Horizontal (or "lateral") gene transfer is an effective mechanism for the exchange of genetic information in bacteria. Classical examples include genes for antibiotic resistance, virulence and enzymes involved in various metabolic pathways. These genes are often carried by plasmids or transposable elements that quite readily shuttle between bacteria. Horizontal gene transfer therefore represents a powerful avenue for bacterial diversification and adaptation to new environments.

Lateral gene transfer also occurs in mammalian cells. Although the frequency of this event is probably not very high, recent work from Lars Holmgren's laboratory 1 demonstrates the possible alarming consequences of horizontal gene transfer for human health. It has been shown previously that DNA can move laterally from apoptotic bodies to normal cells. Bergsmedh et al. tested a hypothesis that the uptake of DNA from apoptotic cells that carry oncogenes might cause transformation of a healthy cell. Two types of cells, easily distinguishable from one another, were used to test this hypothesis. The donor cells were rat embryonic fibroblasts, either wild type (REF) or expressing the activated human oncogenes c-myc, H-ras^V12 and a green-fluorescent protein (GFP) fused with a hygromycin resistance gene (REFrm). The recipient cells were mouse embryonic fibroblasts from wild-type (MEF) or p53-deleted mice (MEF p53^-/-). The authors first established that the addition of apoptotic (but not necrotic) REFs to healthy MEFs resulted in the rapid internalization of apoptotic bodies by MEFs. Interestingly, the addition of apoptotic REFrm also induced foci formation in MEF cells, although they eventually became senescent or died. When the MEF p53^-/- cells were used as recipients, however, numerous foci were detected. Furthermore, the DNA of H-ras^V12 and c-myc could be detected in the MEF p53^-/- cells, confirming that indeed the DNA of these oncogenes was transferred horizontally from rat to mouse cells, although it gradually disappeared. The frequency of the transfer and long-term integration of DNA could be achieved only by selecting for hygromycin-resistant recipient MEFs.

The authors next tested their hypothesis in vivo by injecting REF, or REFrm-induced MEF or MEF p53^-/- foci into SCID mice. Within three weeks, the mice developed tumors, but only from the foci where the donor cells carried oncogenes (REFrm). Strikingly, cells isolated from these tumors could now be propagated for over three months in vitro and could be injected into mice again, forming tumors. Furthermore, fluorescent in-situ hybridization (FISH) analysis of cells derived from tumours confirmed the presence of rat as well as hybrid mouse–rat chromosomes and human c-myc and H-ras^V12 DNA in mouse cells.

Several important conclusions can be made from this work. First, the authors demonstrated that entire chromosomes could be transferred from apoptotic bodies and integrated into the "host" genome. This further confirms earlier evidence that the entire genome is not fragmented during apoptosis. Second, the integration of laterally transfered DNA is not stable unless it would confer selective advantage to the recipient cell. Significantly, however, the presence of oncogenes seemed to offer this selective advantage in vivo. Lastly, the authors demonstrated the oncogenic potential of apoptotic bodies carrying oncogenes.

The role of apoptosis in a multicellular organism is to remove unwanted, damaged and potentially harmful cells. So why don't we acquire bad mutations and genes from all those dying cells in our bodies? Fortunately for us, cells have evolved potent safeguard mechanisms, such as the protection of the genome by p53, that prevents normal cells from undergoing unwanted transformation. The acquisition of mutations in genes involved in the control of genome integrity would therefore make cells more vulnerable to the effects of foreign DNA. Compromised genome integrity is frequently associated with cancer cells. Horizontal gene transfer could therefore be one mechanism that contributes significantly to the further accumulation of genes necessary for malignant transformation and tumorigenesis.