CRISPR kickstarts combination drug discovery for neuroblastoma

One of the reasons cancer is so hard to treat is because it changes rapidly. Cancer cells mutate their DNA more readily than healthy cells while also changing to survive and grow uncontrolled. These rapid changes are one of the reasons cancer is so difficult to treat. Even when treated with a single drug, some cancer cells may already have or can develop mutations that enable them to resist therapy. After treatment, only those resistant cells remain and remake a tumor that is even harder to treat than before.

The late Donald Pinkel, MD, inaugural St. Jude CEO and director, was among the first to recognize that combining multiple chemotherapeutic drugs into a treatment plan could overcome resistance. That insight led to the increase in survival of pediatric acute lymphoblastic leukemia (ALL) patients from 20% to 50%. That treatment method still serves as the basis of modern therapy, which has achieved 94% survival rates at St. Jude.

Neuroblastoma is a type of childhood solid tumor that begins in early nerve cells called neuroblasts and can affect multiple organs. Unlike ALL, neuroblastoma has lower survival rates and can be more resistant to treatments. St. Jude researchers may have found a way to identify novel therapies in an approach published recently in Nature Communications.

Finding a path towards a better combination therapy for neuroblastoma

“Relapsed neuroblastoma is very tough to treat,” said co-corresponding author Adam Durbin, MD, PhD, St. Jude Department of Oncology, a physician-scientist who treats the disease. “For the 60% who survive and are cured of their initial disease, the combination of multi-agent chemotherapy, radiation, surgery, two stem cell transplants and immunotherapy has a lot of toxicity and side effects.”

Since this solid tumor presents as a disseminated disease (affecting multiple tissues), this difficult therapeutic regimen — one of the most intensive treatments in pediatric oncology — is necessary. Finding novel drug combinations that could improve treatment for neuroblastoma is paramount.

Potentially thousands of drugs and compounds are therapeutically available however, making it difficult to screen for which combinations would improve the current standard of care. The sheer volume of experiments needed is impractical. To address this, St. Jude researchers used a clever approach fueled by gene editing technology to identify these potential combinations faster.

“This study shows a proof-of-principle for a scalable drug combination screening strategy,” said senior corresponding author Paul Geeleher, PhD, St. Jude Department of Computational Biology. “Instead of testing each of the hundreds of thousands of unique drug-drug combinations, we’re using CRISPR gene editing to knock out genes that we already know to be druggable, then giving those neuroblastoma cells a known chemotherapeutic to look for synergistic effects.”

“Classical screening that combines different drugs in many cell lines is challenging to do at scale,” said co-corresponding author Jun Yang, MD, PhD, St. Jude Department of Surgery. “CRISPR can significantly speed up the discovery by reducing the amount of labor needed.”

CRISPR screens can be performed in a fraction of the time as a standard screen, even by a single individual. The benefits don’t stop there.

“Maybe the biggest advantage to CRISPR prescreening is that it is genetically clean,” Yang explained. “When you combine drugs together, the results can be murky because they may have multiple impacts. The CRISPR system knocks out a specific gene, giving you greater insight into the exact mechanism of action.”

A promising combination to improve neuroblastoma treatment

The study found multiple promising hits. One example was using the common chemotherapy drug doxorubicin in combination with a knockout (inactive or removed form) of the DNA-dependent protein kinase catalytic subunit (PRKDC) gene. The scientists validated the hit, finding that the combination of doxorubicin and a PRKDC inhibitor controlled tumor growth more than either drug alone in mouse models.

“We found that when we combined doxorubicin with a PRKDC inhibitor, we saw a very strong synergistic effect,” Yang said. “The impact was greater than the standard of care therapy in that system and greater than simply adding the effects of the two drugs.”

To understand how that synergistic effect was happening, the researchers looked at the mechanism of action of doxorubicin and the role of PRKDC in the cell. Doxorubicin creates DNA damage to trigger cell death pathways, such as apoptosis. PRKDC is involved in a major DNA damage repair pathway called non-homologous end-joining. After doxorubicin damages DNA, PRKDC inhibitors prevent repair of that damage, causing cell death that may have otherwise been avoided.

Knowing the mechanism of action will help further develop the combination treatment in preclinical work. While many drugs have shown promise in such studies only to be toxic in practice, the screen was designed to account for that possibility.

Proactively taking out the toxicity of potential neuroblastoma treatments

“The goal, ultimately, is to find better treatments that are more targeted with less toxicity,” Durbin said. “Plus, we want to find effective treatments for children with relapsed disease. So that kids can grow up and not suffer from all of the problems related to high-intensity neuroblastoma therapy.”

“We wanted to identify less toxic therapies by finding drug combinations that are synergistic in cancer cells, but not normal cells,” Geeleher explained. “It’s a big problem with combining drugs — it’s hard to predict how they will affect cancer cells versus how they affect normal cells and how toxic they will be in combination.”

The screening process included an “outgroup” to account for cells that weren’t the targeted cancer. While the researchers were most interested in how drug-gene knockout combinations would affect neuroblastoma cells, they also treated various non-neuroblastoma cell lines simultaneously with the same combinations. These non-neuroblastoma cells were the outgroup — a group of cells to look for side effects and off-target toxicity that was not neuroblastoma-specific.

Using these outgroup cells as a point of comparison, the researchers could deprioritize drug and gene knockout combinations that would likely cause excessive toxicity in both the desired targets and healthy cells. More importantly, they could identify those special few combinations that only impacted the targeted neuroblastoma.

Through this work, the researchers created a robust system capable of recapitulating known drug combinations and identifying potentially novel ones for pediatric neuroblastoma. This was possible while accounting for potential toxicity — one of the major barriers that hinder potential therapeutics. The system itself may apply to other cancers and targets, which could result in further discoveries.

“Ultimately, we’ve designed a new strategy to enhance the activity of the drugs that we’re currently using,” Geeleher concluded. “And since these are already approved drugs, we hope we can accelerate the adoption of new combinations into the clinic.”