Making the Cut

There are no surefire options for making large, targeted deletions in mammalian cells. In murine embryonic stem (ES) cells, it's possible to insert two Flp recombinase target (FRT) or loxP sites in the genome and then introduce Flp or Cre recombinase to remove the intervening sequences. Homologous recombination with BAC-borne replacement sequences has also been successful in these cells. For human cells, these techniques have not worked. Thus, a report by Lee et al. in Genome Research describing targeted genomic deletions in 293T cells—a human embryonic kidney line—should generate considerable interest. The authors turned to zinc finger nucleases (ZFNs), which consist of zinc finger DNA binding domains fused to an endonuclease to make large targeted deletions in mammalian cells. Methods are emerging to target ZFNs to endogenous sequences, and to date the technology has been used to introduce site-specific mutations in human genes. In their article, Lee et al. show that co-transfection of pairs of ZFNs can delete genomic sequences between the ZFN cut sites. Sizes of deletions ranged from less than 1 Kb to 14.9 Mb. Sequence analysis suggested the repair mechanism is nonhomologous end-joining, leaving minor sequence variations at the junction. Although the best deletion frequency was only 10%, this was sufficient to prepare clonal populations of mutated cells. Because the method sidesteps construction of gene targeting vectors and does not leave foreign DNA sequences in the manipulated genome, it may gain preeminence even in organisms for which genome engineering methods already exist. A critical factor will be continued improvements in ZFN engineering, but as those strategies become more robust, knocking out genes and homologous gene clusters may become routine.

Lee et al. Targeted chromosomal deletions in human cells using zinc finger nucleases. Genome Res. [Epub ahead of print, December 1, 2009, doi:10.1101/gr.099747.109].

Letting Go

Although tagging proteins can improve solubility and stability, tag removal may be necessary before functional and structural studies. The traditional solution involves adding a protease recognition site, but this approach can be time-consuming, lead to off-target cleavages, and require additional purification steps. Self-cleaving tags have been described, but existing systems do little to improve solubility, limiting their applicability. Now, in PLoS ONE, Shen et al. describe a self-cleaving tag that enhances recovery of intact, soluble protein. The tag is a cysteine protease domain (CPD) of the cholera toxin protein that has highly specific, inducible activity. After a target protein–CPD-His fusion is captured by Ni²⁺ affinity chromatography, a small molecule inducer of CPD activity is added, triggering autocleavage at a site upstream of CPD. Depending on the expression construct, the target protein is released with between one and four extra residues at its C terminus. Crucially, off-target protease activity is undetectable, even when the target protein contains many potential cleavage sites. Equally important, the CPD tag enhanced the expression levels of all targets tested, including proteins from organisms that express poorly in Escherichia coli. Moreover, the new tag improved recovery of two test proteins chosen for their poor stability and solubility, respectively. To produce target protein–CPD fusions, users may choose one of the seven pET-based expression vectors newly created by Shen et al., and can perform the subsequent cleavage at a variety of temperatures (4–37°C) and buffer conditions (including in the presence of protease inhibitor cocktails and in the absence of salt).

Shen et al. 2009. Simplified, enhanced protein purification using an inducible, autoprocessing enzyme tag. PLoS ONE. 4(12):e8119.

Dead or Alive

Flow cytometry is a powerful means to isolate small populations of cells. However, if the markers used for identifying the cells are intra-rather than extracellular, membrane permeabilization and fixation must be done prior to staining, which precludes many RNA analysis methods. Because assessing gene expression in fixed and sorted cell subsets would provide insights into the quantitative differences between distinct cell lineages within a tissue, a team interested in islet cell subsets investigated whether nuclease protection assays (NPAs) could deliver this information. Writing in Nature Biotechnology, Pechhold et al. describe how they adapted high-throughput, small-scale quantitative NPAs to this application. Fixation should not prevent short cDNA probes from binding to paraformaldehyde-treated RNA, so 50-mer oligonucleotides were incubated with lysates of sorted cells. After S1 nuclease digestion to remove single-stranded nucleic acids, they hybridized the samples with an array of capture probes for genes of interest, after which sequence-specific linkers and horseradish peroxidase (HRP)-conjugated detection probes were annealed. Following supplementation with a substrate that produces luminescence in the presence of HRP, luminescence signal was measured as a proxy of gene expression level. Pilot tests determined that gene expression profiles were indistinguishable between mRNA prepared from fresh or fixed cells. Likewise, permeabilization, staining, and cell sorting led to minimal, unbiased losses. Harnessing the technique to assess gene expression levels in islet cell subtypes from mice at different stages of development, the authors used lineage-specific expression profiles to obtain new biological insights. The method, which requires just a few thousand sorted cells and can compare 15 genes at once, should help characterize subpopulations of cells in tumors or tissues for which intracytoplasmic or nuclear markers are known.