The 2007 Nobel Prize for Physiology or Medicine went to scientists who created the gene-targeting tools that are now used routinely for precisely eliminating and modifying chromosomal genes in mouse embryonic stem (ES) cells. But ready means of making site-specific gene modifications in human ES cells remain elusive. One such method1 has proved difficult to use. Two new techniques just published should prove useful additions to the stem-cell handbook.

Researchers at Vita Salute San Raffaele University in Milan, Italy, and Sangamo BioSciences in Richmond, California, used synthetic zinc-finger nucleases that introduce double-strand breaks at specific DNA sequences. When cells repair the broken DNA, they can incorporate sequences from a similar DNA. By finding a more efficient way to introduce such a DNA template sequence and the nuclease into cells, the team was able to genetically modify human ES cells at specific sites at success rates of up to 5% (ref. 2).

While Sangamo researchers had previously shown that the zinc-finger nucleases were capable of gene editing, the techniques used were specific to cultured cell lines. They now show a more widely applicable delivery method. This uses the HIV-1-derived integrase-defective lentiviral vectors, which do not insert themselves into the genome, but which serve as vehicles for the nuclease and the DNA sequence for insertion. (Other viruses are incorporated into the human genome in ways that are hard to control; this poses a wealth of problems.)

Engineering the nucleases on a site-by-site basis is still labor intensive, however. Researchers at Invitrogen in Carlsbad, California, and stem-cell company Cellartis in Göteborg, Sweden, took a simpler approach for adding genes to the genome, albeit one that allows less control over the site of DNA integration. They used the integrase from the bacteriophage φC31 to incorporate DNA into human ES cells3.

Assimilation into the genome via φC31 integrase tends to occur at transcriptionally active 'hotspots' for integrase action. Using the gene for green fluorescent protein (GFP) as the marker to be inserted, the researchers found that integration was successful in about 20% of the cells. And when it did occur, about half the insertions were at fewer than 10 hotspots. They created stably modified human ES cell lines from the transfected cells and found that the modified ES cells could still differentiate into all three germ layers. This technique has the advantage that it escapes the size limitations of using a viral vector. However, it may be limited to modification of well-expressed genes.

While both these techniques will need to be monitored for unintended genetic modifications, they should allow researchers to introduce site-specific genetic changes and follow the expression of genes in real time in human ES cells more easily than before.