eScholarship provides open access, scholarly publishing
services to the University of California and delivers a dynamic
research platform to scholars worldwide.
University of California
Deletion mapping in Xp21 for patients with complex glycerol kinase deficiency using SNP mapping
Stanczak, Christopher M
Nelson, Stanley F
McCabe, Edward R B
in http://www.wiley.com/WileyCDA/WileyTitle/productCd-HUMU.html. (http://
Cells, Cultured, Chromosome Deletion, Chromosome Mapping, Chromosomes, Human, X,
Female, Gene Dosage, Genetic Screening, Glycerol Kinase, Humans, Male, Microarray Analysis,
Polymorphism, Single Nucleotide, Random Allocation, Syndrome
Infantile or complex glycerol kinase deficiency (cGKD) is a contiguous gene deletion syndrome
caused by a loss of GK (MIM# 300474), along with its neighboring genes, Duchenne muscular
dystrophy (DMD; MIM# 300377) and/or Nuclear Receptor Subfamily 0, Group B, Member 1
(NR0B1; MIM# 300473). Patients with cGKD present with glyceroluria and hyperglycerolemia
in association with DMD and/or adrenal hypoplasia congenita (AHC). The purpose of these
investigations was to determine whether the Affymetrix GeneChip Mapping Array (SNP chip)
could be utilized to detect and map breakpoints in patients with cGKD. Genomic DNAs from
several primary lymphoblastoid cell lines from patients with cGKD were analyzed on the
Affymetrix platform. The Affymetrix SNP chip is a high-density oligonucleotide array that allows
a standardized, parallel interrogation of thousands of SNPs across the entire genome (except
for the Y chromosome). Analysis of the array features' hybridization intensities enabled clear
delineation of the patient deletions with a high degree of confidence. Many of these patient
deletions had been mapped by PCR and their breakpoints confirmed by sequencing. This study
demonstrates the utility of the Affymetrix Mapping GeneChips for molecular cytogenetic analysis,
beyond the SNP genotyping for which the arrays were initially designed. With one out of 160 live
births (approximately 25,000 U.S. neonates annually) reported to have cytogenetic disorders, we
envision a significant need for such a standardized platform to carry out rapid, high-throughput,
genomic analyses for molecular cytogenetics applications.
HUMAN MUTATION 0,1^8,2006
Deletion Mapping in Xp21 for Patients
With Complex Glycerol Kinase Deficiency
Using SNP Mapping Arrays
Christopher M. Stanczak,1Zugen Chen,1Yao-Hua Zhang,2Stanley F. Nelson,1
and Edward R.B. McCabe1–4?
1Department of Human Genetics, David Geffen School of Medicine at the University of California, Los Angeles (UCLA), Los Angeles,
California;2Department of Pediatrics, David Geffen School of Medicine at UCLA, Los Angeles, California;3Molecular Biology Institute, UCLA,
Los Angeles, California;4Mattel Children’s Hospital at UCLA, Los Angeles, California
Communicated by Nancy Spinner
Infantile or complex glycerol kinase deficiency (cGKD) is a contiguous gene deletion syndrome caused by a loss
of GK (MIM] 300474), along with its neighboring genes, Duchenne muscular dystrophy (DMD; MIM]
300377) and/or Nuclear Receptor Subfamily 0, Group B, Member 1 (NR0B1; MIM] 300473). Patients with
cGKD present with glyceroluria and hyperglycerolemia in association with DMD and/or adrenal hypoplasia
congenita (AHC). The purpose of these investigations was to determine whether the Affymetrix GeneChip
Mapping Array (SNP chip) could be utilized to detect and map breakpoints in patients with cGKD. Genomic
DNAs from several primary lymphoblastoid cell lines from patients with cGKD were analyzed on the Affymetrix
platform. The Affymetrix SNP chip is a high-density oligonucleotide array that allows a standardized, parallel
interrogation of thousands of SNPs across the entire genome (except for the Y chromosome). Analysis of the
array features’ hybridization intensities enabled clear delineation of the patient deletions with a high degree
of confidence. Many of these patient deletions had been mapped by PCR and their breakpoints confirmed
by sequencing. This study demonstrates the utility of the Affymetrix Mapping GeneChips for molecular
cytogenetic analysis, beyond the SNP genotyping for which the arrays were initially designed. With one out
of 160 live births (approximately 25,000 U.S. neonates annually) reported to have cytogenetic disorders,
we envision a significant need for such a standardized platform to carry out rapid, high-throughput, genomic
analyses for molecular cytogenetics applications. Hum Mutat 0, 1–8, 2006.Published 2006 Wiley-Liss, Inc.y
KEY WORDS: contiguous gene syndrome; deletion mapping array; glycerol kinase deficiency; molecular cytogenetics;
SNP mapping array
Glycerol kinase deficiency (GKD) is an inborn error of
metabolism that was initially described by McCabe et al. ;
affected individuals manifest hyperglycerolemia and glyceroluria.
GKD is an X-linked recessive disorder that results from mutation
of the glycerol kinase (GK; MIM] 300474) gene [McCabe, 2001].
GK catalyzes the phosphorylation of glycerol to glycerol
3-phosphate in an ATP-dependent reaction.
The GK gene has been mapped to Xp21 and lies in a region
containing the following genes: (Xpter) . . . IL1RAPL1 (MIM]
300206) – MAGE cluster (MIM]s 300097, 300152, 300153) –
NROB1 (MIM] 300473) – GK – DMD (MIM] 300377) . . .
(Xcen) [McCabe, 2001]. Two testis-specific genes, GK2 (MIM]
137028) and GK3, are located at 4q13 and 4q32 and two
pseudogenes are located at 1q41 and Xq23 [Sargent et al., 1994].
GK was originally described with 19 exons [Sjarif et al., 1998] and
two alternately spliced exons were subsequently reported, bringing
the total to 21 exons [Sargent et al., 1994]. The genomic locus is
approximately 50kb, and the major isoform mRNA is 2,581bp and
encodes a 553–amino acid protein [Sargent et al., 1994].
GKD presents with diverse phenotypes and is categorized into
three different clinical presentations [McCabe, 2001]:
1. Infantile or complex GKD (cGKD), the most common form
of GK mutation, is a contiguous gene syndrome caused by
a microdeletion of GK and its neighboring genes, Duchenne
Published online inWiley InterScience (www.interscience.wiley.com).
yThis article is a US Government work and, as such, is in the public
domain in theUnited States of America.
The Supplementary Material referred to in this article can be
accessed at http://www.interscience.wiley.com/jpages/1059-7794/
Received 20 March 2006; accepted revised manuscript 15
?Correspondence to: Edward R.B. McCabe, MD, PhD, Department
of Pediatrics, David Ge¡en School of Medicine at UCLA, 10833
LeConte Ave., 22-412 MDCC, Los Angeles, CA 90095-1752.
Grant sponsor: National Institute of Child Health and Human
Development;Grant number: RO1HD 22563.
PUBLISHED 2006 WILEY-LISS, INC.
muscular dystrophy (DMD) and/or Nuclear Receptor Subfamily
0, Group B, Member 1 (NR0B1), the causative gene for adrenal
hypoplasia congenita (AHC). Patients present with hypergly-
cerolemia and glyceroluria, along with DMD and/or AHC.
2. Juvenile or symptomatic GKD results from point mutations in
GK that cause episodic vomiting, acidemia, and central nervous
system (CNS) deterioration, including stupor and coma.
3. Adult or asymptomatic GKD, also resulting from point
mutations, is a benign form detected incidentally with
The severity of symptomatic GKD varies greatly and does not
seem to be correlated with enzyme activity [Dipple et al., 2001;
Dipple and McCabe, 2000]. We have postulated that modifier
genes and multiple functions of the same protein (moonlighting
activities) may contribute to the lack of a simple genotype–
phenotype relationship [Sriram et al., 2005].
We have collected lymphoblastoid cell lines (LCLs) from
patients with GKD, including a number with cGKD. Many of
these patients’ deletions have been mapped by PCR of sequence-
tagged sites (STSs) [Zhang et al., 2004].
We hypothesized that it would be possible to utilize Affymetrix
Gene Chips (Affymetrix, Santa Clara, CA; http://affymetrix.com)
for copy number estimation by comparing the intensities for each
SNP relative to a control. Affymetrix 10K GeneChips (SNP
Chips) contain 11,555 SNPs covering the entire human genome,
except the Y chromosome. The median physical distance between
SNPs is about 105kb, and the average distance is about 210kb.
To be able to achieve a 490% call rate with 99.96%
reproducibility, each SNP has 40 different 25-bp oligonucleotides
tiled on the slide, each with variations of perfect matches,
mismatches, and flanking sequences (per Affymetrix). The
Affymetrix 10K GeneChip contains multiple markers in Xp21
that may be deleted in cGKD. With the 100K GeneChip, the
coverage increases approximately 10-fold.
This study was undertaken to determine if the Affymetrix
SNP Chip would be applicable to mapping the deletions in
patients with cGKD.
DNA Application and SNP Chip Analysis
Following the Affymetrix protocol for the 10K and 100K human
Mapping GeneChips and reagents, we sought to map the break-
points of the deletions in 12 cGKD patient cell lines. We grew the
patients’ lymphoblastoid cell lines and harvested and purified the
DNA using commercially available kits and reagents. Each
patient’s genomic DNA was then subjected to restriction digestion
with XbaI (New England Biolabs, Ipswich, MA; www.neb.com/
nebecomm) for 2hr and the enzyme was deactivated by heating to
701C for 20 minutes. Next, proprietary adaptor sequences
(Affymetrix) containing XbaI compatible ends and primer binding
sites were ligated to the digested genomic DNA fragments using
T4 DNA ligase (New England Biolabs) at 161C for 2hr and heat
inactivated at 701C for 20 minutes. A single-primer PCR
amplification was performed for 35 cycles with an annealing
temperature of 591C using AmpliTaq Gold (Applied Biosystems,
Foster City, CA; www.appliedbiosystems.com). The products were
purified using QIAquick PCR Purification spin columns (Qiagen,
Valencia, CA; www1.qiagen.com), quantitated by spectrophoto-
metry and qualitatively checked by agarose gel electrophoresis.
Each sample was fragmented by DNase I (Affymetrix) at 371C for
30 minutes, again checked by gel electrophoresis, and end-labeled
using terminal deoxynucleotidyl transferase (Affymetrix) for 2hr at
371C. The samples were then added to hybridization buffer,
denatured at 951C for 10 minutes, and hybridized to the Affymetrix
SNP Chip for 18hr at 481C. Finally, the SNP Chips were washed and
stained with streptavidin (Pierce, Rockford, IL; www.piercenet.com),
biotinylated goat anti-streptavidin antibody (Vector Laboratories,
Burlingame, CA; www.vectorlabs.com), and streptavidin phycoery-
thrin (Molecular Probes, Carlsbad, CA). The arrays were scanned in
the UCLA Microarray Core Facility and analyzed using GCOS and
GDAS Affymetrix software and copy number was determined using
the Chromosome Copy Number Analysis Tool (Affymetrix).
The probability that a stretch of n ‘‘No Calls’’ would happen
by chance was determined to be (1 – the average call rate)n
[De Veaux et al., 2004]. Normalized copy number in the patients’
deleted regions was calculated by setting the average X
chromosome copy number to 1 (or 2 in the case of the females)
by dividing the average copy number within the deletion by the
average copy number of the X chromosome(s) for each patient.
The data on controls were obtained from experimental results
on four patients’ with no known cytogenetic abnormality on their
10K SNP Chip
Deletions in the DNA from these 12 patients were identified by
two methods: 1) the strings of Affymetrix ‘‘No Call’’ genotypes, and
2) low perfect match–minus mismatch (PM-MM) intensity readings.
The PM-MM model corrects for spurious nonspecific binding by
subtracting the average oligonucleotide intensity for the mismatches
from the average element intensity of the matching sequences.
With the PM-MM method negative relative intensity readings may
be observed due to background subtraction. The deletions were
unique in each patient (Supplementary Tables S1 and S2; available
online at http://www.interscience.wiley.com/jpages/1059-7794/supp-
mat) and were reproducible when DNA from the same patient was
tested in duplicate (Supplementary Fig. S1; Patient 13155-1).
Mapping relative PM-MM intensity across a patient’s X
chromosome clearly indicated the location of the deleted segment
by its low intensity (Fig. 1). The variability in PM-MM intensities
may correlate with the binding affinities of the SNP sequences. In
the deleted regions, the PM-MM intensities were variable but were
uniformly diminished relative to the nondeleted regions (Table 1).
The genomic locations of the ‘‘No Call’’ genotypes were
distributed randomly in the normal controls and in the nondeleted
regions of our patient’s genomic DNAs. Analysis of the genotype
call percentages for the SNPs on the array and of the SNPs within
each deletion showed that the average call percents declined from
an overall average of 94.19 to 4.92% in the region of the deletions
(Supplementary Fig. S1). The fact that the call percentage in the
unaffected controls remained at 94% indicated that the decreased
call frequency in the deleted regions of the patients was due to the
deletion and not to a defect in the SNP features on the chips
in the region. The higher call rate (28.6%) in Patient 3122-1
(Supplementary Fig. S1) resulted from the fact that he had a small
deletion that spanned only seven SNPs, and therefore two spurious
calls resulted in a relatively increased call percentage. The results
on the DNA from Patient 3122-1 illustrated the usefulness of
the combined approach of ‘‘No Calls’’ and PM-MM intensity in
2 HUMAN MUTATION 0,1^8,2006
Human Mutation DOI 10.1002/humu
determining the robustness of the data at this location. The
average PM-MM intensity in the deleted region of Patient 3122-1
was –140.5 vs. 5065.4 for the entire array, and the intensities
of the two called SNPs were also low (144.7 and –109.8).
The probability of getting seven ‘‘No Calls’’ in a row by chance
was calculated to be 2?10–9. The average Affymetrix PM-MM
intensities of all the SNPs on the array varied in each patient
sample (Supplementary Fig. S2), but within the deleted region of
each of our patients, the average PM-MM intensity was markedly
diminished. This ablation of signal was evident in each of the
patient cell lines, but not in the controls.
By comparing the ratio of SNP intensities between the patients
and reference controls, we estimated copy number. Representative
data from our determination of copy number along the X
chromosome (Supplementary Fig. S3) clearly indicated the deleted
region in Xp21. Patient 13155-1’s calculated copy number average
Position along X Chromosome
FIGURE 1. Patient 17219-1 deletion detection by SNP intensity.The PM-MM SNP intensity, calculated by subtracting the average
intensity of the oligonucleotides that had a mismatched base from the average intensity of the oligonucleotides that correspond to
the perfect sequence,was graphed relative to theXchromosome physical position (bp) of the SNPs. Despiteconsiderablevariation,
the deleted region in Xp21 was clearly recognized by the consecutive low PM-MM intensities. Negative readings occurred where
background was greater than signal intensity. Bars extending above 0 PM-MM intensity indicate presence of oligonucleotides,
while barsextending below 0 PM-MM intensity indicate absenceofoligonucleotides.
TABLE 1. Summary ofCallPercentages andIntensities inEachof thePatients, and theNormalMale andFemaleControls
‘‘No calls’’ Call %Averageintensity‘‘No calls’’ Call %Averageintensity
HUMAN MUTATION 0,1^8,20063
Human Mutation DOI 10.1002/humu
for his X chromosome was 1.17, while the average calculated copy
number within his deletion was 0.40. (The same patient’s DNA
analyzed on the 100K SNP array yielded a average calculated copy
number of 1.16 for his X Chromosome and 0.28 with the deleted
region.) The summary of each of the patients’ and controls’ copy
number determinations is provided in Table 2, and the mean
normalized copy number was 0.38 with a range of 0.19–0.58.
The data on controls were obtained from experimental results
on four patients with no known cytogenetic abnormality on their
X chromosome (DiGeorge patients).
By mapping the deleted SNPs using the UCSC Genome Browser
(www.genome.UCSC.edu), we were able to determine the
minimum and maximum sizes of the deletions in the patients’
cell lines (Fig. 2). The breakpoints of the deletions were between
the last present (maximum) and first absent (minimum) SNP .
These breakpoints confirmed previously performed breakpoint
sequencing results in the patients for whom these data were
available (Zhang Y-H, Ho JC, Huang B-L, McCabe LL, August G,
Kern I, Morris M, McCabe ERB, unpublished results).
We compared individual SNP intensities telomeric, within, and
centromeric of Xp21 GK to other GK SNPs (Supplementary Fig.
S4). The reduction of Xp21 GK specifically (SNP_A-1512978,
SNP_A-1515463, and SNP_A-1507601) and not of either the
testis-specific GK2 (SNP_A-1510447 and SNP_A-1511576)
intensities or the GK3 (SNP_A-1515104 and SNP_A-1512733)
genes on chromosome 4 indicated the specificity of the SNP
hybridization. The high intensity (5222.5) of SNP_A-1512978
in Patient 4512-1 confirmed that the deletion breakpoint was
between this SNP and SNP_A-1515463.
100K SNP Chip
While our 10K SNP Chip results succeeded in detecting
deletions in our GKD patient samples, the location of the
deletion breakpoints could not be precisely determined (Fig. 2).
In the course of these investigations Affymetrix released
a 100K SNP Chip with 10 times the number of SNPs. Our
successful results with the 10K SNP array suggested that we
should attempt to use the 100K SNP Chip to take advantage
of the greater resolution offered by more SNPs in the vicinity
of our patients’ deletions to determine precisely the location
of their breakpoints. We mapped denser SNP coverage of the
100K SNP Chip with 75 SNPs in Xp21.2 (UCSC Genome
Our finer mapping results allowed us to confirm the patients’
deletions and validate our 10K results. The 100K SNP chip
contained a sufficient number of SNPs to permit us to vary the
bin size used to determine the global copy number. In the
representative analysis (Supplementary Fig. S5), increasing the bin
size resulted in reduced noise in the data, allowing us to eliminate
false deletions caused by spurious low SNP intensities, but at the
expense of clearly defined deletion breakpoints. Conversely, using
a smaller bin size allowed us to map the patients’ deletion
endpoints with accuracy.
The confidence we could assign to our copy number
determination also assisted in detecting true deletions and their
endpoints. The copy number and associated probability was
calculated for Patient 13155-1 using copy number analysis tool
(CNAT) for each SNP along the X chromosome (Fig. 3). The
deleted region was clearly indicated by the reduction in copy
number and P-value. By including the associated probability in the
analysis of the calculated copy number, the number of false-
positive deletions was reduced and more refined deletion mapping
We demonstrated that SNP arrays can be utilized to detect X
chromosomal deletions responsible for cGKD in affected males.
The mean normalized X chromosomal copy number was
approximately two in females, approximately one in males and
o0.4 in the regions of the cGKD deletions. The observation that
signal intensity remained detectable in these patients was not
due to mRNAs responsible for transcription of other GK isoforms
and remains undetermined at this time.
The first microarrays were spotted with cDNA to monitor
gene expression levels. Pollack et al.  attempted to utilize
these arrays for copy number analysis. However, this technique
had many shortfalls. The microarray was not sufficiently sensitive
to detect deletions due to the individual features having homology
between paralogous genes and lacking intervening sequences.
Additionally, the genomic representation was limited to only ‘‘gene
rich’’ regions, and, most importantly, copy number did not
correlate well with hybridization strength.
While the comparative genomic hybridization (CGH) techni-
que has several limitations, it provides more specific breakpoint
delineation, and is less labor-intensive and locus-specific than
fluorescence in situ hybridization (FISH). The most significant
problem for array CGH, using large clones, such as BAC clones,
for its arranged features, is high background noise. The relatively
low signal to noise ratio results from several factors, including
imperfect spotting and cross-hybridization. Despite the use of
copious amounts of cot-1 DNAto suppress nonspecific hybridization,
large genomic clones contain many common repeats that lead to
spurious binding and background [Kooperberg et al., 2002].
Furthermore, the large clones limit the spatial resolution needed to
map smaller deletions accurately [Ren et al., 2005]. We therefore
sought to utilize another array methodology to improve the detection
sensitivity and resolution of molecular cytogenetic analyses.
As the potential for this technology became more apparent,
several BAC and cosmid clone arrays were created to detect
TABLE 2. AverageCalculatedCopyNumber for theEntire
XChromosome andin theRegions of theDeletions in the
NormalMale andFemaleControls and thecGKDPatients?
4HUMAN MUTATION 0,1^8,2006
Human Mutation DOI 10.1002/humu
FIGURE 2. PatientdeletionsmappedusingSNPlocationsandtheUCSCGenomeBrowser.Thepatients’deletedgenomicregionswere
represented by black bars.The minimum and maximum size of the deletions were determined by the locations of the last present
and the ¢rst deleted SNPs (shown above blackbars), respectively.The genes in this region are shown below the blackbars.The data
compared well with breakpoint sequencing results in patients for whom these data were available (ZhangY-H, HoJC, Huang B-L,
McCabe LL, August G, Kern I, Morris M, McCabe ERB, unpublished results) and validated the use of the SNP Chip for molecular
HUMAN MUTATION 0,1^8,20065
Human Mutation DOI 10.1002/humu
cytogenetic changes [Ishkanian et al., 2004]. As one of the major
difficulties was background noise due to cross-hybridization to
common repeats in the clones, an array was spotted with repeat-
free, nonredundant DNA [Tanabe et al., 2003]. While this was a
more successful approach, it was costly and labor-intensive to
PCR-amplify thousands of fragments, and, more importantly, it was
limited in the number of genomic regions to which it could be
applied. In an effort to increase the specificity and resolution
of array CGH with genomic clones Lucito et al.  designed
an array using 70-mer oligonucleotides.
Copy number estimation with Affymetrix SNP Chips offers an
attractive alternative to spotted arrays that contain BACs or
cosmids. The SNP Chips contain a proprietary array of 25-mer
oligos in quartets of perfectly matching sequence (PM) and
sequence that had a single basepair mismatch (MM) for each
DNA strand (Affymetrix). The shorter sequence and ability to
subtract cross-hybridization theoretically should provide improved
signal-to-noise ratio and greater mapping resolution. We demon-
strate that it is possible to detect gene loss in patients with cGKD
using SNP chips. Since the arrays were designed for genotyping,
the output string of ‘‘No Call’’ reading was anticipated in the
regions of deleted DNA. But such a result could also occur due to
high background or a poorly hybridizing sequence surrounding
the SNP and resulting in low signal. The purpose of a quantifiable,
more direct method to detect deletions was to look for regions
with low PM-MM intensity readings. Our data indicate excellent
concordance of low PM-MM intensity readings with the ‘‘No Call’’
readings within previously identified deletions.
Normalized SNP intensity readings can immediately highlight a
region of deleted DNA, but there still is considerable variation of
intensity levels. By standardizing against control intensity at each
SNP in determining copy number, the noise is greatly reduced
since the variability of each SNP’s PM-MM intensities is
normalized relative to its feature’s unique binding affinity for that
particular sequence of DNA. The smallest deletion that we
detected was about 0.5 MB (Patient 4512-1), but the detection
limit varies greatly depending on the density of SNPs in the region
of the deletion. Unfortunately for SNP array CGH, most deletion
breakpoints are in areas of low copy number repeats that provide
a substrate for recombination [Stankiewicz and Lupski, 2002] and
this is also true for cGKD (Zhang Y-H, Ho JC, Huang B-L,
McCabe LL, August G, Kern I, Morris M, McCabe ERB,
unpublished results). We therefore tried a higher resolution array
for some of our samples to validate these results and to determine
more precisely the patient’s breakpoints. The concurrent analysis
of copy number and associated P-value, as well as breakpoint
sequencing (Zhang Y-H, Ho JC, Huang B-L, McCabe LL, August
G, Kern I, Morris M, McCabe ERB, unpublished results) provided
additional evidence that these deletions were accurately mapped.
The SNP Chip–based method suffers from the inability to
detect balanced translocations or inversions. As in all array CGH
techniques, if genomic content is present in another genomic
region it still can bind to the array elements. In other words,
physical position cannot be determined by array CGH, only the
absence or presence of DNA in the genome. Despite this
limitation, SNP-based arrays have been utilized to detect genomic
aberrations [Zhou et al., 2004]. SNPs have been used to detect
regions of loss of heterozygosity [Mei et al., 2000]. Affymetrix
platforms have been used to detect copy number changes in cancer
and other acquired cytogenetic abnormalities as well as constitu-
FIGURE 3. Copy number and probabilities of copy number determinations for Patient 13155-1. Copy number was calculated using
CNATand plotted light (gray) for each SNP along the X chromosome.The associated probability was graphed (darker gray) below.
ThedeletedregionwasclearlyindicatedbythereductionincopynumberandP-value. [Color ¢gurecanbeviewedintheonlineissue,
whichis available at www.interscience.wiley.com.]
6HUMAN MUTATION 0,1^8,2006
Human Mutation DOI 10.1002/humu
tional and congenital abnormalities, and copy-number polymorph-
isms [Huang et al., 2004; Wong et al., 2004; Slater et al., 2005;
Zhao et al., 2005; Ming et al., 2006; Wirtenberger et al., 2006].
Many feel that microarray-based molecular cytogenetic ap-
proaches will replace more traditional chromosome visualization
methods over the next decade [Shaffer and Bejjani, 2004; Smeets,
2004]. It is anticipated that these newer methods will identify an
increased proportion of patients who will have subtle molecular
cytogenetic abnormalities compared with current approaches
[Hwang et al., 2005; Lapierre et al., 1998; Oostlander et al.,
2004]. The number of individuals with cytogenetic disorders
recognized by the current methods are quite impressive: 1 in 160
live births (approximately 25,000 U.S. neonates annually);
approximately 2% of prenatal diagnoses in women 35 years or
older; and 50% of all first trimester spontaneous abortions
[Nussbaum et al., 2001]. Indications for ordering a chromosome
analysis include infertility and cancer, indicating a significant
opportunity for the appropriate microarray platform.
The method we describe permits molecular cytogenetic analyses
on DNA without requiring intact chromosomes for karyotype or
FISH analysis. Since DNA can be analyzed in newborn screening
dried blood specimens [McCabe et al., 1987; Jinks et al., 1989;
Bhardwaj et al., 2003], this SNP Chip–based approach would
permit newborn screening for cytogenetic disorders if this
approach passes rigorous validation processes, including compar-
isons with FISH analysis or quantitative PCR, and if the cost can
be reduced. Such a program would require considerable discussion
of the ethical and social issues, however, before it could be
In conclusion, the Affymetrix SNP Chip can be utilized for the
detection of genomic deletions in patient cell lines with cGKD.
Our work provides proof of principle that the SNP chip can be
used for molecular cytogenetic analysis beyond the SNP genotyp-
ing for which the arrays were initially designed. We envision the
use of such a platform for rapid, high-throughput, genomic analysis
for molecular cytogenetic applications in the future.
Bhardwaj U, Zhang Y-H, McCabe ERB. 2003. Neonatal hemoglobinopathy
screening: Molecular genetic technologies. Mol Genet Metab 80:
De Veaux RD, Vellman PF, Bock DE. 2004. Introductory statistics. Boston:
Pearson Addison Wesley. 648p.
Dipple KM, McCabe ERB. 2000. Phenotypes of patients with ‘‘simple’’
Mendelian disorders are complex traits: thresholds, modifiers and
systems dynamics. Am J Hum Genet 66:1729–1735.
Dipple KM, Zhang Y-H, Huang B-L, McCabe LL, Dallongeville J, Inokuchi
T, Kimura M, Marx HJ, Roederer GO, Shih V, Yamaguchi S, Yoshida I,
McCabe ERB. 2001. Glycerol kinase deficiency: evidence for complexity
in a single gene disorder. Hum Genet 109:55–62.
Huang J, Wen W, Zhang J, Liu G, Bignell G, Stratton M, Futreal A, Wooster R,
Jones K, Shapero M. 2004. Whole genome DNA copy number changes
identified by high density olgionucleotide arrays. Hum Genome 13:
Hwang KS, Pearson MA, Stankiewicz P , Lennon PA, Cooper ML, Wu J, Ou
Z, Cai WW, Patel A, Cheung SW. 2005. Cryptic unbalanced
translocation t(17;18)(p13.2;q22.3) identified by subtelomeric FISH
and defined by array-based comparative genomic hybridization in a
patient with mental retardation and dysmorphic features. Am J Med
Genet A 137:88–93.
Ishkanian AS, Malloff CA, Watson SK, DeLeeuw RJ, Chi B, Coe BP ,
Snijders A, Albertson DG, Pinkel D, Marra MA, Ling V, MacAulay C,
Lam WL. 2004. A tiling resolution DNA microarray with complete
coverage of the human genome. Nat Genet 36:299–303.
Jinks DC, Minter M, Tarver DA, Vanderford M, Hejtmancik JF, McCabe
ERB. 1989. Molecular genetic diagnosis of sickle cell disease using dried
blood specimens from newborn screening blotters. Hum Genet 81:
Kooperberg C, Fazzio TG, Delrw JJ, Tsukiyama T. 2002. Improved back-
ground correction for spotted DNA microarrays. J Comput Biol 9:55–66.
Lapierre JM, Cacheux V, Collot N, Da Silva F, Hervy N, Rivet D, Romana
S, Wiss J, Benzaken B, Aurias A, Tachdjian G. 1998. Comparison of
comparative genomic hybridization with conventional karyotype and
classical fluorescence in situ hybridization for prenatal and postnatal
diagnosis of unbalanced chromosome abnormalities. Ann Genet 41:
Lucito R, Healy J, Alexander J, Reiner A, Esposito D, Chi M, Rodgers L,
Brady A, Sebat J, Troge J, West JA, Rostan S, Nguyen KC, Powers S, Ye
KQ, Olshen A, Venkatraman E, Norton L, Wigler M. 2003.
Representational oligonucleotide microarray analysis: a high-resolution
method to detect genome copy number variation. Genome Res 13:
McCabe ERB, Fennessey PV, Guggenheim MA, Miles BS, Bullen WW,
Sceats DJ, Goodman SI. 1977. Human glycerol kinase deficiency with
hyperglycerolemia and glyceroluria. Biochem Biophys Res Commun 78:
McCabe ERB, Huang SZ, Seltzer WK, Law ML. 1987. DNA microextrac-
tion from dried blood spots on filter paper blotters: potential applications
to newborn screening. Hum Genet 75:213–216.
McCabe ERB. 2001. Disorders of glycerol metabolism. In: Scriver C,
Beaudet A, Sly W, Valle D, editors. Metabolic and molecular basis of
inherited diseases, 8th edition. New York: McGraw-Hill. p 2217–2237.
Mei R, Galipeau PC, Prass C, Berno A, Ghandour G, Patil N, Wolff RK,
Chee MS, Reid BJ, Lockhart DJ. 2000. Genome-wide detection of allelic
imbalance using human SNPs and high-density DNA arrays. Genome
Ming J, Geiger E, James A, Ciprero K, Nimmakayalu M, Zhang Y, Huang
A, Vaddi M, Rappaport E, Zackai E, Shaik T. 2006. Rapid detection of
submicroscopic chromosomal rearrangements in children with multiple
congenital anomalies using high density oligonucleotide arrays. Hum
Nussbaum RL, McInnes RR, Willard HF. 2001. Thompson & Thompson
genetics in medicine, 6th edition. Philadelphia: W.B. Saunders. 444p.
Oostlander AE, Meijer GA, Ylstra B. 2004. Microarray-based comparative
genomic hybridization and its applications in human genetics. Clin
Pollack JR, Perou CM, Alizadeh AA, Eisen MB, Pergamenschikov A,
Williams CF, Jeffrey SS, Botstein D, Brown PO. 1999. Genome-wide
analysis of DNA copy-number changes using cDNA microarrays.
Nat Genet 23:41–46.
Ren H, Francis W, Boys A, Chueh AC, Wong N, La P , Wong LH, Ryan J,
Slater HR, Choo KH. 2005. BAC-based PCR fragment microarray:
high-resolution detection of chromosomal deletion and duplication
breakpoints. Hum Mutat 25:476–482.
Sargent CA, Young C, Marsh S, Ferguson-Smith MA, Affara NA. 1994.
The glycerol kinase gene family: structure of the Xp gene, and related
intronless retroposons. Hum Mol Genet 3:1317–1324.
Shaffer LG, Bejjani BA. 2004. A cytogeneticist’s perspective on genomic
microarrays. Hum Reprod Update 10:221–226.
Sjarif DR, Sinke RJ, Duran M, Beemer FA, Kleijer WJ, Ploos van Amstel
JK, Poll-The BT. 1998. Clinical heterogeneity and novel mutations in the
glycerol kinase gene in three families with isolated glycerol kinase
deficiency. J Med Genet 35:650–656.
Slater H, Bailey D, Ren H, Cao M, Bell K, Nasioulas S, Henke R, Choo
KH, Kennedy G. 2005. High resolution identification of chromosomal
abnormalities using oligonucleotide arrays containing 116,204 SNPs.
Am J Hum Genet 77:709–726.
Smeets DF. 2004. Historical prospective of human cytogenetics: from
microscope to microarray. Clin Biochem 37:439–446.
Sriram G, Martinez JA, McCabe ERB, Liao JC, Dipple KM. 2005. Single-
gene disorders: what role could moonlighting enzymes play? Am J Hum
HUMAN MUTATION 0,1^8,20067
Human Mutation DOI 10.1002/humu
Stankiewicz P , Lupski JR. 2002. Genome architecture, rearrangements and Download full-text
genomic disorders. Trends Genet 18:74–82.
Tanabe C, Aoyagi K, Sakiyama T, Kohno T, Yanagitani N, Akimoto S,
Sakamoto M, Sakamoto H, Yokota J, Ohki M, Terada M, Yoshida T,
Sasaki H. 2003. Evaluation of a whole-genome amplification method
based on adaptor-ligation PCR of randomly sheared genomic DNA.
Genes Chromosomes Cancer 38:168–176.
Wirtenberger M, Hemminki K, Burnwinkel B. 2006. Identification of
frequent chromosome copy-number polymorphisims by use of high-
resolution single-nucleotide-polymorphism arrays. Am J Hum Genet 78:
Wong KK, Tasng YT, Shen J, Cheng RS, Chang YM, Man TK, Lau CC.
2004. Allelic imbalance analysis by high-density single-nucleotide
polymorphic allele (SNP) array with whole genome amplified DNA.
Nucleic Acids Res 32:e69.
Zhang Y-H, Huang B-L, Niakan KK, McCabe LL, McCabe ERB, Dipple
KM. 2004. IL1RAPL1 is associated with mental retardation in patients
with complex glycerol kinase deficiency who have deletions extending
telomeric of DAX1. Hum Mut 24:273.
Zhao X, Weir BA, LaFramboise T, Lin M, Beroukhim R, Garraway L,
Beheshti J, Lee JC, Naoki K, Richards WG, Sugarbaker D, Chen F,
Rubin MA, Janne PA, Girard L, Minna J, Christiani D, Li C, Sellers WR,
Meyerson M. 2005. Homozygous deletions and chromosome amplifica-
tions in human lung carcinomas revealed by single nucleotide
polymorphism array analysis. Cancer Res 65:5561–5570.
Zhou X, Mok SC, Chen Z, Li Y, Wong DT. 2004. Concurrent analysis of
loss of heterozygosity (LOH) and copy number abnormality (CNA) for
oral premalignancy progression using the Affymetrix 10K SNP mapping
array. Hum Genet 115:327–330.
8HUMAN MUTATION 0,1^8,2006
Human Mutation DOI 10.1002/humu