anti-immunoglobulins or rheumatoid factor, that might block complement-fixing sites on the antiviral antibody. The appearance of viral antigens might in turn be related to other factors, such as the phase of the cell cycle or coinfection with a second virus. If a particular virus produces few or no new anti- genic sites on the surface of cells or if these antigenic sites are far apart, complement-mediated cell de- struction may not occur. If, however, the density of virus-specific antigens on the cell surface should rise during the course of an infection, this would increase the likelihood of complement-mediated cell destruction. Fluctuations in the density of viral antigens on the infected cell surface might contribute heavily to the pathogenesis of lesions associated with " slow virus " infections (Porter, 1971).
The attachment of antiviral antibody in sublytic concentrations to the surface of infected cells may conceivably be instrumental in deciding the fate of the cell. On the one hand, the attachment of anti- viral antibody might accelerate phagocytosis of the infected cell by activated macrophages. On the other hand, antiviral antibody might prevent sensitized lymphocytes from recognizing or reacting with the viral antigens and thereby inhibit the cell-mediated immune response. In virus infections, this might prove to be the counterpart of " blocking " or " enhancing " antibody.
Under certain circumstances the destruction of virus-infected cells by antiviral antibody and comple- ment may be more beneficial than harmful to the host. Antibody-mediated cell destruction may be one of the mechanisms by which the host combats those viruses that tend to elude neutralization by spread- ing directly from cell to cell. Moreover, the destruc- tion of cells that are actively producing virus would slow down viral replication and release or expose the infectious virus within the cell to neutralizing antibody. Thus, in virus infections, antibody- mediated cell destruction may fulfil many of the functions that have been postulated for cell-mediated immunity and may serve as a complementary or supplementary defence mechanism. In addition, in vitro experiments suggest that the release of one or more chemotactic-generating factors (Brier et al., 1970) from infected cells and/or the interaction between antiviral antibody and viral antigens can activate the complement sequence and cause the release of mediators able to attract polymorpho- nuclear and mononuclear leucocytes.
(1) Although many investigators have speculated that immunologic injury may contribute to the pathological picture in certain virus infections, it has been difficult to isolate and evaluate this pheno- menon in vivo. The release of 51Cr from virus- infected cells by antiviral antibody and complement provides a simple, objective, and quantitative tech- nique for studying immunologic injury in vitro. With this technique it should be possible to (a) investigate a variety of viruses; (b) evaluate virus- induced immunologic injury in different types of cell; (c) compare the roles of cytolytic and non- cytolytic antibody in the serum of patients during the course of various virus infections; (d) determine whether biological mediators are released or acti- vated as a result of the interaction of antiviral antibody and complement with viral antigens; and (e) investigate the relationship between antibody- mediated and cell-mediated destruction of infected cells.
(2) In vivo studies should be extended and experi- mental models developed. Additional studies should be conducted to evaluate the results of the passive administration of cytolytic antiviral antibody to infected animals with normal and depleted levels of complement. Efforts should be made to demonstrate the presence of antiviral antibody and complement on the surface of injured cells at the site of the lesion. Further thought should be given to the potential beneficial or harmful effects of passive protection with immunoglobulins containing cytolytic antiviral antibody or with vaccines (e.g., rabies vaccine) that might induce cytolytic antibody.
(3) Attempts should be made to compare the in vitro and in vivo effects of antibody and comple- ment on the lysis of virus-infected cells. Whether the attachment of nonlytic antibody to infected cells can inhibit the cell-mediated immune response should be investigated.
C. ALLISON, Clinical Research Centre, Northwick Park Hospital, Harrow, Middlesex, England
I. B. BEVERIDGE, Veterinary Public Health Consultant, World Health Organization, Geneva, Switzerland
C. COCKBURN, Chief Medical Officer, Virus Diseases, World Health Organization, Geneva, Switzerland
JUNE EAST, Department of Environmental Carcino- genesis, Imperial Cancer Research Fund, Mill Hill, London, England