Rotavirus double and triple layered viral particles: correlative characterization using electron microscopy, disc centrifuge and capillary electrophoresis

Abstract

Rotavirus (Reoviridae) viral particles consist of capsids with 3 concentric protein layers surrounding the virus genetic material. The outer layer is composed of the VP4 and VP7 proteins, which independently elicit neutralizing antibodies and induce protective immunity. Rotarix™, GSK live attenuated Rotavirus vaccine, was characterized by EM. The observations show that the vaccine contains both triple‐layered particles (TLPs) containing all 3 protein layers and double‐layered particles (DLPs) which lack the outer capsid proteins VP4 and VP7. As DLPs are non‐infectious, it may be of importance to address the question of their abundance vs . TLPs in bulk preparations. Although such information can be obtained by EM, the method is relatively impractical for this purpose, particularly for routine analysis, as high number of particles should be observed and sized to get accurate relative abundances. To determine the size and quantify the relative abundance of TLPs and DLPs on large populations of viral particles in bulk preparations, two size‐based, orthogonal analytical techniques were applied to detect these particles: disc centrifuge (DC) and capillary electrophoresis (CE). To confirm the capacity of both methods to distinguish clearly the two types of capsids, TLPs and DLPs were partially purified using CsCL gradient. They were characterized by EM negative staining (Fig. 1) and cryo‐TEM (Fig. 2). Their apparent diameter difference of about 10 nm was well correlated in both techniques. The profiles obtained by DC and CE (Fig. 3) show the well‐resolved identification. In addition, further confirmation was obtained by incubation of the preparations with EDTA, to remove the calcium required for VP7 protein stabilisation and thus to convert TLPs into DLPs [ 1 ]. The shift from TLPs to DLPs is clearly visible with CE (not shown), DC and also EM (Fig. 4). Dynamic Light Scattering (DLS) was also evaluated and works fine with the partially purified samples of either TLPs or DLPs. The shift from TLPs to DLPs by this method is clearly observable as well (not shown). However, DLS does not allow distinguishing clearly the two populations of capsids in real samples with mixed TLPs and DLPs. These orthogonal structural and sizing analysis methods can therefore be combined to assess and measure accurately DLPs/TLPs ratio. They thus provide a tool to understand the possible link of this ratio with the immunological potency of vaccine preparations.