Until recently, in situ detection of nucleic acids has been limited to the proverbial "snapshot," generally achieved by the hybridization of probes to thin sections of fixed tissue. From these static data, it is difficult to determine dynamic information; additionally, interpretation can be confounded by issues of artifacts caused by the fixation procedure, which may distort the results leading to misinterpretation of DNA localization, abundance, or chromatin activity. Live cell imaging can alleviate this problem by providing a real-time view of cell function. However, it comes with its own inherent problems, such as being limited to the detection of naturally single-stranded RNA (while DNA detection would require high temperature denaturation, a situation not compatible with cell survival). In the review on p. 489, Dirks and Tanke lead the reader through the available in vivo RNA (and telomere) detection methods, including the use of FRET with oligonucleotide probes, peptide nucleic acids, locked nucleic acids, molecular beacons, and recognition of nucleic acids by fluorescently tagged proteins. They also summarize selected recent results, giving the methods some context. The take-home message appears to be that, although it is abundantly clear that significant advances have been made in recent years, it is equally obvious that there is still much to be done before in vivo detection technology comes of age.
Get a Grip
Assaying cell adhesion yields important information for investigating processes such as tumorigenesis and in design of materials for tissue engineering. At present, high-throughput assays are performed in multiwell plates. Despite the small scale, these experiments nonetheless require inconveniently large amounts of extracellular matrix (ECM) proteins or other adhesion molecules in order to coat the wells. Although some cell adhesion microarrays have been described, existing formats do not facilitate accurate comparisons of relative cell adhesion. That is where Kuschel et al. come in, and on p. 523 they describe an experimental strategy for comparative profiling of cellular adhesion to different ECM proteins. To make the arrays, ECM proteins (spiked with fluorescently labeled BSA) are spotted with a piezoelectric arrayer onto nitrocellulose-coated slides. Agitation conditions and cell density are carefully controlled to ensure that the number of cells that adhere to each spot (as counted by DAPI-stained nuclei) accurately indicate binding avidities. With this protocol, only 20,000 cells are required for an adhesion profiling experiment using 14 ECM proteins. As a result, this new cell adhesion microarray is equally adept at conserving cells and ECM proteins and should be widely applicable to any studies investigating cell-substrate interactions.
Mammalian Markup
Recombination techniques utilizing the Cre and Flp systems are now accepted as standard procedures in genome engineering, particularly in manipulating embryonic stem cells, replacing less flexible restriction enzyme approaches. Notwithstanding their popularity, the frequently lackluster efficiency of integration/recombination is still a confounding factor with these approaches. The use of phage-derived serine recombinases has been an attractive alternative to Cre/Flp tyrosine recombinases, but their frequent requirement for cofac-tors and poor efficacy has reduced their appeal. The Bxb1 integrase, however, has recently proven its worth as a robust serine recombinase that is active with fewer restrictions than other such integrases. In the article on p. 460 of this issue, Russell et al. describe for the first time the use of a codon-optimized Bxb1 integrase in mammalian cells, showing that it is able to catalyze unidirectional recombination between attP and attB sites with high efficiency. This knowledge gives researchers more tools and flexibility in their recombineering experiments and makes this alternative integrase available to those workers using all manner of mammalian cells, including stem cells.
Crossing OverTemplate switching during PCR results in pseudorecombination, or the generation of a hybrid product derived from the two parental templates. This phenomenon is critical to the use of sexual PCR (DNA shuffling) in in vitro evolution procedures. However, PCR-mediated recombination can also complicate experiments, as when a pseudorecom-binant triggers incorrect scoring in genotyping or population genetics studies. Thus, an estimate of the preponderance of pseudorecombination in a given system would be useful in order to accurately interpret the results. An RFLP-based method for measuring template switching exists in the literature, but this technique remains vulnerable to several confounding factors. Yu et al. (p. 499) revisit the problem and describe a fluorescence assay that accurately assays recombination frequency. Disconcertingly, they find that recombination frequency under run-of-the-mill PCR conditions routinely hits 20%–35%. This rate far exceeds the estimates provided by the original method. Interestingly, they find the lowest recombination frequency for the template corresponding to the most structured RNA; the most obvious correlate for recombination frequency is thus template length. Researchers who PCR amplify samples containing mixtures of genotypes should be aware of the general prevalence of pseudorecombination and this elegant strategy for calculating recombination frequency in experimental systems.
Metal Detector
Atomic absorption spectrometry (AAS) is a highly accurate way of measuring zinc amounts in serum or plasma. Unfortunately, the method gives no indication of whether the ion is in a labile pool or complexed within the metalloprotein a-2-macroglobulin. The distinction matters, since metalloprotein-associated zinc represents a fixed pool of the metal that is unaffected even in the presence of clinically relevant zinc deficiency. Thus far, zinc levels are implicated in incidence of infectious disease and may influence chronic conditions and childhood growth rates. Clearly, a fuller understanding of the physiological significance of zinc levels depends upon a robust assay for labile zinc pools in body fluids. Zalewski et al. (p. 509) suggest an attractive zinc measurement strategy that is based upon the use of a zinc-responsive fluorophore, zinquin, a reagent that has previously been used to visualize zinc ions in sub-cellular compartments. The authors demonstrate that zinquin offers a straightforward, rapid, and cost-effective means to quantify labile zinc. The technique is sensitive enough to detect a reduction in plasma zinc levels in mice fed a zinc-poor diet for just 7 days; a small analysis of human populations also suggested a correlation between zinc levels in urine and presence of type II diabetes. Finally, although the assay is likely to be widely employed in clinical studies, the zinquin-based measurement in equally useful for measuring zinc in tissue culture supernatants, indicating utility for both in vivo and in vitro analyses.
Getting Over an RNA Fixation
Formalin-fixed paraffin-embedded (FFPE) tissue continues to be a rich sample source for many gene expression profiling studies, in no small part due to the long history and continued use of formalin fixation as an effective means to preserve histological specimens. However, the fixation process is not kind to the less-than-robust nucleic acids in the samples, particularly RNA, which undergoes significant degradation due to both the fixation process and prolonged storage. This has led researchers to develop a variety of means to increase the sensitivity and efficiency of RNA detection in FFPE samples. One such protocol is described by Yang et al. on p. 481 of this issue, in which the previously used branch DNA assay is applied to FFPE tissue. The authors demonstrate that the main advantages of this technique are the ability to quantify mRNA in the homog-enate of FFPE tissues without the need for RT-PCR and its superior sensitivity over real-time PCR analysis. The procedure, which involves deparaffinization of the tissue, destruction of the cross-linking, release of the RNA by pro-teinase K treatment, and finally hybridization with the branch DNA probes, is convincingly shown by the authors to be unaffected by the fixation process or by the presence of genomic DNA, while both of these factors negatively impact standard real-time RT-PCR techniques.
