Viruses used for gene therapy have generally been genetically engineered to prevent viral replication and to deliver transgenes to achieve a therapeutic benefit. However, viruses engineered to remain replication-competent may be exploited for cancer gene therapy because replication within cancer cells results in oncolysis. We have examined a strategy that targets herpes simplex virus 1 (HSV-1) replication specifically to diffuse liver metastases after delivery into the portal vein. In initial proof-of-principle experiments, we used the HSV-1 mutant hrR3, which has an E. coli b-galactosidase gene inserted into the ICP6 locus. This inactivates the HSV-1 viral ribonucleotide reductase gene and renders it replication-defective. However, cellular ribonucleotide reductase can complement the absence of its viral counterpart to allow hrR3 productive replication in infected cells. We have demonstrated that cellular ribonucleotide reductase levels are extremely low in normal liver and hepatocytes and high in liver metastases. Accordingly, titers of hrR3 infectious virions recovered after infection of colon carcinoma cells are two to three log orders higher than those recovered after infection of hepatocytes. In contrast, wild-type HSV-1 replicates equally well in hepatocytes and colon carcinoma cells.

hrR3 selectively replicates in liver metastases and not in normal liver, unless ribonucleotide reductase levels are experimentally raised in the normal liver. This replication is cytolytic; hrR3 destroys colon carcinoma cell lines infected in vitro with multiplicities of infection as low as one viral unit per ten tumor cells. When mice bearing diffuse liver metastases are treated with a single intra-portal administration of hrR3, the anti-tumor efficacy is striking. Histochemical staining of liver sections for b-galactosidase expression reveals significant viral replication in liver metastases and no viral replication in normal liver tissues. hrR3-mediated tumor inhibition is equivalent in immunocompetent and nude mice, suggesting that the host immune response is not the primary mechanism of tumor destruction. Infection of diffuse liver metastases with a replication-incompetent HSV-1 mutants d120 (ICP4-defective) and d27 (ICP27-defective) does not produce any measurable anti-neoplastic effects. We have also demonstrated that administration of hrR3 produces significant anti-tumor activity even in mice that have been previously vaccinated against HSV-1. These results indicate that replication-competent HSV1 mutants hold significant promise as cancer therapeutic agents.

We have constructed several replication-conditional HSV-1 mutants for both gene therapy and oncolysis of liver tumors. Several of these mutants are based on the design of hrR3, such that they are defective in viral ribonucleotide reductase expression which results in replication preferentially in liver tumors rather than normal liver. Other HSV-1 mutants possess engineered alterations in the g134.5 gene, which also results in HSV-1 replication preferentially in tumor rather than normal cells. We are also characterizing the anti-neoplastic efficacy of several mutants that are engineered to express therapeutic transgenes, such as the yeast cytosine deaminase gene or the rat CYP2B1 gene.

Genetic Engineering, Characterization, and Animal Model Testing of Herpes Simplex Viral Vectors for Treatment of Liver Metastases
Principal Investigator: Kenneth K. Tanabe, MD
Consultant: David Knipe, PhD, Harvard Medical School
Group Members: James Donahue, MD; Hideki Kasuya, MD; Soundararajalu Chandrasekhar, PhD; Hiroshi Kawasaki, MD, PhD