Josh P. Roberts has an M.A. in the history and philosophy of science, and he also went through the Ph.D. program in molecular, cellular, developmental biology, and genetics at the University of Minnesota, with dissertation research in ocular immunology.
Aberrant genetic structure and human disease have been connected since the late 1950s, when an extra chromosome was found in children with Down syndrome, and sex-chromosome abnormalities were associated with Turner and Klinefelter syndromes. These were followed almost immediately by reports of trisomies 13 and 18 in children with other developmental syndromes [1]. Since that time, more, improved cytogenetic tools have been used to discover the chromosomal abnormalities found in disease states, and in some instances even to use that correlation to find disease etiology and suggest for prognoses and treatments.
A modern cytogenetics lab carries out its search for chromosomal abnormalities by far more than just Giesma staining and fluorescence in situ hybridization (FISH) of chromosomes. Such tools “are not going to go away, because at this point for any structural variation or balanced translocation, inversion, even low-level mosaic conditions, [these] are the only methods that will give you the spatial organization of the entire genome within one nuclei,” says Alka Chaubey, director of the Greenwood Genetic Center cytogenetics lab. But for the most common abnormalities, like insertions and deletions that lead to copy number variation (CNV), the standard of care has become the microarray.
Here we look at how some of these tools and applications—including next-generation and third-generation sequencing techniques—are being used to uncover structural genetic abnormalities that may lead to disease.
Array of sunshine
A trained cytogeneticist can find a chromosomal aberration on a Giesma-stained karyotype with a resolution on the order of several megabases. Certain FISH techniques (like fiber FISH, which queries stretched, deproteinated chromatin) can resolve down to a kilobase or so, but they can only look at one or very few sequences at a time.
To achieve that kind of resolution for CNV targets on a larger scale, cytogeneticists turn to microarrays. These make use of an indexed series of short (approximately 60-mer) oligonucleotides bound to a solid surface (beads, glass slides, 96- or 384-well plates), and to which the fragmented chromosome samples will hybridize. In array-comparative genomic hybridization (aCGH), a green fluorophore-labeled test sample is mixed with a red fluorophore-labeled control sample (or vice versa), and the two compete for binding to the oligo probe. A green signal indicates a gain of copy number at that locus, a red signal points to a loss of that stretch of DNA, and a yellow signal says there is no change. Some arrays instead hybridize just a test sample to the probe, with the intensity of the signal indicating copy number. These latter arrays typically are compared to a reference database.
With Roche discontinuing its NimbleGen microarray line, three major vendors now supply the North American market: Agilent, Illumina and Affymetrix, says Kim Caple, Affymetrix’s senior vice president, clinical business, Genetic Analysis Business Unit. Oxford Gene Technology (OGT) is a provider primarily in Europe, she says. (“I would say OGT and Agilent are essentially the same, in that OGT prepares the design, but it is actually printed by Agilent,” notes Chaubey.)
There are some differences between the companies’ arrays—in terms of basic technology and the platforms required to use them, density of probes and variety of offerings—but “from a general process standpoint, the workflows are similar on all of the array products,” Caple says.
CNV microarrays can cover the entire genome, or they can be targeted to look at a specific set of genes—either an off-the-shelf or a custom-produced set. For example, Greenwood runs the Affymetrix CytoScan® arrays “if a patient comes in with autism and seizures,” says Chaubey. “The Affy platform, with 2.7 million probes, has very good coverage of the genes, of the backbone; the SNPs [single nucleotide polymorphisms] allow for identification of uniparental disomy and autosomal recessive conditions. But there is one thing that platform lacks: coverage of every exon of every gene.” So if an X-linked abnormality is suspected, Greenwood sometimes runs the more targeted OGT CytoSure™ Chromosome X Arrays, with which “all the genes on the X chromosome, and every exon of those genes, are covered,” notes Chaubey.
Although 60-mer oligos may be able to query whether the test sample has the complementary sequences, they cannot tell where in the genome those sequences came from. A duplication located a megabase away on the same chromosome, for example, would look the same on a microarray as would a duplication on an entirely different chromosome. Similarly, a translocated sequence would be detected as if it were in its normal locus.
Sequencing
Mike Lyons, associate clinical geneticist and co-director of clinical services at Greenwood, uses the microarray as the first-line genetic test, often supplemented by a karyotype “to make sure there isn’t some rearrangement that isn’t picked up on the microarray.” If that doesn’t give an answer, he may order a specific NGS test—an autism panel would look at 80-plus genes known to be associated with autism, for example—and hope to find an aberration of one of the genes already known to be associated with the clinical presentation. But in the case where “there is just nonsyndromic or nonspecific developmental or intellectual disability, where there is not a suspicion about a particular clinical presentation, then exome [sequencing] is more likely to find a diagnosis in that scenario,” states Lyons.
As the exome databases improve, and there are better algorithms to filter out the variants of unknown significance (VOUS) that complicate interpretation, “exome may become more of a first-line test,” Lyons says. As for whole genome sequencing (WGS), on the other hand, “just the sheer amount of data is so much that it’s somewhat daunting in a clinical setting to do that testing on a lot of different patients, so clinically we’re not doing that yet.”
NGS can detect deletions and duplications at more than an order-of-magnitude better resolution than other molecular cytogenetic techniques, and with read-pair-based preparation and analytical techniques NGS can also be used to detect breakpoints and balanced events including inversions and translocations, notes Evan Eichler, professor of genome sciences at the University of Washington. Yet when these fall into large, highly repetitive genomic sequences—as is often the case—NGS generally is unable to accurately map those breakpoints.
That’s where so-called third-generation or next-next-generation sequencing comes in (at least, in the research lab). Pacific Biosciences’ (PacBio’s) latest P6-C4 chemistry allows for an average read length of greater than 10,000 bases from a single molecule, with many greater than 30,000 bases, says chief scientific officer Jonas Korlach. These long reads enable PacBio to “span a bridge between what has traditionally been viewed as cytogenetics, and molecular genetics—SNP calling and the short-read sequencing.”
BioNano Genomics’ technology can also “essentially bridge that gap,” says Chaubey. Instead of sequencing, that company’s optical NanoChannel Irys system examines long high-molecular-weight DNA molecules labeled with sequence specific fluorescent probes, to “give information of not only copy number changes but also structural variation.”
Work to be done
Someday, de novo assembly of the genome will make traditional cytogenetics obsolete—with not only every SNP but also every structural aberration laid out to be read right from the sequence (although for this to happen Eichler thinks PacBio’s throughput will have to increase, and sequencing costs drop, by more than an order of magnitude. Also, he says, much longer read lengths will be required to completely assemble the complex regions of the human genome). For now, diagnoses and clinical research will still be performed mostly by microarrays karyotyping and NGS, sometimes with a sign on the door reading: “Gone FISHing.”
[1] Ferguson-Smith, MA, “History and evolution of cytogenetics,” Mol Cytogenet, 8:19, 2015. [PubMed ID: 25810762]