The Sheltzer Lab


Genome imbalances and tumorigenesis

Human cancers exhibit a diverse array of genomic gains and losses that alter the dosage of hundreds or thousands of genes at once. The prevalence of aneuploidy in cancer – first noted more than 100 years ago – has led to a widespread belief that genomic imbalances play a crucial role in tumor development. Indeed, in the early 20th century, Theodor Boveri speculated that abnormal karyotypes altered the balance between pro- and anti-proliferative cellular signals, and were therefore sufficient to induce transformation. “Boveri’s hypothesis” has motivated decades of research into the origins and consequences of aneuploidy, but the precise relationship between abnormal karyotypes and tumorigenesis remains unclear.

We are developing novel models of aneuploidy to explore the impact of genome dosage alterations on tumor development and progression. Using a variety of techniques, including CRISPR/Cas9, microcell-mediated chromosome transfer, and small-molecule mitotic accelerants, we are changing chromosome copy number in human cells. We can then study how these aneuploidies impact a number of cancer-related phenotypes, including metastasis, chemotherapy resistance, and cell cycle progression.

While aneuploidy is a ubiquitous feature of human tumors, it occurs rarely in somatic cells. Thus, differences between aneuploid and euploid cells may represent crucial therapeutic vulnerabilities in cancer. By identifying phenotypes that are shared among cancer cells with different chromosomal imbalances, we hope to discover pathways that can be manipulated to selectively eliminate aneuploid cells. Drugs that target these pathways may have broad utility against a wide range of aneuploid cancers, while exhibiting minimal toxicity in euploid tissue.

Sheltzer, J.M., Ko, J.C., Replogle, J.M., et al. (2017) Single-chromosome gains commonly function as tumor suppressors.  Cancer Cell 31, 1-16.

Sheltzer, J. M. and Blank, H. M. and Pfau, S. J. and Tange, Y. and George, B. M. and Humpton, T. J. and Brito, I. L. and Hiraoka, Y. and Niwa, O. and Amon, A. (2011) Aneuploidy drives genomic instability in yeastScience 333.

cancer dependencies - corrected.png

Identifying cancer dependencies and the true targets of anti-cancer drugs

Substantial progress has been made in the treatment of certain malignancies by targeting cancer ‘addictions’, or genetic dependencies that encode proteins required for the survival and/or proliferation of cancer cells. Therapeutic agents that block the function of a cancer dependency – like the kinase inhibitor lapatinib in Her2+ breast cancer – can trigger apoptosis and durable tumor regression. Discovering and characterizing druggable cancer dependencies is a key goal of preclinical research.

We are using CRISPR/Cas9 to identify genetic addictions in different cancer types. While screening several cancer cell lines, we discovered that many genes previously reported to be both cancer-essential and the target of anti-cancer drugs are actually dispensable for cancer growth. For instance, we found that MELK, a putative “addiction” in multiple cancer types, could be eliminated using CRISPR without any detectable loss in cancer cell fitness. Additionally, we demonstrated that OTS167, a small-molecule inhibitor of MELK undergoing phase II clinical trials, continued to kill MELK-knockout cancer cells with no decrease in potency. This indicated that an anti-cancer agent had entered clinical trials in human patients due to flawed preclinical data and based on an incorrect understanding of that drug’s mechanism of action.

We are working to discover how preclinical cancer research can be conducted in a robust and reproducible manner to prevent similarly-flawed results from driving the treatment of human patients. Moreover, we are applying a variety of genetic and biochemical approaches to uncover the true mechanisms-of-action of potent anti-cancer drugs that have been mis-characterized.

Giuliano, C.J., Lin, A., Smith, J.C., Palladino, A.C., and Sheltzer, J.M. (2018) MELK expressed correlates with tumor mitotic activity but is not required for cancer growtheLife 7:e32838.

Lin, A., Giuliano, C.J., Sayles, N.M., and Sheltzer, J.M. (2017) CRISPR/Cas9 mutagenesis invalidates a putative cancer dependency targeted in on-going clinical trials. eLife, 6:e24179.

Screen Shot 2019-02-21 at 4.24.56 PM.png

Discovery and characterization of genes affecting survival time in cancer

Cancers that arise from the same tissue can exhibit vast differences in clinical behavior. For instance, among individuals diagnosed with early-stage colorectal cancer, about 60% of patients will be cured by surgery alone, while the remaining 40% will experience a recurrence that is frequently fatal. Biomarkers that can successfully differentiate between patients with benign and aggressive cancers could lead to improved risk prediction, better clinical management, and a decrease in dangerous and unnecessary over-treatment.

In order to gain a global understanding of the genomic features in a primary tumor that influence cancer prognosis, we are collecting and analyzing molecular profiles from tens of thousands of patients with known clinical outcomes. These data have revealed that mutations in almost all cancer driver genes contain remarkably little information on patient prognosis. However, copy number alterations in these same driver genes harbor significant prognostic power. Focal CNAs are associated with worse outcomes than broad alterations, and CNAs in many driver genes remain prognostic when controlling for stage, grade, TP53 status, and total aneuploidy. The biological importance of cancer copy number alterations is unexpected – and a topic of continuing investigation.

More information on our cancer survival analysis can be found here:

Smith, J.C., Sheltzer, J.M. (2018) Systematic identification of mutations and copy number alterations associated with cancer patient prognosiseLife e39217.

Sheltzer, J. M. (2013) A transcriptional and metabolic signature of primary aneuploidy is present in chromosomally unstable cancer cells and informs clinical prognosisCancer Res 73 (21).