Surface display of N-terminally anchored invasin by Lactobacillus plantarum activates NF-κB in monocytes

Abstract

The probiotic lactic acid bacterium Lactobacillus plantarum is a potential delivery vehicle for mucosal vaccines because of its generally regarded as safe (GRAS) status and ability to persist at the mucosal surfaces of the human intestine. However, the inherent immunogenicity of vaccine antigens is in many cases insufficient to elicit an efficient immune response, implying that additional adjuvants are needed to enhance the antigen immunogenicity. The goal of the present study was to increase the proinflammatory properties of L. plantarum by expressing a long (D1 to D5 [D1-D5]) and a short (D4-D5) version of the extracellular domain of invasin from the human pathogen Yersinia pseudotuberculosis. To display these proteins on the bacterial surface, four different N-terminal anchoring motifs from L. plantarum were used, comprising two different lipoprotein anchors, a transmembrane signal peptide anchor, and a LysM-type anchor. All these anchors mediated surface display of invasin, and several of the engineered strains were potent activators of NF-κB when interacting with monocytes in cell culture. The most distinct NF-κB responses were obtained with constructs in which the complete invasin extracellular domain was fused to a lipoanchor. The proinflammatory L. plantarum strains constructed here represent promising mucosal delivery vehicles for vaccine antigens.

Fig 1

Schematic overview of the expression cassette for N-terminal anchoring of invasin in L. plantarum. The vectors are based on previously described secretion vectors (19) in which the cassette is translationally fused to the inducible P_sppA promoter. All parts of the cassette are easily exchangeable using the introduced restriction sites: the NdeI site at the translational fusion point, the SalI site between the N-terminal anchor and invasin, and the downstream multiple cloning site (MCS, including EcoRI). Three principally different N-terminal anchoring motifs were used, all containing a signal peptide (SP) for secretion: two different lipoanchors were generated using lipobox fragments from Lp_1261 and Lp_1452, one transmembrane anchor was generated by fusing invasin to C-terminally truncated Lp_1568, which contains an SP but no signal peptide cleavage site, and one LysM anchor was generated by fusing invasin to Lp_3014.

Fig 2

Cell growth of invasin-expressing L. plantarum strains. OD₆₀₀ values were measured at the induction point (black bars) and 2 h after induction (gray bars). The data are means from triplicate experiments and presented as the means ± SD.

Fig 3

Surface localization of invasin in modified L. plantarum cells. (A and B) Representative flow cytometric (A) and microscopic (B) analyses of L. plantarum cells harboring plasmids designed for N-terminal surface anchoring of invasin. The cells were probed with rabbit anti-invasin polyclonal antibody and, subsequently, FITC-conjugated anti-rabbit IgG antibodies. The L. plantarum strains are denoted by different colors in the flow cytometry histograms and different numbers in the micrographs. L. plantarum harboring a vector without the inv gene was used as negative control (pEV, black, 8) and is depicted in all three flow cytometry histograms. L. plantarum strains harboring the following invasin-encoding plasmids are shown: pLp_1261Inv (light blue, 1), pLp_1261InvS (dark blue, 2), pLp_1452Inv (pink, 3), pLp_1452InvS (red, 4), pLp_1568InvS (brown, 5), pLp_3014Inv (green, 6), and pLp_3014InvS (dark green, 7). (C) Magnifications of representative cells depicted in panel B: pLp_1261Inv (1), pLp_1452InvS (4), pLp_1568InvS (5), and pLp_3014Inv (6).

Fig 4

NF-κB-directed luciferase expression in monocytes. Various L. plantarum cells were coincubated with U937 monocytes stably transfected with a plasmid (3x-κB-luc) encoding a NF-κB-inducible luciferase gene (9). The intensities of the resulting luciferase-generated chemiluminescence signals were measured, and relative values are shown as the ratio of results for individual samples and the negative control (U937). The data come from triplicate experiments and are presented as the means ± SD. Statistically significant differences (P < 0.01) compared to the results for L. plantarum/pEV are denoted by an asterisk (*). Signals labeled with two additional asterisks (**) are significantly stronger (P < 0.01) than the signals labeled with only one asterisk. LPS, bacterial lipopolysaccharides (positive control).