Lactobacillus gasseri LA39 and L. reuteri LA6 isolated from feces of the same human infant were found to produce similar cyclic bacteriocins (named gassericin A and reutericin 6, respectively) that cannot be distinguished by molecular weights or primary amino acid sequences. However, reutericin 6 has a narrower spectrum than gassericin A. In this study, gassericin A inhibited the growth of L. reuteri LA6, but reutericin 6 did not inhibit the growth of L. gasseri LA39. Both bacteriocins caused potassium ion efflux from indicator cells and liposomes, but the amounts of efflux and patterns of action were different. Although circular dichroism spectra of purified bacteriocins revealed that both antibacterial peptides are composed mainly of ␣-helices, the spectra of the bacteriocins did not coincide. The results of D-and L-amino acid composition analysis showed that two residues and one residue of D-Ala were detected among 18 Ala residues of gassericin A and reutericin 6, respectively. These findings suggest that the different D-alanine contents of the bacteriocins may cause the differences in modes of action, amounts of potassium ion efflux, and secondary structures. This is the first report that characteristics of native bacteriocins produced by wild lactobacillus strains having the same structural genes are influenced by a difference in D-amino acid contents in the molecules.

Gassericin A, produced by Lactobacillus gasseri LA39, is a hydrophobic circular bacteriocin. The DNA region surrounding the gassericin A structural gene, gaaA, was sequenced, and seven open reading frames (ORFs) of 3.5 kbp (gaaBCADITE) were found with possible functions in gassericin A production, secretion, and immunity. The deduced products of the five consecutive ORFs gaaADITE have homology to those of genes involved in butyrivibriocin AR10 production, although the genetic arrangements are different in the two circular bacteriocin genes. GaaI is a small, positively charged hydrophobic peptide of 53 amino acids containing a putative transmembrane segment. Heterologous expression and homologous expression of GaaI in Lactococcus lactis subsp. cremoris MG1363 and L. gasseri JCM1131 T , respectively, were studied. GaaI-expressing strains exhibited at least sevenfold-higher resistance to gassericin A than corresponding control strains, indicating that gaaI encodes an immunity peptide for gassericin A. Comparison of GaaI to peptides with similar characteristics found in the circular bacteriocin gene loci is discussed.

Lactobacillus gasseri LA39 and LA158 isolated from human-infant feces produce bacteriocins named gassericins A and T, respectively. Both gassericins have high heat stability (121 degrees C, 10 min), good pH tolerance (pH 2-11), and strong bactericidality against many gram-positive bacteria, especially lactic acid bacteria, and thus are expected to be effective food preservatives. A microwell plate assay against 12 strains of custard cream spoilage bacteria showed that the gassericins had broader antibacterial spectra than nisin A. Although the gassericins allowed gram-negative isolates to grow, they successfully inhibited the growth of all tested bacterial strains in microwells with the addition of glycine. Glycine was bacteriostatic against many strains except lactic acid bacteria. For practical use, gassericin A was efficiently produced by cultivation in a food-grade medium improved using cheese whey, nourishing proteose peptone, and surfactant yolk lecithin. The practical preservative effect of gassericin A and glycine was verified from the viability of 4 isolated strains, Bacillus cereus, Lactococcus lactis ssp. lactis, Achromobacter denitrificans, and Pseudomonas fluorescens, in custard creams. Custard cream containing 123 arbitrary units of gassericin A per milliliter entirely growth-inhibited the 2 gram-positive strains. In custard cream containing an insufficient amount of gassericin A (49 arbitrary units/mL), the gram-positive strains gradually grew but were completely inhibited by the addition of 0.5% (wt/wt) glycine. The 2 gram-negative strains did not multiply even in the additive-free custard cream, probably because of the unsuitable growth environment. This is the first report showing the combined effect of bacteriocin and glycine and their application for food preservation, which may be helpful for future use in the food industry.

Aim:  The study aimed for the complete purification and recharacterization of the highly hydrophobic circular bacteriocins, gassericin A and reutericin 6. Methods and Results:  Gassericin A and reutericin 6 were purified to homogeneity using previously described method and reverse‐phase HPLC with an octyl column and eluents of aqueous acetonitrile and 2‐propanol. Mass analysis, N‐terminal sequencing and bacteriocin assay of the HPLC‐purified bacteriocins showed the two bacteriocins had identical seamless circular structures with the same m/z value (5651) of [M + H]+ and both had the same specific activity. d/l‐amino acid composition analysis using two distinct methods with the chiral fluorescent derivatization reagents (+)‐1‐(9‐fluorenyl)ethyl chloroformate and o‐phthalaldehyde/N‐acetyl‐l‐cystein revealed neither gassericin A nor reutericin 6 contained d‐alanine residues contrary to our previous results. Conclusion:  Purified gassericin A and reutericin 6 are chemically identical circular molecules containing no d‐alanine residues. Significance and Impact of the Study:  The HPLC conditions developed in this study will facilitate advanced purification and correct characterization of other highly hydrophobic bacteriocins.

Gassericin A, a bacteriocin produced by Lactobacillus gasseri LA39, has a cyclic structure linking N-and C-terminal amino acids. Gassericin A was expressed in Escherichia coli JM109 as a biotinylated fusion protein on the basis of the DNA sequence of mature bacteriocin. A positive clone accumulated the bacteriocin, with no activity, as a soluble fusion protein in the cytoplasm. After release of an N-terminal tag with factor Xa protease, gassericin A was converted into an active peptide having N-and C-termini. The total amount of purified bacteriocins (expressed and native) was 480 m g/L and 370 m g/L, respectively. However, the specific activity of expressed gassericin A was 15 AU/mg lower than that of native bacteriocin (2600 AU/mg). Although the actual Mr (molecular weight) of the expressed bacteriocin should be 5666, the peptide showed the same mobility (Mr 3800) in sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE) as native cyclic gassericin A, suggesting that the expressed peptide retains compact folding of the molecule similar to that of native gassericin A.