Modifications of human and viral deoxyribonucleic acid by formaldehyde fixation

Abstract

Background: Formaldehyde reacts with human and viral DNA through interaction with hydrogen bonds, fixation of DNA-protein, and hydroxymethylation of the nucleic acids. Even though most archival tumor tissues are fixed with formaldehyde, little has been done to analyze the consequences of the reaction of formaldehyde with DNA. Misleading results can be obtained from fixed tissue using polymerase chain reaction (PCR) typing or restriction fragment length polymorphism analyses.

Experimental design: We have studied variations in fixation time in various tissues obtained at autopsy and in prostatic carcinoma biopsies to analyze the effects of the formaldehyde fixation. Different PCR-products were studied after different fixation times.

Results: DNA from fixed tissues appears to be no more fragmented than the native DNA. Changes in the DNA structure is more important than DNA quantity for performing PCR on fixed tissues. PCR products longer than 2 to 300 bp was difficult to amplify from some tissues. Only 8 hours of fixation can be enough to inhibit amplification of more than 421 bp. Tissue fixed for longer than 215 hours cannot be amplified for more than 200 basepair products unless excessive numbers (50-80) of PCR-cycles are used.

Conclusions: The loss of PCR product is related to fixation time and PCR-product-length, probably because of the rate of denaturation followed by modification of DNA. Contrary to what has previously been assumed, formaldehyde neither fragments nor reduces the quantity of DNA, but rather changes the structure of DNA. Different tissues may also react differently with formaldehyde, in part because of different tissue fixation gradients. When the PCR product is shorter than 200 bp, DNA isolated from paraffin-embedded tissues fixed with 4% formaldehyde can be useful to any kind of PCR product analysis.