Anatomic Pathology / UNLABELED PROBE AND AMPLICON MELTING ASSAYS
Closed-Tube SNP Genotyping Without Labeled Probes
A Comparison Between Unlabeled Probe and Amplicon Melting
Michael Liew, PhD,1 Michael Seipp,1 Jacob Durtschi,1 Rebecca L. Margraf, PhD,1 Shale Dames, MS,1
Maria Erali, MS,1 Karl Voelkerding, MD,1,2 and Carl Wittwer, MD, PhD1,2
Key Words: LCGreen; Unlabeled probe; Amplicon melting; Lactose intolerance; Temperature standards; Human platelet antigen
DOI: 10.1309/N7RARXH3623AVKDV
Abstract
Two methods for closed-tube single nucleotide
polymorphism (SNP) genotyping without labeled
probes have become available: unlabeled probe and
amplicon melting. Unlabeled probe and amplicon
melting assays were compared using 5 SNPs: human
platelet antigens 1, 2, 5, and 15 and a C>T variant
located 13,910 base pairs (bp) upstream of the lactase
gene. LCGreen Plus (Idaho Technology, Salt Lake City,
UT) was used as the saturating DNA dye. Unlabeled
probe data were readily interpretable and accurate for
all amplicon lengths tested. Five targets that ranged in
size from 42 to 72 bp were well resolved by amplicon
melting on the LightScanner (Idaho Technology) or
LightTyper (Roche, Indianapolis, IN) with no errors in
genotyping. However, when larger amplicons (206 bp)
were used and analyzed on lower resolution
instruments (LightTyper and I-Cycler, Bio-Rad,
Hercules, CA), the accuracy of amplicon genotyping
was only 73% to 77%. When 2 temperature standards
were used to bracket the amplicon of interest, the
accuracy of amplicon genotyping of SNPs was
increased to 100% (LightTyper) and 88% (I-Cycler).
In conjunction with real-time polymerase chain reaction
(PCR), melting analysis is commonly used for genotyping.
Sequence-specific, fluorescently labeled hybridization probes
generate melting curves that distinguish different genotypes.
The method was introduced using a labeled primer and a
labeled probe for factor V Leiden genotyping.1 More commonly, 2 adjacent hybridization probes are used, initially
demonstrated with the hemochromatosis single nucleotide
polymorphisms (SNPs) C282Y and H63D.2 The method was
further simplified using a single fluorescein-labeled probe that
changes fluorescence on hybridization, demonstrated on several clinical targets, including factor V Leiden and the cystic
fibrosis deletion, F508del.3
All of the aforementioned methods require fluorescently
labeled probes with unique designs and added cost compared
with unlabeled oligonucleotides. DNA melting curves can
also be generated using fluorescent double-stranded DNA
binding dyes. The most widely used dye is SYBR green I.4,5
Recently, a new family of saturating LCGreen dyes has been
introduced.6,7 These dyes allow genotyping with unlabeled
probes that have no fluorescent label, as demonstrated for
SNPs found in the cystic fibrosis and RET genes.8,9
Genotyping by melting usually depends on allele discrimination by probe melting temperature (Tm). The Tm difference between SNP alleles is determined by the probe length
and the base mismatch but is usually 2°C to 8°C and easily
detected on standard real-time instrumentation. Unlabeled
probe genotyping is based on this Tm difference and requires
only 3 unlabeled oligonucleotides.8 An even simpler method
for genotyping uses only the 2 primers required for PCR.
Initially developed for small amplicons, direct “amplicon
genotyping” requires no probes.7 Heterozygotes are easily
Am J Clin Pathol 2007;127:341-348
© American Society for Clinical Pathology
341
DOI: 10.1309/N7RARXH3623AVKDV
341
341
Liew et al / UNLABELED PROBE AND AMPLICON MELTING ASSAYS
detected by a change in melting curve shape, not a shift in Tm.
However, differentiating between the 2 possible homozygotes
can be problematic when probes are not used.
In the majority of biallelic SNPs, the 2 possible homozygotes have a Tm difference of about 1.0°C within small amplicons.7 In some SNPs, the difference is smaller, and mixing of
unknown samples with a known genotype may be necessary.10
The difference between homozygotes also becomes smaller as
the amplicon gets larger. At some point, the temperature resolution required for amplicon genotyping may exceed the ability of the technique and/or the instrument used.11
Apparent DNA melting temperatures are subject to 2 primary sources of variation. Tms vary with the salt concentration of different buffers, or variation can even result from variable evaporation during processing. Apparent variations are
also caused by instrumentation, for example, spatial temperature variation across a block used to heat a 96- or 384-well
plate.11 Any experimental variation will have the greatest
impact on methods that require the greatest resolution, such as
amplicon melting.
In the present study, the genotyping accuracy of unlabeled probe and amplicon melting are demonstrated on the
high-resolution LightScanner (Idaho Technology, Salt Lake
City, UT), a 96/384-well instrument dedicated to melting
analysis.12,13 Five SNPs were studied. Four were SNPs
involved in determining the human platelet antigens (HPAs)
1, 2, 5, and 15.14 One target was a C>T SNP (rs4988235)
found 13,910 base pairs (bp) upstream of the lactase gene
(LCT) associated with lactase intolerance.15 Furthermore,
some larger amplicons were analyzed on lower-resolution
instruments (LightTyper, Roche, Indianapolis, IN, and ICycler, Bio-Rad, Hercules, CA) so that errors in amplicon
genotyping were expected.11 To increase the accuracy of the
amplicon melting assay, internal temperature standards were
incorporated into the PCR.
Materials and Methods
DNA Samples
Human blood specimens submitted to ARUP (Salt Lake
City, UT) for routine clinical genotyping were deidentified
according to a global ARUP protocol under institutional
review board No. 7275. DNA from 123 samples of unknown
HPA and LCT C>T-13910 genotype were extracted using the
same manufacturing lot number of the blood DNA minikit
(Qiagen, Valencia, CA). The final DNA concentration
ranged from 15 to 35 ng/µL as determined by absorbance at
260 nm. Sequencing to confirm genotypes was performed at
ARUP by standard dideoxy methods using previously
described primers.16 Oligonucleotides for primers and probes
342
342
Am J Clin Pathol 2007;127:341-348
DOI: 10.1309/N7RARXH3623AVKDV
were obtained from Integrated DNA Technologies (Coralville,
IA) or the University of Utah core facility.
Unlabeled Probe Genotyping Assays
PCR was performed in 10-µL volumes in the presence of
0.5 µmol/L of an unlabeled oligonucleotide probe, 1×
LCGreen Plus (Idaho Technology), 2 to 4 mmol/L of magnesium chloride (MgCl2), and 1 µL (15-25 ng/µL) of genomic
DNA in a DNA Engine (Bio-Rad). The unlabeled probe was
blocked at its 3' end with a C6 amino modifier (Glen
Research, Sterling, VA). Before amplification, samples were
overlaid with 20 µL of mineral oil, a 96-well adhesive cover
was applied, and the plate was centrifuged for 30 seconds at
2,000g at room temperature to remove any air bubbles. All
amplifications were initiated with a 10-minute hold at 25°C,
followed by a 10-minute hold at 95°C. During temperature
cycling, denaturation was performed at 94°C for 15 seconds,
annealing for 15 seconds, and extension at 72°C for 15 seconds. The annealing temperature, number of cycles, and
MgCl2 concentrations are given for each locus in the following paragraphs. After temperature cycling, a 1-minute hold at
72°C and a 15-second hold at 94°C were performed.
The HPA targets were amplified with 1× Master
Hybridization Probe Master Mix (Roche) and 0.5 U of
AmpErase (Applied BioSystems, Foster City, CA). Primer
ratios were asymmetric at 1:6 (0.08:0.48 µmol/L). The 188-bp
HPA 1 amplicon was amplified in the reaction mix containing
2 mmol/L of MgCl2 by previously described primers14 and
analyzed with the unlabeled probe AGCGAGGTGAGCCCAGAGGCAGGGCCTGTA. Thermal cycling consisted of 50
cycles with annealing at 58°C. The 206-bp HPA 2 amplicon
was amplified in the reaction mix containing 3 mmol/L of
MgCl2 by previously described primers14 and detected with
the unlabeled probe CCCCAGGGCTCCTGACGCCCACACCCAAGC. Thermal cycling was for 40 cycles with annealing
at 65°C. The 222-bp amplicon for HPA 5 was detected with
the unlabeled probe GTCTACCTGTTTACTATCAAAGAGGTAAAAAAAAAAAAATAAACTAATAG and amplified by
primers ATGAGTGACCTAAAGAAAGAGG and GGGGACATCCTCAAAAATGA in the reaction mix containing 4
mmol/L of MgCl2. Thermal cycling was for 40 cycles with
annealing at 60°C. The 102-bp amplicon for HPA 15 was
detected by the unlabeled probe AAATTCTTGGTAAATCCTGTAACTGAAGTCAAGATAATAA and amplified by
primers TCAGTTCTTGGTTTTGTGATGTTT and CCCAAGAAGTGATAGAATCAGG in the reaction mix containing 2
mmol/L of MgCl2. Thermal cycling was for 50 cycles with
annealing at 60°C.
The 264-bp PCR product for genotyping the LCT C>T13910 SNP was amplified with 0.4 U of platinum Taq in 1×
platinum Taq PCR buffer (Invitrogen, Carlsbad, CA), 0.01
U/µL of UNG (Roche), 0.49 µmol/L of forward primer
© American Society for Clinical Pathology
Anatomic Pathology / ORIGINAL ARTICLE
GCTTTGGTTGAAGCGAAGAT, 0.06 µmol/L of reverse
primer CCATTTAATACCTTTCATTCAGGA, and the unlabeled probe GGCAATACAGATAAGATAATGTAGCCCCTGGCCTCAAAGGAACTCTCC. Thermal cycling was for
49 cycles with annealing at 56°C and 3.5 mmol/L of MgCl2.
Amplicon Genotyping Assays
For HPAs 1, 2, 5, and 15, duplex PCRs were run as
described previously14 with the following modifications. The
primers for HPA 15 were TCAGTTCTTGGTTTTGTGATGTTT and TCCTAAATTCTTGGTAAATCCTG. For HPA 2
and 5, the primer concentrations were 0.25 and 0.3 µmol/L,
respectively. The number of cycles for the HPA 1 and 5 PCR
was 38, and the MgCl2 concentration was 2.5 mmol/L. The
UNG, thermal cycler, and cycling conditions of the unlabeled
probe assays described above were used, except that the
annealing temperature was 60°C and the post-PCR conditions
were a 10-second hold at 95°C followed by cooling to 15°C.
Two different PCR products were used for LCT C>T13910 SNP genotyping, 47 bp and 206 bp. The 47-bp product
was amplified with 0.25 µmol/L of each primer, AGTTCCTTTGAGGCCAGG and GCTGGCAATACAGATAAGATAATGTA, and 3.5 mmol/L of MgCl2. PCR was performed
as described above for 34 cycles with an annealing temperature
of 58°C. The 206-bp PCR product was amplified with 0.4 U/µL
of KlenTaq1 in 1× PC-2 buffer (AB Peptides, St Louis, MO)
with 0.25 µmol/L of each primer, CCTCGTTAATACCCACTGACCTA and CCATTTAATACCTTTCATTCAGGA. PCR
was performed as for the 47-bp product.
The 206-bp HPA 2 amplicon used in the unlabeled probe
assay was also used for amplicon melting, except with primer
concentrations of 0.25 µmol/L each and 0.2 U of AmpErase in
20-µL reaction volumes. Samples were amplified and melted
on an I-Cycler under the same conditions used for unlabeled
probe genotyping.
In all amplicon melting experiments, 0.1 µmol/L
(LightScanner and LightTyper) or 0.4 µmol/L (I-Cycler) of 2
temperature standards was included so that the amplicon(s)
of interest melted between them. These temperature standards
were selected from 4 alternatives. The highest temperature
standard (~93.0°C) was formed from equal concentrations of
GCGGTCAGTCGGCCTAGCGGTAGCCAGCTGCGGCA
CTGCGTGACGCTCAG and its complement with the complement incorporating locked nucleic acids at the underlined
positions. Without locked nucleic acids, the duplex had a
melting temperature of approximately 86.5°C. A lower temperature standard (~68.0°C) was formed from equal concentrations of ATCGTGATCTCTAGAGTTATCTAAGTCGTTATATA and its complement. The lowest temperature standard had a 3-bp deletion in the complementary strand at the
underlined positions to decrease the melting temperature to
approximately 62.5°C.
Melting Curve Acquisition and Analysis
Samples were melted on LightTyper, LightScanner, or ICycler instruments. LightTyper analysis was between 60°C
and 95°C at a ramp rate of 0.1°C/s, exposure times of 48 ms,
and a data collection frame interval of 500 ms. LightScanner
analysis used default settings between 55°C and 98°C.
Turnaround time per plate for both of these instruments was
15 minutes. I-Cycler melting analysis was between 55°C
and 98°C, with holds every 0.1°C for 10 seconds and
required 75 minutes.
Data analysis was performed with custom software using
exponential background subtraction.17 Genotyping for unlabeled probe assays was based on probe Tm, obtained from negative derivative plots of the LightTyper or LightScanner data.
Genotyping by amplicon melting was based on normalized
melting curves7,18 and how they clustered with control samples
run in the same plate. Homozygous samples were identified by
a single transition, with different alleles being separated by
0.3°C to 1°C. Heterozygous samples showed a characteristic
broad transition with an altered curve shape resulting from heteroduplex contributions. When temperature standards were
present, the melting curves were optionally aligned before
analysis using the 2 internal standards to position and stretch
the curves as necessary, using linear interpolation.
Results
Derivative melting curves of all unlabeled probe assays
are shown in ❚Figure 1❚. The HPA targets were melted on a
high-resolution LightScanner and the LCT SNP was melted
on a lower resolution LightTyper. All genotypes are distinct
and easily resolvable without interference from the no-template control. There are 2 melting transitions (Figures 1D and
1E), one for the unlabeled probe at a lower temperature and
another for the amplicon at a higher temperature.
Homozygous alleles are represented by a single probe melting
peak, whereas heterozygous samples have 2 peaks corresponding to both alleles separated by approximately 4°C. For the
HPA samples, 13 samples are shown for each target, and all
genotypes are present except the rare HPA 5, A/A homozygote. A total of 100 samples were screened for the LCT13910C>T SNP, and 42 T/T homozygotes, 28 C/T heterozygotes, and 30 C/C homozygotes were identified.
The amplicon sizes used in the initial amplicon melting
assays ranged from 42 to 72 bp with good separation between
all genotypes ❚Figure 2❚. The HPA targets were melted on a
LightScanner, and the LCT SNP was melted on a LightTyper.
There was 100% concordance between the amplicon melting
and the unlabeled probe assays. However, the amplicon melting assay for the LCT-13910C>T SNP had an interfering
primer dimer that complicated result interpretation.
Am J Clin Pathol 2007;127:341-348
© American Society for Clinical Pathology
343
DOI: 10.1309/N7RARXH3623AVKDV
343
343
Liew et al / UNLABELED PROBE AND AMPLICON MELTING ASSAYS
A
B
6
C/C
C
C/T
6
T/T
6
C/C
G/G
3
T/T
–dF/dT
–dF/dT
–dF/dT
G/A
3
3
T/C
0
70
80
0
70
90
80
Temperature (˚C)
0
55
90
Temperature (˚C)
D
65
75
Temperature (˚C)
E
6
6
A/A
T/T
3
–dF/dT
–dF/dT
C/C
C/A
3
C/T
C/C
0
0
60
70
70
80
78
86
Temperature (˚C)
Temperature (˚C)
❚Figure 1❚ Derivative melting curves of unlabeled probes for genotyping of human platelet antigen 1 (A; 188 base pairs [bp]), 2 (B; 206
bp), 5 (C; 222 bp), 15 (D; 102 bp), and LCT-13910 (E; 264 bp). Each graph shows 13 samples. Genotypes are labeled with arrows.
A
B
100
100
C/C
T/C
T/T
0
78
83
C/C
C/T
50
T/T
0
79
88
Fluorescence
Fluorescence
Fluorescence
100
50
C
84
89
Temperature (˚C)
Temperature (˚C)
D
G/G
G/A
50
0
67
72
77
Temperature (˚C)
E
100
C/A
50
0
68
Fluorescence
Fluorescence
100
C/C
A/A
73
Temperature (˚C)
78
C/C
C/T
50
T/T
0
70
78
86
Temperature (˚C)
❚Figure 2❚ Normalized melting curves for amplicon genotyping of human platelet antigen 1 (A; 42 base pairs [bp]), 2 (B; 51 bp),
5 (C; 69 bp), 15 (D; 72 bp), and LCT-13910 (E; 47 bp). Each graph shows 13 samples. Genotypes are labeled with arrows.
344
344
Am J Clin Pathol 2007;127:341-348
DOI: 10.1309/N7RARXH3623AVKDV
© American Society for Clinical Pathology
Anatomic Pathology / ORIGINAL ARTICLE
A
Amplicon
28
18
Tm std #1
Tm std #2
No template
8
–2
63
72
81
90
Temperature (˚C)
❚Figure 3❚ Derivative melting curves of amplicon melting for
genotyping of the LCT gene C>T-13910 single nucleotide
polymorphism. One representative curve of each genotype is
shown. The peaks seen at the lowest and the highest melting
temperatures are the temperature standards. The peaks seen
in the middle (ca 81°C) are the amplicon melting peaks used
for genotyping.
2 heterozygotes and 9 T/T homozygotes were incorrectly
genotyped as C/C. In contrast, all samples were genotyped
correctly when the samples were temperature-corrected using
the temperature standards.
B
100
100
Fluorescence
Fluorescence
T/T
C/T
C/C
–dF/dT
A larger 206-bp amplicon was designed for LCT13910C>T to eliminate the interfering primer dimer and to
assess the effect of PCR product size on genotyping by amplicon melting. Internal temperature controls were selected to
bracket the amplicon as shown in ❚Figure 3❚. In the absence of
template, the peaks of the temperature standards appear larger
because all curves are normalized to 100% fluorescence
before derivatives are taken. The lower temperature control
melted at 68°C, the higher control at 86.5°C, and the amplicon
in between at about 80°C. The homozygous C/C genotype on
average was only 0.3°C more stable than the T/T sample, and
the heterozygous sample had a broader melting profile, very
similar to the amplicon peaks in the unlabeled probe assays.
However, when a full 96-well plate of samples was analyzed
on the LightTyper, the curves overlapped significantly, making genotyping difficult ❚Figure 4A❚. Calibration using temperature controls (off scale in Figure 4) allowed much better
separation of genotypes ❚Figure 4B❚. Indeed, the Tm SD within a genotype decreased more than 50% from 0.10°C to
0.16°C to 0.04°C to 0.05°C ❚Table 1❚.
To assess the need for temperature controls and correction,
a blinded study of amplicon genotyping on the LightTyper was
performed on 48 samples, including 24 T/T, 14 C/T, 5 C/C,
and 5 no-template controls. Without temperature correction,
75
50
25
75
C/C
50
C/T
25
T/T
0
79.5
80.5
81.5
82.5
0
79.5
80.5
Temperature (˚C)
81.5
82.5
Temperature (˚C)
❚Figure 4❚ Normalized melting curves for amplicon genotyping of the LCT gene C>T-13910 single nucleotide polymorphism
obtained on a 96-well LightTyper. We tested 34 T/T homozygotes, 28 C/T heterozygotes, and 30 C/C homozygotes identified by
the unlabeled probe assay, along with 4 no-template controls. The amplicon melting transitions before (A) and after (B) temperature correction with internal temperature controls are shown. Genotypes are labeled based on the genotype determined by the
unlabeled probe assay. Genotypes could be placed accurately only after temperature correction.
❚Table 1❚
Melting Temperature Mean and SD of C>T-13910 Genotypes From 92 Samples*
T/T Homozygous (n = 34)
Original data
Corrected
*
C/T Heterozygous (n = 28)
C/C Homozygous (n = 30)
Mean (°C)
SD (°C)
Mean (°C)
SD (°C)
Mean (°C)
SD (°C)
81.21
81.02
0.10
0.05
81.20
81.04
0.16
0.05
81.52
81.30
0.10
0.04
The melting temperatures were obtained from the derivative melting curves using the temperature at the highest point of the peak.
Am J Clin Pathol 2007;127:341-348
© American Society for Clinical Pathology
345
DOI: 10.1309/N7RARXH3623AVKDV
345
345
Liew et al / UNLABELED PROBE AND AMPLICON MELTING ASSAYS
To further demonstrate the usefulness of temperature
standards, a 206-bp HPA 2 amplicon was melted on the ICycler using 13 samples of each genotype and 4 no-template
controls ❚Figure 5❚. Without temperature standards, 27% of
the homozygotes were incorrectly genotyped, whereas all heterozygotes were correctly identified. After temperature correction, 88% accuracy was obtained.
Both unlabeled probe and amplicon genotyping detected
1 sample with an aberrant melting profile for the LCT13910C>T locus, suggesting a unique sequence variation
under the probe and within the amplicon. Sequencing identified this sample as a compound heterozygote C>A-13909,
C>T-13910 ❚Figure 6❚.
Discussion
Unlabeled probe8 and amplicon melting7 are recently
described methods for closed-tube genotyping that do not
require labeled probes or processing after PCR. Only standard
oligonucleotides and post-PCR melting are used; no real-time
A
data19 or allele-specific amplification20 is necessary. Both
methods are simple to design and promise to be cost-effective
alternatives to other genotyping assays that require more complicated probe systems (closed-tube) or post-PCR processing
(open-tube).
Melting analysis with unlabeled probes is similar to conventional HybProbes2 (Roche) except that only 1 probe is needed
and a saturating DNA dye is used instead of covalently attached
fluorescent labels. SNP allele Tms are usually 2°C to 8°C apart
and are easily separated. Unlabeled probe genotyping has been
demonstrated on the LightCycler (Roche), the LightTyper, and
the high-resolution instrument HR-1 (Idaho Technology).8,9
High-resolution melting does not seem to be required because of
the large temperature separation between alleles. Exponential
background subtraction can be used to remove high background
fluorescence observed at low temperatures.17
Amplicon melting is the simplest possible implementation of PCR-based genotyping.7 No probes and only 2 standard oligonucleotides for PCR primers are used. Genotyping
by amplicon melting has been applied to several targets,
including HPAs14 and P-450 2C9 alleles.21 In the rare cases
B
100
Fluorescence
Fluorescence
100
50
0
87
89
C/C
50
C/T
T/T
0
87
91
Temperature (˚C)
89
91
Temperature (˚C)
❚Figure 5❚ The 206-base-pair human platelet antigen 2 amplicon analyzed on an I-Cycler. Normalized curves are shown before (A)
and after temperature correction (B). Genotype regions are indicated by arrows.
A
B
29
75
C/A, C/T
–dF/dT
Fluorescence
100
C/C
50
C/T
25
19
C/A, C/T
T allele C allele
9
T/T
0
79
80
81
Temperature (˚C)
82
–1
69
73
77
81
85
Temperature (˚C)
❚Figure 6❚ Detection of the compound heterozygote LCT C>A-13909, C>T-13910 by amplicon melting (A) and the unlabeled probe
assay (B).
346
346
Am J Clin Pathol 2007;127:341-348
DOI: 10.1309/N7RARXH3623AVKDV
© American Society for Clinical Pathology
Anatomic Pathology / ORIGINAL ARTICLE
in which homozygotes cannot be differentiated, mixing samples with a known genotype and quantitative heteroduplex
analysis is effective.10 Many different,22 but not all,9 heterozygotes within an amplicon can be differentiated.
However, most of these studies were performed on high-resolution melting instruments with limited throughput. In a
recent study, different instruments for amplicon melting were
compared.11 Predicted error rates for genotyping depended
on the Tm difference between homozygotes and the temperature precision of the instrument. Plate (96/384-well) instruments were particularly variable, potentially limiting the
throughput of small amplicon genotyping by melting to instruments with a lower throughput.
When the amplicon size is small (<100 bp), results from
unlabeled probe and amplicon genotyping are concordant,
and the accuracy of both methods is high. However, with
larger PCR products (>200 bp), the risk of error with amplicon genotyping increases, and instrument resolution becomes
critical. For example, with the 206-bp amplicons studied in
the present study, the Tm difference between homozygous
alleles was only 0.3°C. In contrast, the allele Tms in the unlabeled probe assays were separated by approximately 4°C,
with distinct peaks for heterozygous samples, similar to
HybProbe melting curves.
To demonstrate the usefulness of temperature standards in
amplicon genotyping, we melted amplicons larger than 200 bp
on the I-Cycler and the LightTyper, typical lower resolution
instruments. The LightTyper was designed for HybProbes23,24
and single-labeled probes (SimpleProbe,12,13,25 Idaho
Technology) that do not require high-resolution melting.
Without temperature correction, the accuracy of amplicon
melting was 77% on the LightTyper and 73% on the I-Cycler.
After temperature adjustment according to the standards, the
accuracy of amplicon genotyping was increased to 100% on
the LightTyper and 88% on the I-Cycler. Internal temperature
standards significantly improve genotyping by amplicon melting. The synthetic double-stranded oligonucleotides used as
standards are selected to avoid interference with the amplification and the melting analysis.
Incorporation of temperature standards into amplicon
melting assays has 2 advantages. First, lower resolution instruments can be used, such as most conventional real-time instruments that use a 96/384-well plate format. Second, larger
amplicons (with smaller Tm differences between homozygotes) can be genotyped with a greater degree of confidence.
Using larger amplicons also increases the Tm difference
between primer dimers and the target amplicon that can overlap with short amplicons (40-70 bp).
One advantage of genotyping by melting assays is that
unexpected sequence variants can often be detected.
Unlabeled probe and amplicon genotyping showed distinctive
melt curves with 1 sample that was sequenced as the compound
heterozygote C>A-13909, C>T-13910. In our population,
it was not a common variant and did not interfere with
either assay.
Five SNPs were used to demonstrate that both unlabeled
probe and amplicon melting are high-throughput, closed-tube
genotyping methods that do not require labeled probes.
Synthetic duplex oligonucleotides can be incorporated as
internal temperature standards to make genotyping by amplicon melting more robust, enabling longer amplicons and the
use of lower resolution instruments.
From the 1Institute for Clinical and Experimental Pathology,
ARUP, and the 2Department of Pathology, University of Utah
School of Medicine, Salt Lake City.
Supported by the ARUP Institute of Clinical and
Experimental Pathology.
Address reprint requests to Dr Liew: ARUP Institute for
Clinical and Experimental Pathology, 500 Chipeta Way, Salt Lake
City, UT 84108-1221.
Aspects of high-resolution melting analysis are licensed from
the University of Utah to Idaho Technology. Dr Wittwer holds
equity interest in Idaho Technology.
References
1. Lay MJ, Wittwer CT. Real-time fluorescence genotyping
of factor V Leiden during rapid-cycle PCR. Clin Chem.
1997;43:2262-2267.
2. Bernard PS, Ajioka RS, Kushner JP, et al. Homogeneous
multiplex genotyping of hemochromatosis mutations with
fluorescent hybridization probes. Am J Pathol. 1998;153:10551061.
3. Crockett AO, Wittwer CT. Fluorescein-labeled
oligonucleotides for real-time PCR: using the inherent
quenching of deoxyguanosine nucleotides. Anal Biochem.
2001;290:89-97.
4. Busi F, Cresteil T. Phenotyping-genotyping of alternatively
spliced genes in one step: study of CYP3A5*3 polymorphism.
Pharmacogenet Genomics. 2005;15:433-439.
5. Oliveira TC, Barbut S, Griffiths MW. Detection of
Campylobacter jejuni in naturally contaminated chicken skin
by melting peak analysis of amplicons in real-time PCR. Int J
Food Microbiol. 2005;104:105-111.
6. Wittwer CT, Reed GH, Gundry CN, et al. High-resolution
genotyping by amplicon melting analysis using LCGreen. Clin
Chem. 2003;49:853-860.
7. Liew M, Pryor R, Palais R, et al. Genotyping of singlenucleotide polymorphisms by high-resolution melting of small
amplicons. Clin Chem. 2004;50:1156-1164.
8. Zhou L, Myers AN, Vandersteen JG, et al. Closed-tube
genotyping with unlabeled oligonucleotide probes and a
saturating DNA dye. Clin Chem. 2004;50:1328-1335.
9. Margraf RL, Mao R, Wittwer CT. Masking selected sequence
variation by incorporating mismatches into melting analysis
probes. Hum Mutat. 2006;27:269-278.
10. Palais RA, Liew MA, Wittwer CT. Quantitative heteroduplex
analysis for single nucleotide polymorphism genotyping. Anal
Biochem. 2005;346:167-175.
Am J Clin Pathol 2007;127:341-348
© American Society for Clinical Pathology
347
DOI: 10.1309/N7RARXH3623AVKDV
347
347
Liew et al / UNLABELED PROBE AND AMPLICON MELTING ASSAYS
11. Herrmann MG, Durtschi JD, Bromley LK, et al. Amplicon
DNA melting analysis for mutation scanning and genotyping:
cross-platform comparison of instruments and dyes. Clin Chem.
2006;52:494-503.
12. Bennett CD, Campbell MN, Cook CJ, et al. The LightTyper:
high-throughput genotyping using fluorescent melting curve
analysis. Biotechniques. 2003;34:1288-1292, 1294-1295.
13. Grannemann S, Landt O, Breuer S, et al. LightTyper assay
with locked-nucleic-acid-modified oligomers for genotyping
of the toll-like receptor 4 polymorphisms A896G and C1196T.
Clin Chem. 2005;51:1523-1525.
14. Liew M, Nelson L, Margraf R, et al. Genotyping of human
platelet antigens 1 to 6 and 15 by high-resolution amplicon
melting and conventional hybridization probes. J Mol Diagn.
2006;8:97-104.
15. Enattah NS, Sahi T, Savilahti E, et al. Identification of a
variant associated with adult-type hypolactasia. Nat Genet.
2002;30:233-237.
16. Rasinpera H, Savilahti E, Enattah NS, et al. A genetic test
which can be used to diagnose adult-type hypolactasia in
children. Gut. 2004;53:1571-1576.
17. Erali M, Palais R, Wittwer CT. SNP Genotyping by Unlabeled
Probe Melting Analysis. Totowa, NJ: Humana Press. In press.
18. Gundry CN, Vandersteen JG, Reed GH, et al. Amplicon
melting analysis with labeled primers: a closed-tube method
for differentiating homozygotes and heterozygotes. Clin Chem.
2003;49:396-406.
348
348
Am J Clin Pathol 2007;127:341-348
DOI: 10.1309/N7RARXH3623AVKDV
19. Wittwer C, Kusukawa N. Diagnostic Molecular Microbiology:
Principles and Applications. Washington, DC: ASM Press; 2004.
20. Wang J, Chuang K, Ahluwalia M, et al. High-throughput SNP
genotyping by single-tube PCR with Tm-shift primers.
Biotechniques. 2005;39:885-893.
21. Hill CE, Duncan A, Wirth D, et al. Detection and
identification of cytochrome P-450 2C9 alleles *1, *2, and *3
by high-resolution melting curve analysis of PCR amplicons.
Am J Clin Pathol. 2006;125:584-591.
22. Graham R, Liew M, Meadows C, et al. Distinguishing different
DNA heterozygotes by high-resolution melting. Clin Chem.
2005;51:1295-1298.
23. Lin Z, Suzow JG, Fontaine JM, et al. A high throughput
beta-globin genotyping method by multiplexed melting
temperature analysis. Mol Genet Metab. 2004;81:237-243.
24. Murugesan G, Aboudola S, Szpurka H, et al. Identification
of the JAK2 V617F mutation in chronic myeloproliferative
disorders using FRET probes and melting curve analysis.
Am J Clin Pathol. 2006;125:625-633.
25. Frances F, Corella D, Sorli JV, et al. Validating a rapid method
for detecting common polymorphisms in the APOA5 gene by
melting curve analysis using LightTyper. Clin Chem.
2005;51:1279-1282.
© American Society for Clinical Pathology