In vitro and in vivo characterization of DNA delivery using recombinant Lactococcus lactis expressing a mutated form of L. monocytogenes Internalin A

Background: The use of food-grade Lactic Acid Bacteria (LAB) as DNA delivery vehicles represents an attractive strategy to deliver DNA vaccines at the mucosal surfaces as they are generally regarded as safe (GRAS). We previously showed that either native Lactococcus lactis (LL) or recombinant invasive LL expressing Fibronectin Binding Protein A of Staphylococcus aureus (LL-FnBPA+) or Internalin A of Listeria monocytogenes (LL-InlA+), were able to deliver and trigger DNA expression by epithelial cells, either in vitro or in vivo. InlA does not bind to its receptor, the murine E-cadherin, thus limiting the use of LL-InlA+ in in vivo murine models. Moreover, FnBPA binds to its receptors, integrins, via fibronectin introducing another limiting factor. In order to avoid the limitations of LL-InlA+ and LL-FnBPA+, a new L. lactis strain was engineered to produce a previously described mutated form of InlA (LL-mInlA+) allowing the binding of mInlA on murine E-cadherin. Results: After showing the expression of mInLA at the surface of LL-mInlA+ strain, in vitro gentamycin survival assay in Caco-2 cells showed that LL-mInlA+ is 1000 times more invasive than LL. LL-mInlA+ invasivity was also validated by fluorescence microscopy. LL and LL-mInlA+ were transformed with pValacBLG, a plasmid containing the cDNA of bovine beta-Lactoglobulin (BLG), resulting in strains LL-BLG and LL-mInlA+ BLG. The plasmid transfer in vitro using LL-mInlA+ BLG was increased 10 times compared to LL-BLG. Moreover, the number of mice producing BLG in isolated enterocytes after oral administration of LL-mInlA+ BLG in vivo was slightly higher than after oral administration of LL-BLG. Conclusions: We confirmed in this study that the production of mInlA at the surface of L. lactis is a promising strategy for plasmid transfer in vitro and in vivo