SUMMARY
The microtubule poison paclitaxel (PTX) is standard-of-care for breast cancer treatment for all breast cancer
subtypes in both early stage and metastatic disease. However, ~50% of breast cancer patients do not benefit
from PTX. Though PTX causes mitotic arrest at high concentrations in cell culture, our data from two biomarker
studies demonstrate that PTX levels in patient tumors are too low to cause mitotic arrest. Instead PTX causes
abnormal multipolar mitotic spindles. Division on multipolar spindles causes unequal chromosome segregation
(termed chromosomal instability, or CIN). While low rates of CIN are common in tumors, increasing the rate of
CIN above a maximally tolerated threshold is lethal. Since PTX increases CIN, these data support a model in
which PTX is effective when it increases CIN over the maximally tolerated threshold. Our long term goal is to
use this novel mechanistic insight to convert PTX from generic chemotherapy into precision medicine by
developing a) a biomarker to predict which tumors will respond to PTX, and b) method(s) to sensitize the ~50%
of resistant cancers. Based on our preliminary data, we hypothesize that there are two major determinants of
PTX sensitivity. First, tumors that have endogenous CIN close to the maximally tolerated CIN threshold are more
sensitive to PTX. Second, some tumors can focus PTX-induced multipolar spindles into near-normal bipolar
spindles by mid mitosis and dramatically reduce PTX-induced cytotoxicity—even when PTX affects spindles in
early mitosis. Since PTX treatment increases the incidence of multipolar spindles in early mitosis, this implicates
mid mitosis focusing of PTX-induced multipolar spindles as a second mechanism of PTX resistance. Aim 1 will
test the hypothesis that pre-treatment CIN correlates with PTX response in primary breast cancer, as determined
in an ongoing biomarker study of neoadjuvant single-agent PTX and a retrospective analysis of I-SPY2 samples.
Additionally, Aim 1 will test whether experimentally inducing CIN enhances PTX response. Aim 2 will enable
state-of-the-art quantitative measure of CIN by a) employing single-cell DNA sequencing (scDNAseq) to quantify
all chromosomes; and b) accounting for dropout of highly aneuploid cells from the tumor population through a
technique known as Approximate Bayesian Computation. Aim 3 will experimentally test how pole focusing
controls PTX response. Additionally it will test the hypothesis that alisertib, an inhibitor of AurkA, can prevent
pole focusing. Preventing pole focusing is expected to sensitize tumors that are resistant to PTX by forcing cells
to undergo multipolar divisions, markedly elevating CIN. Finally, a CRISPR screen will be used to identify genes
that confer this mechanism of primary resistance. A gene signature for pole focusing could improve a CIN-based
predictive biomarker of PTX response. Together, this translational work based on the conceptual advance
provided by our previous translational studies will provide basic knowledge of the biological mechanisms
regulating cell death in response to a cornerstone of anti-cancer treatment. This will provide a sound basis for
the future development of a predictive biomarker of PTX and strategies for sensitizing PTX-resistant cancers.