Biomedical Engineering Reference
In-Depth Information
Table 7 Synthetic DNA derived from sequences of Vimentin (VIM), avim (artificial derivate
of VIM) or MLC-2v
Target
Size
Sequence
avim
53 bp
ACCTGCTCAATGTTAAGAT GGCCCTTGACATTGAGATTGCCGCGGCC
GCAAAA
VIM 1
60 bp
CCAGCAGCTTCCTGTAGGTGGCAATCTCAATGTCAAGGGCCATCTTAA
CATTGAGCAGGT
VIM 2
111 bp
CCCTTGACATTGAGATTGCCACCTACAGGAAGCTGCTGGAAGGCGAG
GAGAGCAGGATTTCTCTGCCTCTTCCAAACTTTTCCTCCCTGAACC
TG AGGGAAACTAATCTGG
MLC-2v
192 bp
ACAGGGATGGCTTCATTGACAAG AACGATCTGAGAGACACCTTTGCT
GCCCTTGGGCGAGTGAAC GTGAAAAATGAAGAAATTGATGAAAT
GATCAAGGAGGCTCCGGGTCCAATTAACTTTACTGTGTTCCTCAC
AATG TTTGGGGA GAAACTTAAGGGAGCGGACCCTGAGGAAACC
ATTCTCAACGCAT
Complementary strands not shown
6.2 Melting Curve Analysis in Solution
One essential requirement for real-time PCR analysis is the identification of the
amplified DNA fragments using melting curve analysis. This procedure requires a
precise temperature regulation in the reaction vessel. We therefore performed a
proof-of-principle experiment to analyze the melting characteristics of DNA with
VideoScan HCU and compared it with results from the real-time iQ5 (Bio-Rad
Laboratories) thermal cycler. A set of four synthetic DNA sequences (Table
7
) was
purchased from IBA GmbH (Germany) and diluted to a final concentration of
2.5 lM in PCR buffer [
3
], including the HRM DNA-binding dye EvaGreen
(Biotium). The melting curve conditions were set to the same conditions (70-95 C,
1 C/step, dwell time 15 s).
The raw fluorescence data were processed utilizing a custom-made plugin for
RKWard (Rödiger et al. [
27
]) and custom-made scripts. In brief, refMFI fluores-
cence values were normalized to zero by subtracting the offset after a robust linear
regression utilizing the lmrob() function. Subsequently, negative first derivatives
of the smooth.spline() function fitted data of the refMFI were plotted against the
temperature (-DrefMFI/DT).
We found that the melting curve shapes were identical between both machines
(not shown). Analysis of melting peaks, however, revealed different melting
temperatures (Fig.
19
a). The iQ5 thermal cycler reported about 80, 76, 80.5, 78 C
whereas the HCU about 82, 78, 82.5, 80 C for avim, VIM 1, VIM 2 and MLC-2v
(Fig.
19
b), respectively. Thus an estimated static offset of 2 C between the two
machines was found, independently of the sequence used. The difference is pre-
sumably due to slight technical differences. Since the offset appears to be constant
there is no limitation in the usefulness of the HCU.
