Our group is interested in the role of cholesterol in the progression of cerebrovascular disease, based on intriguing data from previous efforts demonstrating that lipid profiles that protect against heart disease seem to worsen the risk of intracerebral hemorrhage (ICH).  Our lab uses genetic variants that predict lipid levels to test casual relationships between cholesterol and ICH.  One way we have done that is to re-examine the role of APOE in ICH risk.  APOE variants encourage amyloid deposition in the vessel wall, but also are strongly associated with lipid levels.  Harnessing this biology, we have demonstrated that at least a portion of the effect of APOE on ICH risk is through this non-amyloid effect on lipid levels.

We are actively working with collaborators, both internally and from other institutions around the world, to conduct genetic studies to better define the molecular basis of the NCL disorders. Although the majority of NCL genes are now identified, enabling a molecular diagnosis in >90% of cases, some patients do not harbor mutations in the known NCL genes, and interpretation of genetic data even in known genes remains a challenge. We are also using our genetic model systems to screen for and test candidate disease modifiers in preclinical studies using our genetic disease models.

We have developed a research partnership with a pediatric neuropsychiatric assessment clinic at MGH. To date, we have collected DNA and cognitive and psychiatric phenotypes on over 1400 youth referred for neuropsychiatric evaluation. We are growing this cohort in order to conduct cross-sectional clinical translational analyses of common genetic risk variants. We also aim to follow these youth over time to relate genetic risk profiles to young adult outcomes.

We have collaborated with another CGM faculty member, Dr. Talkowski, who has developed a strategy for whole genome sequencing based on large-insert jumping libraries. We used this method to characterize a pilot sample of 30 youth with severe neuropsychiatric presentations (including early onset psychosis or mood disorders or ADHD or ASD with notable comorbidities and/or a history of psychiatric hospitalization).

Potential for pharmacogenetics in specific populations

Understanding the pro-inflammatory signaling in the TRPML1 deficiency and modulation of neuroinflammation to treat mucolipidosis IV.

Using AAV gene therapy for CNS delivery of MCOLN1 gene (the gene that encodes TRPML1). Can we reach enough TRPML1 expression in the CNS (particularly in the retina and in the brain) to correct the major manifestations of the disease?

Advancing experimental therapeutic trials with repurposed drugs. Given the time frame, cost, and failure rate of traditional drug development efforts with new chemical entities, one expedient path toward therapeutics for neuropsychiatric disorders may be to repurpose an already existing approved drug. To this end, through our chemical genomic screening platform in conjunction with our patient-specific stem cell models, rodent neuron, and biochemical studies, we have identified multiple F.D.A.-approved drugs and compounds currently under clinical investigation that are candidates for repurposing studies. For example, for bipolar disorder, based upon our studies on the regulation of WNT signaling by the mood stabilizer lithium (Pan et al., Neuropsychopharmacology, 2011) and our high-throughput screens for lithium ‘enhancers and mimetics’ (James et al. Science Signaling 2009; Biechele et al. Chemistry & Biology, 2010; Zhao et al. Journal of Biomolecular Screening, 2012), we have discovered that HMG-Co-A reductase inhibitors (e.g. simvastatin (Zocor)) and the drug riluzole (Rilutek) used to treat amyotrophic lateral sclerosis are both capable of enhancing WNT signaling, a major pathway controlling adult neurogenesis in the hippocampal dentate gyrus. Similarly, with the goal of rapidly advancing the translation of novel therapeutics for frontotemporal dementia (FTD), using our human iPSC models that we recently developed (Silva et al. Stem Cell Reports, 2016) we have screened for and identified multiple F.D.A. approved drugs capable of causing tau clearance in human neurons through autophagy-dependent and independent mechanisms. In collaboration with our clinical colleagues at MGH we are planning to perform pilot clinical trials to advance testing of these therapeutic concepts in patients.

Genomic medicine. We have assessed the predictive value of DNA-based risk assessment for heart attack and evaluated the role for lifestyle and pharmacologic intervention to reduce risk.

The cause of HD has been known for more than 20 years, but effective mechanisms-based treatments are yet to be developed due to the possibility that huntingtin protein is involved in many biological processes and expansion mutations may significantly alter many of those. The recognition that mutant huntingtin is the trigger of a complete dominant disease has stimulated therapeutic application of HTT gene silencing. However, most of the HTT-lowering strategies by targeting mutant HTT mRNA do not provide perfect allele-specificity. In addition, they suffer from difficulties in maintaining dosages of silencing reagents and controlling mutant HTT levels, necessitating repeated treatments. This may pose a significant barrier for chronic progressive neurodegenerative diseases. Alternatively, a single treatment targeting DNA to permanently inactivate the mutant allele may provide better treatment efficacy with less adverse consequences caused by repeated treatments through brain or spinal cord. All these considerations make DNA targeting strategies highly promising. With some beneficial effects of RNA-lowering approaches in HD model systems, development a personalized DNA targeting strategy to permanently inactivate the mutant allele will overcome limitations of RNA-targeting approaches and maximize therapeutic efficacy.

Genotype-phenotype: Huntington’s disease families, clinicians, researchers

Discover biomarkers and therapeutic targets for mitochondrial disease. A genome-wide CRISPR screen for genes that survive a mitochondrial toxin highlighted the endogenous hypoxia response as a potential therapeutic for mitochondrial disease (A).  Mice engineered with a genetic mitochondrial disease (Ndufs4^-/-) that were raised in low oxygen conditions show strikingly increased lifespan (B) and weight (C) compared to mice in normal oxygen conditions (blue vs red curves).  These murine studies support the possibility of hypoxia as a potential therapeutic for mitochondrial disease.

We are developing an in-vitro high-throughput platform for drug screening and determination of dose-response curves of gene therapy strategies for monogenic vascular disorders using CRISPR CAS9, siRNA and lentiviral gene-editing in endothelial and smooth muscle cells.

We are actively recruiting patients who are starting medications associated with risk of sudden cardiac death due to a ventricular arrhythmia called torsade de pointes in order to determine the role that common genetic variants play in risk of this life-threatening drug toxicity.

We are developing methods for modeling microglia-mediated synaptic pruning in vitro, and are using them to understand pruning abnormalities in schizophrenia and to identify novel treatments.

We are developing new tools for visualizing risk in clinical populations, allowing clinicians to make better use of emerging clinical and genomic predictors.

Our in vitro studies showing aberrant activation of mTORC1 signaling and treatment of NF2-null meningioma cells with mTORC1 inhibitor rapamycin led to clinical trials for NF2 patients (Fig. 2).

Furthermore, our more recent studies using dual mTORC1/mTORC2 inhibitor AZD2014 establish that inhibition of both mTORC1 and mTORC2 is more effective (Fig.3) (Beauchamp, 2015). This work is leading to clinical trials for NF2-related and sporadic meningiomas in collaboration with Dr. Scott Plotkin at MGH.

Our longitudinal studies follow patients over time, calling them every six months, in order to determine how brain function after stroke is affected by the genes we carry, by the way we manage our health, and by the way we manage our diet and lifestyle.

CT scans and MRI help us to measure the effects of different genes on the structure of the brain and the damage caused by strokes.

Charting a course from gene discovery to translation: type 2 diabetes-associated melatonin receptor variation. We collaborate with leading chronobiologists and nutritionists on genotype-targeted physiologic studies, in-vitro human tissue studies, and gene x environment studies in shift-workers (SHIFT) and late-night eaters (ONTIME) to investigate mechanisms linking food timing, circadian rhythms and melatonin signaling to glucose metabolism and to evaluate possible interventions.

We were awarded a PCORI grant to build a virtual Down syndrome clinic and test in a randomized control trial.

We were awarded a grant to investigate a combination of factors which might predict if a patient with Down syndrome has obstructive sleep apnea.

We are working with a team of computer engineers to design a mobile app to assist people with Down syndrome making healthy decisions when they are out to eat.

Mucolipidosis Type IV is caused by mutation of the mucolipin-1 lysosomal membrane channel. This is a devastating disease that causes severe psychomotor and mental retardation in infants, and progressive retinal disease leads to blindness by the time patients are in their early teens.  Our goal is to develop gene therapy for both the retinal and CNS dysfunction.

The New England Precision Medicine Consortium

The Harvard Medical School hospital consortium of MGH, Brigham and Women’s Hospital and Children’s Hospital has been recently selected as one of six new sites comprising a nationwide Undiagnosed Diseases Network. As Chief of Medical Genetics at MGH, and the MGH site director for the UDN, Dr. Sweetser is coordinating a team of expert clinicians and researchers, and is using whole exome and whole genome sequencing, and a variety of investigative resources to identify the underlying basis of a variety of challenging human diseases.

Implementation of Genomic Medicine: Whole-genome Sequencing of a Fetus in Prenatal Diagnostics Detects a Balanced Rearrangement

We are using two complementary mass-spectrometry-based approaches to define (i) the proteome and phosphoproteome and (ii) the activated kinome signature of NF1-deficient Schwann cells (SCs) derived from plexiform neurofibromas from patients with NF1. Together these data sets will provide a detailed view of the cellular signaling networks and the molecular targets controlled by neurofibromin in a cell-type relevant to the disease. Small molecule inhibitors to both known and novel signaling pathways will be used to examine the adaptive kinome in treated cell lines.

Improving the targeting range and activities of CRISPR genome editing technologies. A few major limitations of genome editing enzymes include: 1) the inability to target sequences of interest, or 2) suboptimal/inadequate activity. We have engineered naturally occurring CRISPR nucleases to enable them to target more frequently and to access previously untargetable sequences in the genome. These engineered enzymes have implications for various genome editing applications including gene knock-out and knock-in, epigenome editing, base editing, and allele specific discrimination.

Defining and minimizing the off-target effects of CRISPR nucleases. Genome editing technologies can exhibit the propensity to target DNA sequences that resemble the intended site. The recognition and cleavage of ‘off-target’ sites can lead to undesirable consequences for research applications, and can have major implications for the therapeutic use of genome editing technologies. To prevent off-target cutting, we have ongoing efforts to develop nucleases with vastly improved genome-wide specificities.

Characterizing and developing alternative technologies with enhanced properties. Beyond the prototypical commonly used SpCas9 nuclease, there is a tremendous variety of enzymes that have distinctive properties. The diverse characteristics of these alternative platforms may therefore be advantageous for biomedical research or potentially for the treatment of disease.

Characterize genetic and non-genetic modifiers of the relationship between rare pathogenic mutations and phenotype.

Identify high-risk patients ‘flying under the radar’ within our MGH clinical practice and facilitate disclosure and tailored care within a preventive genomics clinical program