| Description |
Fluorescently-labeled oligonucleotides are often used to detect DNA alterations. Most require several steps and manual transfer of samples following PCR. Sample transfer can be eliminated using homogeneous genotyping by melting analysis. The amplified DNA is slowly heated, causing the double-stranded, fluorescently-labeled DNA to dissociate, or melt. This is accompanied by a change in fluorescence, which can be monitored on a temperature-controlled fluorimeter. Fluorescence resonance energy transfer (FRET) was used to monitor 5'-Cy5-labeled probes hybridized adjacently to a fluorescein-labeled probe during melting analysis. One hundred five samples were genotyped for the cystic fibrosis mutations F508del and I507del. All heterozygous and homozygous mutations were clearly distinguishable. An unexpected F508C heterozygote was found by an intermediate melting transition. The inherent fluorescence quenching of deoxyguanosine residues can also be used to monitor the melting of the entire amplicon if a fluorophore-labeled primer is added before PCR. Low temperature transitions appearing as shoulders or separate peaks result from the melting of heteroduplex DNA formed after re-annealing the heterozygous amplicon. Heteroduplex DNA is maximized by rapid heating, rapid cooling, and low ionic strength. Different homozygotes were differentiable by melting temperature. Cystic fibrosis, hemoglobin S, C, and E, and HTR2A amplicons ranging from 44 to 304 bp were genotyped. To better understand the effects that varying sequence and fluorophores have on quenching, some of the variables related to fluorescence change were analyzed in a system consisting of probes and blunt-ended complementary DNA. Probe/unlabeled complement duplexes displayed differing directions and magnitudes of fluorescence change upon melting even when the same fluorophore was attached to the same 5' or 3' base. In 5 out of 12 groups of probe/complement duplexes, sorted by fluorophore type, the number of Gs on the same strand within five bp of the fluorophore correlated to magnitude of negative fluorescence change upon melting with an R of at least 0.23 (average R = 0.4739). In six of these 12 groups, the number of Gs on the unlabeled strand within five bp of the fluorophore correlated to magnitude of positive fluorescence change upon melting with an R2 of at least 0.25 (average R2 = 0.480). |
