Date of Award

January 2018

Document Type


Degree Name

Medical Doctor (MD)



First Advisor

Kristopher T. Kahle

Second Advisor

Richard P. Lifton


Congenital hydrocephalus (CH) is a major cause of childhood morbidity and mortality, affecting 1 in every 1,000 newborns [1], and exerts a tremendous burden on the United States health care budget, consuming over $2 billion annually [2]. Classically regarded as an imbalance between cerebrospinal fluid (CSF) production and reabsorption, CH is treated with invasive, morbid CSF diversion surgeries with high rates of failure and complications. Despite evidence that genetic factors play a major role in the pathogenesis of CH – an estimated 40% of human CH has a genetic etiology – our knowledge of specific CH-causing mutations and their pathogenic mechanisms remains primitive [3, 4].

The work presented within this thesis represents the first, and largest, whole-exome sequencing study of patients with well-phenotyped, primary congenital hydrocephalus: 125 parent-offspring trios and 52 additional probands with CH. Analysis of the burden of rare de novo and transmitted mutations identified four novel genes, each highly intolerant to heterozygous loss of function, that surpassed genome-wide significance thresholds. The let-7 microRNA target TRIM71 had three de novo missense mutations (p = 2.15 x 10-7), with two occurring at the identical base. The two different mutation sites occurred at homologous positions in blades of the NHL domain and altered completely conserved arginine residues directly involved in binding to target mRNAs. SMARCC1, a core subunit of the BAF (yeast SWI/SNF) chromatin remodeling complex, had two de novo damaging mutations and three rare transmitted LOF variants (p = 8.15 x 10-10). PTCH1 encoding Patched 1 had two de novo and one rare transmitted LOF function mutations (p = 1.06 x 10-6). Additionally, we identified two de novo duplications at the SHH locus encoding the Patched 1 canonical ligand Sonic Hedgehog.

To functionally validate candidate mutations and establish gene causality, we developed a novel method using quantitative Optical Coherence Tomography (qOCT) imaging in live Xenopus tropicalis to visualize ventricular flow in real-time. We then utilized CRIPSR/Cas9 gene editing of CH-causing candidate mutations to recapitulate essential aspects of human CH pathophysiology in a live, complete ventricular system.

Our results implicate four new genes in the pathogenesis of human CH that collectively explain 10% of non-syndromic congenital hydrocephalus cases. Strikingly, all four genes are required for neural tube development, are highly expressed in the ciliated neuroepithelium, and regulate the balance between neural progenitor cell proliferation and differentiation. These findings suggest that the ventriculomegaly of a subset of CH patients may be a secondary epiphenomenon related to decreased neurogenesis and cortical thinning, and not a primary pathological accumulation of CSF. Together, these results provide fundamental novel insight into the pathogenesis of hydrocephalus, have immediate implications for genetic testing and counseling for affected families, and demonstrate an innovative, rapid, and cost-effective platform for future CH gene discovery that combines WES, CRISPR/Cas9 genome editing, and qOCT to create live Xenopus models of human disease for the purposes of target validation, disease modeling, and drug screening.


This thesis is restricted to Yale network users only. It will be made publicly available on 06/25/2100