Douglas Marchuk, PhD

James B. Duke Professor
Director, Division of Human Genetics

Headshot of Douglas Marchuk, PhD265 CARL Building
Box 3175 DUMC
Durham, N.C. 27710
Phone: (919) 684-1945
Fax: (919) 684-2790


research • lab memberspublicationswebsite

Scientific Approach:  My laboratory studies the genetics of cardiovascular disease using both the human and the mouse as a model system. Much of our work over the past two decades has been focused on inherited diseases of vascular dysplasia. The genes responsible for these genetic syndromes encode proteins that act at critical points of the pathways that control vascular morphogenesis. Our objectives are twofold: (1) to provide basic knowledge on the role of these genes and gene products in vascular morphogenesis and (2) to gain specific knowledge of the pathology observed in these disorders.   The first step in this approach is to identify the genetic loci that underlie these disorders. These genetic endeavors provide the basis for future molecular biological studies on the role of the mutant protein in the pathology of the disease, and the role of the normal proteins in vascular development.   These studies invariably involve biochemical and cell biological analyses of the mutant protein and its role in the cell. Future investigations also require an in vivo model, usually a knockout or transgenic mouse model. The animal model serves as a tool to investigate the pathophysiology of the disease, and a more tractable system to begin to understand the biology of the gene product in vascular morphogenesis.   Coming full circle, we can determine if the additional factors identified in the animal model also modify the clinical phenotype in the human disease.

Cerebral Cavernous Malformations. As an example of the above mentioned approach, links below highlight some of our recent work on Cerebral Cavernous Malformations, congenital vascular anomalies of the brain comprising focal, thin-walled, grossly dilated vascular spaces. CCMs are responsible for significant neurologic disability, in particular, intractable migraine, seizures, and hemorrhagic stroke. Our laboratory has identified two of the three known genes causing this syndrome. While the causative genes have been identified, the molecular events initiating lesion formation are still being elucidated. We had hypothesized that CCM lesion genesis occurs due to second-site somatic mutations, following a “two-hit” mutation model. Patients harboring a germline mutation would require a somatic mutation in the wild-type allele to initiate lesion formation. We have been analyzing CCM lesion from human patients harboring a germline (inherited) mutation. In a number of cases, we have identified a somatic mutation in the wild-type allele, thus fulfilling the two-hit hypothesis. Laser Capture Microdissection of the lesion shows that the somatic mutations arise only in the endothelial cells of the grossly-dilated vessels. However, human CCM lesion tissue availability is restricted to large, multi-vessel, late stage lesions. This limits our ability to perform molecular analyses at the earliest stages of lesion development. Therefore, we have created mouse mutant models for Ccm1, Ccm2, and Ccm3. These mouse models faithfully recapitulate the pathology and disease course seen in the human patients. We are now using these models to elucidate the mechanism behind the development and maturation of the CCM lesion.   We are also using our mice to investigate a number of novel therapeutic approaches that have been suggested by biochemical studies of the CCM proteins.

Sturge Weber Syndrome. We have other studies of diseases of vascular dysplasia. One other example is Sturge-Weber Syndrome. Sturge Weber Syndrome (SWS) is a rare, congenital, but sporadic (non-inherited) disease characterized by a facial port-wine birthmark, choroidal and leptomeningeal vascular malformations, epilepsy, stroke-like episodes, headache, and cognitive impairment. Over two decades ago Rudolf Happle hypothesized that SWS and certain mosaic phenotypes are due to a somatic mutations in a critical developmental genes. He postulated that each mutation would be lethal if passed through the germline (hence no familial segregation), and would only ever appear in the mosaic state. He coined the term “paradominant inheritance” to describe a mosaic phenotype resulting from an otherwise lethal mutation that can only survive through somatic mosaicism. For SWS, Happle further postulated that the relatively common port-wine stain (PWS, a nevus flammeus, or capillary malformation ofthe skin) and the rarer Sturge Weber syndrome, which includes the port-wine statin as part of the phenotype, were due to somatic mutation in the samegene, with the extent and severity of the phenotype determined by the time during development when the somatic mutation was acquired. For many years, Happle’s hypothesis had remained unproven for any syndrome. We used genome-wide sequence analysis of affected and unaffected tissues to identify the SWS gene. We identified the same somatic mutation in all three affected tissue samples that was absent in unaffected tissue from the same patients. Further analysis of a large set of port-wine stain and SWS tissue samples showed that the great majority harbored the identical somatic mutation in the GNAQ gene, leading to p.R183Q in the Gαq protein. Gαq is a G protein subunit that modulates a wide spectrum of downstream signaling pathways, depending on the biochemical and cellular context. We are now investigating the biochemical, cell biological, and in vivo consequences of this mutation.

New Directions: We have also become interested in harnessing the power of mouse forward genetics to map novel genes that affect the severity and progression of disease in mouse models of disease. We begin with an animal model of the disease, such as a surgically induced or transgenic model of disease. The surgical intervention or the transgene acts as a sensitizer to create the disease in the animals. These sensitized models often exhibit drastically different rates of disease progression or outcome depending on the inbred strain employed. Using genetic crosses with the sensitizer, we can map the loci that influence disease outcome. We call these risk factors “modifier loci” since they modify disease progression or outcome. Importantly, given the ancestral relatedness of the commonly used inbred mouse strains, we can use ancestral haplotype mapping to rapidly identify the gene and sequence variant underlying the modifier locus. The orthologs of the modifier genes discovered in the mouse models are then investigated in the corresponding human disease populations.   Below, we highlight two different projects using this approach, one on heart disease, and another on ischemic stroke.

Heart Disease Modifier Genes.  One example of this approach is the Calsequestrin model of cardiomyopathy (heart disease). The Calsequestrin (CSQ) mouse model exhibits extreme variation in the phenotype progression and severity highly dependent on the genetic background. Genetic mapping in the context of the transgenic sensitizer has yielded seven heart failure modifier (Hrtfm) loci that modify disease progression.  We have now identified the gene for one of these – Hrtfm2. A single gene in the interval shows a significant difference in transcript levels between strains. Strains with a susceptible phenotype show high levels of transcript while strains that express low transcript levels are protected. In strains showing reduced message levels, we have identified an intronic SNP that activates a cryptic splice site leading to aberrant splicing, followed by nonsense-mediated decay of the message. We are currently investigating the role of the human ortholog of this gene in heart disease progression. In addition, we are seeking a small molecule inhibitor of the encoded protein as a novel therapeutic for heart disease.

Ischemic Stroke Modifier Genes. A second project relates to genes that modify the outcome of ischemic stroke. Although epidemiological studies have provided substantial evidence for genetic influences in the development of stroke, genetic determinants of the extent of ischemic tissue damage remain unknown. Through the use of a surgically-induced mouse model of focal cerebral ischemia we have shown large, reproducible differences in infarct volume among different inbred mouse strains. A genome-wide linkage scan in various cross of inbred mouse strains identified a small number of loci that account for the majority of the phenotypic variance in infarct volume. Measurement of infarct volume in additional inbred strains allowed an ancestral SNP haplotype-based approach to fine-mapping each locus, thereby reducing the possible number of candidate genes. In vitro and in vivo studies are currently underway to investigate the role of these genes in modulating tissue damage after ischemia. The results of this study may uncover novel genes that modulate the severity of human ischemic stroke, as well as provide new targets for therapeutic intervention.