01 Aug 2002

The use of DNA arrays, or chips, to simultaneously measure thousands of mRNA transcripts has revolutionized the way biologists quantify gene expression in vitro. Now, in tomorrow’s Science, researchers at the Albert Einstein College of Medicine in New York demonstrate the use of fluorescence in situ hybridization, or FISH, to measure multiple transcription events in vitro.

Working in Robert Singer’s lab, first author Jeffrey Levsky and colleagues combined in situ hybridization with fluorescent oligonucleotide probes to visualize transcription by color. Using a palette of five fluorophores, each emitting at a distinct part of the spectrum, the authors designed probe sets that would bind to specific nascent transcripts and yield a unique intermediary color, or “barcode,” upon excitation. They found that a minimum of two probes per barcode improved the specificity of the technique because emissions from single probes that bind to nucleotides non-specifically could be ignored.

Using the mix-and-match barcodes Levsky et al. measured transcription of ten known genes, including β- and γ-actin, simultaneously. They found a direct correlation between the size of nuclei and the number of transcription events in living adenocarcinoma cells, and they demonstrated that transcription could occur at none, one, or two alleles-interleukin-8, for example, was transcribed from one allele in about 20 percent of nuclei, but from two alleles in about five percent of cells. The authors also followed transcription events in real time and showed that in fibroblasts the kinetics of β-actin and c-fos activation were almost identical.

While the problem of spectral overlap is likely to keep the numbers of usable barcodes in the tens, rather than the thousands, the power of the technique lies in its spatial and temporal resolution. Thus, it can be used to quantify allelic differences in transcription, or identify the order of activation in a given pathway. In addition, because it paints a picture of active transcription, this FISH technique is likely to serve as a complement to chip analysis, which measures the total amount of transcript in a given sample. Indeed, this may explain why the activation of fibroblast β-actin, γ-actin, and c-jun, detected by FISH, failed to show up in the chip analysis. So for now, at least, it seems the answer to the question “FISH” or “chips?” should be “yes.”—Tom Fagan