Efrfect of stress on trehalose production: Information was provided on the effect of aerobiosis on growth and trehalose accumulation in Propionibacterium freudenreichii subsp. shermanii NIZO B365. Growth under aerobic conditions (10% oxygen) at pH 7.0 had a drastic impact on the maximal intracellular concentration of trehalose (540 mg/g of protein in the late exponential phase). Additionally, the effect of other environmental conditions (NaCl concentration, growth temperature, pH, carbon source) on the same parameters was examined. Trehalose accumulates in response to salinity, oxygen or low pH, up to levels of 40% of the cell dry mass. Pathways for the synthesis of trehalose in Propionibacterium: Two pathways for the synthesis of trehalose were identified in Propionibacterium freudenreichii subsp. shermanii NIZO B365: i) the two-step pathway involving the enzymes trehalose 6-phosphate synthase (TPS) and trehalose 6-phosphate phosphatase (TPP) and ii) the single step pathway in which trehalose is synthesised from maltose by the action of trehalose synthase (TS). The genes encoding TPS, TPP, and TS enzymes were cloned and expressed in E. coli and the recombinant enzymes characterised in detail. TPP was highly specific for trehalose 6-phosphate and was Mg2+ dependent. The kinetic parameters were KM 0.92 mM and Vmax 133 U/mg of protein. TPS was highly specific for glucose 6-phosphate but was able to use UDP-, ADP-, GDP- or TDP-glucose as glucosyl donors although there was a clear preference for UDP-glucose. The kinetic parameters of TS obtained for maltose were KM 6.4mM and Vmax 48U/mg of protein whereas KM 58.9mM and Vmax 158U/mg were determined for trehalose. The expression of the two pathways under different environmental conditions was assessed from immunoassays. A manuscript describing our results is nearly ready for submission (see Articles). Trehalose production in fermented milk: A natural strain accumulating high levels of trehalose (about 20% of cell mass) was identified among 18 examined strains of Propionibacterium spp. isolated from dairy sources (obtained either from DSM or NIZO culture collections). This strain was used to evaluate the production of trehalose in skim milk supplemented with casitone and lactate, to mimic the actual conditions in dairy fermentations. Trehalose accumulated to similar levels (250mg/g of cell protein). This was an encouraging result for the practical applications of this strain, aiming at trehalose production in situ. During the mid-term review of this project it was decided that the implementation of a trehalose-containing fermented dairy drink should be cancelled since the number of total Deliverables proposed was exaggerated.
Optimal production of low-calorie sugars In collaboration with partner 3 (UCL), we have developed a fermentation method to produce sorbitol and mannitol in a pH controlled batch culture. With the optimised production strains delivered by UCL (Group Hols), we were able to obtain 30% rerouting towards polyols in growing cultures. The feasibility of using homofermentative LAB (the major players in food fermentation) for polyol production has been demonstrated in this project. The fact that genetically modified strains were used in this task hampers the commercial application of the method.
A glucose Ylactose+ L. lactis strain was delivered. This strain produces glucose as a natural sweetener in milk products and has the additional benefit of reducing the lactose content of those products. Experimental dairy products using this strain have beeb produced at NIZO food research. The results will be combined and a manuscript will be submitted shortly. Deletion of the genes coding for glucokinase, mannose-PTS (PTSman) and cellobiose-PTS (PTScel) is required and sufficient to completely block glucose metabolism in L. lactis. The PTScel was shown to be an alternative glucose uptake system in L. lactis. Contribution of the different glucose transport system was evaluated by studying strains with different deletion combinations of glk (glucokinase), ptnABCD (PTSman) and ptcBA (PTScel). Glucose metabolism was studied by in vivo NMR, several biochemical assays (14C-glucose uptake assays, enzyme activities, etc) and transcriptome analysis of some strains was performed to investigate molecular mechanisms governing glucose uptake in this organism. A manuscript reporting this data is in late stage of preparation. Analysis of glucokinase-deficient strains indicated an interaction between glucokinase and the PTSman that restricts transport through the latter under certain conditions. The nature of this interaction will be explored in a follow-up project.
Biosynthesis of a pneumococcal polysaccharide in L. lactis L. lactis was successfully engineered for the production of a new polysaccharide. By introduction of the serotype specific part of the Streptococcus pneumoniae gene cluster encoding capsular polysaccharide (CPS, serotype 14) in a L. lactis background that expressed the general, regulatory polysaccharide biosynthesis genes, the resulting L. lactis strain produced a polysaccharide that is biochemically identical to the native pneumococcal polymer. Remarkably, the cassette cloning combination described above was the only combination that led to the production of pneumococcal polysaccharides, indicating a very strict host specificity of the regulatory units involved in polysaccharide biosynthesis. Notably, L. lactis released the type 14 polysaccharide in the medium while Pneumococcus produces a capsular form of this polymer, which is more difficult to isolate and separate from the producing cells. These findings have great potential for the production of polysaccharide ingredients for vaccin preparations that are derived from pathogenic hosts by food-grade bacteria (L. lactis), providing the basis for a patent application. Technical implementation of this deliverable includes the filing of a patent application related to this work in December of 2004, publication in a scientific journal (in preparation) and presentation of these findings at several meetings. Currently, the possibilities to obtain funding to continue this work are evaluated.
High level B-vitamin production by natural food bacteria: We have developed, validated and applied new methods to select natural B-vitamin producing food grade bacteria. From the NIZO culture collection, Propionibacterium freudenreichii strains were selected with relatively high folate production. These P. freudenreichii are natural producers of vitamin B12. Subsequently, these strains were treated with riboflavin (vitamin B2) analogues to obtain natural riboflavin overproducers with a stable phenotype. These strains were deposited in the NIZO culture collection. The triple B-vitamin (B2, B11 and B12) producing P. freudenreichii strain B2336 has been used in pilot yoghurt production trials. We have developed a method for co-fermentation of P. freudenreichii with a traditional yoghurt starter culture. The method consists of a pre-fermentation with a propionibacterium strain followed by fermentation with the classical yoghurt starter. The method of sequential inoculation yields good viable counts of the propionibacterium, which results in increased levels of B-vitamins in the end-product. If the strain P. freudenreichii NIZO B2336 was used in pilot production trial, a doubling of the amount of vitamin B2 was found in the product. NIZO food research has initiated contract research with one of the partners in the NutraCells consortium to bring the prototype production process for making a fermented dairy product with increased B-vitamin levels into the product development trajectory.
Lactococcus lactis with overexpressed rib genes: This deliverable was achieved and generated a strain with an extremely high level of riboflavin production. This finding, together with additional characterisation of the riboflavin operon in this bacterium, was published in the peer-reviewed journal Applied and environmental Microbiology (2004, vol.70: pp. 5769-5777)1. In addition, a multivitamin producing L. lactis strain was constructed which overproduced both riboflavin as well as folic acid. The latter results were published in the journal Metabolic Engineering (2004, vol. 6: pp. 109-115)2. Furthermore, the mechanism by which L. lactis internalises riboflavin from its growth medium, thereby reducing the amount of riboflavin in foods, has been elucidated and the scientific work describing these findings will soon be submitted to the journal Molecular Microbiology. Mutant lactic acid bacteria with high riboflavin production: This deliverable was achieved using a number of lactic acid (and propionic acid) bacteria. By exposing Lactococcus lactis, Lactobacillus plantarum, Leuconostoc mesenteroides and Propionibacterium freudenreichii to roseoflavin, riboflavin-overproducing variants were obtained. The methodology to produce such mutants was proven very useful, allowing the isolation of vitamin B2-overproducing starter bacteria, which can be used for the in situ production of this vitamin in dairy foods. This work will soon be submitted for publication to the peer-reviewed scientific journal Applied Environmental Microbiology. High production of riboflavin with food-grade engineered LAB: Both genetically engineered as well as roseoflavin-induced LAB were isolated that produced high levels of riboflavin. Several of such LAB were directly derived from commercially used starter cultures. Some of these strains were used for the animal trials, while the technical know-how of generating such strains was applied by one of the industrial partners, which has lead to a number of commercially exploitable strains. Some of these results have been accepted for publication in the British Journal of Nutrition3, while a second publication on the dairy-based products containing increased levels of in vivo produced riboflavin is currently under submission in the Journal of Dairy Science.
In this Workpackage, the research focus was on isolation and/or construction of strains with high ƒÑ-galactosidase activity. Partner 05 was responsible for the targeted expression of a-galactosidase (encoded by the melA gene) in Lactococcus lactis, Partner 06 looked for sources of enzymes with high a-galactosidase activity and performed all the animal trials and Partner 07 screened their culture collection for lactic acid bacteria with high ƒÑ-galactosidase activities. This partner also supplied the soy material that can be pre-fermented for a-galactoside-removal. Newly isolated and constructed strains have been used successfully in animal experiments to test for raffinose and stachyose removal leading to reduced flatulence (gas formation).
L. lactis with high Leloir activity A strain with significant increase in the galactose consumption rate was delivered. L. lactis strains with high Leloir activity (overexpression of galP or galPMKT) did not show improved galactose utilisation, but instead poor growth on galactose was observed. Using in vivo NMR, accumulation of the Leloir pathway intermediates galactose-1-phosphate and glucose-1-phosphate was observed, suggesting a bottleneck at the level of ƒÑ-PGM. It was expected that this obstruction could be overcome by overexpressing a gene encoding ƒÑ-PGM, which was unknown in L. lactis (new deliverable). Overexpression of S. thermophilus pgmA together with galPMKT resulted in reduced amounts of the phosphorylated metabolites and increased galactose consumption rates, showing that £\-PGM was indeed a major bottleneck in galactose metabolism. A manuscript describing our results is under preparation (see Articles). This work is the result of collaboration with partner 2 (RUG, Oscar Kuipers). Identification and characterization of L. lactis £\-PGM. A unique bacterial phosphoglucomutase was identified and fully characterized. Lactococcus lactis contains ƒÑ-phosphoglucomutase (ƒÑ-PGM) activity, which catalyzes the specific conversion of ƒÑ glucose 1P into glucose 6P, reversibly. However, overexpression of femD, the only homologue in the sequenced genome of this organism with significant identity to ƒÑ-D-phosphohexomutases, did not result in the expected enhanced ƒÑ-PGM activity. The ƒÑ-PGM activity was purified and its N-terminal determined. A gene coding for the isolated protein was identified by homology search. Cloning and overexpression of that gene under the nisin promoter in pNZ8048 led to a 30-fold increase of ƒÑ-PGM activity. Biochemical characterization of the enzyme was performed and the effect of its overproduction on glucose and galactose metabolism was investigated. A manuscript describing our results is under preparation (see Articles). This work is the result of collaboration with partner 2 (RUG, Oscar Kuipers).
A focus was placed on increasing the vitamin B12 content in yoghurt using selected strains. Propionibacterium and Lactobacillus reuteri are known vitamin B12 producers and the performance of selected strains, obtained from NIZO food research and CERELA, was tested with respect to vitamin B12 production in yoghurt trials. Strains Propionibacterium B374 and B369 from the NIZO strain collection are relatively high folic acid producers (about 20mg/L x OD), but their capacity to produce vitamin B12 is yet unknown. The strains were pre-cultured in skimmed milk + 0.1% (w/v) casiton to support their growth. Yoghurt was prepared at 30 degrees Celsius using the MUH306 yoghurt culture with varying dosages of Propionibacterium B374 and B369 pre-culture. The results show that during yoghurt fermentation no growth occurred of Propionibacterium. The total vitamin B12 content in prepared yoghurt s was determined by a microbiological assay using Lactobacillus delbrueckii ATCC 7830. Vitamin concentration in milk and yoghurt samples is measured before fermentation (t=0), after fermentation (t= pH4.5) and after 14 days (P+14) storage at 4 degrees Celsius. It is well known that during preparation of yoghurts the vitamin B12 content is reduced because the yoghurt bacteria consume vitamin B12 for growth purposes. In addition, this process of vitamin B12 consumption continues during storage of the yoghurt at 4 degrees Celsius. The results of the reference yoghurt MUH306 confirm these literature reports. Co-cultivation of yoghurt culture with Propionibacterium strains results in an increased level of vitamin B12 in the final product. The results show that Propionibacterium does not grow during yoghurt fermentation, but is nevertheless able to actively produce vitamin B12. In addition, both Propionibacterium strains were shown to produce vitamin B12 at 4 degrees Celsius and consequently vitamin B12 concentration is enhanced in yoghurt during storage at 4 degrees Celsius for 14 days. In a second series of yoghurt trials we have tested the ability of selected Lactobacillus reuteri strains NIZO B1533 and CRL 1098 to produce vitamin B12 during yoghurt fermentation at 37 degrees Celsius and during subsequent storage for 14 days at 4 degrees Celsius. Lb. reuteri shows a ten-fold increase in cell numbers during yoghurt fermentation, but no growth is observed during storage at 4 0C. Both tested Lb. reuteri strains were able to enhance the vitamin B12 concentration level in yoghurt. Furthermore, the results clearly show that Lb reuteri strains produce vitamin B12 in yoghurt during storage for 14 days at 4 degrees Celsius. This is the first report that describes the production of vitamin B12 in milk and yoghurt using Lb. reuteri strains. In the literature it has been suggested that Lb.reuteri CRL1098 might be a good candidate to increase vitamin B12 content in fermented food (Taranto, 2003). However, Taranto and co-workers only describe vitamin B12 production by Lb. reuteri using synthetic laboratory growth media. Furthermore, these authors provide evidence that exogenous added vitamin B12 to the growth medium may repress endogenous vitamin B12 biosynthesis in Lb. reuteri, as has been observed in other micro-organisms. Remarkably, our results clearly show that Lb. reuteri produces vitamin B12 in milk, despite the fact that milk itself also contains vitamin B12. Grading results show that 1% (v/v) inoculum of Lb. reuteri strains produces a mild, sweet flavour to the yoghurt in comparison with reference yoghurt. Similar results were observed for 5% (v/v) inoculum with this exception that strain B1533 in addition produces an extra off-flavour described as "old" and "carton" in the yoghurt.
High folate-producing Lactococcus lactis Through the strategy of metabolic engineering, high folate-producing Lactococcus lactis variants were constructed. In the first year, threefold increase in folate production was achieved by overexpression of the first genes in folate biosynthesis, folKE. Also, reduced folate production was achieved by overexpression of another gene, folA. In the third year, much higher (200-fold) folate production was achieved by overexpression of all folate genes, except folA. This high folate production (10 - 15mg/L) only occurred in growth medium containing para-aminobenzoic acid, a precursor for folate. Another goal of the metabolic engineering strategy was to control bioavailability of the produced folate. By cloning of the human (and rat) folate deconjugase gene in L.lactis, the polyglutamyl-taillength of the produced folate was reduced, leading to more efficient excretion of the folate in the growth medium. By overexpression of the folC gene in L. lactis, coding for polyglutamyl folate synthetase, the produced folate was had much longer polyglutamyl tails and the folate was no longer excreted into the medium. The gene vectors that were used to achieve high folate production in L. lactis, were also transfered to other lactic acid bacteria. In this way, the non-folate producer Lactobacillus gasseri, was transformed into a folate-producer. An additional scientific result of the metabolic engineering was the discovery of a new gene involved in folate biosynthesis, folQ. Finally, a transcriptome analysis was performed on the folate-overproducing Lactococcus lactis. This provided some leads for future metabolic engineering strategies. These results have been described in thirteen scientific publication, three contributions (chapters) to text books, one PhD Thesis (Wilbert Sybesma) and several popular publications and have been presented at a large number of international scientific meetings. The high-folate-producing Lactococcus lactis strains are not (yet) commercialised, for the obvious reason that they are GMO's. However, since we are dealing with homologous recombination in almost all results, these strains do not need to labeled as GMO's in some non-European countries such as the USA, and could reach these markets in the near future. For this reason, it was checked, in thefinal year of the project, if these high-folate-producers are an effective source of folate. In animal trials using rats fed on a folate-depleted diet, it was convincingly shown that the bacterial folate is a good, and bioavailable, source of folate. As to other implementation of the results, the metabolic engineering strategy has clearly indicated how high folate production can be achieved. It also presented some possibilities to increase folate by changes in the medium composition (para-aminobenzoic acid!) or by controlled fermentation (slow growth leads to high folate!)