Analysis of chemical modification of RNA from formalin-fixed samples and optimization of molecular biology applications for such samples.

Formalin-fixed archival samples are known to be poor materials for molecular biological applications. We conducted a series of experiments to understand the alterations in RNA in fixed tissue. We found that formalin-fixed tissue was resistant to solubilization by chaotropic agents. However, proteinase K completely solubilized the fixed tissue and enabled the extraction of almost the same amount of RNA as from a fresh sample. The extracted RNA did not show apparent degradation. However, as reported, successful PCR amplification was limited to short targets. The nature of such 'fixed' RNA was analyzed using synthetic homo-oligo RNAs. The heterogeneous increase in molecular weight of the RNAs, measured by MALDI-TOF mass spectrometry, showed that all four bases showed addition of mono-methylol (-CH(2)OH) groups at various rates. The modification rate varied from 40% for adenine to 4% for uracil. In addition, some adenines underwent dimerization through methylene bridging. The majority of the methylol groups, however, could be removed from bases by simply elevating the temperature in formalin-free buffer. This demodification proved effective in restoring the template activity of RNA from fixed tissue. The improvement in PCR results suggested that more than half of the modification was removed by this demodification.