Looking backward and forward at cancer-associated mutations
with precision genome editing
Name
gould-samgould-phd-biology-2026-thesis .pdf
Size
130.01 MB
Format
Adobe PDF
Checksum (MD5)
8cb246e70910b741724b096b0cf30795
Author(s)
Gould, Samuel I.
Advisor(s)
Sánchez-Rivera, Francisco J.
Date Issued
February 2026
Publisher
Massachusetts Institute of Technology
Abstract
Every cancer patient is unique. Their tumors harbor thousands of mutations layered onto an already unique genome and continually evolve into genotypically distinct clones under the selective pressures imposed within the tumor microenvironment. While clinical sequencing efforts have generated libraries of millions of cancer-associated mutations, the overwhelming majority of these mutations have not been functionally characterized. Determining if and how specific mutations drive malignant phenotypes – including proliferation, metastasis, and therapeutic resistance – is essential to improving cancer treatment and anticipating mechanisms of drug failure.
Precision genome editing now enables the direct installation of patient-derived mutations in their native genomic context, offering a powerful approach for functional interpretation at scale. However, variability in editing efficiency and other technical challenges have limited systematic endogenous mutational analysis. In this thesis, I develop experimental and computational platforms to enable large-scale, endogenous functional characterization of cancer-associated mutations, and use these technologies to discover novel phenotypes mediated by alterations in tumor suppressor genes and oncogenic drivers.
First, I apply prime editing to generate >1,000 clinically observed TP53 mutations, establishing a “prime editing sensor” strategy in which an integrated reporter cassette simultaneously measures prime editing guide RNA (pegRNA) efficiency and variant fitness. Filtering low-efficiency pegRNAs enables accurate identification of known and novel pathogenic alleles in TP53, the most frequently mutated gene in cancer. I further show that creating variants within their native genomic context reveals phenotypes masked in exogenous overexpression systems, underscoring the necessity of endogenous mutation engineering for accurate functional characterization.
Second, to dissect and anticipate resistance to precision therapeutics, I perform base editing tiling mutagenesis across the cyclin-dependent kinase (CDK) family to probe residue essentiality and identify resistance alleles across 15 targeted therapies, including ATP-competitive inhibitors, PROTACs, and molecular glue degraders. This work uncovers both established and novel resistance variants, which I validate extensively in vitro and in vivo. Leveraging evolutionary relationships among the CDKs, I identify shared and divergent resistance mechanisms within the family, and perform iterative mutagenesis to model the sequential mutation acquisition observed in patients treated with multiple targeted therapies. Finally, I demonstrate the clinical relevance of this work by identifying patients with CDK6 mutations whose response to CDK4/6 inhibitors is consistent with our experimental findings.
Altogether, my work establishes scalable platforms for endogenous mutational interrogation and reveals functional consequences of clinically relevant mutations in TP53 and the CDK family. By integrating clinical genomics, genome editing, and high-throughput screening, my work illuminates how existing – and future – cancer-associated mutations confer malignant traits and resistance, advancing efforts to predict tumor evolution and rationally design more durable therapies.
MIT Department
Massachusetts Institute of Technology. Department of Biology
Terms of Use
In Copyright - Educational Use Permitted
Copyright retained by author(s)
Persistent DSpace Link
