Lipid selectivity in novel antimicrobial peptides: Implication on antimicrobial and hemolytic activity.

Antimicrobial peptides (AMPs) are small cationic molecules that display antimicrobial activity against a wide range of bacteria, fungi and viruses. For an AMP to be considered as a therapeutic option, it must have not only potent antibacterial properties but also low hemolytic and cytotoxic activities [1]. Even though many studies have been conducted in order to correlate the antimicrobial activity with affinity toward model lipid membranes, the use of these membranes to explain cytotoxic effects (especially hemolysis) has been less explored. In this context, we studied lipid selectivity in two related novel AMPs, peptide 6 (P6) and peptide 6.2 (P6.2). Each peptide was designed from a previously reported AMP, and specific amino acid replacements were performed in an attempt to shift their hydrophobic moment or net charge. P6 showed no antimicrobial activity and high hemolytic activity, and P6.2 exhibited good antibacterial and low hemolytic activity. Using both peptides as a model we correlated the affinity toward membranes of different lipid composition and the antimicrobial and hemolytic activities. Our results from surface pressure and zeta potential assays showed that P6.2 exhibited a higher affinity and faster binding kinetic toward PG-containing membranes, while P6 showed this behavior for pure PC membranes. The final position and structure of P6.2 into the membrane showed an alpha-helix conversion, resulting in a parallel alignment with the Trps inserted into the membrane. On the other hand, the inability of P6 to adopt an amphipathic structure, plus its lower affinity toward PG-containing membranes seem to explain its poor antimicrobial activity. Regarding erythrocyte interactions, P6 showed the highest affinity toward erythrocyte membranes, resulting in an increased hemolytic activity. Overall, our data led us to conclude that affinity toward negatively charged lipids instead of zwitterionic ones seems to be a key factor that drives from hemolytic to antimicrobial activity.

Neutrophil elastase treated dendritic cells promote the generation of CD4(+)FOXP3(+) regulatory T cells in vitro.

We have previously shown that neutrophilic elastase converts human immature dendritic cells (DCs) into TGF-ß secreting cells and reduces its allostimulatory ability. Since TGF-ß has been involved in regulatory T cells (Tregs) induction we analyzed whether elastase or neutrophil-derived culture supernatant treated DCs induce CD4(+)FOXP3(+) Tregs in a mixed lymphocyte reaction (MLR). We found that elastase or neutrophil-derived culture supernatant treated DCs increased TGF-ß and decreased IL-6 production. Together with this pattern of cytokines, we observed a higher number of CD4(+)FOXP3(+) cells in the MLR cultures induced by elastase or neutrophil-derived culture supernatant treated DCs but not with untreated DCs. The higher number of CD4(+)FOXP3(+) T cell population was not observed when the enzymatic activity of elastase was inhibited with an elastase specific inhibitor and also when a TGF-ß1 blocking antibody was added during the MLR culture. The increased number of CD4(+) that express FOXP3 was also seen when CD4(+)CD25(-) purified T cells were cocultured with the TGF-ß producing DCs. Furthermore, these FOXP3(+) T cells showed suppressive activity in vitro. These results identify a novel mechanism by which the tolerogenic DCs generated by elastase exposure contribute to the immune regulation and may be relevant in the pathogenesis of several lung diseases where the inflammatory infiltrate contains high numbers of neutrophils and high elastase concentrations.

Replacement of methionine-161 with threonine eliminates a major by-product of human glutamate decarboxylase 65-kDa variant expression in Escherichia coli.

Most insulin-dependent diabetes mellitus patients gen-erate conformational autoantibodies to the islet-cell 65-kDa variant of human glutamate decarboxylase (GAD65), and several immunochemical tests for the early detection of type-1 diabetes rely on GAD65 antibody (GADA) assessment using properly folded recombinant GAD65 as the antigen. In addition, preventive therapies based on tolerization by GAD65 administration may be available in the near future. Therefore, there exists a strong interest in a facile and economically sound expression procedure for this antigen. Several attempts to produce, in native form, wild-type GAD65 in Escherichia coli have failed. However, this difficulty was recently surmounted in our laboratory by expressing GAD65 as a fusion protein with thioredoxin [Papouchado, Valdez, Ghiringhelli, Poskus and Ermácora (1997) Eur. J. Biochem. 246, 350-359]. In this work, a new GAD65 hybrid gene was prepared by joining engineered cDNA obtained from human and rat tissues. The new gene was modified additionally to finally code for human GAD65 with a single amino-acid substitution: Met-161-->Thr. This change impeded the co-expression of a 48-kDa by-product from an internal translation site. Also, a second 58-kDa by-product was identified as a GAD65 C-terminal proteolytic fragment that co-purifies with thioredoxin-M161T GAD65. The new GAD65 variant was expressed and easily purified, yielding an antigen that performed equally or better than wild-type GAD65 in the reference radiobinding assay for GADA. The procedure provides an inexpensive source of large amounts of fully active and immunochemically competent GAD65.