Clinical Chemistry 46: 1464-1470, 2000;
(Clinical Chemistry. 2000;46:1464-1470.)
© 2000 American Association for Clinical Chemistry, Inc.
Screening for Single-Nucleotide Polymorphisms Using Branch Migration Inhibition in PCR-amplified DNA
Alla Lishanski1
1
Advanced Diagnostics Group, Dade Behring Inc., San Jose, CA 95135.
 |
Abstract
|
|---|
Background: New methods are required for the exploration of
the human genome by discovering sequence variations. This study
evaluated the performance of a new method for screening a large number
of samples for several DNA polymorphisms.
Methods: We used a homogeneous method based on inhibition
of spontaneous branch migration by any sequence difference between two
molecules of PCR-amplified DNA. A set of four PCR primers is required:
a forward primer, either biotinylated or labeled with digoxigenin, and
two reverse primers that share a priming domain but have different
"tail" sequences at their 5' ends. After PCR amplification,
denaturation and reannealing of the single DNA strands produce doubly
labeled cruciform structures, which dissociate by strand exchange. The
presence of two different alleles in a sample causes complete
inhibition of dissociation, and the association of biotin and
digoxigenin is homogeneously detected using luminescent oxygen
channeling immunoassay.
Results: The 90 samples of the Human Variation Panel (Coriell
Cell Repositories) were screened for nine known single-nucleotide
polymorphisms (SNPs) and one 5-bp deletion. The average
signal-to-background ratio varied from 
10 to 20. The frequency of
the predominant allele for different SNPs varied from 51% to 88%
overall. For some SNPs, it varied among the nine ethnic groups, e.g.,
25–85% (average, 51%) for one SNP. The average heterozygosity varied
from 0.17 to 0.54 and as much as 0.2–0.9 (average, 0.54) for one of
the SNPs.
Conclusion: The method allows simple and rapid screening of a
large number of samples for the presence of multiple alleles.
|
Introduction
|
|---|
Detection of single-nucleotide polymorphisms
(SNPs)1
is central to the emerging discipline of pharmacogenomics,
both for research applications such as genomics and for clinical
applications such as patient stratification. These applications require
the ability to screen very large numbers of samples quickly and
inexpensively. We recently described branch migration inhibition (BMI),
a homogeneous mutation detection method based on spontaneous BMI in
PCR-amplified DNA (1). This method makes use of the fact
that spontaneous strand exchange is inhibited by any sequence
difference between two DNA molecules (2)(3)(4). It offers a
simple approach to detecting all possible polymorphisms in an amplicon
in a single reaction. This study describes the application of BMI to
screening the Human Variation Panel for several known polymorphisms.
|
Materials and Methods
|
|---|
dna and primers
Genomic DNA samples were purchased from Coriell Cell
Repositories. Two panels were used: M08PDR (eight individual DNA
samples from the Human Polymorphism Discovery Resource Panel), and the
Human Variation Panel. The latter consisted of nine subpanels: HD01,
HD02, HD03, HD04, HD09, HD06, HD07, HD08, and HD27 (5).
The SNP information can be found in the National Center for
Biotechnology Information (NCBI) SNP database (6). Nine SNPs
located in chromosome 21 were chosen (NCBI assay identification nos.
3989, 4141, 4212, 4213, 4214, 4215, 4216, 4030, and 4031,
respectively). They were submitted to the database by Drs. M. Olivier
and D. R. Cox (Stanford Human Genome Center, Stanford, CA).
The primer sequences for these SNPs were the same as indicated in the
NCBI SNP database above.
The primer sequences for amplifying the region that contains a 5-bp
deletion were 5'-TCA.AAT.TGT.TGG.CTA.ACA.CCA-3' (forward) and
5'-TAC.TGG.TGT.ACC.GTC.CAT.GT-3' (reverse).
All forward primers were 5' end-labeled with biotin and
digoxigenin, respectively. The same tail sequences, t1 and t2, were
added to the 5' ends of all reverse primers: t1,
5'-ACC.ATG.CTC.GAG.ATT.ACG.AG-3'; t2, 5'-GAT.CCT.AGG.CCT.CAC.GTA.TT-3'.
The universal labeled primer sequence was
5'-TGC.CAC.CTG.ACG.TCT.AAG.AA-3'. This sequence constituted the 5'
domains of all of the adapter primers, whose 3' domains were the
sequences of the primers listed in the NCBI SNP database
(6).
The oligonucleotides were purchased from Operon Technologies or from
Oligos Etc.
pcr and branch migration
PCR amplifications were carried out using T3 thermocyclers
(Biometra). Thirty-five PCR cycles were performed with 30 s of
denaturation at 94 °C, 1 min of reannealing at 62 °C, and 1 min
of extension at 72 °C. The thermocycling was preceded by a 10-min
incubation at 95 °C to activate the AmpliTaq
GoldTM DNA polymerase and was followed by 2 min
of denaturation at 95 °C and a 30-min incubation at 65 °C
(reannealing and branch migration). The reaction mixtures (10 µL)
contained 10 ng of genomic DNA, 0.25 U of AmpliTaq Gold DNA
polymerase (PE Bioscience), 200 µmol/L each dNTP, 62.5–250 nmol/L
each primer, and 0.5 µg/mL ethidium bromide in the commercial Taq
buffer (10 mmol/L Tris-HCl, pH 8.3, 50 mmol/L KCl, 1.5 mmol/L
MgCl2, 0.1 g/L gelatin).
The ratio of universal-to-adapter forward primer was 19:1 (total
concentration, 125 nmol/L), and 40 PCR cycles were performed as
described above.
The presence of amplified product was verified by the increased
fluorescence of ethidium bromide in the presence of double-stranded DNA
by placing PCR tubes on an ultraviolet transilluminator.
signal detection
For luminescent oxygen channeling immunoassay
(LOCITM) detection, 2-µL aliquots of
amplified material subjected to the BMI conditions were mixed with
50-µL suspensions containing 1.25 µg of streptavidin-coated donor
beads and 0.625 µg of anti-digoxigenin monoclonal antibody-coated
acceptor beads (7). The mixtures were incubated for 30 min
at 37 °C and irradiated at 680 nm; the chemiluminescent emission was
measured using a custom-made reader that accommodated eight-tube strips
(the automated AlphaQuestTM microplate reader is
now commercially available from Packard BioScience).
The above protocol detects only heterozygotes. To detect homozygotes
for the minor allele, 1 µL of each sample was mixed with 5 µL of
the PCR buffer containing 1 µL of an amplicon that corresponds to DNA
homozygous for the predominant allele. The mixture was denatured and
subjected to branch migration (2 min at 95 °C, followed by 30 min at
65 °C). LOCI beads were added to the reaction mixture, and the
signal was read as above.
|
Results
|
|---|
experimental procedure
A brief outline of the experimental protocol is illustrated
in Fig. 1
[for a more detailed description of BMI, see Lishanski et al.
(1)]. Genomic DNA is amplified using a set of four
specifically designed primers. The two forward primers have the same
sequence, but one is 5' end-labeled with biotin, and the other is
labeled with digoxigenin. The two reverse primers have the same priming
sequence but two different tail sequences, t1 and t2, each 20
nucleotides in length, that are not complementary to the genomic target
and become incorporated into duplex PCR products upon amplification.
PCR is followed by heat denaturation and reannealing of the single
strands to eventually form, among other structures, a doubly labeled
four-stranded cruciform DNA structure. When the two arms of this
structure are identical (no mutation or SNP is present), strand
exchange via spontaneous branch migration leads to its complete
dissociation into two duplex molecules, and no signal is observed. When
there is a mutation or a SNP present, branch migration in the presence
of Mg2+ is inhibited, and the cruciform structure
remains unresolved. The stable association of biotin and digoxigenin in
this structure can be detected by standard ELISA or homogeneously
detected by LOCI (7) as shown schematically in Fig. 2
.
View larger version (13K):
[in this window]
[in a new window]
|
Figure 1. The principle of polymorphism detection by BMI.
Only one of the two detectable cruciform structures is shown. All
intermediate steps that lead to its formation are omitted and are
described in detail in Ref. (1). B, biotin;
D, digoxigenin; T1 and T2,
sequences complementary to t1 and t2,
respectively. The block that a single base-pair difference presents for
spontaneous branch migration is symbolized by two open
squares.
|
|
View larger version (12K):
[in this window]
[in a new window]
|
Figure 2. Detection of cruciform DNA structures by LOCI.
Sensitizer, streptavidin-coated latex bead loaded with a
photosensitizer; Acceptor, anti-digoxigenin monoclonal
antibody-coated (Anti-Dig Mab) latex bead loaded with a
chemiluminescent olefin (5). The doubly labeled cruciform
structure links the two beads together. Irradiation at 680 nm
(h 680 nm) produces singlet oxygen
(1O2), which reacts with
the olefin to produce a chemiluminescent emission (h
550 nm) from the acceptor particle. B,
biotin; D, digoxigenin.
|
|
detection of several known dna polymorphisms by bmi
The smallest subset of the National Human Genome Research
Institute’s DNA Polymorphism Discovery Resource (8) was
screened for nine known SNPs in chromosome 21 using BMI. The panel
consists of eight individual genomic DNA samples from different groups
representative of the genetic diversity found in the US population. All
information about the SNPs can be found in the NCBI SNP database
(6). The primer sequences and the PCR conditions were
exactly the same as indicated in the NCBI SNP database, except that the
tail sequences t1 and t2, which are necessary for performing BMI were
added to the 5' ends of the reverse primers. These tail sequences were
the same for all the amplicons studied. The forward primers were 5'
end-labeled with biotin or digoxigenin, respectively. A 5-bp deletion
in the Down syndrome region of chromosome 21 (9) was also
included in the panel. The lengths of the 10 studied amplicons varied
from 122 to 245 bp.
The results of BMI screening for 10 known DNA polymorphisms are
shown in Fig. 3
. Heterozygous samples were immediately revealed as positives by
virtue of two different alleles present in the same PCR reaction.
Homozygotes for both alternative alleles appeared as negatives and were
indistinguishable at this stage. To reveal homozygotes for the minor
allele (Fig. 3
, filled symbols), all amplified samples were mixed with
an equal amount of a sample homozygous for the predominant allele, the
mixtures were denatured, and branch migration was repeated. All
previously negative samples that became positive after this procedure
were identified as homozygotes for respective minor alleles. The
average heterozygote-to-average homozygote signal ratio in the first
screen (no reference added) varied from 9 (SNP 4212) to 22 (5-bp
deletion). Only minimal optimization that consisted of increasing the
primer concentration was required to achieve this degree of
discrimination. The correctness of genotype identification by BMI was
confirmed by sequencing of selected samples.
View larger version (23K):
[in this window]
[in a new window]
|
Figure 3. Detection of 11 known polymorphisms by BMI.
The signal is normalized for each polymorphism separately, with average
signal for heterozygotes taken as 1. The horizontal
lines are at 3 SD above the mean negative. Open
symbols correspond to BMI performed without a reference.
Homozygotes for minor alleles (filled symbols) are shown
twice: the lower symbols correspond to BMI performed
without a reference, and the upper symbols correspond to
BMI performed with a reference. Indicated at the bottom
of Fig. 3
are (top to bottom) NCBI assay identification
numbers, alternative alleles for each SNP, and amplicon lengths
(including the tails).
|
|
screening of the human variation panel for known dna polymorphisms
After ascertaining the adequacy of BMI analysis for the chosen
polymorphisms, we screened a larger panel, the Coriell Cell
Repositories Human Variation Panel (5) for all of them as
described above. Fig. 4
shows an example of screening of this panel for one of the SNPs
above (SNP 4213). To avoid false negatives attributable to
amplification failure, the presence of amplified DNA was confirmed by
increased fluorescence of the ethidium bromide that was included in the
PCR reaction mixture (10). In many cases, the samples that
failed to amplify could be pinpointed easily because they generated
abnormally low (3- to 10-fold lower than true negatives) LOCI signals.
View larger version (23K):
[in this window]
[in a new window]
|
Figure 4. Screening of the Human Variation Panel for SNP 4213.
The signal was normalized as described in the legend for Fig. 3
. Each
vertical section corresponds to one of the nine ethnic
groups. Control samples (right) are positive or negative
by sequencing.
|
|
Allele frequencies and heterozygosities were calculated for
each polymorphism. The results are summarized in Tables
1and
2, respectively. For some SNPs (e.g., 3989 and 4031), the two
alternative alleles were present in approximately equal amounts
overall, but in certain groups one of the alleles predominated. For
other SNPs (e.g., 4216 and 4030), one of the alleles was clearly
predominant in all of the groups (Table 1
). However, the relatively
small number of samples in each group did not allow us to reach any
statistically significant conclusions about allele distribution. The
overall percentage of heterozygotes varied from 17% to 54% for
different SNPs, and as much as from 20% to 90% (average, 54%; SNP
3989) among the nine groups (Table 2
).
using a generic labeled primer for bmi
To avoid making expensive labeled primers for each amplicon to be
analyzed, we developed a different protocol (Fig. 5
). The forward primer was a 9:1–20:1 mixture of a universal
labeled primer and a sequence-specific adapter primer. The universal
primer sequence was derived from a bacterial cloning vector. The 3'
proximal domain of the adapter primer was complementary to the target
genomic DNA, and its 5' proximal domain was identical to the universal
primer. In the first few rounds of PCR, the adapter primer generated
enough amplicon to serve as a target for the universal primer. The
final PCR product was suitably labeled for subsequent LOCI detection.
The use of a universal primer for similar purposes has been reported by
others (11).
View larger version (7K):
[in this window]
[in a new window]
|
Figure 5. Use of a universal labeled primer in BMI.
Forward primer is 20:1 mixture of a universal primer [50%
biotinylated (B) and 50% labeled with digoxigenin
(D)] and an adapter primer. The 5' domain of the
adapter is identical to the universal primer. A single PCR reaction
produces tagged amplicons, which serve as substrates for BMI.
|
|
The results for four different SNPs using the universal primer approach
are shown in Table 3
. The discrimination was comparable to that achieved using the
respective sequence-specific labeled forward primers (Fig. 3
).
|
Discussion
|
|---|
The aim of this study was to evaluate the performance of BMI in a
relatively high-throughput application in terms of its robustness and
accuracy. The additional goal was to simplify the method and make it
more user-friendly. Ideally, one should be able to use the existing PCR
primers without the necessity to redesign or chemically modify them
other than to add the tail sequences and the tags. To find whether this
degree of simplicity is possible for BMI, 10 arbitrarily chosen DNA
sequences that are known to contain polymorphic sites were screened.
The published primers that have been designed without considering BMI
requirements were used, and PCR conditions were exactly as recommended
by the SNP submitter. None of the 10 amplicons presented serious
problems in terms of high background that may be caused by nonspecific
priming. Most of them required no optimization at all, and for those
that initially produced rather low signal-to-background ratios
(3–5), a simple doubling or quadrupling of the primer concentration
improved this ratio to 10–20. The priming specificity afforded by
the AmpliTaq Gold DNA polymerase with its built-in hot-start feature
was sufficient to perform BMI in the cases examined in this study
without recourse to the other hot-start methods described previously
for Pfu polymerase (1). Implementation of a
hot-start is essential because conventional AmpliTaq polymerase
generates unacceptably high background. Surprisingly, the hot-start
provided by the Platinum Pfx enzyme (Life Technologies) has not yet
been sufficient in our hands.
One of the potential problems in automatable procedures such as BMI can
be false-negative samples that result from PCR failures. Therefore, it
is important to verify the presence of a sufficient amount of amplified
product in each reaction. Because gel electrophoresis is undesirable,
it is fortunate that for most of the tested amplicons failed
amplification manifested itself by an abnormally low LOCI signal
(severalfold lower than the true negatives). For those amplicons in
which this effect was absent or less pronounced, the nonappearance of
increased fluorescence of ethidium bromide in the presence of
double-stranded amplified DNA was indicative of failure to amplify. The
incidence of such failures was typically <2%.
One of the limitations of BMI, apart from its inability to identify
polymorphisms or determine their exact location within an amplicon, is
that it cannot distinguish between the homozygotes for two alternative
alleles. It detects only heterozygotes, and if finding that more than
one allele of a sequence exists is not deemed sufficient, an additional
step is required, which consists of adding a reference amplicon that
corresponds to one of the two possible homozygotes to each amplified
sample and repeating denaturation and branch migration. This step
allows identification of homozygotes for the second allele. The
inability of BMI to differentiate between the two types of homozygotes
is less of a disadvantage if BMI is used for SNP discovery.
There exists a possibility of multiplexing for BMI. Each amplicon in a
multiplex mixture would have to be given a different pair of tails to
prevent the amplicons from signal-generating interactions with each
other. An additional technical challenge would be the necessity to use
a different pair of tags corresponding to bead pairs with
non-overlapping spectral characteristics for each amplicon. We have not
yet explored the multiplexing potential of BMI.
The results of screening of the Human Variation Panel by BMI hint at
possible differences in the allele frequency distribution between
various ethnic groups for some of the biallelic SNPs. However, the
relatively small number of samples included in this study did not allow
us to reach any statistically significant conclusions (nor was such
comparison intended). The main conclusion of this study is that BMI is
suitable for rapid prescreening of large numbers of samples for the
presence of DNA polymorphisms with subsequent characterization by
sequencing of the positives that are revealed. The possibility of
reducing the cost by using a single generic tagged primer enhances the
potential of the method for SNP discovery.
|
Acknowledgments
|
|---|
I wish to thank Drs. Sam Rose and Alan Dafforn for their support,
many helpful discussions, and critical reading of the manuscript; Drs.
Ted Ullman and Nurith Kurn for their input in the evolution of BMI; Dr.
Nurith Kurn for contributing to the idea of universal primer; and Yen
Ping Liu (Dade Behring Inc., San Jose, CA) for the gift of LOCI
beads.
|
Footnotes
|
|---|
Address for correspondence: Dade Behring Inc., PO Box 49013, San Jose,
CA 95161-9013. Fax 408-239-2707; e-mail
alla_lishanski{at}dadebehring.com
1 Nonstandard abbreviations: SNP, single-nucleotide polymorphism; BMI, branch migration inhibition; NCBI, National Center for Biotechnology Information; and LOCI, luminescent oxygen channeling immunoassay. 
|
References
|
|---|
-
Lishanski A, Kurn N, Ullman EF. Branch migration inhibition in PCR-amplified DNA: homogeneous mutation detection. Nucleic Acids Res 2000;28:e42.[Abstract/Free Full Text]
-
Panyutin IG, Hsieh P. Formation of a single base mismatch impedes spontaneous DNA branch migration. J Mol Biol 1993;230:413-424.[Web of Science][Medline]
[Order article via Infotrieve]
-
Hsieh P, Panyutin IG. DNA branch migration. In: Eckstein
F, Lilley DMJ, eds. Nucleic acids and molecular biology. Berlin:
Springer-Verlag, 1993;9:42–65..
-
Biswas I, Yamamoto A, Hsieh P. Branch migration through DNA sequence heterology. J Mol Biol 1998;279:795-806.[Web of Science][Medline]
[Order article via Infotrieve]
-
Coriell Cell Repositories (Human Variation
Panels). http://locus.umdnj.edu/nigms/nigms_cgi/panel.cgi?id =
1 (accessed May 2000)..
-
NCBI. A database of single nucleotide polymorphisms.
http://www.ncbi.nlm.nih.gov/SNP/index.html (assessed May 2000)..
-
Ullman EF, Kirakossian H, Singh S, Wu ZP, Irvin BR, Pease JS, et al. Luminescent oxygen channeling immunoassay: measurement of particle binding kinetics by chemiluminescence. Proc Natl Acad Sci U S A 1994;91:5426-5430.[Abstract/Free Full Text]
-
Collins FS, Brooks LD, Chakravati A. A DNA polymorphism discovery resource for research on human genetic variation. Genome Res 1998;8:1229-1231.[Free Full Text]
-
Lishanski A, Bell M, Scherrer S, Rine J. Development of
polymorphic markers in chromosome 21 using double-strand conformation
polymorphisms (DSCP) [Abstract]. Cold Spring Harbor Symposium on
Genome Mapping and Sequencing, May 11–15, 1994, Cold Spring Harbor,
NY: 148..
-
Higuchi R, Dollinger G, Walsh PS, Griffith R. Simultaneous amplification and detection of specific DNA sequences. Biotechnology 1992;10:413-417.[Medline]
[Order article via Infotrieve]
-
Nazarenko IA, Bhatnagar SK, Hohman RJ. A closed tube format for amplification and detection of DNA based on energy transfer. Nucleic Acids Res 1997;25:2516-2521.[Abstract/Free Full Text]
The following articles in journals at HighWire Press have cited this article:
|
|
|
|
 
Q. Yang, A. Lishanski, W. Yang, S. Hatcher, H. Seet, and J. P. Gregg
Allele-Specific Holliday Junction Formation: A New Mechanism of Allelic Discrimination for SNP Scoring
Genome Res.,
July 1, 2003;
13(7):
1754 - 1764.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
