May 2010: Volume 36, Number 5
Array-based Cytogenetic Testing
Poised for a Transformative Role in the Lab and Clinical Genetics
Bassem A. Bejjani, MD, and Marvin R. Natowicz, MD, PhD
The gain or loss of genetic material causes many human genetic disorders. Today, we also know that most cancers acquire DNA copy number changes during the life of the patient or the tumor. These genomic gains and losses may result in congenital anomalies, dysmorphic features, developmental disabilities, or other neurological abnormalities, as well as tumor initiation or progression. Until recently, traditional chromosome analysis was the primary tool for the cytogenetic assessment of patients with these clinical concerns.
For decades, cytogeneticists have been limited to visually examining whole genomes with a microscope, a technique known as chromosome analysis or karyotyping. In the 1970s and 1980s, the development and application of molecular diagnostic methods such as PCR, Southern blots, and fluorescence in situ hybridization (FISH) enabled researchers to make many important advances in genetics, including clinical cytogenetics.
The recent development of DNA array methodologies, such as microarray-based comparative genomic hybridization (aCGH) arrays and single-nucleotide polymorphism (SNP) arrays, now allows cytogeneticists to simultaneously examine all loci in the genome with unprecedented resolution. These technical innovations and their clinical applications are transforming the practice of genetic diagnostics in clinical labs and are already playing a key role in revolutionizing the knowledge base and clinical practice in medical genetics.
Here we describe the principles and practice of array-based cytogenetic testing, including the strengths and limitations of these methods.
A Brief History of Clinical Cytogenetics
To understand the importance of DNA microarray technology to clinical practice, it is helpful to look at these advances in the context of the evolution of clinical cytogenetics. The era of modern clinical cytogenetics is often said to have begun in 1956 with the observation by Joe-Hin Tijo and Johan Levan that normal human cells contain 46 chromosomes.
Within a few years of this breakthrough, researchers described several numerical chromosome anomalies associated with specific abnormal phenotypes. In 1959, Jérôme Lejeune and colleagues discovered that Down syndrome was caused by trisomy 21, and in 1960 scientists identified the chromosomal bases of trisomy 13 and trisomy 18. Following these discoveries, Torbjörn Caspersson and associates reported in 1970 that applying quinacrine mustard to human chromosomes caused them to exhibit light and dark bands, giving rise to the technique known as chromosome banding. This discovery made it possible for the first time to determine the cytogenetic breakpoints associated with structural chromosomal abnormalities, such as chromosomal translocations or deletions. Within a few years, Jorge J.Yunis described a method to study human chromosomes with still higher resolution, enabling increasingly subtle chromosomal abnormalities to be discovered.
However, both routine and high-resolution karyotyping are time consuming, labor intensive, and require cultured cells. Typically, complete analysis takes 2 weeks, from sample preparation to analysis and interpretation. Furthermore, for most clinical applications, karyotyping cannot identify genetic imbalances that are smaller than 5 million basepairs (Mb).
The introduction of molecular cytogenetic methods in the 1980s and 1990s, including FISH and aCGH, enabled scientists to examine the genome at still greater resolution. FISH circumvents some of the limitations of traditional cytogenetics, making it possible to determine the number and location of specific DNA sequences, both in metaphase chromosomes and in interphase nuclei. It also significantly simplifies preparation and evaluation of samples.
The resolution of FISH also far surpasses that of conventional banding and enables detection of chromosomal deletions and some duplications that are not visible by routine cytogenetics. Applications of FISH in the clinical setting include screening for aneuploidy in prenatal specimens, searching for microdeletions in contiguous gene syndromes, evaluating rearrangements of the subtelomeric regions in non-specific mental retardation, and defining gene rearrangements in leukemias and lymphomas. Efficient use of FISH, however, requires that the patient either exhibits features consistent with a clinically recognizable syndrome with a known chromosomal etiology or demonstrates an abnormal karyogram requiring further molecular characterization, such as a marker chromosome. Because the technique uses individual probes, it reveals DNA gains, losses, or rearrangements of only the probe-targeted segments but does not provide any information about the rest of the genome. In other words, FISH analyses do not detect abnormalities distinct from the genomic segments for which the probes have been designed.
Despite these limitations, single-locus and telomere FISH applications led to a rapid succession of discoveries of new clinical syndromes and established subtelomere abnormalities as an important cause of developmental delays and dysmorphic features. The fascinating history of clinical cytogenetics and the development of molecular cytogenetics are discussed in greater detail elsewhere (1,2).
More recently, scientists have developed techniques that integrate aspects of both traditional and molecular cytogenetic techniques. These techniques fall under the umbrella term chromosomal microarrays (3,4). Despite the short time that they have been available for clinical use, chromosomal microarrays have enabled clinicians to diagnose numerous subtle chromosomal abnormalities that were previously undetectable and have made discoveries of new contiguous gene deletion and duplication syndromes commonplace.
The high resolution capability of microarray analysis has also allowed researchers to find many duplications or deletions of part or all of a single gene. Such information is useful for describing new monogenic disorders and improving the understanding of previously described syndromes. Chromosomal microarray analysis has also contributed to a better understanding of the pathogenetic mechanisms of many chromosomal aberrations.
Array-based CGH Analysis
The earliest chromosomal CGHs compared the relative fluorescent signal intensities of two differentially labeled DNAs, a test and a reference, when hybridized to a reference metaphase spread. This scheme identifies relative copy number changes across the genome of the test DNA (5). In today’s aCGH, the reference metaphase chromosome spread is replaced with small segments of DNA (probes) from known genomic regions of interest that have been immobilized on a solid platform, such as a microscope slide, and arrayed in an ordered fashion as targets for analysis (6–10).
In aCGH, two genomic DNAs, a test and a reference, are fluorescently labeled with different dyes and competitively hybridized to the arrayed probes (Figure 1). To conduct the test, the laboratorian first labels DNA from a test sample, such as DNA from white blood cells, skin fibroblasts, or cells from amniocentesis, with a fluorescent dye. The normal control DNA (reference) sample is also labeled, but with a different-colored fluorescent dye. The laboratorian then denatures the two labeled DNAs and applies them to the array where they are allowed to competitively hybridize with the arrayed probes. Digital imaging systems then capture and quantify the relative fluorescence intensities of the labeled DNA probes that are hybridized to each target. To analyze the results, the laboratorian determines the fluorescence intensity ratio of the test and reference hybridization signals at distinct positions throughout the genome. These ratios define the relative copy number of sequences in the test genome compared to the normal reference genome.
Click here for Figure 1
Probes for aCGH may be either synthetic oligonucleotides (25–85 bp) designed to represent areas of interest, or larger genomic clones such as bacterial artificial chromosomes (BACs) of 80–200 kb. In either case, the probes are orders of magnitude smaller than metaphase chromosomes, and the probe size, number, and genomic distance between DNA probes determines the theoretical resolution of an array. For example, a microarray with BAC probes selected from regions 1 Mb apart on the genome will be unable to detect copy number changes in the intervening sequence, whereas an oligonucleotide microarray with coverage of the genome at 35 kb intervals will detect smaller abnormalities.
aCGH: Strengths and Limitations
aCGH has important advantages over traditional cytogenetic and molecular cytogenetic techniques. Key among these is its unbiased ability to simultaneously detect aneuploidy, deletions, duplications, or amplifications at any locus represented on the array. Basically, the technique is equivalent to thousands of FISH experiments (Figure 2). Because its resolution is limited only by the size and spacing of probes, aCGH can detect submicroscopic alterations below the level of the light microscopic resolution in addition to detecting numerical and large structural alterations evident by traditional karyotyping.
Click here for Figure 2
aCGH also allows much more accurate mapping of cytogenetic abnormalities compared to older cytogenetic methods or most applications of FISH. Current iterations of oligonucleotide-based aCGH, as well as SNP arrays, routinely detect deletions and copy number gains of >100 kb and, depending on the design of the probes and their density on the platform, even < 50 kb. The simplicity and reproducibility of this method, its ability to be automated, and no need for cell culture are other important advantages.
As is the case for any diagnostic technique, aCGH has analytic and postanalytical limitations. First, because abnormalities detected by aCGH are based on ascertainment of a net imbalance of DNA dosage between a control and patient sample, aCGH will not detect rearrangements in which there is no net imbalance of DNA dosage, such as balanced translocations and inversions. Second, even though the coverage of aCGH is genome-wide and is at several orders of resolution higher than traditional cytogenetics, the ultimate resolution for genetic disease is at the base pair level, a level of resolution not attainable by any of the current aCGH platforms. Third, although aCGH detects DNA copy number changes, seemingly identical array results may be caused by distinct molecular mechanisms. For example, a copy number gain identified by microarray analysis may be due to any one of several pathogenetic mechanisms: a duplication; an insertion; a marker chromosome; or an unbalanced translocation. FISH analysis is then often necessary to determine the specific mechanism, which allows complete interpretation and accurate genetic counseling, as well as follow-up studies (Figure 3).
Click here for Figure 3
While aCGH has many advantages, one might predict that it would also be of limited value in detecting chromosomal mosaic states, clinical situations in which there are two or more cytogenetically distinct cell lines present. Reports from several laboratories, however, unexpectedly indicated that aCGH may be superior to traditional karyotyping in detecting mosaicism, although further studies are needed (11,12).
There are also important challenges in interpreting aCGH results. Copy number variants (CNVs) are unexpectedly common in the human genome and many are without apparent clinical consequence (13). Finding a novel or rare CNV, which occurs with regular frequency in the clinical application of aCGH, requires careful evaluation of databases of pathologic and benign CNVs and, often, analysis of parental samples to determine whether the CNV is a de novo occurrence or inherited trait. Labs may have difficulty assessing the clinical significance of novel or rare CNVs if parental samples are not available. In addition, even if parental samples are available, the well-documented precedent of marked phenotypic variability and incomplete clinical penetrance of some CNVs also makes interpreting rare CNVs challenging.
SNP Arrays: Strengths and Limitations
SNP arrays are another approach to comprehensive molecular cytogenetic analysis. SNPs are DNA sequence variations in which a single nucleotide in the sequence of the genome —adenine, thymine, cytosine, or guanine—differs between individuals or between paired chromosomes in an individual. These differences create two distinct alleles, usually labeled “A” and “B”, that appear in distinct proportions in the normal population. Because an individual usually inherits one copy of each nucleotide position from each parent, a normal genotype for that nucleotide is either “AA,” “AB,” or “BB.” Researchers already have identified more than 10 million SNPs in the human genome.
Like aCGH, SNP arrays take advantage of hybridization of patient-derived, denatured, single-strand DNA to arrays, each with hundreds of thousands of probes representing unique nucleotide sequences. However, unlike CGH-based microarrays that directly compare control and reference samples, SNP-based microarrays quantitatively determine relative copy number for a region within a single genome without simultaneous comparison to a reference genome. Instead, the array compares the CNV of the test sample to a database of historic reference controls. Specialized software then aligns the SNPs in chromosomal order, generating a virtual karyotype. Whereas early SNP arrays were mainly used in research settings, particularly in the study of cancer genomics, recent improvements in algorithms for SNP data analysis now allow for clinical applications of SNP arrays.
In contrast to aCGH, SNP arrays are capable of identifying single nucleotide substitutions because the analysis uses allele-specific oligonucleotide probes. Furthermore, SNP arrays can be used for detecting loss of heterozygosity (LOH), a form of allelic imbalance that results from the loss of an allele or from an increase in copy number of one allele with the simultaneous loss of the other. In addition to copy number imbalance information, SNP arrays provide high resolution genotype information in a single experiment. For example, a single copy number deletion causes LOH in the deleted region; concordance of LOH and copy number loss for a genomic interval provide independent evidence for deletion of that region.
In addition to copy number imbalances, SNP arrays have the added advantage of being able to detect copy number neutral loss of heterozygosity (CNN-LOH) by comparing segments of homozygosity to segments of copy number loss. In some genetic conditions, CNN-LOH may be caused by uniparental disomy of a genomic interval, in which an allele or whole chromosome from one parent is missing and is replaced by a gain of the other parental chromosome.
SNP arrays also have limitations, both in the analytic and post-analytic processes. Similar to aCGH, SNP arrays cannot detect structural genomic abnormalities that are not associated with genomic imbalance, such as inversions or balanced translocations. Also, SNP arrays, like aCGH, can provide information of unclear clinical significance. Generally speaking, the likelihood of this increases with the density of SNPs spotted on the array. Detecting novel or uncommon segmental CNN-LOH afforded by the SNP arrays may also complicate interpretation of the results. Early designs of SNP arrays suffered from non-uniform distribution of SNP probes across the genome, resulting in relatively large gaps in the coverage of certain regions where SNPs were either rare or where probes had not been developed. To provide more uniform genome coverage, some recent SNP arrays use a more dense distribution of SNPs. Other recent SNP arrays use a combination of probes, with some (nonpolymorphic probes) designed to detect CNVs and others designed to detect SNPs. Work is ongoing to determine the optimal arrays for different clinical applications.
Diagnosing Cytogenetic Disease: The Future
Diagnosis of an underlying cytogenetic abnormality as the basis for an individual’s developmental disability, congenital anomaly, or other phenotypic abnormality has significance for both the affected individual and his or her family. It provides an explanation for the individual’s physical, developmental, or neurological differences, and an end to what is usually a diagnostic odyssey. More accurate discussions of the natural history and treatment of the clinical condition are often possible once the etiology of the condition is determined. Such knowledge also allows for accurate reproductive recurrence risk counseling and the possibility of antenatal diagnostic testing, if desired.
aCGH has already proven to be a powerful tool for etiological diagnosis of many abnormal phenotypes. Recent data indicate that aCGH provides 2–5 times greater diagnostic yield than traditional cytogenetics in evaluations of individuals with unexplained mental retardation, autism, or multiple congenital anomalies.
Although only a few prospective studies have been performed to evaluate the diagnostic utility of SNP arrays in suspected genetic conditions, the promising results suggest that SNP arrays will have great utility for clinical diagnostics (14–16). Despite only being in clinical use for a few years, the large number of newly discovered microduplication and microdeletion syndromes is testimony to the diagnostic power of these tools (3,4).
In the near future, we anticipate that these approaches will replace traditional cytogenetics as the first line method of chromosomal evaluation in most instances. However, the ramifications of the clinical use of these arrays extend beyond their diagnostic importance. These techniques have already revealed new complexities of the architecture of the human genome and mechanisms for genomic rearrangements. The development and application of still higher resolution arrays and newer nucleic acid sequencing methods, improved algorithms for the analysis of huge amounts of genomic data, and expanded human genomic variation databases portend a new era in the diagnosis and understanding of human genetic disease and, hopefully, the associated development of effective therapies.
- Trask BJ. Human cytogenetics: 46 chromosomes, 46 years and counting. Nat Rev Genetics 2002;3:769–778.
- Speicher MR, Carter NP. The new cytogenetics: Blurring the boundaries with molecular biology. Nat Rev Genetics 2005;6:782–792.
- Bejjani BA, Shaffer LG. Clinical utility of contemporary molecular cytogenetics. Ann Rev Genomics Hum Genet 2008;9:71–86.
- Vissers LE, de Vries BB, Veltman JA. Genomic microarrays in mental retardation: From CNV to gene, from research to diagnosis. J Med Genet 2009; doi:10.1136/jmg.2009.072942.
- Kallioniemi A, Kallioniemi OP, Sudar D, Rutovitz D, et al. Comparative genomic hybridization for molecular cytogenetic analysis of solid tumors. Science 1992;258:818–821.
- DeRisi J, Penland L, Brown PO, Bittner ML, et al. Use of a cDNA microarray to analyze gene expression patterns in human cancer. Nat Genet 1996;14:457–460.
- Lucito R, Healy J, Alexander J, Reiner A, et al. Representational oligonucleotide microarray analysis: A high-resolution method to detect genome copy number variation. Genome Res 2003;13:2291–2305.
- Pinkel D, Segraves R, Sudar D, Clark S, et al. High resolution analysis of DNA copy number variation using comparative genomic hybridization to microarrays. Nat Genet 1998;20:207–211.
- Schena M, Shalon D, Davis RW, Brown PO. Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 1995;270:467–470.
- Solinas-Toldo S, Lampel S, Stilgenbauer S, Nickolenko J, et al. Matrix-based comparative genomic hybridization: Biochips to screen for genomic imbalances. Genes Chromosomes Cancer 1997;20:399–407.
- Ballif BC, Rorem EA, Sundin K, Lincicum M, et al. Detection of low-level mosaicism by array CGH in routine diagnostic specimens. Am J Med Genet A 2006;140:2757–2767.
- Cheung SW, Shaw CA, Scott DA, Patel A, et al. Microarray-based CGH detects chromosomal mosaicism not revealed by conventional cytogenetics. Am J Med Genet A 2007;143A:1679–1686.
- Sharp AJ. Emerging themes and new challenges in defining the role of structural variation in human disease. Hum Mutat 2009;30:135–144.
- Wagenstaller J, Spranger S, Lorenz-Depiereux B, Kazmierczak B, et al. Copy-number variations measured by single-nucleotide-polymorphism oligonucleotide arrays in patients with mental retardation. Am J Hum Genet 2007;81:768–779.
- Bruno DL, Ganesamoorthy D, Schoumans J, Bankier A, et al. Detection of cryptic pathogenic copy number variations and constitutional loss of heterozygosity using high resolution SNP microarray analysis in 117 patients referred for cytogenetic analysis and impact on clinical practice. J Med Genet 2009;46:123–131.
- McMullan DJ, Bonin M, Hehir-Kwa JY, de Vries BB, et al. Molecular karyotyping of patients with unexplained mental retardation by SNP arrays: A multicenter study. Hum Mutat 2009;30:1082–1092.
Bassem A. Bejjani, MD, is director of the Medical Genetics Training Program at Providence Sacred Heart Medical Center, and co-founder and chief medical officer of Signature Genomic Laboratories, both in Spokane, Wash. Email: firstname.lastname@example.org
Marvin R. Natowicz, MD, PhD, is a clinical pathologist and medical geneticist at the Institutes of Pathology and Laboratory Medicine, Genomic Medicine, Neurology and Pediatrics at Cleveland Clinic and professor of pathology in the Cleveland Clinic Lerner College of Medicine of Case Western Reserve University, both in Cleveland, Ohio. Email: email@example.com
Disclosures: Dr. Bejjani has received salary/consultant fees from Signature Genomic Laboratories and owns stock in the company.