Research Highlights from the ASHG 2013 Annual Meeting

October 28, 2013

By Aaron Krol 

For more highlights from the ASHG Meeting, see our piece on research with immediate clinical relevance at Clinical Informatics News!

October 28, 2013| The American Society of Human Genetics held its annual meeting in Boston last week, bringing together thousands of genetics professionals to share their research, explore new product offerings from the industry, and discuss the pressing scientific, social, and ethical issues that have converged on this rapidly developing field. Here are some highlights you may have missed:

Jeff Murray, the ASHG President and a geneticist and pediatrician at the University of Iowa, delivered his keynote address on Tuesday. Murray recalled the long history of the ASHG, beginning in 1950 when President H.J. Mueller declared in his own address that “we shall never be able to demonstrate with certainty that a hereditary human disease arises from mutation,” through the hopeful period in 1983 when Murray, with Arno Motulsky, was able to publish a paper on the strength of identifying just one single-nucleotide polymorphism (SNP), to the present day when Murray estimates that around one billion SNPs are identified every day. Murray’s greatest passion, however, was for taking the practice of science outside the lab and to those who need it most. “All of us need to think not only about the science that’s in front of us,” he said, “and not only about the patient that’s in front of us,” but also about applying the medical knowledge available right now to advance public health around the world.

In the blockbuster finding of the 2013 meeting, Jeanne Lawrence (University of Massachusetts Medical School) presented her lab’s work on in vitro gene therapy techniques for addressing trisomy 21, better known as Down syndrome. Because trisomy 21 is a chromosomal disorder, it has been widely assumed that it would not be amenable to present-day gene therapy, as hundreds of genes are involved in its pathology, it does not arise from deleterious variants, and the molecular pathways causing the disease are still only dimly understood. The Lawrence lab, however, took inspiration from a natural form of chromosome-level regulation: the XIST gene on the X chromosome, which is responsible for preventing both X chromosomes from being expressed in women. XIST codes for a giant strand of RNA that renders an entire X chromosome non-functional. Using a zinc-finger nuclease, Lawrence and her colleagues inserted XIST – at 21 kilobases, the largest transgene ever to be inserted with this method – into one copy of chromosome 21 in stem cells with trisomy 21. The researchers then used eight different methods to measure the expression of chromosome 21 in the cell, finding at every point that a single copy of the chromosome had been silenced and transcription levels were virtually normal. Cellular differences were observed within a week of this therapy, with treated cells proliferating faster and forming neural progenitor cells earlier than untreated cells. Although there are many formidable barriers to using this technique to treat Down syndrome in vivo, Lawrence does expect it to yield enormous insights into the specific pathways by which changes in gene expression lead to the disorder’s physiological symptoms – including the expression of genes not on chromosome 21. The lab is also preparing to administer the therapy in mouse models of Down syndrome.

Scott Carter (The Broad Institute) presented a novel use of genetic testing to track the origins and causes of brain metastases. Banking on the heterogeneous genetic profiles of cancer cells to provide a sort of miniaturized evolutionary history, Carter and his collaborator Priscilla Brastianos collected cell samples from 101 patients of their brain metastases, primary tumors, and normal cells. They then performed whole exome sequencing on the samples – and importantly, looked not just for which somatic mutations had arisen, but also for their frequencies in each cell population. Invariably, the metastases showed direct descent from the primary tumor, but also carried, at 100% frequency, unique mutations not found in the primary tumor. This argues heavily for a single-clone origin of brain metastases, in which a cancer cell with a novel mutation migrates from the tumor to the brain and multiplies there, producing a distinct cellular population with a sibling relationship to the tumor.

David Reich discussed new findings from the Neandertal Genome Consortium’s ongoing studies of the genetic relationships between ancient human species. Using nearly-complete genome sequences of both Neandertals and the recently-discovered Denisovans, the Consortium has described five distinct gene flow events between different species of humans and retraced the speciation events that divided them. Their analyses show that while Neandertals and Denisovans are more closely related to one another than they are to us, Neandertals have left a larger genetic imprint on modern humans through interbreeding, while Denisovans show evidence of a genetic exchange with an unknown, even earlier human species.

Ran Blekhman (University of Minnesota) presented on the correlation between human genetics and the composition of the human microbiome. The Human Microbiome Project (HMP), performing genetic analyses of bacteria collected from human participants, is starting to yield insights into the flora that together comprise the majority of cells and genes found in the human body. However, as the project does not collect human samples, the relationship between our genomes and microbiomes has remained mysterious. Blekhman’s lab took an innovative approach to overcoming this lack of data, collecting the “host contamination” data from the HMP’s shotgun sequences of bacterial colonies. Using reads contaminated by human cells, Blekhman was able to call 4 million SNPs from 100 individuals at an average 10X coverage, enough to perform population analyses. His tentative results suggest some degree of evolutionary response to local microbiomes, as the fixation index of SNPs tended to be elevated when they were associated with the presence of bacterial species.

Elaine Lim, from the Sequencing Initiative Suomi Project, described how the Finnish population bottleneck provides a unique opportunity to examine rare genetic variants. Because the Finnish population is relatively isolated, and descends from a small founder population, the proportion of rare variants among Finns is skewed relative to other Europeans: very rare variants, at under .5% frequency, are almost completely absent in the Finnish genome, while variants between .5% and 5% frequency are more common than expected. In addition, Finland keeps a comprehensive medical record system, allowing extensive studies of correlations between rare variants and medical histories. Lim described a series of findings already emerging from such studies, including overturning earlier research that linked variants in the FANCM gene causally to Fanconi Anemia, when seven Finns were found homozygous for the variant with no symptoms.

Fasil Tekola-Ayele, from the Center for Research on Genomics and Global Health at the NIH, reviewed the progress of the African Genome Variation Project (AGVP). The AGVP is a multi-institutional project from partners including the Wellcome Trust Sanger Institute, MalariaGEN, and the African Partnership for Chronic Disease Research, to perform genome-wide association studies (GWAS) of a diverse population of Africans. So far, the AGVP has genotyped almost 1500 individuals from 18 different ethnic groups across Africa, and begun tracing their historical relationships, finding that the Africans so far measured can be placed in five main population clusters associated more strongly by geography than by ethnicity. As Africa represents the largest spectrum of human genetic diversity of any continent, and only 10 GWAS have so far been performed on Africans, the AGVP and projects like it are vital to understanding human demographic history.

Matthew Nelson (GlaxoSmithKline) presented evidence that successful drug discovery is associated positively with known gene associations, whether or not drug trials had their origins in well-understood genetic pathways. Nelson overlapped proposed drug targets with information on gene-disease associations from publicly-available GWAS datasets, comparing the gene associations of approved drug targets with those that did not pass approval. Approved targets were statistically more likely to address a molecular pathway connected with a known gene-disease association. This correlation was greatest for nutritional and metabolic diseases, while for some categories of drug targets, like behavioral diseases, there was no correlation. Nelson also found that the size of a known genetic effect was not nearly as important as whether a gene was involved in the relevant epidemiology of a drug target. He suggests that these findings may provide some guidance for future drug development, although the correlation is too weak to make or break a given drug trial.

Amy Breman (Baylor College of Medicine) compared cell-based non-invasive prenatal testing (NIPT) with the presently used cell-free NIPT methods. To assess the efficacy of cell-based NIPT in detecting chromosomal disorders, Breman’s lab performed whole genome amplification (WGA), followed by microarray analysis, on in vitro lymphoblast cell lines with four disorders caused by varying sizes of chromosomal abnormalities: trisomy 21, the large-deletion diGeorge Syndrome, a somewhat smaller duplication of the MeCP2 gene, and Charcot-Marie-Tooth disease. Three different WGA methods were tested, of which Ampli1 proved most sensitive. When Apli1 was performed on cells fixed in 2% paraformaldehyde, all four conditions were correctly called three out of three times. Breman hopes to move forward with this line of testing in blind in vitro trials, as the cell-based method was able to identify copy number variations smaller than current NIPT methods can reliably call.