One emerging field in molecular virology is the identification and expression of viral proteins that are not required for the synthesis of new virions, but allow a virus to inhibit or mediate elements of the host immune system [45–49]. Some of these mediatory proteins are very host-specific, while others are not. In most eases studied to date, the viral proteins target unique regulatory or effect or components of the immune system. These host immune elements range from the cytokine network, to signal transduction pathways, to complement cascade, to cytotoxic lymphocyte killing mechanisms, to the innate apoptosis response to infections [5, 46, 50]. Of all the mammalian viruses, poxviruses have evolved to encode a larger number of these viral anti-immune molecules than any other virus group [51]. And because variola virus is the only highly pathogenic orthopoxvirus to have evolved in humans alone, its own unique array of anti-immune proteins is believed to be adapted to molecules of the human immune system. Accordingly, there are compelling reasons to predict that expressed variola proteins of this class will exhibit mediatory activities specifically against human immune pathways, only some of which are understood or possibly even as yet discovered.

Viral proteins are generally multifunctional. They form networks—complexes with other proteins, viral and cellular. The functions of these proteins are frequently determined by post-translational modifications (e.g., phosphorylation, nucleotidylation, poly(ADP-ribosylation). A single viral protein expressed in transfected cells may have totally different functions than if expressed in productively infected cells. For example, the localization of herpes simplex virus ICP22 by itself is different from that of the same protein in infected cells. Also, the unphosphorylated protein binds a different set of proteins than the processed

(phosphorylated) protein. Live variola virus would be required to verify the function of viral proteins coded by the variola genome.

Sequence information suggests that five kinds of variola proteins could be expressed for research purposes:

• Secreted proteins that function as paracrine ligand mimics, or affect humeral systems such as the cytokine network or the complement cascade.
• Cell surface proteins that mimic cellular receptors or regulate immune recognition molecules, such as MHC-I or CD4.
• Intracellular proteins that regulate cellular signal transduction pathways, such as apoptosis.
• Virion component proteins that can alter immunogenic aspects of the virus particles.
• Proteins that can control or favor the ability of a virus to grow in selected tissues.

Some aspects of variola protein synthesis could be accomplished using only cloned genes, while others would require expression by the live virus. The following sections address first, the kinds of strategies that could be used for synthesis of variola virus proteins, and second, the potential usage of these expressed proteins for the study of human immunology and for the development of novel drugs to treat human immune-based diseases.

All poxviruses have evolved to replicate in the cytoplasm of infected cells and utilize genetic regulatory sequence elements that are virus-specific [52]. Poxvirus messenger RNAs are not spliced, but may contain sequences that preclude efficient expression in heterologous vector systems. In other words, the success of authentic expression of any poxvirus protein in the absence of live virus cannot always be ensured. Some variola proteins can readily be produced in standard expression systems (e.g., bacterial, yeast, baculovirus), but others may be refractory to such expression. In general, there are three major strategies for expressing poxvirus proteins, which differ in their requirements for availability of live variola virus:

• Expression of single cloned genes. This strategy utilizes standard expression vectors and requires only cloned variola DNA fragments well characterized in terms of sequence variation. This strategy works only for some poxvirus genes, and does not allow for virus-specific modification events such as post-translational processes under virus control. The strategy also does not fa-

• cilitate identification of activities that rely on interactions between more than one viral protein or complexes between distinct viral proteins.
• Co-expression of multiple proteins. Generally, this strategy requires composite expression vectors or the creation of chimeric viruses that contain larger variola DNA fragments encoding for multiple genes. Given the uniqueness of poxvirus gene regulation, this latter approach would most easily be accomplished with recombinant viruses derived from other poxviruses, such as vaccinia. However, the uncertainty concerning variola genes that determine pathogenesis would mandate strict regulatory restrictions on such recombinant viruses, and it is likely that they would require the same containment and restrictions as live variola virus itself. Moreover, the authenticity of such proteins in comparison with live variola virus would need to be demonstrated.
• Full genomic expression. Expression of the full spectrum of variola proteins, which includes all the events of synthesis, post-translational modification, and assembly into potentially active complexes and requires infection of cultured cells with live variola virus. This strategy would probably be the only way to reconstitute variola virus particles for authentic protein functional and structural studies aimed at the analysis of host cell receptors used by the virus or dissemination of the virus through the body.

The two most important potential uses for variola proteins are as research tools to investigate the human immune system and as a source of new drugs to treat human diseases caused by aberrant immune functions.

First, variola proteins are novel and unique probes of the human immune system. At present, it is estimated that between 10 and 50 percent of the genomic coding capacity of poxviruses is devoted to the expression of proteins that interact with host defense molecules. Many proteins of this class have already been discovered and partly characterized for other poxviruses, with host targets extending across a broad range of immune cell functions. These targets include cytokines (e.g., viral inhibitors of interferon, tumor necrosis factor, interleukin-l); growth factors (e.g., viral mimics of epidermal growth factor, vascular endothelial growth factor); complement (e.g., viral analogs of C4B-binding proteins); apoptosis (e.g., viral inhibitors of the caspase cascade); and various immune cell functions (e.g., viral inhibitors of antigen presentation, macrophage activation or cytotoxic lymphocyte killing mechanisms) [53].

This is an area of active research on many poxvirus systems, and one can only speculate as to how many such anti-immune proteins are expressed by variola virus. As noted earlier, given the long history of variola as an exclusively human-adapted virus, it can be predicted that a substantial number of these viral proteins are targeted to inhibit human immune molecules, some of which un-

doubtedly remain to be identified. Thus, the expressed variola proteins collectively represent an untapped resource of experimental probes with which it may be possible to identify and characterize new and potentially still-unknown human immune system components.

Second, purified variola proteins might be used as potential drugs to treat human diseases. A number of poxvirus proteins that function to inhibit immune pathways in the context of viral infection also inhibit the same immune molecules when purified and tested in the absence of virus. The discovery and analysis of the biological properties of poxvirus-encoded serpins have yielded a wealth of knowledge about how viruses can modulate inflammation [54]. For example, a secreted serine proteinase inhibitor from a rabbit-specific poxvirus and a similar but different orthopoxvirus homologue of the rabbit-specific poxvirus inhibit human proteinases in vitro [55, 56]. Rabbit poxvirus proteins can prevent inflammatory cell-dependent atherosclerosis in an animal model of vascular restenosis [57]. Similarly, a variety of viral apoptosis, or programmed cell death, inhibitors, such as crmA of cowpox, offer novel avenues for approaching the therapy of diseases associated with excessive cell death [58]. A homologue gene of serp2 found in the rabbit-specific myoxoma poxvirus inhibits some of the molecules involved in controlling apoptosis, but cannot substitute for crmA [59]. Such examples suggest that some viral proteins may be uniquely specific in function. Finally, selected viral proteins from variola could perhaps stimulate immune tolerance pathways, which could lead to improved methods for blocking transplant rejection or achieving increased effectiveness of gene therapy vectors [60].

Although the above uses remain hypothetical, the opportunity to investigate such avenues of research in the future will be dependent on the results of the World Health Assembly vote on the future of variola virus.