WO2006093408A2

WO2006093408A2 - Bacteria that naturally overproduce folate

Info

Publication number: WO2006093408A2
Application number: PCT/NL2006/050042
Authority: WO
Inventors: Hendrikus Wegkamp; Eilt Johannes Smid; Willem Meindert De Vos
Original assignee: FrieslandCampina Nederland Holding BV; CSM Nederland BV
Current assignee: FrieslandCampina Nederland Holding BV; Corbion Group Netherlands BV
Priority date: 2005-03-01
Filing date: 2006-03-01
Publication date: 2006-09-08
Anticipated expiration: 2007-09-01
Also published as: RU2431662C2; ES2351150T3; US20080118943A1; RU2007136115A; JP2008531043A; ATE482291T1; EP1698704B1; WO2006093408A3; CN101203599A; EP1698704A1; BRPI0607843A2; DE602005023704D1

Abstract

The present invention relates to the mutant bacteria comprising high levels of folate and being resistant to methotrexate. Also provided are food and food supplement compositions comprising these bacteria and methods for isolating mutant bacteria.

Description

Bacteria that naturally overproduce folate

FIELD OF THE INVENTION

The present invention relates to methods for selecting mutant bacteria producing increased folate levels. In these methods the antifolate methotrexate (MTX) is used as selection agent. Provided are also food and food supplement compositions comprising the mutant bacteria or derivatives or extracts of the bacteria.

BACKGROUND OF THE INVENTION Folate analogues and in particular inhibitors of dihydro folate reductase have been the topic of research for many years because of their potential role in cancer chemotherapy and as agents against the malaria parasite. One of the first folate analogues described in literature is methotrexate from Lederle Laboratories. This compound has already been used in 1948 for the treatment of acute leukemia (Hitchings, G. H., Jr., 1989, In Vitro Cell Dev Biol 25:303-10).

The activity of methotrexate (MTX) is based on competitive inhibition of the target enzyme dihydrofolate reductase. This enzyme is needed for the production and recycling of tetrahydrofolate (THF) using dihydrofolate (DHF) as a substrate. The specificity of the methotrexate can be explained by small structural differences of dihydrofolate reductases in different species (Chang et ah, 1978, Nature 275:617-24). Methotrexate prevents the cell from producing sufficient tetrahydrofolate for the biosynthesis of methionine, DNA and RNA. However in the presence of purines, thymidine, glycine, methionine and pantothenic acid (all metabolites that need tetrahydrofolate for their own biosynthesis) the demand of cells to synthesize tetrahydrofolate can be decreased (Harvey, R. J., 1973, J Bacteriol 114:309-22). Nevertheless cells will always be in some way hampered in growth by MTX, because THF is essential for formation of methionyl-tRNA^fmet, a compound that is indispensable for the initiation of protein synthesis, regardless of the presence of folate dependent metabolites (Baumstark et al, 1977, J Bacteriol 129:457-71).

Lactic acid bacteria such as Lactococcus lactis, Lactobacillus plantarum and Lactobacillus casei show a high degree of variability in their response to methotrexate and another folate analogue, trimethoprim. L. lactis is known be insensitive to both antifolates (Leszczynska et al., 1995, Appl Environ Microbiol 61:561-6). It has been shown for numerous organisms that prolonged exposure to antifolates results in selection of resistant cell types. Several underlying mechanisms have been described to explain the resistance mechanism of both eukaryote cell lines as well as bacteria to methotrexate (MTX). Tamura et al. (1997 ^', Microbiology 143: 2639-46) investigated the mechanisms leading MTX resistance in Enterococcus hirae. They found several phenotypic effects including (i) increased folic acid reductase (FAR) activity (ii) increased dihydro folate reductase (DHFR) activity, (iii) decreased synthesis and intracellular retention of MTX containing two glutamyl residues, (iv) decreased uptake of MTX accompanied by decreased uptake of folates; and (v) reduction of folate- binding capacity. The form of folate present in the media during the development of resistance affected DHFR and FAR activities and the transport of folates.

However, although it is known that exposure to MTX can be used to select cells having increased resistance to MTX, no link between MTX resistance and folate levels of the cells has been disclosed in the prior art. In contrast to the MTX resistance mechanisms described above, the present inventors have surprisingly found that bacterial cells have another way of generating MTX resistance. This mechanism is the production of high levels of folate. This finding can be exploited in the screening and selection of (spontaneous) mutant bacteria producing high folate levels, by using methotrexate as selection agent. Such bacteria, especially food grade bacteria, can be used to fortify food products with folate. Folate is a soluble B vitamin, required for proper cell growth and functioning. Many food products, such as bread, cereals, dairy products, etc. are supplemented with folic acid to ensure adequate intake of folate. The mutant bacteria of the present invention enable fortification of food products with natural folate by for example fermentation using the mutant bacteria.

DEFINITIONS "Lactic acid bacteria" as used herein, are bacteria, which produce lactic acid as an end product of fermentation, such as bacteria of the genus Lactobacillus, Lactococcus, Streptococcus and Bifidobacterium. A bacterial "strain" or "isolate" is used herein interchangeably and refers to a bacterium which remains genetically unchanged when grown or multiplied. The multiplicity of identical bacteria are included.

A "mutant bacterium" or a "mutant bacterial strain or isolate" refers to a natural (spontaneous, naturally occurring) mutant bacterium or an induced mutant bacterium comprising one or more mutations in its genome (DNA) which are absent in the wild type DNA. An "induced mutant" is a bacterium where the mutation was induced by human treatment, such as treatment with chemical mutagens, UV- or gamma radiation, etc. In contrast, a "natural" or "spontaneous mutant" has not been mutagenized by man. Mutant bacteria are herein non-GMO, i.e. not modified by recombinant DNA technology.

"Wild type strain" or "wild type isolate" refers to the non-mutated form of a bacterium, as found in nature.

"Folate dependent metabolites" refers to metabolites or precursors of folate biosynthesis or catabolism which are able to mask MTX sensitivity of a cell when present in the surrounding medium.

"Inhibitory amount of MTX" refers to the amount required to substantially inhibit growth of a cell. Generally, for prokaryotic cells an amount of at least about 1.25mg/L growth medium is inhibitory for MTX sensitive cells. "Total folate" refers to the extracellular (secreted) and intracellular folate levels produced by a strain.

"Food-grade" micro-organisms are in particular organisms, which are considered as not harmful, when ingested by a human or animal subject.

The term "comprising" is to be interpreted as specifying the presence of the stated parts, steps or components, but does not exclude the presence of one or more additional parts, steps or components. A composition comprising bacterial strain X, may thus comprise additional strains, other components, etc.

In addition, reference to an element by the indefinite article "a" or "an" does not exclude the possibility that more than one of the element is present, unless the context clearly requires that there be one and only one of the elements. The indefinite article

"a" or "an" thus usually means "at least one". DETAILED DESCRIPTION OF THE INVENTION

While testing whether recombinant folate overproducing bacteria have modified susceptibility to MTX, it was found that a number of wild type strains of L. plantarum, which were incubated for more than about 40 hours on inhibitory amounts of MTX (1,25 mg/L) were not inhibited by MTX, but had a growth rate similar to that on medium lacking MTX. When analysing the folate levels of these strains, approximately 5% of these colonies displayed a significant higher folate production compared to the (non mutated) wild type strain. The total folate pools were 4% to 63% higher compared to (average) wild-type levels. This observation demonstrated that MTX can be used to efficiently select natural (non-GMO) folate overproducing bacteria. The resistance to MTX was maintained, even when the strains were grown for more than 50 generations on medium lacking MTX, which indicated that the resistance is caused by (spontaneous) stable mutations in the bacterial genome.

The same effect, that increased folate production leads to high resistance against MTX, was seen in recombinant L. plantarum strains, which were transformed with the complete folate biosynthesis gene cluster under the control of a strong promoter. These recombinant strains produced high levels of folate and were resistant to MTX when grown in a medium lacking folate dependent metabolites. Without limiting the scope of the invention, it is speculated that the intracellular dihydrofolate levels compete with the accumulated MTX for the active site on the enzyme dihydrofolate reductase (DHFR).

This finding was applied to develop a rapid screening method, based on growth in the presence of MTX, to select folate overproducing bacterial strains. The folate overproducers that were identified in this study displayed a high increase in folate levels compared to the wild type folate producers. The total folate levels ranged from

20 to 60 times the level of the wild type. The intracellular folate pools found in the folate overproducers were 10 to 20 times higher compared to the levels found in the wild type strain. These increased intracellular folate levels cause a decreased sensitivity

(i.e. resistance) towards MTX. This explained why growth of the folate overproducing strains was not significantly affected by MTX. Sensitivity of the wild- type towards MTX was only observed when all the metabolites which are associated with folate were omitted from the medium.

Bacteria according to the invention Therefore, in one embodiment of the invention mutant bacteria are provided which comprises an increased amount of intracellular and/or extracellular folate compared to the wild type and which have a growth rate (μ) of at least 0.1 h^"1 when grown on medium comprising (at least) 1.25 mg/1 methotrexate and lacking folate dependent metabolites (i.e. they are said to be "resistant" to MTX). Preferably under these growth conditions, the growth rate of the mutant bacterium is not significantly different than that of the wild type strain when the wild type strain is grown on the same medium lacking MTX (i.e. "normal growth"). Most importantly, the mutant bacterium shows significant growth at an MTX concentration which is inhibitory to the wild type strain. An "Inhibitory concentration" is the concentration of MTX which results in substantially no growth of the wild type strain. Figure 3, for example, shows that at concentrations of 1.25, 1.5, 2.0 and 2.5 mg/1 MTX, the wild type has a growth rate close to zero (μ < 0.05 h^"1), while the folate overproducing strains have a growth rate (μ) of above 0.1 h^"1, especially around about 0.15 h^"1. The growth medium should not be supplemented with folate dependent metabolites, as in their presence no significant difference in MTX resistance can be seen, probably because the wild type strain compensates for its lower folate levels by using the folate metabolites of the medium. Suitable growth medium is for example modified CDM (Chemically Defined Medium), which lacks glycine, inosine, orotic acid, thymidine, guanine, adenine, uracil and xanthine.

Growth rate can be measured by various means. For example, at certain time points following inoculation of the medium with the bacterial strain or isolate spectrophotometric readings can be taken and the growth rate over time calculated.

The mutant bacterium preferably comprises an increased amount of intracellular and/or extracellular folate compared to the wild type strain. Preferably, the amount of total folate is at least about 4%, 5%, 10%, 15%, 20%, 25%, 30%, 40%, 50%, more preferably at least about 60%, 63% (or more) higher than the average amount of intracellular folate of the wild type strain(s). As in some bacterial species the wild type strains show significant variation in intracellular folate levels, it is preferred that a number of different wild type strains (e.g. 2, 3, 4 or more strains) are analysed for intracellular folate levels and the average folate level is calculated. The total, intracellular or extracellular amount of folate produced by a bacterial isolate or strain can be determined and quantified by methods known in the art, such as the L. casei microbiological assay described by Sybesma (see example 1.3) or variants thereof, by HPLC analysis or other known methods. A micobiological test may for example use an indicator strain (e.g. an L. casei strain), which is grown on folate produced by the test strain. If large amounts of folate are produced by the test strain, the indicator strain grows well, resulting in a high OD measurement after a specified period of growth. In contrast, if low amounts of folate a produced by the test strain, lower OD values will be measured. The mutant bacterium is preferably a spontaneous (natural) mutant, selected for example using the method according to the invention (see further below). The bacterium is of any species of which the wild type strain shows natural sensitivity to MTX, such as L. plantarum, L. casei, or other species of the group of lactic acid bacteria. Being sensitive to MTX means that growth of the bacterium is substantially inhibited by at least 1.25mg/L MTX, when the strain is grown in a medium lacking folate dependent metabolites. Whether a bacterial species shows sensitivity to MTX can be easily tested by growing the bacterium on medium supplemented with MTX and lacking folate dependent metabolites. The bacterium is preferably a food grade bacterium and, in one embodiment, it belongs to a genus selected from the group consisting of Lactobacillus, Lactococcus, Streptococcus, Bifidobacterium, Leuconostoc and Streptococcus.

Preferably it is a food grade lactic acid bacterium. It may be a species selected from the group consisting of Lactobacillus reuteri, L. fiermentum, L. acidophilus, L. crispatus, L. gasseri, L. johnsonii, L. casei, L. plantarum, L. paracasei, L. murinus, L. jensenii, L. salivarius, L. minutis, L. brevis, L. gallinarum, L. amylovorus, Lactococcus lactis, Streptococcus thermophilus, Leuconostoc mesenteroides, Lc. lactis, Pediococcus damnosus, P. acidilactici, P. parvulus, Bifidobacterium bifidum, B. longum, B. infantis, B. breve, B. adolescente, B. animalis, B. gallinarum, B. magnum, and B. thermophilum. In an alternative embodiment the mutant bacterium is an induced mutant having increased folate and being resistant to MTX. Such a mutant may be obtained using the same method as used in the selection of natural mutants (described herein below), except that prior to the screening method a mutagenesis step is included. A mutagenesis step may include one or more known mutagenesis methods, such as exposure to a chemical and/or physical mutagen (e.g. N-methyl-N'-nitro-N-nitrosoguanidine; UV radiation, gamma-radiation, etc.). Following mutagenesis, the bacteria are grown in/on medium comprising inhibitory amounts of MTX and isolates which do grow are selected and analysed for folate content.

In yet another embodiment, the mutant bacteria according to the invention can be characterized by their gene expression pattern. This is illustrated by the finding that methotrexate resistant L. plantarum strains producing at least 50% more folate compared to the wild-type show a 2-fold (or more) overexpression of one or more of the folC genes in the genome. The gene folC codes for the enzyme dihydrofolatesynthase. In one embodiment the bacteria according to the invention show overexpression compared to the wild-type of one or more genes coding for enzymes of the folate biosynthesis pathway. In addition to the gene directly involved in folate biosynthesis, also higher transcription levels were found for the following genes: FK (coding for fructokinase), POX (pyruvate oxidase), LOX (lactate oxidase) and PDH (pyruvate dehydrogenase). Compared to the wild-type, a significantly lower transcript level was found for: DAK (dihydroxyacetone phosphotransferase or glycerone kinase). Thus, in one embodiment of the invention, the bacterium/bacteria according to the invention comprise significantly enhanced mRNA transcript levels of one or more genes of the group encoding dihydrofolate synthase (EC 6.3.2.17), fructokinase (EC 2.7.1.4), pyruvate oxidase (EC 1.2.3.3), and 6-phospho-β-glucosidase (EC 3.2.1.86). Additionally or alternatively, the bacteria comprise significantly reduced mRNA transcript levels of the gene encoding for dihydroxyacetone phosphotransferase (= glycerine kinase) (EC. 2.7.1.29). The term "upregulated'Or "downregulated" refers to an increased mRNA transcription level or a decreased mRNA transcription level compared to the gene in the wild type strain.

A "significantly enhanced transcript level" or a "significantly enhanced transcription level" refer to an amount of mRNA transcript of at least about 2 fold, preferably at least about 3 fold, or even about 4 fold or more compared to the transcript level found in the wild type. Likewise, a "significantly reduced transcript level" or "transcription level" refers to an amount of mRNA transcript of at least about 2 fold, 3 fold, etc. less than found in the wild type. mRNA transcript levels of one or more genes can be measured and quantified using routine molecular biology methods, such as for example Northern analysis and quantitative reverse transcriptase (RT-) PCR. As DNA/cDNA sequence information is available for these genes, primers and probes suitable for hybridizing to the target mRNA or cDNA are available and only routine experimentation is required for the analysis of transcript levels. For example, the nucleic acid sequences of L. plantarum (the genome of which has been sequenced) may be used to identify (e.g. by in silico analysis), clone (e.g. using PCR based methods, hybridization based methods, etc) and/or sequence the orthologous genes from other species of bacteria. The nucleic acid sequences of these genes, or parts thereof, may then be used to make primers or probes for detection and quantification of the target genes.

Optionally, the transcription pattern (i.e. the enhanced or reduced transcript levels of one or more of the genes mentioned above) may be used as a selection criterion. For example, bacteria may first be screened for altered expression levels in a first screen, and subsequently analysed for methotrexate resistance and/or folate levels (as described below). Likewise, it may be used as an additional screening/selection step in the method described below (for example between steps (b) and (c) or after step (d), or it may even replace steps (a) and (b).

The transcription profile of one of the methotrexate resistant strains described herein {Lactobacillus plantarum NIZO B2550) has been compared with that of the parental strain {Lactobacillus plantarum WCFSl (the full sequence of L. plantarum WCFSl has been deposited in the EMBL database under accession no. AlΛ>352(->3)). This was done by using whole genome DNA micro-arrays of Lactobacillus plantarum WCFSl. Total mRNA samples of the parental strain and the methotrexate resistant mutant were isolated and subsequently treated with reverse transcriptase to synthesize fiuorescently labelled cDNA. The samples of parental strain and mutant were differentially labelled. Subsequently, a mixture of the labelled cDNA samples was hybridized on the DNA- microarray to determine the relative abundance of the mRNA of all genes in the mutant relative to the parental strain of L. plantarum.

In one embodiment the mutant bacteria are the bacterial strains (or any derivatives thereof) deposited by NIZO Food Research, P.O. Box 20, 6710 BA Ede, the

Netherlands at the CBS (Centraalbureau voor Schimmelcultures, Uppsalalaan 8, 3584 CT Utrecht, The Netherlands) under the Budapest Treaty under Accession number CBS 117120.

Compositions according to the invention

A further aspect of the invention relates compositions comprising mutant bacteria as described herein above or derivatives or extracts thereof, as well as to methods for the production of such compositions. The bacteria of the invention are cultured under appropriate conditions, optionally recovered from the culture medium and optionally formulated into a composition suitable for the intended use. Methods for the preparation of such compositions are known per se.

Preferably, the mutant bacteria, derivatives or extracts thereof, are used to make food or food supplement compositions comprising natural folate.

A preferred composition according to the invention is suitable for consumption by a subject, preferably a human or an animal. Such compositions may be in the form of a food supplement or a whole food or food composition, which besides the bacteria of the invention also contains a suitable food base. A food or food composition is herein understood to include not only solid and semi-solid compositions, but also liquids for human or animal consumption, i.e. a drink or beverage. The food or food composition may be a solid, semi-solid and/or liquid food or food composition, and in particular may be a dairy product, such as a fermented dairy product, including but not limited to a yoghurt, a yoghurt-based drink or buttermilk. Such foods or food compositions may be prepared in a manner known per se, e.g. by adding one or more mutant bacteria of the invention to a suitable food or food base, in a suitable amount.

Also food supplements can be made, comprising suitable amounts of one or more mutant bacterial strains, or extracts therefrom (e.g. partially or substantially purified natural folate). Food supplements include for example vitamin tablets, pills, capsules or powders, comprising recommended daily amounts of various vitamins, including folate. The synthetic folic acid normally used in such preparations may be replaced with the natural folate according to the invention.

In a further preferred embodiment, the live bacteria are used in or for the preparation of a food or food composition, e.g. by fermentation. In doing so, the bacteria of the invention may be used in a manner known per se for the preparation of such fermented foods or food compositions, e.g. in a manner known per se for the preparation of fermented dairy products using lactic acid bacteria. In such methods, the bacterial cells of the invention may be used in addition to the micro-organism usually used, and/or may replace one or more or part of the micro-organism usually used. For example, in the preparation of fermented dairy products such as yoghurt or yoghurt-based drinks, a food grade mutant bacterium of the invention may be added to or used as part of a starter culture or may be suitably added during such a fermentation. The "growth medium" of the bacteria in this case is a food-grade medium, such as milk-based.

In a further embodiment the bacteria used may be dead (e.g. lysed) or non-viable, or lyophilized.

The compositions manufactured using one or more mutant bacterial strains according to the invention or supplemented with a suitable amount of one or more of the strains, comprise an enhanced amount of natural folate, produced by the bacteria. One or more strains which are high overproducers of folate may be mixed or may be added sequentially to the composition or during its production. The total amount of folate present in the final product can be analysed using for example HPLC analysis. Methods and uses according to the invention

Methotrexate is used for the selection of bacteria having increased intracellular and increased total folate levels.

In one embodiment a method for selecting bacteria having increased intracellular and/or total folate levels is provided. The method comprising the steps of:

(a) growing bacteria on (or in) medium comprising methotrexate, wherein the medium is not supplemented with folate dependent metabolites,

(b) selecting bacteria having a growth rate (μ) of at least 0.1 per hour, (c) determining the folate level of the bacteria selected in step b

(d) and selecting bacteria from having increased folate levels compared to the wild type or another suitable control,

(e) optionally repeating steps (b), (c) and/or (d).

"Growing" in step (a) may also mean contacting the bacteria with such a medium.

The medium may comprise various concentrations of MTX. For example a first round of selection may be carried out on a relatively low concentration, such as 1.25, 1.5, 2.0, 2.5mg/l medium. Isolates growing on one or more of these concentrations may then be screened directly for folate content or may be subjected to one or more further rounds of MTX selection, using for example a higher concentration of MTX, such as 4, 5, 10 or more mg/liter. The stringency of the selection process can be varied by selecting more stringently, e.g. only isolates with a high growth rate at high concentration, or less stringently, e.g. isolates showing moderate growth at relatively low MTX concentrations. Preferably, in each MTX selection step, the isolates are grown on or in the medium for at least 40 hours, more preferably at least 50 hours or more. Suitable controls, e.g. wild type isolates, should always be included, as the mutants are selected based on the difference in growth between the mutant and the control (e.g. the wild type). The control need not necessarily be a wild type, but can also be a previously selected strain. Optimum incubation temperature, pH of the medium, and other parameters can be easily determined by the skilled person. The exact conditions depend on the bacterial species used. Preferably, the method is set up in such a way that large numbers of isolates can be screened simultaneously, e.g. in 96-well microtitre plates or the like.

The method can be applied on all bacteria which show a natural sensitivity for methotrexate. Using the folate analogue methotrexate, one can select and isolate efficiently natural folate overproducing bacteria which can be used for the production of fermented food products with increased folate concentration (natural fortification or fermentation fortification) as described above.

Any bacterium obtainable according to the method and having significantly increased folate levels is comprised herein.

Unless stated otherwise, the practice of the invention will employ standard conventional methods of molecular biology, virology, microbiology or biochemistry. Such techniques are described in Sambrook et al. (1989) Molecular Cloning, A Laboratory Manual (2^nd edition), Cold Spring Harbor Laboratory, Cold Spring Harbor Laboratory Press; in Sambrook and Russell (2001) Molecular Cloning: A Laboratory Manual, Third Edition, Cold Spring Harbor Laboratory Press, NY; in Volumes 1 and 2 of Ausubel et al. (1994) Current Protocols in Molecular Biology, Current Protocols, USA; and in Volumes I and II of Brown (1998) Molecular Biology LabFax, Second Edition, Academic Press (UK); Oligonucleotide Synthesis (N. Gait editor); Nucleic Acid Hybridization (Hames and Higgins, eds.), all incorporated herein by reference.

Figure legends Figure 1:

Folate production of the L. plantarum wild type strain (1) and the L. plantarum strains harbouring pNZ7019 (2) and pNZ7021 (3) cultivated on CDM. Legend: extra-cellular folate levels; (white bar), intracellular folate levels; (grey bar), total folate levels; (black bar).

Figure 2:

The growth rate of the wild type L. plantarum strain (1), and of two L. plantarum strains harbouring the vector pNZ7019 (2) and pNZ7021 (3). The growth performance of the three L. plantarum strains was tested on CDM (a) and the modified CDM (b), containing an increasing concentration of methotrexate 0, 0.3125, 0.625, 1.25 and 2.5 mg/L (MTX).

Figure 3:

Folate production levels of 15 folate overproducing strains, selected on MTX resistance. The folate production of the 15 strains is compared to the wild type folate production

EXAMPLES

1. Materials and methods

1.1 Strains and cultivation conditions

Wild type L. plantarum WCFSl strains and its derivatives were grown at 37 ⁰C on Chemically Defined Media (CDM), containing 19 g/L β-phosphoglycerate (6) or modified CDM. In this modified CDM, glycine, inosine, orotic acid, thymidine, guanine, adenine, uracil and xantine were omitted from the normal CDM. Additionally 8 ng/ml chloroamphenicol was added to all the media used. Methotrexate (Sigma, aminopterin) was added to both types of CDM, in concentrations ranging from 0 to 10 mg/1.

1.2 DNA techniques, constructing of plasmids and transformations The plasmids used in this study are listed in Table 1.

Table 1.

The vector pNZ8148 is an empty vector containing the nisin inducible promoter, a multiple cloning site, and a chloramphenicol resistance marker. The pNZ8148 derivatives and the L. plantarum genomic DNA was isolated using established procedures. PCR was performed using Pfx DNA polymerase (Invitrogen, Paisley, UK) in a Mastercycler PCR apparatus (Eppendorf, Hamburg, Germany) using the following cycles: denaturation at 94 ⁰C for 30 s (3 minutes the first cycle), annealing at 45 ⁰C for 25 s, and elongation at 68 ⁰C for 1 minute per 1000 bps for a total of 30 cycles.

The vector pNZ8148 was digested using Sphl and BgIII, thereby excising the nisin promoter. In a parallel experiment the constitutive promoter pepN of L. lactis (Van Alen-Boerrigter et al., 1991. Appl. Environ. Microbiol. 57:2555-61) was digested from the vector pNZ7017 using Sphl and BgIII. Hereafter both fragments were ligated using T4 DNA ligase (Invitrogen), the resulting plasmid was designated as vector pNZ7020. The folate gene cluster of L. plantarum was amplified from chromosomal DNA using the following primers: (i) folBKpn-F (5'-

GAAAGAGGCTGGGT ACCATTATGGGCATGATTC-S'), which was extended at the 5' with a Kpnl restrictions site, thereby overlapping the start codon of the fo IB gene and (ii) the reverse primer LpfPxba-R (5 '-CTTAACCCCATCTAGACGTAATATCG- 3') which was extended at the 5' end to create an Xbal restriction site, allowing a fusion between the stop codon offolC and the vector. The amplified DNA fragment was digested with Xbal and Kpnl, resulting in a truncated version of the fo IB-fo IP gene cluster, (folP contains an Xbal restriction site). The vector pNZ7020 was also digested with Xbal and Kpnl, hereafter were both fragments ligated with T4 DNA ligase, the resulting vector was designated as pNZ7020a. The missing part of the folP gene was amplified by PCR using the following primers lpfP-F (5'- CATGGCATCGAT ATTGAACGAATTG-3') and the lpfPxba-R primer (5'- CTTAACCCCATCTAGACGT AAT ATCG-3'). The amplified DNA was digested with Xbal. Then the vector pNZ7020a was digested with Xbal and subsequently phosphorilised using Alkaline phosphatase (Pharmacia Biotech) to prevent self- ligation. Both Xbal digested pieces of DNA were ligated using T4 DNA ligase, the orientation of the inserted fragment was checked using PCR. The resulting vector was named pNZ7021. The three vectors listed in table 1 where, hereafter transferred into competent cells of L. lactis NZ9000 in an ylgG background. Subsequently, midiprep of the three vectors were transformed into L. plantarum using established procedures.

1.3 Folate quantification using the microbiological assay

The folate levels of the three genetically engineered L. plantarum strains were analyzed using the Lactobacillus casei microbiological assay as described previously (Sybesma et al, 2003. Appl. Environ. Microbiol. 69:3069-76.)

1.4 Monitoring growth of the strains

Microtiter plates were used to determine the growth rate of the three L. plantarum strains harbouring the different vectors. The growth of the three strains was monitored in a 96 well microtiter plate for 35 hours using a spectrophotometer (SPECTRAmax®, Molecular Devices, Sunnyvale, CA, USA). The growth was tested on the two types of media, and each construct was analyzed in triplicate on both media for the following MTX concentrations: 0, 0.3125, 0.625, 1.25 and 2.5 mg/L. 1.5 Selection for natural folate overproducers

The wild type L. plantarum WCFSl containing pNZ8148 was again cultivated in modified CDM containing 2.5 mg/L MTX, but this growth medium was split-up into 96 wells of the microtiter plates. This microtiter plate was incubated for 3 days, and hereafter aliquots from each of the 96 wells were dispersed into two fresh 96 well microtiter plate containing the modified CDM with this time 10 mg/L MTX. Through each step the original microtiter plate was mixed with 60 % glycerol and stored in a -80 ⁰C freezer. The glycerol stored microtiter plates were used to inoculate the mutated or adapted cells on fresh modified CDM medium. The folate levels of these 288 MTX resistance cultures were checked using a quick screening method. This method is based on the normal microbiological assay as described previously, however some changes were adopted. In this quick screening method the folate produced by the wild type is compared, to the folate produced by the mutant strains. Cultivation of the indicator strain on the folate produced by the wild type leads to a relative low optical density. The optical density of the indicator strain on folate overproducers will be higher then of the wild type. Strains producing an increased optical density of the indicator strain were reanalyzed in by using the extended Lb. casei microbiological assay.

2. Results

2.1 Recombinant strains

The folate production levels of L. plantarum wild type carrying pNZ8148 and the two genetic constructs pNZ7019 and pNZ7021 were studied on normal CDM medium. Folate production of L. plantarum WCFSl strain containing the empty vector pNZ8148 was found to be 100 ng fo late/ml culture (Fig 1). L. plantarum carrying pNZ7019 (the folate biosynthesis gene cluster of L. lactis) produces up to 2000 ng/ml folate in the culture (Fig. 1). This corresponds to approximately 20 times overproduction compared to the wild type. Transformation of L plantarum WCFSl with vector pNZ7021, carrying the folate biosynthesis gene cluster of L. plantarum, results in a strain producing folate levels of 6000 ng/ml culture (Fig. 1).

Next, the specific growth rates of the two folate overproducing strains and the wild type strain, were compared on CDM with and without folate dependent metabolites in the presence of a range of methotrexate (MTX) concentrations (Fig. 2). In the presence of the folate dependent metabolites, none of the tested strains was significantly inhibited by MTX (Fig. 2a). The growth rate of L. plantarum carrying plasmid pNZ7021 is lower compared to the other two strains, at all tested MTX concentrations. On a medium in the folate dependent metabolites are absent, a completely different picture arises. On this medium, the folate overproducers are clearly more resistant to MTX compared to the wild type strain (Fig 2b). At a concentration of 2.5 mg/L MTX, complete growth inhibition of the wild type strain was observed in the modified CDM, while both folate overproducing strains grew well on this medium. Between 0 and 2.5 mg/L MTX, dose dependent growth inhibition was observed for the wild type L. plantarum strain. These results show that, if analysed in a proper growth medium, folate overproduction leads to MTX resistance. Therefore, MTX can be used to discriminate folate overproducing strains from wild type strains.

2.2 Wild type strains

Prolonged incubation (>40 hours) of the wild type L. plantarum WCFSl on modified CDM supplemented with 2.5 mg/L MTX, resulted in significant growth of the culture. If this culture was diluted in fresh modified CDM supplemented with 2.5 mg/L MTX, immediate growth of the culture was observed. Single colonies were isolated from this culture to test if natural mutants with increased MTX resistance were selected or if some kind of temporary adaptation takes place. Modified CDM containing 2.5 mg/1 MTX was inoculated with a single colony derived from an MTX resistant culture. Next, the growth medium was split into two cultures; one culture was grown for 50 generation in the modified CDM with 2.5 mg/1 MTX, the other culture was grown for 50 generations in the same medium in the absence of MTX. After the 50 generation both cultures where transferred again to modified CDM containing the 2.5 mg/1 MTX. Both strains with a different history revealed a comparable growth rate on this medium indicating that a stable mutation causes the MTX resistance phenotype. We isolated 288 colonies from cultures containing 2.5 mg/L or 10 mg/L MTX.

These colonies were analysed for folate overproduction using a rapid pre-screening method. Based of this rapid screen, 5 out of 288 colonies gave raise to cultures with higher folate levels compared to the wild type. These 5 cultures were subsequently plated on modified CDM supplemented with 2.5 mg/L MTX. From each plate 3 single colonies were isolated, stored and analyzed again for folate production in the following way. Folate levels in cultures derived from 15 positive colonies were monitored (Fig. 3). The average folate levels are 20% to 60% higher than that observed in the wild type. To confirm that all colonies were still L. plantarum, we have performed RAPD analyses. All colonies tested showed a high similarity with the parental strain.

2.3. Transcription profiling.

Using oligonucleotide microarray technology for genome-wide transcription profiling, we have identified genes in the methotrexate resistant mutant L. plantarum NIZO B2550 (derived from L. plantarum WCFSl) which are specifically up- or down regulated compared to the wild-type. Cultures of L. plantarum WCFSl and NIZO B2550 were grown on modified CDM and at a turbidity (OD600) of 1, cells were harvested and mRNA was isolated. The mRNA samples were processed for analysis on DNA micro-arrays. The relative abundance of specific mRNA's was expressed as the fold change in expression of genes in the methotrexate resistant strain compared to the corresponding genes in the parental strain. Analysis of the transcriptome profile reveals that in comparison with the parental strain, fo ICl, the gene coding for a dihydro folate synthase is specifically upregulated 4-fold (see Table 2) in the methotrexate resistant mutant compared to the wild-type. This enzyme couples L-glutamate to dihydropteroate to form dihydro folate, which is a structural analogue of methotrexate and serves as the substrate for the enzyme dihydro folate reductase. More dihydrofolate synthase through overexpression oϊfolCl is expected to lead to higher intracellular dihydrofolate pools, which in turn could compete with methotrexate for conversion by dihydrofolate reductase. This observation reveals the mechanism of methotrexate resistance in the isolated mutants. It also explains why these mutants produce higher levels of folate in the absence of methotrexate.

L. plantarum WCFSl has two folC genes (folCl and folC2). Interestingly, only folCl is upregulated in the methotrexate resistant mutant. This gene is not located in a functional cluster, this in contrast to folC2, which is located in the folate biosynthesis cluster.

In addition ϊofolCl, a number of other metabolic genes (coding for fructokinase, pyruvate oxidase and 6-phospho-beta-glucosidase) were found to be upregulated in the methotrexate resistant strain (see Table 2). Three genes (daklA, daklB and dakl) coding for a glycerone kinase (dihydroxyacetone phosphotransferase) were found to down-regulated in the mutant as compared to the parental strain.

Table 2. Genes specifically up- and down regulated in L. plantarum NIZO B2550 compared to L. plantarum WCFSl.

Claims

1. A mutant bacterium, characterized in that the bacterium comprises an increased amount of intracellular and/or extracellular folate compared to the wild type and having a growth rate of at least 0.1 μ per hour when grown on medium comprising 1.25 mg/1 methotrexate and lacking folate dependent metabolites.

2. The bacterium according to claim 1, wherein said bacterium is a naturally occurring mutant or an induced mutant.

3. The bacterium according to claim 1 or 2, wherein said bacterium shows overexpression compared to the wild-type of one or more genes coding for enzymes of the folate biosynthesis pathway.

4. The bacterium according to any of the preceding claims, wherein the expression of one or more of the genes selected from the genes coding for dihydro folate synthase, fructokinase, pyruvate oxidase and 6-phospho-β-glucosidase are upregulated at least about 2-fold compared to the wild-type.

5. The bacterium according to claim 1, 2 or 4, wherein the expression of the gene coding for dihydroxyacetone phosphotransferase is downregulated at least about 2-fold compared to the wild type.

6. The bacterium according to any of the preceding claims, wherein the amount of total folate of the mutant bacterium is at least 10% higher than that of the wild type strain.

7. The bacterium according to any of the preceding claims, wherein the bacterium is of the species Lactobacillus plantarum, L. casei, L. reuteri, L. fermentum, L. acidophilus, L. crispatus, L. gasseri, L. johnsonii, L. paracasei, L. murinus, L. jensenii, L. salivarius, L. minutis, L. brevis, L. gallinarum, L. amylovorus,

Lactococcus lactis, Streptococcus thermophilus, Leuconostoc mesenteroid.es, Lc. lactis, Pediococcus damnosus, P. acidilactici, P. parvulus, Bifidobacterium bifidum, B. longum, B. infantis, B. breve, B. adolescente, B. animalis, B. gallinarum, B. magnum, and B. thermophilum.

8. Lactobacillus plantarum-stmin CBS 117120.

9. A composition comprising a bacterium according to any of the preceding claims.

10. The composition according to claim 9, wherein the composition is a food composition or food supplement composition.

11. The composition according to claim 10 wherein the food composition is a fermented food composition.

12. Use of methotrexate for the selection of bacteria having increased intracellular and increased total folate levels.

13. A method for selecting bacteria having increased intracellular and/or total folate levels, said method comprising the steps of:

(a) growing bacteria on or in a medium comprising methotrexate, wherein the medium is not supplemented with folate dependent metabolites,

(b) selecting bacteria having a growth rate (μ) of at least 0.1 per hour,

(c) determining the folate level of the bacteria selected in step (b) and selecting bacteria from having increased folate levels compared to the wild type, (d) optionally repeating steps (a), (b) and/or (c).

14. The method according to claim 13, wherein said medium comprises at least

1.25 mg/1 methotrexate and wherein the bacteria are grown on said medium for at least 40 hours.