Researchers use chemistry to solve a nagging problem that has challenged the analysis of RNA from formaldehyde-fixed tissue specimens.
In hospitals far and wide, tissue samples taken during routine biopsies and surgeries are prepared for histology analysis by being fixed in formaldehyde and embedded in paraffin. Hundreds of millions of these preserved samples are banked and indexed by disease, treatment and outcome. Collectively, they serve as a resource with enormous potential for yielding genetic links to disease and for identifying potential drug targets, says Eric Kool, a chemist at Stanford University.
Extracting molecular information about RNA, DNA and proteins from formaldehyde-fixed specimens, however, has been fraught with challenges. Formaldehyde causes extensive molecular cross-linking and the formation of aminal and hemiaminal adducts, hindering the analysis of fixed samples by PCR, sequencing, immunohistochemistry and mass spectrometry. A number of protocols exist for de-paraffinization and reversal of formaldehyde cross-links and adducts, but they require the application of biomolecule-degrading heat and remain far from efficient.
Kool and his team recently tackled the challenge of reversing formaldehyde cross-links in RNA using chemistry—and without applying high heat. “A few years ago a researcher at a local genomics company came to me to see if I had any ideas for solving their 'formaldehyde problem,'” Kool recounts. At that time, he didn't have a solution for the researcher, but the problem stayed lodged in the back of his mind until recently, when his lab was working on an unrelated project that involved developing catalysts to speed the formation of imine bonds. “It occurred to me that these catalysts could also be useful in breaking imine and imine-like bonds, and suddenly the formaldehyde problem came back to me,” he says.
Using monomeric nucleotides as a model system, Kool's group tested the ability of a series of arylacid and amine catalysts to reverse formaldehyde adducts. The winning catalyst was a bifunctional phosphanilate compound that could reverse both aminal and hemiaminal adducts. They next tested an extensively cross-linked 16-mer RNA strand consisting of a central self-complementary sequence flanked by overhanging adenosines on both sides. Heating the sample to 60 °C resulted in the removal of formaldehyde cross-links and adducts, but it also caused RNA degradation. The phosphanilate catalyst, however, efficiently removed cross-links and adducts at the much milder temperature of 37 °C.
As a final test, the researchers prepared a model tissue specimen consisting of a pellet from a cultured cell line fixed with formaldehyde and embedded in paraffin. They used a standard protocol for de-paraffinization and then added their phosphanilate catalyst, optimizing the incubation time and temperature. Kool's team found that in comparison to a commercial kit and a literature protocol that required heating at higher temperatures (80 °C and 70 °C, respectively), the phosphanilate catalysts, applied at an optimal temperature of 55 °C, improved RNA yields from the model specimen 13-fold on average.
Kool expects that a modified version of his team's mild approach to reversing cross-links and adducts in RNA could also be applied to DNA, and to proteins, which could potentially be used for proteomic analysis of formaldehyde-fixed specimens. His group has also begun to apply the method to real clinical specimens, with promising results. “We hope that catalytic methods [such as this] will be applied broadly and will help large numbers of cancer patients as their doctors make medical decisions about treatment,” he says.