Heat shock enhances outer-membrane vesicle release in Bordetella spp

Abstract

Pertussis, also known as whooping cough, is caused by the Gram-negative bacterium Bordetella pertussis, an obligate human pathogen. Despite high vaccination rates in high-income countries, resurgence of pertussis cases is an occurring problem that urges the necessity of developing an improved vaccine. Likewise, the efficacy of vaccines for Bordetella bronchiseptica, which causes similar disease in pigs and companion animals, is debatable. A promising approach for novel vaccines is the use of outer membrane vesicles (OMVs). However, spontaneous OMV (sOMV) release by Bordetella spp. is too low for cost-effective vaccine production. Therefore, we investigated the influence of growth in various media commonly used for culturing Bordetella in the Bvg⁺, i.e. virulent, phase and of a heat shock applied to inactivate the cells on OMV production. Inactivation of the bacterial cells at 56 °C before OMV isolation greatly enhanced OMV release in both Bordetella spp. without causing significant cell lysis. The growth medium used barely affected the efficiency of OMV release but did affect the protein pattern of the OMVs. Differences were found to be related, at least in part, to different availability of the nutrient metals iron and zinc in the media and include expression of potentially relevant vaccine antigens, such as the receptors FauA and ZnuD. The protein content of OMVs released by heat shock was comparable to that of sOMVs as determined by SDS-PAGE and Western blot analysis, and their heat-modifiable electrophoretic mobility suggests that also protein conformation is unaffected. However, significant differences were noticed between the protein content of OMVs and that of a purified outer membrane fraction, with two major outer membrane proteins, porin OmpP and the peptidoglycan-associated RmpM, being underrepresented in the OMVs. Altogether, these results indicate that the application of a heat shock is potentially an important step in the development of cost-effective, OMV-based vaccines for both Bordetella spp.

Graphical abstract

Fig. 1

Influence of medium and heat shock on OMV release. B. pertussis (Bp) and B. bronchiseptica (Bb) were grown in Verwey, SS or THIJS medium and either killed by heat shock (HS) or not (no HS) at 56 °C for 1 h before centrifugation. OMVs were isolated from the supernatants of equal amounts of cells, based on OD₆₀₀, from B. pertussis (A) or B. bronchiseptica (B) and analyzed by SDS-PAGE. OMV production was scaled-up by growing larger cultures of B. pertussis (C) or B. bronchiseptica (D) in baffled flasks with an air:liquid ratio of 5:1 for two days. In these experiments, the Verwey medium was not supplemented with starch to allow for quantification of the results. For growth of B. pertussis, the media were supplemented with heptakis. For SDS-PAGE analysis, hOMVs were four- (B. bronchiseptica) or five-fold (B. pertussis) diluted relative to sOMVs as indicated below the lanes. In panels A-D, LPS (lipid A plus core moiety), which is also stained with the Bradford reagent, is indicated with an arrow at the right and molecular weight markers are shown at the left. Protein content in B. pertussis (E) and B. bronchiseptica (G) OMVs was quantified using a Lowry assay. Values are depicted as the amount of protein per liter of bacterial culture per OD₆₀₀ unit. LPS content in B. pertussis (F) and B. bronchiseptica (H) OMVs was quantified using the purpald assay. Values are depicted as the amount of LPS per liter of bacterial culture per OD₆₀₀ unit. (E-H) Bars represent mean values with standard deviations of three biological replicates. Significant differences were determined with multiple t tests using GraphPad Prism 6 and are indicated by asterisks (*, p ≤ 0.05; **, p ≤ 0.01; ***, p ≤ 0.001; ****, p ≤ 0.0001).

Fig. 2

Comparison of protein content of OMV preparations. (A) SDS-PAGE analysis of OMVs from B. pertussis (Bp, left panel) and B. bronchiseptica (Bb, right panel) grown in Verwey medium isolated after heat shock (+ HS) or without heat shock (- HS). sOMVs were ∼10-fold (left panel) or ∼30-fold (right panel) concentrated compared to hOMVs. Proteins were stained with the Bradford reagent (left panel) or with the more sensitive silver stain (right panel). LPS (lipid A plus core moiety), which is also stained with both reagents, is indicated with an arrow at the right and molecular weight markers are shown at the left. (B) Western blot analysis of B. pertussis whole-cell (TC) lysates, sOMVs and hOMVs. sOMVs were 10-fold concentrated relative to hOMVs. Membranes were incubated with antibodies directed against pertactin (Prn) or BrkA. The two bands detected with the anti-BrkA antiserum presumably correspond to the full-length protein and the processed passenger domain of this autotransporter.

Fig. 3

Morphology of B. pertussis cells and OMVs visualized by transmission electron microscopy. Bacterial cells were grown in Verwey medium and either exposed to heat shock for 1 h at 56 °C (HS) or not (no HS). The dashed box in the middle panel is ∼4-fold magnified and depicted in the right panel. Cell-associated OMVs (filled arrowheads) and released OMVs (open arrowheads) are indicated.

Fig. 4

Heat modifiability of OMV proteins analyzed by semi-native SDS-PAGE. sOMVs (- HS) and hOMVs (+ HS) from bacteria grown in Verwey medium were either boiled or not in sample buffer before SDS-PAGE as indicated above the lanes. sOMVs of B. pertussis (Bp) and B. bronchiseptica (Bb) were ∼10-fold and ∼30-fold, respectively, concentrated compared to the hOMVs. (A) Proteins were stained with the Bradford reagent (left panel) or with the more sensitive silver stain (right panel). LPS (lipid A plus core moiety), which is also stained with both reagents, is indicated with an arrow at the right and molecular weight markers are shown at the left. (B) Western blot analysis of OMVs. Membranes were incubated with antisera directed against porin OmpP, which can be detected as a monomer (*) or as an oligomer (**).

Fig. 5

Influence of incubation period at 56 °C on cell viability, lysis and OMV release. B. pertussis (Bp) and B. bronchiseptica (Bb) cultures grown for one day in Verwey medium were incubated for the indicated time periods at 56 °C. (A) Viability of bacteria after incubation for different periods at 56 °C determined by plating 10-fold serial dilutions of cultures on BG plates. (B) OD₆₀₀ of bacterial cultures expressed as percentage of the OD₆₀₀ at t = 0. Mean values with standard deviations of three biological replicates are shown. Significant differences were determined using one-way ANOVA followed by Dunnett's multiple comparisons test using GraphPad Prism 6 and are indicated by asterisks (**, p ≤ 0.01). (C) Whole-cell lysates of bacterial cultures were analyzed by SDS-PAGE and Western blotting with an antiserum directed against GroEL. (D) SDS-PAGE analysis of OMVs isolated from the medium of bacterial cultures. OMVs were isolated from equal amounts of cells (based on OD₆₀₀). LPS (lipid A plus core moiety), which is also stained with the Bradford reagent, is indicated with an arrow at the right and molecular weight markers are shown at the left.

Fig. 6

Preferential loading of OMPs into OMVs. (A) Zoomed-in ∼85-kDa range of the last three lanes of Fig. 1C (upper panel) and of Fig. 1D (lower panel). Bp = B. pertussis, Bb = B. bronchiseptica. (B) Western blot analysis of whole-cell lysates of B. pertussis grown in Verwey, SS or THIJS medium. Membranes were incubated with antisera directed against FauA or ZnuD as indicated. (C) Western blot analysis of whole-cell lysates of B. bronchiseptica grown in Verwey medium supplemented or not with 100 µM FeSO₄•7H₂O (left panel) or in SS medium supplemented or not with 1 µM ZnCl₂ (right panel). Membranes were incubated with antisera directed against FauA or ZnuD as indicated. (D) Western blot analysis of whole-cell (TC) lysates and hOMVs of B. pertussis grown in Verwey, SS or THIJS medium. Membranes were incubated with antisera directed against porin OmpP, FauA or ZnuD. The ratio of the signals for FauA and ZnuD was determined relative to that of porin OmpP and is indicated. (E) A purified OM fraction (fraction 8 from Fig. S2A) and hOMVs of B. pertussis were analyzed by SDS-PAGE, and proteins were stained with the Bradford reagent (left panel) or blotted (right panel). Blots were incubated with antisera directed against OmpP or RmpM as indicated. Lysozyme, which is used during spheroplasting, is found in the OM fraction because it associates with the OM by electrostatic interactions. In all panels, the positions of molecular weight markers are shown at the left.