Validation of an FFF-MALS Method to Characterize the Production and Functionalization of Outer-Membrane Vesicles for Conjugate Vaccines

Abstract

With the ongoing development of conjugate vaccines battling infectious diseases, there is a need for novel carriers. Although tetanus toxoid and CRM197 belong to the traditional carrier proteins, outer-membrane vesicles (OMVs) are an excellent alternative: in addition to their size, OMVs have self-adjuvanting properties due to the presence of genetically detoxified lipopolysaccharide (LPS) and are therefore ideal as a vaccine component or antigen carrier. An essential aspect of their development for vaccine products is characterization of OMVs with respect to size and purity. We report on the development of a field-flow fractionation multiangle light-scattering (FFF-MALS) method for such characterization. Here, we introduced NIST-traceable particle-size standards and BSA as a model protein to verify the precision of the size and purity analysis of the OMVs. We executed a validation program according to the principles provided in the ICH Guidelines Q2 (R1) to assess the quality attributes of the results obtained by FFF-MALS analysis. All validation characteristics showed excellent results with coefficients of variation between 0.4 and 7.32%. Estimation of limits of detection for hydrodynamic radius and particle concentration revealed that as little as 1 μg OMV still yielded accurate results. With the validated method, we further characterized a full downstream purification process of our proprietary OMV. This was followed by the evaluation of other purified OMVs from different bacterial origin. Finally, functionalizing OMVs with N-γ-(maleimidobutyryl)oxysuccinimide-ester (GMBS), generating ready-to-conjugate OMVs, did not affect the structural integrity of the OMVs and as such, they could be evaluated with the validated FFF-MALS method.

Figure 1

Elution profiles for the separation of BSA, OMV, and particle-size standards. The left-side y axes pertain to the relative signal from the 90° light-scattering detector. (A) Method C for the elution of BSA (blue) and OMV (red) and (B) Method D for the elution of particle-size standards (green) with radii of (1) 11.5 nm, (2) 25.5 nm, (3) 50.0 nm, and (4) 101.5 nm. The crossflow for both figures is plotted as the black dotted line, of which the first 6 min represent the steps to flush the channel and inject the sample (Flu/Inj).

Figure 2

FFF-MALS results (n = 3): fractions 4 and 8 collected from DSP steps as described in Table 1 (for fractions 1–3 and 5–7), see Figure S12. Crossflow for both figures is shown as the black dotted line, of which the first 6 min represent the steps to flush the channel and inject the sample (Flu/Inj).

Figure 3

Different OMVs analyzed with FFF-MALS method C (OMV-BSA, n = 3). Blue: B. pertussis—R_h(Q)_z = 55.0 ± 0.4 nm; red: E. coli—R_h(Q)_z = 58.2 ± 0.7 nm.

Figure 4

FFF-MALS analysis: OMV, OMV buffer-exchanged, and OMV-GMBS-modified; no differences in size distributions were found for OMV/GMBS ratios of 3:1, 2:1, or 1.3:1.