Project Summary/Abstract
Most tumors are not homogeneous, but rather comprised of highly distinct and heterogenous cancer cell
populations. Adjacent cells within a tumor may harbor different genetic alterations and have different phenotypes.
Thus, complex tumors are more 'than the sum of their parts', which contributes to aggressiveness, metastatic
potential, and drug resistance. This complexity of intratumor heterogeneity is a major challenge in cancer
therapy. In contrast to normal cells in which chromosome segregation is tightly controlled to maintain a diploid
chromosome set, cancer cells are frequently aneuploid. Indeed, abnormal gains or losses of chromosomes are
amongst the most common characteristics of cancer cells with nearly all solid cancers exhibiting some type of
aneuploidy. A high degree of tumor aneuploidy also correlates with poor clinical outcomes. Although this strong
correlation between aneuploidy and cancer is well known, causal relationships are incompletely understood.
Especially in normal tissues and during early stages of carcinogenesis, chromosome segregation errors are rare
and transient events that are notoriously difficult to study. To advance our understanding how chromosome
segregation errors drive cancer initiation, development, evolution, and heterogeneity and enable novel types of
model systems to, for example, explore aneuploidy as a therapeutic target in patient-derived cancer organoids,
this pilot project proposes to develop a molecular and imaging toolkit to observe and manipulate chromosome
segregation dynamics with high spatial and temporal accuracy in three-dimensional model and patient-derived
cancer organoids. In Aim 1, we develop methods to identify, count and track specific chromosomes during
multiple cell divisions in cancer organoids. Two proposed approaches involve endonuclease-deactivated dCas9
combined with gRNAs to create unique chromosome-specific spectral barcodes or using dCas9 imaging of
repetitive DNA sequences to generate chromosome-specific fluorescence patterns, with the long-term goal of
enabling "live karyotyping" in cancer organoids using deep learning networks. In Aim 2, we develop three
alternative strategies to control chromosome dynamics and integrity using optogenetics. The proposed
optogenetic actuators include a light-controlled kinetochore-microtubule interface, an optogenetic chromosome
trap to immobilize specific chromosomes during cytokinesis, and photoactivated Cas9 to produce acentric
chromosome arms or induce chromothripsis. The effectiveness of these strategies will be evaluated in mammary
epithelial and breast cancer models, with the goal of understanding how acutely induced segregation errors or
chromothripsis of specific chromosomes affects cancer cell dynamics and tumor evolution.