As discussed in last month's special focus on mouse genomic technology (see below), new techniques are constantly emerging for switching on the site-specific recombinase Cre in spatially and temporally controlled ways. Now, two papers report new approaches to delivering the recombinase to undifferentiated and terminally differentiated cells — a self-deleting Cre-lentiviral vector that overcomes the cytotoxic effects of prolonged cre expression and a cell-permeable form of Cre. Because its uptake depends on protein trafficking, this cell-permeable Cre will probably both facilitate future gene-function studies and provide cell biologists with a useful new tool.

Traditionally, there have been two main ways to deliver Cre to target cells — by using cre-expressing transgenic mice or cre-carrying viral vectors. Although such vectors have their advantages, they predominantly infect dividing, undifferentiated cells. So to overcome this limitation, Alexander Pfeifer and colleagues created a new Cre-lentiviral delivery vector (LV-Cre) because lentiviruses can infect both dividing and non-dividing cells. When the authors injected LV-Cre into Rosa26–Cre (R26R) reporter mice, which ubiquitously express a Cre-activatable form of lacZ, β-gal staining was evident in both undifferentiated and terminally differentiated cells. However, when injected into the brains of R26R mice, this vector caused brain abnormalities to develop after several weeks. On closer inspection of Cre's effects, Pfeifer et al. found that its prolonged presence causes cell-cycle arrest and apoptosis in cells in vitro and in vivo — possibly because of Cre-mediated illegitimate recombination at 'pseudo' loxP sites in the mouse genome. So to prevent this cytotoxic activity, the authors developed a self-deleting form of LV-Cre (LV-Cre-SD), by inserting a single loxP site into the vector's 3′ UTR. When the vector is reverse-transcribed and inserted into the genome, this 3′ UTR region is duplicated and introduced into the 5′ UTR, creating a vector that is flanked with loxP sites. LV-Cre-SD mediates recombination both in vitro and in vivo at the same efficiency rates as LV-Cre but without its cytotoxic effects, proving that this self-regulating Cre vector is likely to be a useful new tool for activating or inactivating gene expression in dividing and differentiated cells.

Jo et al. took a different approach to delivering Cre to cells by developing a cell-permeable form of the protein that carries a membrane-trafficking sequence. When this protein is repeatedly injected intravenously or intraperitoneally into R26R mice over 3–5 days, it causes widespread lacZ expression throughout organs such as the brain, kidney and liver, without adversely affecting the mice. However, there are a couple of disadvantages to using this modified Cre — its widespread dissemination limits its use in tissue-specific studies and it mediates recombination in fewer cells than does Cre that is delivered by transgene expression. Nevertheless, this cell-permeable Cre is likely to become a useful tool for future gene-function studies, given its ease of use and systemic effects. Furthermore, because it provides a stable record of protein trafficking and uptake, it could provide a new approach to developing protein-based therapies for treating human disease and equip cell biologists with a handy new tool.