ľ¹Ï¸£ÀûÓ°ÊÓ

My older brother is a scientist and his enthusiasm for the field definitely helped me lean the same way, but for a long time, I was still unsure. Until my first internship, which sparked my fascination with fundamental biological mysteries, and in particular, the complexities of DNA damage and cell division. Little did I know that one day, I would have the chance to discover something that could directly impact patients with cancer.  

Defining the problem

It started with investigating the fundamental inner workings of cell division. It’s well-established that normal cell division is exquisitely controlled and requires the separation of our chromosomes (our DNA). This is facilitated by microtubules, structures inside the cell that act as a scaffold for the chromosomes, pulling them apart into identical daughter cells.

Cancer is the perfect example of cell division gone wrong. Mutations occur in the genes that control cell division allowing cancer cells, for example, to divide more quickly and evade stop signals. If we can interrupt microtubule formation, we can prevent cancer cells from dividing and making more copies.

Unraveling the molecular target of an anti-cancer therapy

Many anti-cancer drugs have been developed, yet few make it to the market, often because of a lack of a clear understanding of how the drug works at the molecular level. One interesting anti-cancer drug is rigosertib, developed by Onconova Therapeutics, Inc. Its recent phase III clinical trial generated disappointing results: there was no overall benefit to patient survival of those who received the drug compared with those who did not. However, the clinical trial did not look into how rigosertib works (the mechanism of action), which may have allowed more critical selection of patients to enroll.

For more than a decade, different mechanisms of action for rigosertib have been reported, and some results appear to be conflicting. Thus, the molecular target and mechanism of action remain challenging and unclear. We were inspired to see if we could develop a novel approach for target identification and elucidating the mechanism of action for rigosertib.

CRISPR/Cas9 may help break the case

We used a modified version of the Cas9 protein, dead-Cas9, which renders the protein inactive, so that it still targets but doesn’t cut DNA. It merely occupies space at a specific location on that DNA. When fused to transcriptional activation or inhibition proteins, this introduces a new method for turning genes on and off without modifying the gene itself (CRISPR-activation and CRISPR-inhibition, respectively).

We’re the first to combine these two technologies (CRISPRi/a) to look at the opposing principles within the context of a single drug. We set out to figure out how some cancer cells become resistant to drug treatment, with the hope of identifying rigosertib’s true mode of action.

How we used CRISPR/Cas9

This is a very simplified version of how we conducted the screens. Details on our methods and results were recently published in Molecular Cell, 2017. Capitalizing on the power of CRISPR/Cas9 to target precise DNA sites, we screened cancer (leukemic) cells for 15,000 genes, and identified genes whose overexpression or inhibition affected rigosertib-induced cell death. We cultured the cells with and without the drug, rigosertib. Cells in which we overexpressed genes protecting them from the drug will be over-represented in the drug-treated cells compared with untreated cells. In the case of the CRISPR-inhibition screen, the exact opposite happens.

Combined CRISPRi/a enabled us to identify genes that are true targets, and weed out the noise. Let me explain. Rigosertib kills a cell at the moment it’s about to divide. Through the CRISPRi screen, say we identify gene X that interferes with cell division at a point prior to when rigosertib intervenes. This interferes with the ability of ribosertib to act, since the cells will not reach the point of division. From this single screen, we see that gene X prevents cell death, however we don’t know whether it’s because gene X interferes with the drug or not.

On the other hand, CRISPRa of gene X does not interfere with cell division, and therefore not affect ribosertib-dependent cell killing. Thus, we then know that gene X is not a true rigosertib target. Thus, genes that are identified in only one of the two screens contribute to noise.

CRISPR/Cas9 helps find new ways to battle cancer

Using CRISPRi/a, we’ve identified a whole set of genes that either antagonize or synergize with rigosertib. Through our investigation of two in particular, we’ve figured out that rigosertib does indeed target microtubule dynamics, and that these genes make cancer cells susceptible to rigosertib. This may provide a new set of biomarkers for the drug and allow better stratification of patients with cancer, such that patients with tumors that overexpress one or more of these genes could truly benefit from receiving rigosertib. 

The power of our combined CRISPRi/a approach is that it’s precise, robust and comprehensible. It’s applicable to other drugs as well, and may be a new method for identifying clinically relevant drug-gene interactions in our pursuit of improving customized therapies for patients. 

Lenno Krenning, PhD
Post-doctoral fellow
Group: Marvin Tanenbaum, PhD
Hubrecht Institute
Science for Life