Human rhinovirus is one of the major causes of the common cold. There are over 100 serotypes of this virus, making it unlikely that there will ever be a vaccine for the common cold using conventional vaccine methods.

About each icosahedral 5-fold axis lies a canyon-like structure. From the atomic structures, the sites of natural escape mutations were found to lie at the top of these canyons. From this result, the dimensions of the canyon, and the conservation of some of the residues lying at the canyon floor, it was thought that the receptor might bind to the bottom of the canyon thus hiding important residues from the immune system. In this way, the virus might be allowed to change its surface residues, thwart antibody binding, but retain crucial residues necessary for viral infectivity.

	In the schematic, VP1 is blue, VP2 is gree, and VP3 is red. The canyon is shown by the black circle. The four antigenic sites (NIm-IA, NIm-IB, NIm-II, NIm-III) are also labeled. The 'C' on the surface diagram shows the position of the receptor (ICAM-1) binding site.
While the mechanism of antibody-mediated neutralization has been extensively studied, it remained highly controversial. In the case of HRV14, Roland Rueckert's group demonstrated that weakly neutralizing antibodies aggregated the virions whereas strongly neutralizing antibodies did not. In the case of the former, aggregation will 'clump' particles together thereby decreasing the number of independent infectious particles. They also found that, in this case, approximately 60 antibodies could bind to the 60 available antigenic sites. In contrast, only about 30 strongly neutralizing antibodies can bind simultaneously to the viral surface. Since these antibodies do not aggregate the virions over a wide range of concentrations, it was believed that these antibodies bound bivalently to the surface of the virion. Such binding might lead to more efficacious neutralization via cross-linking the particles together or by inducing conformational changes in the virion.
	Shown here are all of the image reconstructions performed to date on human rhinovirus/antibody complexes.  Fab17 is the Fab fragment from mAb17 and can neutralize as either an intact antibody or as an Fab fragment.  Fab12 is the Fab fragment from a strongly neutralizing antibody, mAb12.  This antibody, it turns out, is probably derived from the same mother cell as mAb17 since the two only differ by a few amino acids.  Both mAb12 and mAb17 do not aggregate the virions very well and appear to bind bivalently to the virion surface.  On the other hand, mAb1 is a weakly neutralizing antibody that aggregates the virion very strongly.  From this reconstruction, it is clear that this antibody is twisted in such a way as to not facilitate bivalent binding (binding with both arms to the surface).
	The main problem with the cryo-EM results is the resolution of the structures. While atomic structures were fit into the cryo-EM envelopes, this was insufficient to unequivically state that antibodies did not cause relative small conformational changes in the virion upon neutralization. Therefore, the Fab17/HRV14 complex was crystallized and its structure was determined to 4 angstrom resolution. From this structure the cyro-EM results (mauve) were shown to match fairly well to the X-ray results (heavy chain yellow and light chain white). Due to flexibility at the elbow, the constant domains were not observable in this structure. Clearly the Fab penetrates the canyon. In this view, the RNA interior is towards the bottom of the figure. From these results, it was clear that the virus structure did not significantly change upon antibody binding.
	Shown here are stereo views of the antigen binding region of the unbound Fab (top) and the structure of the antigen binding region of the Fab when it is bound the virus surface (bottom). This is in contrast to the virus structure that is not changed upon antibody binding. Shown below is a composite figure of all of the above results. We have now shown that none of the antibodies tested induced large conformational changes in the virus upon antibody binding. We propose that such conformational changes are not required for neutralization. Instead, the data strongly suggests that all antibodies abrogate cell attachment and we believe that this is due to steric interference. Secondly, we have clearly shown that antibodies penetrate deep into the receptor binding region and therefore the canyon hypothesis is unlikely to be correct. This is consistent with all of the other antibody/virus structures determined to date. Indeed, viruses do not avoid our immune system in the same way that parasites do - they simply move onto a naive host.

[Return to top]