“This is a good example of how the intersection of cutting-edge genetic and epigenetic profiling with the latest genome- and epigenome-editing technologies can be used to advance our understanding of how sequence variants can impact diseases,” says J. Keith Joung, MD, PhD, associate chief for Research in the MGH Department of Pathology and a co-senior author of the report. “We believe that this is an important next frontier for advancing diagnosis and treatment of diseases influenced by genetics and that CAUSEL provides a blueprint for how to proceed with these types of studies.”
Co-senior author Matthew Freedman, MD, of the DFCI Center for Cancer Genome Discovery and Center Functional Cancer Epigenetics explains that the effort to determine the precise DNA variants that increase disease risk and how they do so is extremely complicated. While the genetic mapping approach known as linkage analysis has enabled the identification of DNA variants in protein-coding genes — like the BRCA1 and 2 genes that, when mutated, cause a clearly increased, inherited risk of breast and ovarian cancer — those variants account for approximately 5 percent of cases of inherited disease risk. The other 95 percent appears to be predominantly influenced by variants mapping to non-protein-coding regulatory elements that control the levels at which protein-coding genes are expressed.
“Now the question is how can we identify the actual pathogenic variant that is driving disease and determine its functional consequences” Freedman says. “The reason that is difficult to answer is that we are not used to working in the non-protein-coding genome. The genetic code — an elegant rule book for how specific DNA sequences code for the amino acids that make up proteins — has been understood since the 1960s, but there is no such code for the vastly greater portion of our genome that does not code for proteins. The genome-wide association studies (GWASs) that are being performed worldwide to find associations with everything from eye color to disease susceptibility usually discover that many variants are correlated with a condition, but it’s very hard to isolate which variant is actually driving the trait.”
The CAUSEL procedure or ‘pipeline’ consists of five steps:
- genetic fine mapping to identify candidate variants,
- epigenomic profiling, in which the candidate variants are intersected with epigenetic data to identify which are most likely to cause the condition,
- epigenomic editing, in this case using reagents developed in Joung’s laboratory, to confirm whether or not the candidate variants may possess regulatory capacity,
- genome editing, to create cell lines with all possible genotypes of the candidate variants,
- phenotypic analysis of those cell lines, to evaluate functional differences relevant to the disease or condition of interest.
The researchers tested this pipeline on a region of chromosome 6 that previous studies have associated with increased prostate cancer risk, probably by controlling expression levels of RFX6, a regulatory protein involved with tumor-associated properties. Fine-mapping that region identified 27 candidate variants, all associated with increased prostate cancer risk, and epigenomic profiling identified one as most likely to be relevant. Epigenomic editing performed at that site, called rs339331, confirmed that the region it lies in could potentially regulate RFX6 expression, so the investigators created three cell lines — one with two copies of the disease-associated variant or allele, one with two copies of the ‘normal’ version, and one with a copy of both versions.
Analysis of these cell lines revealed that, while cells with two normal alleles had the appearance of normal cells, both lines containing the cancer-associated variant had an appearance more typical of cancer cells. Cells carrying two cancer-related alleles were more likely to adhere to surfaces, a property typical of cancer cells, and those cells lines also exhibited changes in the expression of genes involved with androgen signaling, a pathway known to be critical in the risk for and treatment of prostate cancer.
Joung notes that, while his team used epigenome- and genome-editing reagents based on engineered transcription activator-like effector (TALE) technology, other approaches, such as the easier to use CRISPR-Cas9 platform, should also work for these steps, making the CAUSEL approach accessible to most laboratories. “As the number of gene variants associated with disease expands, it will become more and more important to identify which ones actually contribute to disease development. When sequence variants are identified by CAUSEL as functional, we can envision that pathologists might rapidly develop tests for those variants, which could then be used to impact clinical care,” he says.
Simon Gayther, PhD, of the Keck School and Cedars Sinai Medical Center, co-senior author of the manuscript explains, “This study and the pipeline it describes represent something of a Holy Grail for the GWAS community, which has been incredibly successful at identifying thousands of novel susceptibility alleles associated with disease but has not yet been able to show how these risk variants cause disease at the cellular level. This pipeline opens up a new-world opportunity to assign biological and clinical significance to risk variants that are, at least in part, responsible for a multitude of cancers and other traits.”
Adds Freedman, who is also an associate member of the Broad Institute, “There is no reason this technology couldn’t be applied to non-cancer-related variants as well. Of the approximately 17,000 gene variants that have been associated with diseases or other conditions, less than 0.1 percent of these associations have rigorously been identified as causal variants. For the more than 99.9 percent that still need causal variant identification, we hope that finding the right cell type and applying our pipeline will close that gap. Now we need to improve the efficiency of our steps and deploy CAUSEL in examining these and the many other variants that are being identified by labs around the world. If this works the way we anticipate, I do believe that the impact will be transformative.”