Finding Cancer Risk in a Gene Desert


The Human Genome Project has had a rough year, from a PR perspective. To mark the 10th anniversary of the completion of a “working draft” for the human genome, many media outlets have chosen the “what have you done for me lately?” angle in assessing the project’s impact upon everyday medicine. Outlets such as the New York Times, The Economist, and The Guardian all ran articles about the as-yet-unfulfilled promises of the Human Genome Project in finding new cures for a wide range of diseases, a reality far from the “complete transformation in therapeutic medicine” promised in 2000 by Francis Collins.

One prominent area of frustration has been with genome-wide association studies (GWAS), which compare a population with a particular disease such as atherosclerosis or lung cancer with a healthy population to identify genes associated with disease risk. Many of the “hits” from these studies offered confusing results; as Nicholas Wade wrote in the Times: “most of the sites linked with diseases are not in genes – the stretches of DNA that tell the cell to make proteins – and have no known biological function, leading some geneticists to suspect that the associations are spurious.”

But Marcelo Nobrega, assistant professor of human genetics at the University of Chicago Medical Center, disagrees with that conclusion. And in a paper published today in the journal Genome Research, Nobrega’s laboratory demonstrates how these “non-genetic” GWAS hits can have dramatic biological effects significant for disease risk.

“The alternative, which we’ve known for a while now, is that the genome is not only comprised of genes,” Nobrega said. “Only one percent of the genome becomes proteins, but there are these regulatory elements everywhere that have important biological roles but are not protein-encoding sequences. Mutations in these proteins presumably will lead to disease or increase the risk of disease, there’s no reason to believe otherwise.”

The paper, authored by Nobrega with Nora Wasserman and Ivy Aneas, demonstrated this principle by tracking down a GWAS hit associated with cancer of the prostate and other organs. Studies identified a segment on the 8th chromosome that was associated with an increased risk of cancer, but it pointed to a region that contained no protein-encoding genes – a stretch known as a “gene desert.” In previous work, Nobrega’s team had tracked down small regulatory elements within these deserts that help choreograph the expression of genes for the development of organs. The new search had the opposite mission: to find not the regulatory elements that build an organ, but the elements that could someday send cell growth awry in those organs.

The GWAS hits landed in a gene desert upstream of a gene called MYC, a known oncogene. So Nobrega’s team engineered an assay to tease out small elements within the gene desert that caused increased or unusual expression of MYC. Bacterial artificial chromosomes (BACs) were created that reproduced three overlapping segments of the gene desert, but with a gene called lacZ attached. When expressed, lacZ produces a blue pigment, allowing the researchers to spot the organs in mice where regulatory elements play an active role.

Two of the three BACs produced blue stain in the prostate, allowing the researchers to narrow their search down from nearly a million base pairs to around 5,000. Within that stretch: a single nucleotide polymorphism (SNP) called rs6983267 associated by GWAS with prostate and colorectal cancer. A second experiment then tested whether the “risk” allele for that SNP – only one nucleotide different from the non-risk allele – changed the expression of lacZ. Despite the minute change, the expression difference was striking, with robust blue staining in the prostate of risk allele mice and little to no detectable prostate staining in the non-risk allele mice.

“Perhaps what this is telling us is that by inheriting the risk allele here, you may drive the overexpression of MYC,” Nobrega said. “It’s not going to cause prostate cancer, but it could increase the risk for prostate cancer.”

The difference was also detectable as early as embryonic stages, suggesting that early over-expression of MYC could raise the risk of prostate cancer much later in life.

“In some ways, it could potentially prime the system for cancer and then, depending on either secondary mutations or environmental injuries or insults, it might or might not develop,” Nobrega said.

Such a finding may still carry a slight taste of the frustration written about over the last month: we may know more about this mechanism for prostate cancer risk, but what can we do about it? Nobrega suggested that the intermediate proteins that connect regulatory elements with gene expression could be potential therapeutic targets for lowering overexpression and cancer risk. An infant known to carry the risk allele could theoretically be treated early in life with an inhibitor of this regulatory pathway to prevent the dangerous overexpression of MYC, thereby reducing their risk of cancer at an older age.

That scenario may not be the easy genetic fix envisioned by many in the original hype surrounding the Human Genome Project. But when looked at through the lens of basic science rather than clinical application, the HGP remains a milestone. Even as it revealed that the genetic story is more complicated than the Central Dogma of Molecular Biology would have us believe, the Human Genome Project gave scientists the essential boost they needed to understand that complex world. As Francis Collins himself wrote in a recent Scientific American commentary, “For those of you who like stories with simple plots and tidy endings, I must confess the tale of the Human Genome Project isn’t one of those.”

About Rob Mitchum (526 Articles)

Rob Mitchum is communications manager at the Computation Institute, a joint initiative between The University of Chicago and Argonne National Laboratory.

%d bloggers like this: