Accordingly the genus of bacterium may be selected from the group consisting of Acetobacter, Azotobacter, Brevibacterium, Burkholderia, Campylobacter, Corynebacterium, Escherichia, Propionibacterium, and Streptomyces. Preferably, the bacterium of the invention is a species of Escherichia, e.g. Escherichia coli.

II A method for producing thiamine using a recombinant microorganism expressing a thiamine mono-phosphate phosphatase

Thiamine can be produced and exported using microbial cells of the invention (e.g. recombinant bacterial cells) by introducing the cells into a culture medium comprising a carbon source for biosynthesis of one or more of thiamine, TMP and TPP; and finally recovering the thiamine produced by the culture, as illustrated in the Examples.

The bacterial cells of the invention will produce thiamine when supplied with a suitable carbon source including glucose, maltose, galactose, fructose, sucrose, arabinose, xylose, raffinose, mannose, and lactose.

III A method of detecting thiamine produced and exported by a recombinant microorganism expressing a thiamine mono-phosphate phosphatase

Methods for detecting and quantifying extracellular and intracellular thiamine produced by a micro-organism of the invention include High Pressure Liquid Chromatography, relative to a thiamine standard. For example, individual thiamine compounds, thiamin, TMP, and TPP can be measured using a modified thiochrome-HPLC assay procedure described previously (Chie et al., 1999). Briefly, 100 μΙ of culture supernatant or intracellular extracts are added to 200 μΙ of 4M potassium acetate. The sample is then oxidized by the addition of 100 μΙ fresh 3.8 mM potassium ferricyanide in 7 M NaOH. The mixture is vigorously mixed and then quenched by addition of 100 μΙ fresh 0.06% H ₂ 0 ₂ in saturated KH ₂ P0 ₄ . Samples are neutralized with 6M HCI and are transferred to HPLC vials and injected onto a Supelcosil LC-18-T column (15cm X 4.6 mm, 3 μΐη) (Supeico-Ref. No 58970-U). Elution is made by a 10%-35% methanol (H ₂ 0 50%-25%) gradient in the presence of 40% 0.1 M K ₂ HP0 ₄ (pH 6.6) and 4 mM tetrabutyl ammonium hydrogen sulfate. Fluorescence is measured at 444 nm after excitation at 365 nm. The chronological order of elution from the column is thiamin, TMP, and TPP. This procedure was utilized to monitor both internal and external thiamine production during fermentation. Alternatively, direct measurement of thiamine and the intermediates HMP and HET in the fermentation broth can be performed by chromatography of samples on a Phenomenex LUNA C18 column, using an Agilent 1100 HPLC system equipped with a thermostated autosampier and a diode array detector (DAD). The column dimensions are 150 x 4.6 mm, particle size 5 micron. The column temperature is kept constant at 20 C. The mobile phase is a mixture of 0.4 g pentane sulfonate in water, pH 2 (A) and methanol (B).

IV Methods for producing a genetically modified bacterium for production and export of thiamine

Integration and self-replicating vectors suitable for cloning and introducing one or more gene encoding one or more a polypeptide having an enzymatic activity associated with thiamine synthesis in a bacterium of the invention are commercially available and known to those skilled in the art (see, e.g., Sambrook et al., Molecular Cloning : A Laboratory Manual, Second Edition, Cold Spring Harbor Laboratory Press, 1989). Cells of a bacterium are genetically engineered by the introduction into the cells of heterologous DNA. Heterologous expression of genes encoding one or more polypeptide having an enzymatic activity associated with thiamine synthesis in a bacterium of the invention is demonstrated in the Example 1 and 3.

A nucleic acid molecule, that encodes one or more polypeptide having an enzymatic activity associated with thiamine synthesis according to the invention, can be introduced into a host cell by means of a self-replicating vector or optionally integrated into the host cell genome using methods and techniques that are standard in the art. For example, nucleic acid molecules can be introduced by standard protocols such as transformation including chemical transformation and electroporation, transduction, particle bombardment, etc. Expressing the nucleic acid molecule encoding the enzymes of the claimed invention also may be accomplished by integrating the nucleic acid molecule into the genome. Genetic modification of the native endogenous thiL gene in a bacterium of the invention is performed to reduce expression of thiamine-phosphate kinase activity (EC 2.7.4.16) as compared to the parent endogenous thiL gene. For example, the native thiL gene may be mutated to encode a polypeptide having reduced thiamine-phosphate kinase activity (EC 2.7.4.16), as described in Example 4.

Genetic modification of one or more of the native endogenous thiBPQ genes in a bacterium of the invention is performed to reduce re-uptake of extracellular thiamine export into the bacterium. The deletion (knockout) of any one of the three genes thiB (or P or Q) in the genome of the bacterium (e.g. E. coli K12 strain), using standard recombineering methods (Datsenko KA, et al. ; 2000), is sufficient to reduce thiamine import as compared to a parent bacterium. For example, the thiB: : Kan construct (tbpA re-named as thiB) can be PCR amplified from strain JW0067 of the Keio collection (thiP from JW0066; thiQ from JW0065) (Baba et al., 2006). The amplified fragment can be transformed by electroporation into competent host cells (e.g. E. coli) carrying the ARed recombinase genes expressed from an inducible promoter (pKD46) (Datsenko KA, et al. 2000). Successful integrants are selected on Kanamycin medium and confirmed by colony PCR. The Kanamycin resistance (kanR) cassette can be eliminated by transforming cells with the temperature- sensitive plasmid pcp20, which expresses the FLP recombinase. Ampicillin- resistant colonies can be isolated at 30°C and then restreaked non-selectively at 42°C. Loss of KanR can be confirmed by colony PCR.

Examples:

Example 1. Identification of a thiamine specific phosphatase

A key precursor for TDP and TPP biosynthesis in many micro-organisms and plants is THI; which in turn is derived from TMP by the action of a phosphatase.

Five different putative phosphatases, predicted to have phosphatase activity towards the substrates TMP or TPP, were identified based on structure/functional prediction. Nucleic acid molecules, whose nucleotide sequences encoded each phosphatase, were individually cloned into an expression vector giving the following five phosphatase expression vectors (+ empty vector control), as shown in Table 1.

Table 1

1.1 Vector construction Vector pBS92 was constructed using the one-step-isothermal DNA assembly method developed by Gibson et al., (2009). The backbone, carrying the SpecR cassette and the SC101 origin of replication, was amplified from PZS4Int-tetR (http://www.expressys.com/main_tools. html) using primers oBS196 and oBS197 and a gene expression cassette comprising an apFAB70 promoter (SEQ ID No. : 148) and apFAB381 terminator (SEQ ID No. : 154) were synthesized and then amplified using primers pBS194 and oBS195. Purified PCR fragments were assembled by one pot isothermal assembly using the NEB Gibson assembly master mix (#E2611) and standard protocol according to the manufacturer's instructions (Gibson et al., (2009). The assembled vectors were transformed into electro-competent DH 10B cells by electroporation and the recovered cells were plated on Luria Broth (LB) Spectinomycin plates. Vectors pBS93-97 were constructed by amplifying pBS92 with the primer pair oBS198 and oBS199; and the purified PCR product was assembled with the respective phosphatase gene using the NEB Gibson assembly master mix according to the manufacturer's instructions . The phosphatase genes cloned in each vector (Table 1) were amplified as follows:

pBS93 : Primers oBS220, oBS221 were used to amplify the synthetic gene encoding Arabidopsis thaliana AT5G32470.1 phosphatase which was codon optimized for E. coli, and synthesized as two gene blocks;

pBS94: Primers oBS202, oBS203 were used to amplify the rsgA gene from S. aureus Newmann genomic DNA;

pBS95 : Primers oBS201, oBS200 were used to amplify the coding sequence for an acid phosphatase (EC 3.1.3.2) gene from Pseudomonas.

pBS96 : Primers: oBS201, oBS202 were used to amplify the S. cerevisiae PH03 gene, which was codon optimized for E. coli, and synthesized as one gene block;

pBS97 : Primers oBS210, oBS211 were used to amplify the nudJ gene (MG1655) from E. coli genomic DNA.

Each of the assembled vectors was transformed into electro-competent DH 10B cells by electroporation and the recovered cells were plated on selective LB plates. This host DH 10B strain (F- mcrA A(mrr-hsdRMS-mcrBC) O80dlacZAM 15 AlacX74 endAl recAl deoR A(ara,leu)7697 araD139 galU galK nupG rpsL λ-) comprises a chromosomal deletion, resulting in the deletion of three genes (thiBPQ corresponding to EcoGenes: EG11574, EG11573, EG11572) respectively encoding : Thiamine ABC transporter periplasmic binding protein; Thiamine ABC transporter permease; and Thiamine ABC transporter ATPase.

Vector pBS140 was constructed following a two-step hierarchical cloning procedure. First the thiC operon and the thiM operon were assembled into the background vectors pGEN49 (comprising an apFAB46 promoter (SEQ ID No. : 147) and an apFAB377 terminator (SEQ ID No. : 153)) and pGEN50 (comprising an apFAB71 promoter (SEQ ID No. : 149)) and an apFAB378 terminator (SEQ ID No. : 152)) respectively, and then the two resulting vectors were combined to form pBS140. The thiC operon was amplified from MG1655 genomic DNA using primers oGEN264 and oGEN227 and the thiM operon was amplified from MG1655 genomic DNA using primers oBS422 and OBS421. Vectors pGEN49 and pGEN50 were amplified with primers oGEN265 and 266 and then assembled with the thiC and thiM operon respectively using one pot isothermal assembly. The resulting vectors were named pBS116 (pGEN49 + thiC operon) and pBS117 (pGEN50 + thiM operon). The assembled vectors were transformed into electro-competent DH 10B cells by electroporation and the recovered cells were plated on LB Kan plates. The sequence of the resulting vectors was confirmed by Sanger sequencing.

The confirmed vector pBS116 was purified and digested for > 8 h using Swal (NEB # R0604). The linearized DNA was gel-purified. Confirmed vector pBS117 was used as a template for PCR with primers oGEN 182 and oGEN 184.

The resulting PCR fragment was gel-purified. The amplified vector and thiM operon were assembled using NEB Gibson assembly master mix, according to the manufacturer's instructions and the mixture was transformed into electro- competent DH 10B cells by electroporation and the recovered cells were plated on LB Kan plates. The sequence of the resulting plasmid pBS140 was confirmed by Sanger sequencing.

Table 2 List of primers Primer Sequence SEQ ID NO. : name

OBS194 GTCCTACTCAGGAGAGCGTTCACCGACAACTCAGGAGAGC 119

GTTCACC

OBS195 CTTTCGTCTTCACCTCGAGGGAAATCAAAATAGGCGTATCA 120

CGAGGCC

OBS196 GATTTCCCTCGAGGTGAAGACGAAAG 121

OBS197 TGTCGGTGAACGCTCTCCTG 122

OBS198 ATTATTACTCGTGTGTTGTCAGAAAG 123

OBS199 CTAGTATTACCTCGCTATTAGTGACGTAATAGGAGGTAAGC 124

OBS200 CTATTACGTCACTAATAGCGAGGTAATACTAGATGTGCCAG 125

CAGCATCCGC

OBS201 CGGAGGCC 1 1 1 U GACAACACACGAG 1 AA 1 AA 1 U AG 1 1 1 1 126

CTAGAGGCAGCGC

OBS202 CTATTACGTCACTAATAGCGAGGTAATACTAGATGAAGACA 127

GGTCGAATAGTG

OBS203 CGGAGGCCTTTCTGACAACACACGAGTAATAATTTAATATC 128

TAACCTTTCTATTTG

OBS210 CCTATTACGTCACTAATAGCGAGGTAATACTAGATGTTTAA 129

ACCGCACGTTACCG

OBS211 CTTTCTGACAACACACGAGTAATAATTTAGATGACACCCTTT 130

GTAAAAGG

OBS220 GTACCTATAATGTGTGGATGTCCCACCGCTTACCTCC 131

OBS221 CATCACCATCATCACCACTGAATTATTACTCGTGTGTTGTCA 132

GAAAG

OGEN 184 AAACCTCTTTATGTTGCAGTCG 133

OGEN 182 AAATTCGCGAGTTCCACTAAGA 134

OGEN227 CCGCTTACCTCCTATTACGTCACTAATAGCTAAGGAGGTAA 135

ATATGTCTGCAACAAAACTGACCCGCC

OGEN264 CGGAGGCCTTTCTGACAACACACGAGTAATAATTCATAGTC 136

TTTGCGAGGCG

OGEN265 ATTATTACTCGTGTGTTGTCAGAAAGGCCTCCG 137 OGEN266 ATTAGTGACGTAATAGGAGGTAAGCGGTGGG 138

OBS421 CGGAGGCCTTTCTGACAACACACGAGTAATAATTCACCACC 139

AGGCGTGGAAG

OBS422 GCTTACCTCCTATTACGTCACTAATAGCTAAGGAGGTAAAT 140

ATGCAAGTCGACCTGCTGG

OBS445 CGGAACAAAGCCGTGGATGTCCAACGTCATTGATAATGGC 141

CCACGCGTGGTATCGCCGCCAATGAGTTGCATATCGTAATA

ATTGAGAAG

OBS446 CCAGAGCGCGTTAAGGCTCGTCCCATCGGAACAAAGCCGT 142

GGATGTCCAACGTCATTGATAATGGCCCACGCGTGGTATC

GCCGCCAATG

OBS456 CATGTGGCGAGTTCTCCCTG 143

OBS457 CAGGTAAACGGTACGCCCAG 144

OBS521 GCAACTGTCGATGGCGAAGC 145

OBS522 CCTGATCAACCGCCACCAC 146

1.2 Expression of putative TMP phosphatase genes in genetically modified E. coli cells

The five candidate TMP phosphatase genes: A. thaliana gene (At5g32470); S. aureus Newmann phosphatase rgsA gene; Pseudomonas phosphatase gene; S. cerevisiae pho3 gene (YBR092C); and an E. coli nudJ gene were each expressed in an E. coli host strain comprising the vector pBS140. The pBS140 vector comprises the thiamine pathway genes thiMD and thiCEFSGH; whose expression in the host cells leads to overexpression of the thiamine biosynthesis pathway.

Cells of transformed E. coli host strains expressing each of the candidate phosphatase genes (and control E. coli host strains) were first pre-cultured in 400μΙ_ Modified MOPS medium (comprising antibiotics for vector maintenance) in deep 96-well cultivation plates at 37 degrees, shaking at 300rpm overnight; and cultures produced were then used to inoculate the same volume of medium to an OD600 nm of 0.0015. These cultures were grown for 24 hours under the same conditions; and then used for thiamine analysis. OD600nm was measured using a plate reader, and then converted to the corresponding OD600nm in a cuvette.

1.3 Extraction and detection of TPP, TMP and thiamine produced by genetically modified E. coli cells

Extracellular and intracellular TPP, TMP and thiamine in each culture was recovered and extracted as follows: 0.4ml_ of each culture was harvested at 4°C by centrifugation in the cultivation plate at 4000 x g for 5 minutes. All remaining steps were performed on ice. 40 μΙ_ of supernatant was gently removed for analysis of extracellular TPP, TMP and thiamine. After decanting the remaining supernatant; the culture plate was inverted to remove residual medium and then vortexed. ΙΟΟμί ice-cold HPLC grade methanol was added to each well of the culture plate; and the cells were vortexed again. After incubation on ice for a minimum of 20 minutes cell debris was pelleted by centrifugation at 4000 x g for 5 minutes. The supernatant was used as intracellular extract for further analysis.

In order to detect TPP, TMP and thiamine using a fluorescence detector, the thiamine compounds produced by each culture were derivatized into thiochromes, which are strongly fluorescent. All steps are performed at room temperature. 40μΙ volumes of the extracellular and intracellular extracts was added to 80μΙ of 4M potassium acetate and mixed by pipetting. 40μΙ of freshly prepared 3.8mM potassium ferricyanide in 7M NaOH was added and mixed. The reaction was quenched by addition of 40μΙ freshly prepared 0.06% H ₂ 0 ₂ in saturated KH ₂ P0 ₄ . The extracts were neutralized by addition of 47μΙ_ 6M HCI and then analyzed by HPLC as described below. All derivatized compounds were quantified using fluorescence standard curves of freshly prepared of TPP, TMP and thiamine standards that were derivatized to thiachromes in parallel with the analyzed extracts.

HPLC was used to quantitate the intracellular and extracellular content of thiamine; TMP and TPP in each culture by a method adapted from the method described by Schyns G et al., (2005).

Column : Hypersil Gold, 3μΐη particle size, dimensions: 150 x 2.1mm (Thermo product number: 25003-152130) with guard column of the same material. Buffer A: lOmM K ₂ HP0 ₄ pH : 7, 4mM t-butyl ammonium hydrogen sulfate Buffer B: Methanol

Buffer C: Distilled water

Gradient: Constant concentration of Buffer A at 40%. 0-2 min : constant concentration of B at 10%; 2-6 min : gradient of B from 10-30%; 6-8 min : constant concentration of B at 30%; 8-8.1 min : gradient of B from 30-10%; 8.1-12 min : constant concentration of B at 10%.

Flow rate 0.8mL/min

Injection volume: 2μΙ_ for intracellular samples, 5uL for extracellular samples. Order of elution : thiamine, TMP, TPP. Detection using fluorescence detector with with Aexcitation : 365nm and Aemission : 444nm.

1.4 A. thaliana protein AT5G32470.1 expressed in genetically modified E. coli cells exhibits TMP phosphatase activity

The composition of the thiamine compounds produced by cells of genetically modified E. coli, engineered to overexpress the thiamine pathway, and expressing one of five different phosphatase genes is shown in Figure 2. The only cells showing a significant shift in thiamine phosphorylation profile towards un-phosphorylated thiamine relative to that of cells comprising the control vector pBS92, were those expressing a gene encoding A. thaliana protein AT5G32470.1. TMP was not detectable in these cells; the amount of TPP was slightly reduced; on which basis it can be deduced that the A. thaliana protein AT5G32470.1 has TMP phosphatase activity, and may also have TPP activity. The TMP phosphatase activity in cells expressing the A. thaliana TMP phosphatase resulted in a striking increase in both total extracellular thiamine compound production, and in extracellular un- phosphorylated thiamine, which accounted for 80% of total thiamine compounds detected in the culture.

Thiamine compounds were barely detectable in cultures of E. coli cells transformed with the empty vectors (pGEN49 and pBS92) that neither expressed a TMP phosphatase, nor thiamine pathway enzymes. A comparison of the levels of thiamine compounds produced by cells expressing thiamine pathway enzymes alone or together with TMP phosphatase, reveals their synergistic effect on thiamine compound production. Example 2 Alignment and structural annotation of the A. thaliana protein (AT5G32470.1)

Sequence alignment and structural annotation of the A. thaliana protein (AT5G32470.1) reveals two structural and functional domains. The first domain is a TenA-like domain located between amino acids 85-292, which based on alignment with the Bacillus subtilis protein TenA, is assigned responsibility for the recognition of phosphorylated thiamine substrates. The crystal structure of B. subtilis protein TenA has been solved alone and in complex with the thiamine precursor HMP (Toms et al. 2005). B. subtilis TenA is an aminopyrimidine aminohydrolase that catalyzes the hydrolysis of 4- amino-5-aminomethyl-2-methylpyrimidine to 4-amino-5-hydroxymethyl-2- methylpyrimidine (HMP). Since HMP corresponds to half of the thiamine molecule, the thiamine binding site residues in the A. thaliana protein, are predicted to correspond to the HMP binding sites in B. subtilis TenA. A sequence alignment shows that key residues, which form H-bond contacts to HMP and that line the active site, are conserved between B. subtilis TenA and AT5G32470.1 (which are identified in figure 5). B. subtilis TenA is capable of hydrolyzing thiamine, albeit at a much slower rate than HMP, strengthening the argument that the TenA domain is responsible for thiamine binding.

A second domain in the A. thaliana protein belongs to a conserved family of haloacid dehalogenase-like hydrolase domains (HAD-like superfamily) (http://www.ncbi.nlm.nih.gov/Structure/cdd/cddsrv.cgi7uid = 277525). This superfamily includes carbon hydrolases and phosphate hydrolases. Members of the HAD-like domain superfamily are known to catalyze a nucleophilic substitution reaction at the carbon or phosphorus in question. A hallmark of this HAD domain (PFAM code PF00702) is the presence of several conserved residues that are part of the of the Rossmann fold (Burroughs, A.M. et al., 2006). A conserved aspartate at the end of sheet 1, a conserved serine or threonine at the end of sheet 2, a lysine on a structure called C2 cap and a conserved aspartate on sheet 4, that together function in catalysis. Accordingly, the HAD-like domain in the A. thaliana protein (Figure 5) was assigned phosphate hydrolase activity. These conserved functional domains allow the identification of polypeptides having TMP phosphatase activity. Example 3 Identification and structural annotation of TMP phosphatase protein families

Members of the TMP phosphatase protein family, capable of enhancing thiamine levels in a bacterial cell, were identified on the basis of sequence, structural and functional homology, employing the following criteria and tools: Four genes were empirically known to encode enzymes with a thiamine monophosphate hydrolysing activity, either promiscuous or physiological. The three genes were the AT5G32470.1 gene, first identified herein as a phosphatase, and three recently reported genes (Hasnain et al. 2016). Sequences that were homologous to these genes, as well as belonging to either the group of TenA-HAD fusion proteins, or alternatively being fused to a thiamine pathway gene or present in operons comprising thiamine pathway genes, was taken as a strong indicator of conservation of their function as encoding a TMP phosphatase. Sequences encoding a haloacid domain, which is a characteristic feature of proteins that catalyse dephosphorylations, was taken as a further indicator of TMP phosphatase function. The PF012710 seed dataset was used as a base dataset for HAD protein diversity. In the case of proteins composed of multiple domains the extra domains were removed based on a preliminary alignment with known HAD domain sequences. The datasets of genes of interest were chosen manually and the headers changed with a small Python script. The genes were then aligned with the Pfam dataset and outgroups chosen accordingly for each group. The sequences were independently aligned for each group with Muscle and were trimmed with Gblocks under permissive settings and the resulting alignments used for tree inference by Maximum likelihood under a WAG model with a CAT distribution.

The characterized TMP phosphatase haloacid domain proteins that catalyze the dephosphorylation of thiamine monophosphate fall into three clades (Figure 3), indicative of their parallel independent evolution.

Clade 1 comprises homologues of At5G32470 and spans the Viridiplantae (figure 4A), and its members all feature a TenA/Teni-4 family domain (PFam : PF03070) at the N-terminus (Figure 5). The members of this clade comprise the proteins: Arabidopsis thaliana NP-198287.3 (AT5G32470.1) [SEQ ID No. : 2] ; Pyrus x bretschneideri XP_009379735.1 [SEQ ID No. :4] ; Brassica napus CDY62623.1 [SEQ ID No.:6]; Glycine max XP_003536133.1 [SEQ ID No.:8]; Nicotiana tomentosiformis XP_009615535.1 [SEQ ID No.:10]; Populus trichocarpa XP_002325785.2 [SEQ ID No.: 12]; Jatropha curcas KDP23738.1[SEQ ID No.: 14]; Citrus sinensis XP_006484613.1 [SEQ ID No.:16]; Prunus persica XP_007199656.1 [SEQ ID No.:18]; Phoenix_dactylifera_XP_008796407 [SEQ ID No.: 20]; Zea mays XP_008678418.1 [SEQ ID No.:22]; Oryza sativa NP_001062539.1 [SEQ ID No.:24]; Picea_sitchensis_ABR16455 [SEQ ID No.:26];

Physcomitrella_patens_XP_001769831 [SEQ ID No.:28]; and Selaginella_moellendorffii_XP_002990363 [SEQ ID No.: 30].

Clade 2 comprises bacterial genes, many of which are fused with thiamine pathway genes (figure 4B). Clade 2 includes an Anaerotruncus colihominis gene (WP_006874980) characterized as a TMP phosphatase (Hasnain et al.

2016). The members of this clade comprise the proteins:

Anaerotruncus_colihominis_\NP_006874980 [SEQ ID No.:32];

Eubacterium_ ven trios um_\N P_005362972 [SEQ ID No.:34];

Coprococcus_eutactus_( ATCC_27759)_EDP27707 [SEQ ID No.:36];

Ruminococcus_bromii_L2-63_CBL14666 [SEQ ID No. 38];

Dorea_/ong/ ^' catena_(DSM_13814)_EDM62146 [SEQ ID No.:40];

Lachnospiraceae_bacterium_EPCO 5128 [SEQ ID No.:42];

Fusicatenibacter_CUQ30753 [SEQ ID No.:44]; Clostridium _. _sp._ _ERI68966

[SEQ ID No.:46]; Eubacterium_hallii_EEG35494 [SEQ ID No.:48];

Eubacterium_sp. CDB67556 [SEQ ID No.:50];

Lachnospira_pectin oschizajC UQ76318 [SEQ ID No.:52];

Peptostreptococcaceae_bacterium_Vl P_009530263 [SEQ ID No.:54];

Peptostreptococcaceae_bacterium_\N P_009527854 [SEQ ID No.:56]; ¾topo6/ ^' um_sp._WP_035427744 [SEQ ID No.:58];

Atopobium_parvulum_\N P_035433109 [SEQ ID No.:60];

/4topob/ ^' um_rimae_WP_003148415 [SEQ ID No.:62];

O/sene//a_u// ^' _WP_013251930 [SEQ ID No.:64]; and

Atopobium_minutum_KRN55115 [SEQ ID No.: 66].

Clade 3 comprises another group of bacterial genes, many of which map adjacent to thiamine pathway genes (Figure 4C); and includes a characterized TMP phosphatase encoded by a Syntrophomonas wolfei gene (WP_011640074.1) (Hasnain et al. 2016). The members of this clade comprise the proteins: Syntrophomonas_wolfei_\NP_011640074 [SEQ ID No. : 68] ; Desulfitobacterium_hafniense_\NP_018212876 [SEQ ID No. : 70] ; Pelotomaculum_thermopropionicum_\NP_012032097 [SEQ ID No. : 72] ; and Desulfotomaculum_ruminis_\NP_013840216 [SEQ ID No. : 74] .

Example 4 Genetically modified E. coli strains engineered for enhanced thiamine production

In order to drive the equilibrium further towards un-phosphorylated thiamine (THI) in a thiamine production strain, the expression of a TMP phosphatase was combined with mutations in the endogenous thiamine or TMP kinases genes, thiK and thiL. Since TPP is essential for cell growth, and since the only enzyme in E. coli capable of producing TPP from TMP is encoded by the thiL gene, full thiL knockout strains are not viable unless provided with a supply of TPP which is un-economic. Instead, kinase activity expressed in the host cells was decreased by introducing a point mutation (codon 133 from GGT to GAC) into the thiL gene causing G133D substitution in the encoded enzyme. This point mutation was known to decrease kinase activity encoded by the mutant thiL927 gene in Salmonella typhimurium (Webb E, et al., 1997). This mutation was introduced into cells of the E. coli host strain using Multiplex Automated Genome Engineering (MAGE).

4.1 Construction of a thiK knockout strain (BS131)

The thiK gene was knocked out in the E. coli host strain DH 10B to produce strain BS131 by using λ-RED recombineering methods (Datsenko K et al., 2000) by the follow steps: A thiK: : Kan construct was PCR amplified from strain JW1092 of the Keio collection (Baba T, et al., 2006). Although the amplified thiK gene was originally named ycfN, in the Keio database, this corresponds to thiK (Melnick J. et al., 2004). The amplified fragment was transformed by electroporation into competent cells of E. coli DH 10B carrying the ARed recombinase genes expressed from an inducible promoter (pKD46) Datsenko et al, 2000). Successful integrants were selected on Kanamycin medium and confirmed by colony PCR using primers oBS521 and oBS522. The Kanamycin resistance cassette was eliminated by transforming cells with the temperature-sensitive plasmid pcp20, which expresses the FLP recombinase. Ampicillin-resistant colonies were isolated at 30°C and then re-streaked non- selectively at 42°C. Loss of KanR was confirmed by colony PCR with primers OBS521 and oBS522.

4.2 Construction of thiL G133D (thiL* strain (BS182))

The thiL gene was partially inactivated by mutation in the E. coli host strain DH 10B to produce strain BS182 using MAGE comprising the transient overexpression of DNA adenine methylase, as described by Lennen et al (2015). Briefly: the pMA7sacB plasmid was introduced into cells of the E. coli DH 10B strain. The resulting strain was grown under inducing conditions (0.2% arabinose) to mid-log phase and electrocompetent cells were prepared using standard conditions. The competent cells were electroporated with a 1 : 1 mixture of oligos oBS445 and oBS446, which both introduce a mutation in the genomic copy of the thiL gene encoding a G133D substitution mutation. After rescue this procedure was repeated once more before cells were plated and successful mutants were identified by sequencing with primers oBS456 and 457.

4.3 Thiamine production is enhanced in genetically modified E. coli having reduced thiamine kinase activity {thiL*) combined with expression of the A. thaliana phosphatase (AT5G32470.1)

The thiK- knockout E. coli strain, BS131, and the mutated ThiL* E. coli strain, BS182, as well as the parent strain DH 10B, were each transformed with the pBS140 vector comprising the thiamine pathway genes thiMD and thiCEFSGH, either alone, or in combination with the pBS93 vector encoding the A. thaliana TMP phosphatase (AT5G32470.1). These genetically modified strains were cultivated as described in Example 1.2, and the thiamine phosphorylation profile and thiamine content of all resulting cultures was determined by HPLC as described in Example 1.3.

As seen in Figure 5 (and Table 3), the expression of the A. thaliana TMP phosphatase alone converts most of the intra- and extra-cellular TMP to thiamine. Mutation of thiL alone significantly decreases the amount of TPP relative to TMP and thiamine. A combination of these two genetic modifications results in almost all of the cellular thiamine being converted to the dephosphorylated state (90% of total, compare with 14% for wild-type kinase strains). This corresponds to a 6.8 fold increase in the amount of THI produced. Furthermore, the amount of extracellular thiamine is also greatly increased, (7 fold) a very desirable for ease of downstream processing of the product. Finally the overall titer of THI + TMP + TPP is also increased by 24%. In strains expressing the A. thaliana TMP phosphatase, the intracellular levels of thiamine compounds are consistently reduced; which in turn may reduce potential feedback inhibition on the thiamine pathway, thereby accounting for the enhanced total thiamine compound production. The thiK- deletion, however, only slightly enhances the production of THI when combined with expression of the A. thaliana TMP phosphatase. Accordingly, the dephosphorylation of thiamine catalyzed by the TMP phosphatase appears to outcompete the phosphorylation of thiamine catalyzed by the thiamine kinase (ThiK), such that deletion of ThiK gene has a lesser effect.

Table 3 Production of thiamine compounds in genetically modified E. coli strains

4.4 Construction of vectors for expression of members of the TMP phosphatase protein family Genes encoding members of clades 1, 2 and 3 of the TMP phosphatase protein family (see Figure 4) were cloned into expression vectors, for expression in a bacterial cell. Vectors pBS791-798 were constructed by amplifying the empty vector, pBS92 (see Example 1.1) with the primer pair oBS198 and oBS1721 (Tables 2 and 4). The phosphatase genes were each synthesized by Gen9, Inc. along with 3' and 5' regions that overlap with vector pBS92 and were supplied in a cloning vector. Each gene was then amplified with primer pair OBS1720, oBS234 (Table 4) having binding sites flanking the gene in the supplied vector. The purified PCR product comprising the respective phosphatase gene was then assembled with the amplified pBS92 backbone using the NEB Gibson assembly master mix according to the manufacturer's instructions.

Each of the assembled vectors was transformed into electro-competent DH 10B cells by electroporation and the recovered cells were plated on selective LB plates. The fully assembled plasmids (Table 5) were then isolated from the resulting strains.

Table 4: List of primers

Primer Sequence SEQ ID name No.:

OBS234 AGGCCTUTCTGACAACACACGAGTAATAATT 155

OBS1720 CGTCACTAATAGCGAGGTAATACTAG 156

OBS1721 CTAGTATTACCTCGCTATTAGTGACGATAATAGGAGG 157

TAAGC

Table 5

Vector Inserted gene Encoded protein / SEQ ID No. pBS791 Jatropha curcas TMP phosphatase KDP23738.1, 14

codon optimized for expression in E. coli

pBS792 Picea sitchensis TMP phosphatase ABR16455, 26

codon optimized for expression in E. coli pBS793 Eubacterium ventriosum TMP phosphatase 34 WP_005362972, codon optimized for expression in

E. coli

pBS794 Anaerotruncus colihominis TMP phosphatase 32

WP_006874980, codon optimized for expression in

E. coli

pBS797 Desulfitobacterium hafniense TMP phosphatase 70

WP_018212876, codon optimized for expression in

E. coli

pBS798 Syntrophomonas wolfei TMP phosphatase 68

WP_011640074, codon optimized for expression in

E. coli

4.5 Extracellular thiamine production is enhanced in genetically modified E. coli having reduced thiamine kinase activity (tAi/Z.*) combined with expression of members of the TMP phosphatase protein family

Representative genes from clade 1, 2 and 3, identified in Example 3 (Figures 3, 4, 5) that were predicted to increase the thiamine production of a thiamine cell factory, were expressed in a production strain to demonstrate their ability to enhance extracellular thiamine levels. The mutated ThiL* E. coli strain, BS182 was transformed with the pBS140 vector comprising the thiamine pathway genes thiMD and thiCEFSGH or control vector pBSlOO, either in combination with control vector pBS92, or in combination with the pBS93 vector encoding the A. thaliana TMP phosphatase (AT5G32470.1), or with one of the TMP phosphatase encoding vectors described in Table 4. These genetically modified strains were cultivated as described in Example 1.2. The supernatant of the cultures was collected after 24 hours and the extracellular thiamine levels were measured by LCMS by MS Omics ApS using the following protocol : All samples were filtered through a 0.2μΜ filter and the samples were analysed using a slightly modified version of the acidic protocol (positive ionization) described by Paglia et al., (Waters application note). For quality control, a mixed pooled sample (QC sample) was created by taking a small aliquot from each sample. Every four-to-five aliquots of the QC sample were analysed. To minimize matrix effects the samples were dilutes 50 times. For quantification, a mixture of the blank media samples was created and the standards were prepared in this mixture. Quantification was performed using the sum of three ions (122.0716, 265.112 and 144.0473 Da).

The resulting extracellular thiamine titers were normalized to the thiamine titers of the non-producing strain control strain (thiL* pBSlOO pBS92) and results are shown in Figure 7. As shown in Figure 6, the expression of the A. thaliana TMP phosphatase (pBS93) greatly increases the extracellular thiamine titer. As seen in Figure 7, expression of all of the representative TMP phosphatases from clade 1, 2 and 3 (PBS791, 792, 793, 794, 797 and 798) in the production strain significantly increased the production of extracellular THI, in therefore share the same properties as the A. thaliana TMP

phosphatase.

References Baba T, Ara T, Hasegawa M, et al. (2006) Construction of Escherichia coli K- 12 in-frame, single-gene knockout mutants: the Keio collection. Mol Syst Biol. ;2 : 2006.0008

Burroughs, A.M. et al., 2006. Evolutionary Genomics of the HAD Superfamily: Understanding the Structural Adaptations and Catalytic Diversity in a

Superfamily of Phosphoesterases and Allied Enzymes. J. Mol. Biol., 361, pp.1003-1034.

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

A. ;97(12) : 6640-5. Gibson DG, Young L, Chuang R-Y, Venter JC, Hutchison C a, Smith HO., (2009) Enzymatic assembly of DNA molecules up to several hundred kilobases. Nat Methods., 6(5) : 343-5.

Goyer A, Hasnain G, Frelin O, Ralat M a, Gregory JF, Hanson AD. (2013) A cross-kingdom Nudix enzyme that pre-empts damage in thiamin metabolism. Biochem J. 454(3) : 533-42. Guzman, L, Belin, D. & Carson, M.J., 1995. Tight Regulation, Modulation, and High-Level Expression by Vectors Containing the Arabinose P BAD Promoter. Journal of bacteriology, 177(14) :4121-4130.

Hasnain G., et al., (2016) Bacterial and plant HAFD enzymes catalyze a missing phosphatase step in thiamine diphosphatase biosynthesis.

Biochemical Journal, 473 (2) 157-166

Komeda Y, Tanaka M, Nishimune T. (1988) A th-1 Mutant of Arabidopsis thaliana is Defective for a Thiamin-Phosphate-Synthesizing Enzyme: Thiamin Phosphate Pyrophosphorylase. Plant Physiol. 88(2) : 248-50. Lennen, R.M . et al., (2015) Transient overexpression of DNA adenine methylase enables efficient and mobile genome engineering with reduced off- target effects. Nucleic acids research, pp.1- 14.

Melnick J, Lis E, Park J-H, et al. (2004) Identification of the two missing bacterial genes involved in thiamine salvage: thiamine pyrophosphokinase and thiamine kinase. J Bacteriol. : 186(l l) : 3660-2

NorranderJ, Tomas Kempe, Joachim Messing, 1983. Construction of improved M 13 vectors using oligodeoxynucleotide-directed mutagenesis, Gene, 26(1), pp. 101-106.

Paglia, G., James Langridge, J., Astarita G., Development of a Metabolomic Assay for the Analysis of Polar Metabolites Using HILIC UPLC/QTof MS, Waters Application Note: Library number: APNT134726984 (2013)

http://www. waters. com/waters/I ibrary.htm?! id = 134726984&locale=en_DK

Schyns G, Geng Y, Barbosa TM, Henriques A, Perkins JB. (2005) Isolation and Characterization of New Thiamine-Deregulated Mutants of Bacillus subtilis. J Bacteriol. ; 187(23) : 8127-8136.

Webb E, Downs D. (1997) Characterization of thiL, Encoding Thiamin- monophosphate Kinase, in Salmonella typhimurium. J Biol

Chem. ;272(25) : 15702- 15707