February 15, 2010 (Vol. 30, No. 4)
Richard A. A. Stein M.D., Ph.D.
Growing Knowledge Has Direct Application to a Better Understanding of Disease
Until recently, single nucleotide polymorphisms (SNPs) were thought to represent the only source of inter-individual variation at the genome level. Subsequently, it became apparent that different individuals may harbor various copy numbers of certain genes or genomic regions.
This newly unveiled source of genetic variability, which became known as copy number variation (CNV), is believed to involve more nucleotides than all SNPs combined. Such extensive inter-individual genetic variation reveals that human genomes are more dissimilar than initially thought, and even raised the question about whether it is still meaningful to talk about a single reference human genome.
“Copy number variations, in addition to DNA mutations assayed by sequencing, are probably the most reliable data in understanding what is involved in tumor formation, and are key for evaluating potential therapeutic targets,” notes Gregory J. Riggins, M.D., Ph.D., professor of neurosurgery and oncology at the Johns Hopkins Medical Institute.
At next month’s CHI conference, “Comprehending Copy Number Variation”, Dr. Riggins will talk about a genome-wide CNV analysis that he and colleagues conducted to investigate the genomic changes in glioblastoma multiforme. This malignant brain tumor was previously shown to be frequently associated with copy number alterations, either as losses or gains of various chromosomal regions.
By using digital karyotyping and Illumina bead arrays, the scientists found that a locus on the large arm of chromosome 12, which contains the GLI1 and CDK4 oncogenes, is amplified in 33% of the patients. Learning about the genes amplified or deleted by CNVs can provide significant clues about the pathogenic mechanisms involved in tumor formation or progression.
“Since the alterations are in the DNA and easy to decipher, it is relatively easy to do genome-wide assays, and this has become a valuable means for assessing gene alterations in large numbers of tumors,” says Dr. Riggins.
In addition to confirming that different techniques used to survey CNVs lead to similar results, Dr. Riggins and colleagues also established a large collection of copy number changes for this brain disease, which will provide a framework for future studies aiming to design therapies based on molecular targets.
Studies on Schizophrenia
While many CNVs are linked to specific medical conditions, others have been described in apparently healthy individuals. Establishing whether a newly found copy number variant is pathogenic or not has emerged as a significant challenge. In a genome-wide survey examining chromosomal changes in schizophrenia, David B. Goldstein, Ph.D., professor of molecular genetics and microbiology and director of the Institute of Genome Sciences and Policy Center for Human Genome Variation at Duke University, together with Anna C. Need, Ph.D., and other collaborators, identified two new deletions (exceeding 2 Mb in size) that were present in a minority of the participants, but were not found in a healthy subject control group.
“We think that such findings can be a suggestion of pathogenicity, even when they are observed in just one or a few cases. One can identify pathogenic copy number variants because of how unusual they are in people who are healthy,” explains Dr. Goldstein.
The study, which was the first to involve an integrated survey of SNPs and CNVs in schizophrenia, demonstrated that rare copy number variants might be more important than more frequently occurring SNPs. Further investigation of the same 2 Mb region in a group of patients with epilepsy uncovered several smaller deletions, pointing toward the possible involvement of a CNV in several neuropsychiatric conditions.
A key difficulty facing future work is being able to accurately survey and characterize CNVs in the genome.
“It is very difficult to figure out, not only the exact copy number status in different parts of the genome, but also the content. These structurally variable parts of the genome are difficult to assay, and the challenge is being able to clearly describe a variation in order to relate it to disease,” continues Dr. Goldstein.
In addition to their link to medical conditions, CNVs are increasingly implicated in shaping the response to pharmaceutical compounds.
“It has become quite an exciting time in pharmacogenomics because it really was a relatively theoretical field up until a year or so ago. That’s when we started seeing more clinical trials which demonstrated some of its utility in the clinic,” says Howard L. McLeod, Pharm D., distinguished professor of pharmacy and medicine at the University of North Carolina at Chapel-Hill.
One of the implications is that, as a result of CNVs, certain individuals might not respond to medication, need larger doses, or face a higher risk for adverse effects. Therefore, if there is an altered copy number for a gene encoding proteins that transport or metabolically degrade a drug, or affect its target, this piece of information can be used to more carefully adjust a drug’s dosage.
“It is not surprising that if there are three copies of a metabolism gene instead of two, the drug is going to be cleared much faster and it cannot have its effect,” explains Dr. McLeod.
Recently, Dr. McLeod and colleagues found that TOPI, which encodes topoisomerase I, an enzyme involved in cleaving and rejoining DNA during replication, recombination, and repair, was amplified in 23% of the individuals with colorectal cancer. In addition, 60% of the tumors analyzed were diploid for both TOPI and TYMS, the gene encoding thymidylate synthase, and this genomic profile was associated with less favorable response to chemotherapy.
Besides their promise in tailoring individual therapy, CNVs have become essential in an additional context. When Dr. McLeod and collaborators recently examined the WHO list of essential medicines, they identified several compounds for which the genes encoding metabolic, transport, or target proteins show copy number polymorphisms among populations.
“The clinical implications are dramatic,” points out Dr. McLeod. “If we know that there is a likelihood of increased risk of adverse effects or a lack of efficacy as a result of a common genetic variant, we can better tailor therapy in a specific country. This involves not individualizing for the person, but individualizing for the country, based on the overall risk. Copy number variation is not only an individual patient issue but also a public health issue.”
How CNVs Emerge
Although CNVs are becoming more relevant in clinical medicine, relatively little is known about how they emerge and what their genomic distribution is within and across species. Recent developments promise to shed light on this issue.
Harris A. Lewin, Ph.D., professor of immunogenetics and director of the Institute of Genomic Biology at the University of Illinois at Urbana-Champaign, and collaborators recently conducted the first multispecies whole-genome comparison of chromosomal organization in 10 amniote species. This approach examined evolutionary breakpoints in chromosomes and facilitated the visualization of genome rearrangements over evolutionary time.
While some regions in the genome appear to be conserved for millions of years, others, known as evolutionary hotspot regions, are characterized by a high rate of gene birth and gene death. They are much more likely to be involved in rearrangements.
The researchers report that for some species, eight to nine times more structural variants are observed within evolutionary breakpoint regions as compared to the rest of the genome.
“We discovered that the evolutionary breakpoint regions have a much higher density of structural variations,” notes Dr. Lewin. Most of these structural variations include changes in gene copy numbers.
“Copy number variants are not randomly distributed in the genome but are heavily weighted toward the genomic locations where chromosome rearrangement is happening during evolution.”
CNVs could thus be a consequence of chromosomal rearrangement and repair events, or they could be part of the mechanism that is driving rearrangements by non-allelic homologue recombination.
“These evolutionary hotspots seem to be the places where, during evolution, new variation is generated on which natural selection can then act,” explains Dr. Lewin.
Previously, Dr. Lewin and his colleagues identified a relationship between evolutionary breakpoint regions and highly recurrent cancer breakpoints frequently described in leukemias and lymphomas, raising the possibility that these chromosomal regions are inherently unstable. This concept could help in understanding how CNVs shape several medical conditions. Examining the location of these chromosomal rearrangements through an evolutionary timeframe provides an intriguing approach to map out the genomic regions that are particularly susceptible to medically relevant CNVs.
“Every human genome is different as is every animal genome. What we don’t understand is how microalterations in genome architecture, such as copy number variation, affect disease,” says Dr. Lewin.
Genetic diseases and chromosome abnormalities could very well represent an unfortunate consequence of our need to generate genetic variation. Ultimately, it is diversity within a species that emerges as perhaps the most powerful factor to guarantee adaptability and survival.
It would be difficult to envision all individuals within any species exhibiting susceptibilities to the same diseases, adverse reactions to the same therapeutic compounds, or sensitivities to the same environmental toxins. CNVs could very well represent an indispensible safety net, instrumental for the successful existence of every species on the planet.