Bystander killing of cancer cells by herpes simplex virus thymidine kinase gene is mediated by connexins.

TLDR

In cancer gene therapy only a fraction of tumor cells are transfected, yet HSV‑tk therapy can kill both transfected and neighboring cells through a bystander effect that likely involves diffusion via gap‑junctional intercellular communication. The study aims to demonstrate that gap junctions mediate the HSV‑tk bystander effect. The authors used HeLa cells, which normally lack gap‑junction communication, transfecting them with HSV‑tk alone or with connexin 43 to restore GJIC, then co‑culturing with non‑transfected cells and treating with ganciclovir. When Cx43 was expressed, both HSV‑tk‑positive and –negative HeLa cells were killed by ganciclovir, whereas without Cx43 only the transfected cells died, confirming that Cx‑mediated GJIC underlies the bystander killing.

Abstract

In gene therapy to treat cancer, typically only a fraction of the tumor cells can be successfully transfected with a gene. However, in the case of brain tumor therapy with the thymidine kinase gene from herpes simplex virus (HSV-tk), not only the cells transfected with the gene but also neighboring others can be killed in the presence of ganciclovir. Such a "bystander" effect is reminiscent of our previous observation that the effect of certain therapeutic agents may be enhanced by their diffusion through gap junctional intercellular communication (GJIC). Herein, we present the evidence, from in vitro studies, that gap junctions could indeed be responsible for such a gene therapy bystander effect. We used HeLa cells for this purpose, since they show very little, if any, ability to communicate through gap junctions. When HeLa cells were transfected with HSV-tk gene and cocultured with nontransfected cells, only HSV-tk-transfected HeLa cells (tk+) were killed by ganciclovir. However, when HeLa cells transfected with a gene encoding for the gap junction protein, connexin 43 (Cx43), were used, not only tk+ cells, but also tk- cells were killed, presumably due to the transfer, via Cx43-mediated GJIC, of toxic ganciclovir molecules phosphorylated by HSV-tk to the tk- cells. Such bystander effect was not observed when tk+ and tk- cells were cocultured without direct cell-cell contact between those two types of cells. Thus, our results give strong evidence that the bystander effect seen in HSV-tk gene therapy may be due to Cx-mediated GJIC.