Gap junctions: the "kiss of death" and the "kiss of life"

Abstract

Cells expressing herpes simplex-thymidine kinase (HSV-tk) can be killed "in vitro" within 5 days of treatment with 20 microM ganciclovir (GCV) and transmit this toxicity to adjacent cells lacking HSV-tk; this phenomenon was termed "bystander effect" or "kiss of death". On testing a large number of cell lines in vitro, a wide range of sensitivity to GCV-mediated bystander killing has been reported. Although intercellular transfer of GCV metabolites through gap junction channels seems to be a likely mechanism for the "kiss of death", some studies suggest that other pathways may contribute to induced apoptosis of neighboring cells. To further investigate the mechanism underlying cell death mediated by HSV-tk and to evaluate the efficacy of gap junction channels formed by different connexins in this process, we have stably transfected a virtually uncoupled mouse neuroblastoma cell line (N2A cells) with different connexin-types expressed by neural cells (Cx32, Cx37, Cx40, Cx43) and co-cultured these cells with N2A cells stably transfected with Cx37 and HSV-tk. Here, we confirm our previous studies and those of others that the extent of cell death and sensitivity to GCV depend on the degree of connexin expression in transfectants. Further, we show that the bystander effect also depends on which connexin is expressed; reported disparities regarding the extent of GCV-mediated cellular apoptosis are likely due both to the degree of functional coupling and the type of connexin expressed. These results support the notion that gap junction hemichannels formed of certain connexins are more likely than others to pair functionally with Cx37, and suggest co-transfection strategies that might prove effective in sensitizing tumor cell populations to GCV. In addition, potential applications are discussed for use of the "good Samaritan effect", a mechanism by which bystander cells have been suggested to prevent cytotoxicity.