Targeted Killing of Herpesvirus-Infected Cells

National Institute Of Allergy And Infectious Diseases

Investigators

Abstract

The first stage of this program is to generate monoclonal antibodies against specific KSHV proteins. Or first efforts are directed against MCD, since this syndrome is characterized by infected B cells that are in the lytic phase of the viral replication cycle. Hence the infected cells are expected to express several KSHV glycoproteins at the surface, including gH/gL, gB, and K8,1A. MCD is a relatively rare disease, and the incidence has been reduced considerably by antiretroviral therapy amongst HIV-infected people. However the disease still appears even in the context of HAART, and also in HIV-negative individuals; the mortality rate is extremely high, with a short life expectancy after diagnosis. Anti-herpesvirus drugs (acyclovir) induce some partial remissions, but more effective therapeutics are needed. Hence the interest in IT development, which we have begun as part of a collaborative effort with Dr. Corey Casper and Larry Corey who oversee a cohort of MCD patients at the Fred Hutchinson Cancer Center, U. Washington. We have generated a panel of murine Mabs against gH, several of which react with the protein expressed at the cell surface. To date, the variable regions of two of these hybridomas have been cloned and sequenced. The corresponding SCFv cDNAs have been constructed and linked to PE38 to give two full-length immunotoxin constructs. These have been expressed on a small scale in E. coli using standard protocols. Large scale preparations are in progress for purification of the ITs and testing for specific cytotoxic activity, initially against a gH/gL stable transfectant cell line that we have generated, as well as against several chronically infected B cell lines that constitutively express KSHV glycoproteins. If these results prove positive (i.e. potent selective killing), we will examine blood samples obtained from patients in the U. Washington MCD cohort.[unreadable] KS poses a different set of challenges. Only a small percentage of cells in KS lesions are infected with KSHV, and these cells are in the early lytic, not the full lytic phase of replication. They express a virally encoded chemokine receptor (vGPCR) that is constitutively active. A variety of studies suggest that the vGPCR is essential for initiation and maintenance of KS lesions. For these reasons, the vGPCR rather than the KSHV glycoproteins would seem to be the most appropriate target for IT development. GPCRs have proven somewhat challenging for Mab generation, though many successes have been reported. We have initiated efforts to generate murine Mabs by immunizing with synthetic peptides from the predicted extracellular regions. ELISA assays have indicated potent antibody responses in many of the mice, and several colonies from spleen cell fusions display good reactivities with the corresponding peptides. Work is in progress to test whether the antibodies react with the vGPCR expressed at the cell surface, again using newly made stable cell lines expressing the transfected protein. If any positives are identified, the corresponding ITs will be produced. We are also pursuing an alternative vGPCR-targeted approach, namely the use of a chemokine ligand instead of an SCFv as the targeting moiety. Such ligand-toxins will not be useful therapeutically because they will certainly give undesired killing of normal host cells expressing corresponding human receptors for the chemokines. However these reagents will be useful in proof-of-concept animal model studies. In particular, we are collaborating with Dr. Silvio Gutkind (NIDCR) who has developed an intriguing KS model in which murine transfectant cell lines expressing the KSHV vGPCR induce tumors in immunodeficient mice. When a minority of these cells are co-injected with an excess of stable transfectant lines expressing other specific KSHV latent genes, tumors are formed with the same ratio of cells as in the injected input. Remarkably, selective elimination of the vGPCR-expressing cells (using an antibiotic that selectively kills them) results in entire tumor clearance, even though the vGPCR-expressing cells constitute only a minor fraction of the tumor. These findings are consistent with a paracrine mechanism wherein the vGPCR is critical for initiation and maintenance of KS tumors. An anti-vGPCR immunotoxin or ligand-toxin can be tested for the ability to induce tumor clearance. Positive results would provide added impetus for generation of true anti-vGPCR ITs from the corresponding Mabs, thus providing the needed specificity for killing of vGPCR-expressing cells in human KS lesions. We have made several ligand-toxins with different chemokines known to react with the vGPCR. These will be tested for killing of vGPCR transfectant cell lines; effective molecules will be tested in the KS mouse model.[unreadable] It is important to note an obvious point, namely that PE, derived from a bacterial protein, is highly immunogenic in humans. For this reason, PE-based ITs can probably be used over only a short period, with the hope that a durable therapeutic benefit can be achieved. This has proven to be the case for some PE-based ITs against human leukemias (developed in the lab of Dr. Ira Pastan, NCI; Dr. Pastan is providing advice and support for our project, though he is not an active collaborator). Because of the strong immunogenicity of the PE moiety of the IT, we believe that the use of murine antibodies is permissible despite their likely immunogenicity. We are also considering related strategies involving targeted killing agents that are likely to be minimally immunogenic (see Future Studies).