by Jeffrey M. Perkel
Most classic cases of human genetic disease result from small mutations in single genes—sickle cell anemia, Huntington’s Disease and cystic fibrosis, for instance. Other diseases, though, stem from larger-scale genomic events—deletion, duplication, inversion or insertion of tracts of nucleotide real estate so large they actually can be seen through a microscope. The classic example is Down syndrome, which results from the presence of a third copy of chromosome 21.
Oftentimes, though, disease results from much smaller chromosomal abnormalities—“copy-number variations” that, though still quite large, fall below the resolution of traditional microscopy-based cytogenetic techniques. Enter molecular cytogenetics. Marrying modern molecular biology with classical cytogenetics, molecular cytogenetics techniques enable researchers to identify chromosomal abnormalities on a genome scale with finer resolution and at greater frequency than ever before.
Classic cytogenetics
Until recently, cytogenetics meant one of a few microscopy techniques—karyotyping, fluorescence in situ hybridization (FISH) [italicize ‘in situ’?] and multicolor FISH (chromosome painting)—all of which are relatively low-resolution. Karyotyping is a method in which metaphase chromosomes are treated with Giemsa stain, which produces a characteristic black-and-white banding pattern on each chromosome. An experienced cytogeneticist can read these patterns to determine whether specific segments of the genome are duplicated, deleted or have moved to different locations. But the resolution of the technique is relatively low, on the order of five to 10 Mb in size, and the data are somewhat subjective.
FISH is a molecular approach in which fluorescently labeled nucleic acid probes are hybridized to a chromosomal spread to determine the copy number and chromosomal location of the DNA that is complementary to the probe. Using FISH, researchers can determine, for instance, whether a specific chromosomal region has been deleted or moved to a different chromosomal location, possibly in a translocation event. Yet FISH can only be applied to one, or a handful of probes, at a time. And it, too, suffers from relatively low spatial resolution. Multicolor FISH is a hybrid technique in which entire metaphase chromosomes are “painted” different colors by hybridization with fluorescently labeled, chromosome-specific probes.
According to Lisa Shaffer, founder of molecular cytogenetics firm Signature Genomics and now chief scientific officer for molecular diagnostics at PerkinElmer (which bought Signature Genomics in 2010), genetic diagnostics laboratories using karyotyping typically have a resolution of about 10 Mb, meaning they can only detect chromosomal abnormalities of 10 Mb or larger. At the same time, the “diagnostic yield” for the test is a paltry 3% to 4%.
The molecular revolution: DNA microarrays and cytogenetics
“The technology [of molecular cytogenetics] has changed my life and that of my patients,” says Shaffer.
The key technology of molecular cytogenetics is the DNA microarray. By hybridizing labeled patient DNA to an array of probes spaced across the human genome as well as concentrated in target regions, researchers can identify amplified or deleted regions and pinpoint breakpoints with kilobase-level accuracy.
Shaffer still recalls spending months as an assistant professor in the early 1990s, mapping and cloning breakpoints on the short arm of chromosome 1. “We had hundreds of FISH probes, and it would take months to finish each case,” she says. Then, with the advent of microarrays in 1999 or so, her world changed. Suddenly, the same work could be accomplished overnight. Since then, she has worked to make molecular cytogenetics a tool for clinical diagnostics, first at Baylor College of Medicine and then at Signature Genomics.
According to Shaffer, molecular cytogenetics has boosted resolution and diagnostic yield by a factor of five, roughly. “Using arrays, we find abnormalities 20% of the time,” she says, which is “five times more often than with [karyotyped] chromosomes.” Of course, that also means that 80% of the time no abnormalities are detected—not because they don’t exist, but because they are too small to be detected with the technology.
Still, says James Clough, vice president of clinical and genomic solutions at Oxford Gene Technology (OGT), a molecular cytogenetics tools and service provider, those benefits of increased diagnostic yield and resolution, coupled with faster turnaround times and greater objectivity, have caused a rapid uptake in adoption of cytogenetic microarrays. Now, he says, the technique “is recommended as a front-line test in a large number of countries,” including the United States.
Indeed, the American College of Medical Genetics in 2010 recommended the use of cytogenetic microarrays “as a first-line test in the initial postnatal evaluation of individuals with . . . A. Multiple anomalies not specific to a well-delineated genetic syndrome; B. Apparently nonsyndromic DD/ID [developmental delay/intellectual disability]; C. Autism spectrum disorders.” [1]
Arrays for clinical applications
Signature Genomics developed an oligonucleotide microarray design that is manufactured by Roche NimbleGen. According to Emily Rorem, international director of product management for clinical applications at Roche NimbleGen, these arrays (called CGX cytogenetics microarrays) are available in three different formats, containing either three (CGX-3), six (CGX-6) or 12 (CGX-12) subarrays with 135,000 probes each (meaning up to 12 samples may be tested in parallel).
The CGX array, Rorem says, offers high-resolution coverage of the entire human genome, with both “backbone” probes—genetic mile markers spaced evenly across the genome—and probes specifically targeted to disease-associated regions, including microdeletion/microduplication, subtelomeric and pericentromeric regions. In this case, the genes are those known to be associated with certain “constitutional” or developmental disorders: birth defects, autism, developmental delay and so on. Thus, although the CGX design has whole-genome coverage, it is not recommended for use in, for instance, cancer studies. “It’s possible there are certain genes on this array that might not have the appropriate high-density coverage [for such a study],” Rorem explains.
The CGX probe design is such that the arrays provide an average resolution of 50 kilobases within targeted regions and 135 kilobases for backbone regions, Rorem says. (That’s more than two orders of magnitude better than karyotyping.) But those probes are entirely nonpolymorphic—that is, they can detect copy-number variations but not single-nucleotide polymorphisms (SNPs). Other vendors, though, include both probe types on their arrays.
More arrays for clinical applications: The power of combining CNVs and SNPs
According to Clough, SNP probes, when combined on the same array with copy-number variation (CNV) probes, provide the ability to determine not only if a segment of DNA is present (and in what amount), but if the patient still retains both the maternal and paternal copies of that segment. Inheritance of two copies of an identical segment of DNA from a single chromosome of one of the parents is called uniparental disomy and results in a loss of heterozygosity (LOH).
LOH is often seen in cases of consanguinity (inbreeding), but it can also occur during cancer progression. In either event, it can cause problems, Clough says, with imprinted genes or in cases where the surviving copy contains a mutation. “LOH means that you don’t have the natural correction for mutations that you would have if you were heterozygous,” he says.
OGT’s CytoSure cytogenetics microarrays, synthesized by Agilent Technologies, are available in a range of formats to suit both resolution and sample requirements. Further, they can resolve chromosomal abnormalities as small as a few kilobases in size, Clough says. The CytoSure ISCA UDP array, for instance, developed in partnership with the International Standard for Cytogenomic Array Consortium, contains some 6,200 SNP and 137,000 CNV probes, spaced every 25,000 bases, on average.
Affymetrix also uses a combined CNV/SNP array format. Richard Shippy, director of strategic product marketing for cytogenetic applications at Affymetrix, says the Affymetrix Cytoscan microarray, based on the company’s industry-leading 6.8 million-probe density, can interrogate some 2.6 million genomic features. About two-thirds of those probes, some 4.5 million, are SNP-focused, targeting 750,000 polymorphisms. The rest report on the status of 1.9 million nonpolymorphic features.
According to Shippy, the advantage of using both SNP and CNV probes is about more than just LOH; SNPs are generally found outside of genes. Thus, a SNP-only approach would tend to bias coverage somewhat, he says.
The Affymetrix Cytoscan array provides resolution “below 50 kb, and often at the individual exon level,” Shippy says. Probes are spaced about every 500 bases for genes of known significance to constitutional abnormalities or cancer, and every 800 bases for other genes. The remaining backbone probes are spaced about every 1,700 bases. According to Shippy, the design was developed to meet the needs of cytogeneticists, who requested both improved genetic coverage and a simplified assay relative to the company’s other GeneChip microarray products.
Other firms developing products for molecular cytogenetics include Agilent Technologies and Illumina.
Embarking upon molecular cytogenetics in your lab
Whichever platform you choose, if you’re looking to delve into molecular cytogenetics, Shaffer recommends at least partnering with an experienced cytogeneticist. It can be difficult to make sense of the data that emerge from a genomic microarray experiment, as everyone’s genome contains copynumber differences, many of which are benign. “I feel strongly a cytogeneticist should be doing the interpretation,” Shaffer says, “because when you see a gain or loss, you have to be able to interpret it in the context of the chromosomes.”
Shaffer further recommends avoiding off-the-shelf, genomic tiling arrays for molecular cytogenetics in favor of cytogenetics-focused arrays, at least when dealing with clinical samples, “because there’s a lot of junk in our genome which can cause interpretation problems.” Those arrays may be used, however, in initial discovery-phase studies, she adds—that is, to identify regions of interest, which then can be studied in greater sample numbers on more refined, custom arrays.
One solution, says Clough, is to consider using a service provider, at least until you can get proficient with using the arrays and interpreting the data yourself. OGT can handle up to 2,000 arrays per week, says Clough, and has the experience to help a new lab get up and running. Between the arrays themselves, the time and the preciousness of patient samples, he says, researchers can waste an awful lot of money spinning their wheels as they get up to speed on microarrays.
“Make sure all your ducks are in a row, because they are relatively expensive ducks,” Clough says.
Reference
[1] Manning, M, and Hudgins, L, “Array-based technology and recommendations for utilization in medical genetics practice for detection of chromosomal abnormalities,” Genet Med 12:742-5, 2010.