Microarray technology can be used to measure expression levels of genes (microarray expression analysis), changes in gene copy number (array comparative genomic hybridization, or aCGH), and gene mutation [mutation microarray analysis by single nucleotide polymorphism (SNP) arrays]. First conceptualized more than 20 years ago and used for whole-genome analysis in yeast around 1997, microarray analysis technology has benefited from continual improvements, including automation and customization. Despite the technology still having been in its infancy, Stephen Friend predicted in 1999: “In the next three to five years, however, the direct impact of these DNA microarrays is likely to enter the doctor's office as some of the harder diagnostic puzzles may be solved using this technology” 1,. In 2003, Pusztal et al. noted that the “true clinical utility and the limits of this exciting new technology are yet to be established” 2,. Although this is still the case, significant progress has been made.
New InputProfessor Chris Q. Doe of the Department of Biology at the University of Oregon (Eugene, OR, USA) and his team have developed thiouracil (TU) tagging, a new method of isolating RNA from specific cells. This method uses uracil phospho-ribosyltransferase (UPRT) to tag RNA with 4-thiouracil, a nucleotide analog that is not represented in normal RNA. The 4-thiouracil–tagged RNA can be purified from untagged RNA and analyzed using microarray technology. The UPRT gene is isolated from the protozoan parasite Toxoplasma gondii and modified, and can then be delivered to cells of interest transgenically or using a retrovirus vector. Doe points out that this tagging method is particularly useful for isolating RNA expressed in cells with long processes like glia, which can't be isolated mechanically (e.g., by micro-dissection) or by cell sorting without being destroyed. Once the RNA is tagged, it can be isolated by grinding up the whole tissue or organism. The tagged RNA is then purified over a streptavid in column after coupling the thiol group to biotin.
In addition to ease of RNA isolation, “the other interesting novel application is that we have temporal control over the time of labeling,” says Doe. For example, one can look at expression in the hippocampal neurons of mice before and after a learning experiment, and compare RNA expression prior to and after memory induction. Likewise, early, middle, and late tumor progression can be observed. “These are the features we are most excited about,” he says. “The other twist,” he adds, “is that we normalize for the uracil number in each transcript to reduce background. That's not a big deal. What's cool is putting the enzyme into cells and grinding up the whole animal after.”
Threat and Opportunity
“I think next-generation sequencing (NGS) platforms will put a dent in the microarray market, but [microarrays are] still cost-effective for large experiments. This reflects my view, but I'm not an outlier,” says Otto Folkerts, Associate Director of Technology Development at the Virginia Bioinformatics Institute (VBI) at Virginia Tech (Blacksburg, VA). Although NGS technologies are finding more widespread application in gene expression analysis (GEP) and in detecting and analyzing SNPs, there are other opportunities for microarrays. Folkerts points out that an analysis of 2600 samples from soybeans infected with a pathogen—which was done at VBI using microarrays—would have been too expensive and time-consuming with current NGS technology. However, NGS technology is becoming more competitive.
Folkerts thinks that as NGS becomes less expensive and easier to do, researchers investigating their favorite, less-common organisms will want to have custom chips for their own use in GEP. He speculates that this may lead to more custom-designed microarrays and more improvements in the technology. For example, VBI is leading the sequencing of the domestic turkey. Researchers interested in turkeys have had to use chicken microarrays, and although these birds have 90% of their genomes in common, the differences are of great interest. They plan to make a custom turkey array once the sequence is finished, and redesign of the chicken array with turkey sequences would also allow creation of a poultry array. “As people sequence more deeply and broadly, this will create demand in the face of the [NGS] threat,” Folkerts says. “I imagine everyone would want a custom array”. “I imagine everyone would want a custom array.”
