No abstract

Aim:  A medium with minimal requirements for the growth of Lactobacillus plantarum WCFS was developed. The composition of the minimal medium was compared to a genome‐scale metabolic model of L. plantarum. Methods and Results:  By repetitive single omission experiments, two minimal media were developed: PMM5 (true minimal medium) and PMM7 [a pseudominimal medium, supporting proper biomass formation of 350 mg l−1 dry weight (DW)]. The specific growth rate of L. plantarum on PMM7 was found to be 50% and 63% lower when compared to growth on established growth media (chemically defined medium and MRS, respectively). Using a genome‐scale metabolic model of L. plantarum, it was predicted that PMM5 and PMM7 would not support the growth of L. plantarum. This is because the biosynthesis of para‐aminobenzoic acid (pABA) was predicted to be essential for growth. The discrepancy in simulated growth and experimental growth on PMM7 was further investigated for pABA; a molecule which plays an important role in folate production. The growth performance and folate production were determined on PMM7 in the presence and absence of pABA. It was found that a 12 000‐fold reduction in folate pools exerted no influence on formation of biomass or growth rate of L. plantarum cultures when grown in the absence of pABA. Conclusion:  Largely reduced folate production pools do not have an effect on the growth of L. plantarum, showing that L. plantarum makes folate in a large excess. Significance and Impact of the study:  These experiments illustrate the importance of combining genome‐scale metabolic models with growth experiments on minimal media.

The pab genes for para-aminobenzoic acid (pABA) biosynthesis in Lactococcus lactis were identified and characterized. In L. lactis NZ9000, only two of the three genes needed for pABA production were initially found. No gene coding for 4-amino-4-deoxychorismate lyase (pabC) was initially annotated, but detailed analysis revealed that pabC was fused with the 3 end of the gene coding for chorismate synthetase component II (pabB). Therefore, we hypothesize that all three enzyme activities needed for pABA production are present in L. lactis, allowing for the production of pABA. Indeed, the overexpression of the pABA gene cluster in L. lactis resulted in elevated pABA pools, demonstrating that the genes are involved in the biosynthesis of pABA. Moreover, a pABA knockout (KO) strain lacking pabA and pabBC was constructed and shown to be unable to produce folate when cultivated in the absence of pABA. This KO strain was unable to grow in chemically defined medium lacking glycine, serine, nucleobases/nucleosides, and pABA. The addition of the purine guanine, adenine, xanthine, or inosine restored growth but not the production of folate. This suggests that, in the presence of purines, folate is not essential for the growth of L. lactis. It also shows that folate is not strictly required for the pyrimidine biosynthesis pathway. L. lactis strain NZ7024, overexpressing both the folate and pABA gene clusters, was found to produce 2.7 mg of folate/liter per optical density unit at 600 nm when the strain was grown on chemically defined medium without pABA. This is in sharp contrast to L. lactis strains overexpressing only one of the two gene clusters. Therefore, we conclude that elevated folate levels can be obtained only by the overexpression of folate combined with the overexpression of the pABA biosynthesis gene cluster, suggesting the need for a balanced carbon flux through the folate and pABA biosynthesis pathway in the wild-type strain.