Successful, safe gene therapy is something of a Holy Grail for researchers. Delivering an exogenous gene into the targeted cells, and ensuring that it will be expressed there, are fraught with difficulties. Viruses are a promising vector for the delivery of a therapeutic gene into a patient's cells, but bring with them a new set of problems; for example, how to avoid an inflammatory or immune reaction to the virus. An alternative, nonviral delivery method is beginning to show potential.

Shi et al. have previously shown that 'pegylated' liposomes — coated with strands of polyethylene glycol — can be used as nanocontainers to carry DNA from the bloodstream into various tissues, including the brain, in rats. In that experiment, the liposomes were conjugated with a monoclonal antibody against the transferrin receptor, and the exogenous gene was driven by the simian virus 40 (SV40) promoter. The DNA was protected inside the liposome until the antibodies bound to the transferrin receptors on the blood–brain barrier and were transported across into the brain. The DNA was then taken up into brain cells by transferrin-mediated endocytosis. Because the transferrin receptor is found in other tissues as well as brain, and the SV40 promoter is not tissue-specific, the gene was also expressed in other organs that express the transferrin receptor.

In their new experiments, Shi et al. have taken the idea an important step further. Working in mice instead of rats, they used an antibody against the mouse transferrin receptor, so the liposomes were still targeted to a number of tissues as well as the brain. But this time, the exogenous gene was driven by the brain-specific glial fibrillary acidic protein (GFAP) promoter. The transgene — β-galactosidase — was expressed only in the brains of the mice, specifically in the astrocytes.

This work is an important proof of the principle that it might be possible to use a nonviral targeting technique, coupled with tissue-specific gene promoters, to achieve the expression of therapeutic genes specifically in the cells that are affected by a genetic disorder. Of course, these are early days; for example, the expression of the exogenous genes in these experiments lasted for only a few days. More work will be needed to find out whether that period can be extended. But it might be possible to use this gene targeting system in conjunction with transgenic mouse models of human diseases to work towards the ultimate goal of this type of study: successful, safe gene therapy in humans.