Abstract

The transcription factor Cone-Rod Homeobox (CRX) is a master regulator of photoreceptor cell fate. Sequence variants in CRX can cause Retinitis Pigmentosa, Cone-Rod Dystrophy, and Leber Congenital Amaurosis, all inherited causes of vision loss and blindness. CRX is the only gene implicated in the pathogenesis of all three of these diseases, which present with both rod- and cone-centric phenotypes of varying age of onset and severity. Several CRX variants have been reported to cause severe dominant disease through antimorphic genetic interactions with wild-type CRX, and yet these mutations are adjacent to variants which are benign or only cause mild, recessive disease. Determining which mutations in CRX are pathogenic and quantifying their effect on functional activity is prerequisite to interpreting patient variation and predicting patient phenotypes. Routine clinical sequencing is on the rise, but our ability to interpret patient variants in genes like CRX has not kept pace. Most are deemed “Variants of Uncertain Significance” (VUS), meaning that insufficient clinical or functional evidence exists to determine their pathogenicity. In ClinVar, the NIH-sponsored database of human genetic variation, CRX alone has over 100 reported VUS, with nearly 95% of all possible CRX variants not yet observed in a patient (and thus also uncharacterized). Without a robust catalog of human genetic variation, advances in patient sequencing cannot be translated into clinical guidance or therapies for patients with uncharacterized variants. There is an unmet need for functional characterization of the thousands of possible variants in CRX, and in all genes with relevance to human disease. One potential solution to this challenge is Deep Mutational Scanning (DMS), which uses massive libraries of variant sequences in multiplexed assays to simultaneously measure the functional consequence of thousands of variants in a gene of interest in a single experiment. In DMS, each gene variant is assigned a functional pathogenicity score based on its activity in a molecular assay. The current system of post-hoc functional variant classification by one-at-a-time molecular studies has been overwhelmed by the rate of variant discovery, but DMS has the potential to classify these variants before they are ever observed in a patient, and at scale. I developed a DMS assay to simultaneously measure the functional consequence of all missense variants in CRX. Using an engineered cell-based landing pad system with a novel synthetic activity reporter, I measured the ability of each CRX variant to transactivate a fluorescent reporter gene. The direct product of this work is a “look-up table” listing the functional consequence of every missense CRX variant on transcriptional activity, which is directly applicable to clinical variant classification and decision-making. In addition to its clinical utility, the CRX DMS also offers insight into the molecular properties of the residues that make up the functional domains of CRX. Distinct patterns of variant substitution effects in the structured DNA binding domain and unstructured transcriptional effector domain reveal rules governing residue positioning and composition. In the DNA binding domain, per-position average variant activity scores closely aligned to an AlphaFold-predicted structure of the CRX homeodomain in complex with DNA. In the transcriptional effector domain, substitutions altering the ratio of acidic and aromatic residues had profound effects on activation, consistent with the established Acidic Exposure Model of transcriptional activation domains. In a complementary series of experiments, I deeply characterized the transcriptional behavior of two known pathogenic CRX variants with varying severity and clinical presentations: the recessive p.R90W substitution variant and the dominant p.E168d2 frameshift variant. In the DMS assay, each variant is tested against a single transcriptional reporter, but in the retina, CRX binds many different cis-regulatory elements (CREs), activating multiple downstream target genes. Using Massively Parallel Reporter Assays (MPRAs) carried out in living retinal tissue derived from mice homozygous or heterozygous for the two pathogenic variants, I systematically profiled the effects of these variants on all known CRX-bound CREs. While the degree of reporter dysregulation caused by the variants largely corresponded with phenotypic severity, the p.E168d2 variant had stronger effects on silencers. In particular, a set of potentially pleiotropic regulatory elements converted from silencers to enhancers in retinas with the p.E168d2 variant. Furthermore, CREs near cone photoreceptor genes were enriched for silencers de-repressed in the presence of p.E168d2. These findings suggest that phenotypically-distinct variants in different domains of CRX have partially overlapping effects on its cis-regulatory function, leading to misregulation of similar sets of enhancers, while having a qualitatively different impact on silencers. Together, this work offers biological insight into CRX as well as a clinical resource for the classification of CRX variants. The technology development efforts that enabled these experiments also have the potential to facilitate numerous future studies. For instance, studies of additional CRX variants, including truncation and frameshift variants, could allow for further characterization and delineation of functional domains. Characterization of CRX variants in multiple environments, such as in the presence of wild-type CRX or other retinal transcription factors like NRL or OTX2, could help to isolate dominant variants and contribute to an understanding of the influence of CRX variants on interactions with partner proteins. And lastly, it may be feasible to combine DMS and MPRA, to measure the effects of variants on many reporters in a single experiment.

Committee Chair

Barak Cohen

Committee Members

Alex Holehouse; Gary Stormo; Robi Mitra; Shiming Chen

Degree

Doctor of Philosophy (PhD)

Author's Department

Biology & Biomedical Sciences (Computational & Systems Biology)

Author's School

Graduate School of Arts and Sciences

Document Type

Dissertation

Date of Award

3-30-2026

Language

English (en)

Author's ORCID

https://orcid.org/0000-0002-3288-7000

Download

Included in

Genetics Commons

Share

COinS