The precise quantification of rare mRNA copies from intronless genes b
y reverse transcription polymerase chain reaction (RT-PCR) requires th
e complete removal of genomic DNA because discrimination of cDNA and D
NA amplification products by differing sizes of PCR products is not po
ssible. Elimination of DNA is achieved by treating the RNA sample with
RNAse-free DNase I before RT-PCR. The lack of a PCR product from DNas
e-treated RNA samples before RT is usually accepted as a proof of effi
cient DNA destruction. However, this may vary depending on the metal c
ofactor used in the DNase I cleavage. Treating DNA-contaminated RNA sa
mples with DNase I and magnesium as a cofactor creates a negative PCR
control after digestion without further RT. Paradoxically, after addit
ional RT-PCR, the original intron-containing DNA fragment size may be
produced again. In the presence of manganese as cofactor; RT-created D
NA fragments do riot appear. This is because in the presence of mangan
ese, DNase I cleaves both DNA strands at approximately the same site,
yielding DNA fragments that are blunt-ended ed or that have protruding
termini of only one or two nucleotides in length. However, overlappin
g fragments with the potential to recombine are created by DNase diges
tion with magnesium as cofactor: Because one cannot differentiate betw
een a PCR signal produced by RNA and one produced by recombined DNA af
ter DNase I digestion and RT: all such DNase I assays should be perfor
med with manganese instead of magnesium.