Human MutationMETHODS
Identifying Sequence Variants in the Human
Mitochondrial Genome Using High-Resolution Melt
(HRM) Profiling
Steven F. Dobrowolski,1 Jesse Gray,1 Trent Miller,1 and Mitch Sears2
1Idaho Technology, Salt Lake City, Utah; 2ARUP Laboratories, Salt Lake City, Utah
For the Focus Section on HRMA Technology
Received 12 December 2008; accepted revised manuscript 10 February 2009.
Published online 3 March 2009 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/humu.21003
ABSTRACT: Identifying mitochondrial DNA (mtDNA)
sequence variants in human diseases is complicated. Many
pathological mutations are heteroplasmic, with the mutant
allele represented at highly variable percentages. High-
resolution melt (HRM or HRMA) profiling was applied to
comprehensive assessment of the mitochondrial genome
and targeted assessment of recognized pathological muta-
tions. The assay panel providing comprehensive coverage of
the mitochondrial genome utilizes 36 overlapping fragments
(301–658bp) that employ a common PCR protocol. The
comprehensive assay identified heteroplasmic mutation in
33 out of 33 patient specimens tested. Allele fraction
among the specimens ranged from 1 to 100%. The
comprehensive assay panel was also used to assess 125
mtDNA specimens from healthy donors, which identified
431 unique sequence variants. Utilizing the comprehensive
mtDNA panel, the mitochondrial genome of a patient
specimen may be assessed in less than 1 day using a single
384-well plate or two 96-well plates. Specific assays were
used to identify the myopathy, encephalopathy, lactic
acidosis and stroke-like episodes (MELAS) mutation
m.3243A4G, myoclonus epilepsy, ragged red fibers
(MERRF) mutation m.8344A4G, and m.1555A4G
associated with aminoglycoside hearing loss. These assays
employ a calibrated, amplicon-based strategy that is
exceedingly simple in design, utilization, and interpretation,
yet provides sensitivity to detect variants at and below 10%
heteroplasmy. Turnaround time for the genotyping tests is
about 1hr.
Hum Mutat 30, 891–898, 2009. & 2009 Wiley-Liss, Inc.
KEY WORDS: melt profiling; HRM; HRMA; mitochon-
dria; heteroplasmy; MERRF; MELAS
Introduction
Mitochondria are cytoplasmic organelles that support oxidative
phosphorylation and the production of ATP. Energy production
in the mitochondria requires genes of both nuclear and
mitochondrial origin. The mitochondria have an autonomously
replicating, 16,569-bp circular DNA genome containing 13 genes
that encode oxidative phosphorylation subunits, two ribosomal
RNA genes, and 22 transfer RNA genes [Anderson et al., 1981]. A
heterogeneous group of human diseases, with a wide spectrum of
clinical manifestations, results from mutation of mitochondrial
genes [Andreu and DiMauro, 2003]. Diseases commonly asso-
ciated with mitochondrial DNA (mtDNA) mutations are as
follows: myopathy, encephalopathy, lactic acidosis, and stroke-like
episodes (MELAS); myoclonus epilepsy, and ragged red fibers
(MERRF); neuropathy, ataxia, and retinitis pigmentosa (NARP);
Leber’s hereditary optic neuropathy (LHON); and Kearns-Sayre
syndrome. Single-nucleotide changes are common causes of
MELAS (m.3243A4G), MERRF (m.8344A4G), NARP
(m.8993T4G/C), and LHON (m.11778G4A), while Kearns-
Sayre syndrome is associated with deletions [Moraes et al., 1989;
Sternberg et al., 1998; Wong, 2007]. Variation in mtDNA is
described in cancerous lesions and assessment of mtDNA is
extensively used in forensic applications [Hatsch et al., 2007;
Mithani et al., 2007; Tobe and Linacre, 2008; Wang et al., 2008b;
Webb et al., 2008].
A variety of means have been utilized to identify mtDNA
variants. Identification of specific variants has been performed by
allele-specific PCR, a variety of real-time PCR procedures, and
real-time amplification refractory mutation system quantitative
PCR [Wang et al., 2008a; Wong et al., 2006; Wong and Bai, 2006;
Wong and Senadheera, 1997]. Mitochondrial DNA sequence
variants have been screened for by indirect means such as
denaturing high-performance liquid chromatography, denaturing
gradient gel electrophoresis, temperature gradient gel electrophor-
esis, and other methods [van Den Bosch et al., 2000; van Eijsden
et al., 2006; White et al., 2005; Wong et al., 2004, 2002; Xiu-Cheng
Fan et al., 2008]. While these procedures have been proven
effective, turnaround times are relatively slow and all require
extensive handling of amplified PCR product, presenting a
contamination hazard within the laboratory.
High-resolution melting (HRM or HRMA) is an effective
method to screen for sequence variation and has been effectively
used to assess genes involving inborn errors of metabolism, cancer
susceptibility, and other genes whose dysfunction is associated
with disease [Bastien et al., 2008; De Leeneer et al., 2008;
Dobrowolski et al., 2007a, 2007b, 2005; Erali et al., 2008; Laurie
and George, 2009]. HRM profiling identifies the presence of
sequence variants by deviation in the shape of a post-PCR melting
profile using probe-based and amplicon-based strategies [Dobro-
wolski et al., 2007a, 2007b; Montgomery et al., 2007; Zhou et al.,
2004]. Recently, incorporation of ‘‘melt calibration’’ to amplicon-
based genotyping enabled the resolution of alternative homo-
zygotes with basepair-neutral changes (e.g., C4G variant that
OFFICIAL JOURNAL
www.hgvs.org
& 2009 WILEY-LISS, INC.
Additional Supporting Information may be found in the online version of this article.
Correspondence to: Steven Dobrowolski Idaho Technology, 390 Wakara Way,
Salt Lake City, Utah, 84108. E-mail: steven_dobrowolski@idahotech.com

inverts the basepair C:G to G:C) despite nearest neighbor pairs
prediction indicating that no melt temperature (Tm) difference
exists between alternative homozygotes [Gundry et al., 2008;
Liew et al., 2004].
Presented herein are means to assess mitochondrial sequence
variants using two permutations of melt profiling. Comprehensive
assessment of the mitochondrial genome (excluding the hypervari-
able displacement loop region) utilizes 36 overlapping PCR
products that function with a common amplification protocol.
Selected fragments from the comprehensive assessment panel
identified heteroplasmic variants in 33 patient specimens and the
complete comprehensive assessment panel was also used to identify
mtDNA DNA sequence variants in 125 specimens from healthy
donors. Melting-based specific genotyping uses a calibrated, short-
amplicon approach to identify the MELAS mutation m.3243A4G,
the MERRF mutation m.8344A4G, and the m.1555A4G variant
associated with aminoglycoside-induced hearing loss. Amplicon-
based genotyping is uncomplicated yet sensitive to identify
heteroplasmic mutations present at and below a 10% allele fraction.
Materials and Methods
DNA Specimens
Anonymously collected DNA specimens from healthy donors
were used in assay development, and subsequently 125 of these
were screened for sequence variants. The protocol to collect and
deidentify these specimens was approved by the Institutional
Review Board of the University of Utah. Patient specimens were
obtained by collaboration with several laboratories and these
specimens were collected with oversight of the respective
institutions’ Ethics Boards.
Design, Development, and Validation of Assays
The mitochondrial reference sequence (GenBank RefSeq
AC_000021.2) was utilized in primer design with the LightScanner
Primer Design software package (Idaho Technology, Salt Lake
City, UT). Primers were designed with theoretical Tm values of
591C to 621C. The actual range of Tms over which the primers
would effectively function was determined by temperature-
gradient PCR using a range of annealing temperatures from
601C to 721C. Products of gradient PCR were assessed by gel
electrophoresis. The majority of primers showed efficacy at an
annealing temperature of 661C, demonstrated by producing a
robust single product of the expected size. Primer pairs that did
not initially function at 661C were modified (e.g., shifting of
placement, addition of nucleotides) to generate robust and
specific product with a 661C annealing temperature. The identity
of PCR products was confirmed by DNA sequence analysis.
Primers used for comprehensive mtDNA analysis have M13 DNA
sequencing tails (see Supp. Table S1) added to their 50 end to
facilitate follow-on DNA sequence analysis. Supporting Table S2
displays primers used to genotype the m.1555A4G, m.3243A4G,
and m.8344A4G mutations. The final PCR reaction cocktail for
the genotyping assays includes double-strand DNA melting
calibrators described by Gundry et al. [2008]. The standard PCR
reaction cocktail is 1PCR buffer with 2mM MgCl2 (Idaho
Technology, Salt Lake City, Utah), 1 LCGreen Plus dye (Idaho
Technology), 200mM each dNTP, 1 unit of Klentaq polymerase
(AB Peptides, St. Louis, MO) in conjugation with Anti-Taq
Monoclonal Antibody (eEnzyme, Gaithersburg, MD), and 10 ng
template DNA. Primers are included at the concentrations
indicated in Supp. Tables S1 and S2. The PCR protocol for all
assays is 2 minutes at 941C followed by 45 cycles of 941C for 30
seconds and 661C for 30 seconds. All PCR reactions are 10 ml
under a 20 ml mineral oil overlay.
HRM Profiling
Post-PCR melt profiling was performed in the 96-well formatted
LightScanner instrument. In each assessment, one or two samples
whose sequence matches that of the mitochondrial DNA reference
sequence were used as the standard to define the wild-type melting
profile. Data was normalized, temperature shifted, and converted to
difference plots as described [Wittwer et al., 2003].
The specific genotyping tests were transitioned over a
temperature window of 501C to 951C. Melt calibration was
performed according to Gundry et al. [2008] and analyzed with
LightScanner software version 2.0 Call-IT (v2.0.0.1331; Idaho
Technology) using the melt calibration module.
Secondary Assessment by Coamplification
During the screening of mtDNA from 125 healthy donors,
samples that generated a wild-type melting profile were reassessed
by coamplification with reference DNA. The coamplification
utilizes 5 ng test specimen and 5 ng of the reference DNA whose
sequence matches the consensus mtDNA sequence (GenBank
RefSeq AC_000021.2). Coamplification will force the formation of
heteroduplex molecules to reveal the presence of a sequence
variant that did not deviate from the melting profile by its own
accord in first pass analysis. The coamplification procedure used
in this study is essentially that used when assessing the X-linked
ornithine transcarbamylase gene in hemizygous male patient
specimens [Dobrowolski et al., 2007b]. After coamplification, melt
profiling is performed as described above. Samples that produced
a deviant melting profile in a coamplification assay were collected
for DNA sequence analysis.
DNA Sequence Analysis
Samples screened with the mtDNAvariant–scanning panel whose
melting profile deviated from that of the controls (both first-pass
analysis and coamplification) were recovered from the melt
profiling plate for DNA sequence analysis. PCR product was
purified using Qiagen MinElute PCR purification kit (Qiagen,
Hilden, Germany) according to the manufacturer’s instructions.
DNA sequencing was primed with the following sequencing
primers: forward strand sequence 50-GTAAAACGACGGCCAGT-
30 and reverse strand sequence 50-CAGGAAACAGCTATGAC-30.
DNA sequencing was performed by the ARUP Reference Labora-
tory DNA sequencing core facility using ABI capillary sequencing
and dye termination chemistry v.1.1 (Applied Biosystems, Foster
City, CA). Sequence data was interpreted using Mutation Surveyor
software v.2.41 (Softgenetics, State College, PA).
Results
Assessing Patient Specimens with the mtDNA Variant
Scanning Fragments
Thirty-three DNA specimens from patients having mitochon-
drial disease and characterized mtDNA mutations were obtained by
the courtesy of collaborating laboratories. Contributing laboratories
identified the mutations and, where appropriate, determined the
892 HUMAN MUTATION, Vol. 30, No. 6, 891–898, 2009