NEW YORK – For the past year, researchers at ChristianaCare's Gene Editing Institute have been working to combine CRISPR genome editing and chemotherapy in non-small-cell lung cancer (NSCLC) patients, in order to improve the efficacy of certain chemotherapy agents.
Their idea had been to use CRISPR-Cas9 gene editing to knock down the NRF2 gene, which protects cells from various types of damage, in order to render patients more sensitive to chemotherapeutic agents such as cisplatin and carboplatin.
Now, the researchers have found a way to make the treatment even more precise, by targeting NRF2 only in tumor cells. Not only could this allow oncologists to use lower doses of chemotherapeutic agents to shrink tumors, but this kind of precision would also leave NRF2 active in normal cells, protecting them from the toxic effects of the chemotherapy.
In May 2019, Gene Editing Institute Director Eric Kmiec and his colleagues were preparing to file an investigational new drug application (IND) with the US Food and Drug Administration for a clinical trial protocol that would combine CRISPR and chemotherapy in KRAS-positive NSCLC patients.
NRF2 is a master regulator of 100 to 200 genes involved in cellular responses to oxidative and/or electrophilic stress. Researchers from the Gene Editing Institute had published a proof-of-concept study in Molecular Therapy Oncolytics in December 2018 demonstrating that using CRISPR to disable the gene in lung cancer cells increased sensitivity to chemotherapeutic agents in tissue cultures and reduced proliferation of cancer cells. They confirmed their observations in xenograft mouse models, where they found that homozygous knockout cells proliferated at a slower rate than the wild-type cells, even in the absence of treatment with chemotherapy.
Kmiec and his colleagues planned to use CRISPR-Cas9 gene editing to disable NRF2 in lung cancer cells, rendering the gene incapable of producing a functional protein and making those cells more sensitive to cisplatin, carboplatin, and vinorelbine (Navelbine). With NRF2 shut down, the genes responsible for the efflux of the anticancer drugs would not be activated, even under the most environmentally stressful conditions.
At that time, the group had completed its animal studies and had engaged in conversations about next steps with the National Center for Advancing Translational Sciences. The institute also had informal, but encouraging, conversations with the FDA that motivated it to pursue a clinical trial for regulatory approval.
While conversations with the FDA continued to be productive throughout the summer and fall of 2019, Kmiec said, the emergence of SARS-CoV-2 and the subsequent pandemic forced the agency to shift its focus. Although this will cause a change in the institute's timeline — as it has for so many research groups — it will also give the researchers more time to explore a hypothesis that emerged during its talks with regulators about how to make the treatment regimen more targeted.
"One of the worries about any biotherapeutic is that it is not going to only act on the tumor cells," Kmiec said. "There have been a thousand different iterations of delivering molecules that are supposed to deliver payloads to tumor cells and not to normal cells, but that hasn't happened. But, sometime late last year, a graduate student in our lab discovered that in certain kinds of cancer cells, a unique site is created early in the mutagenesis pathway that creates a unique activity site for CRISPR-Cas9."
What the student in Kmiec's lab discovered was that the protospacer adjacent motif (PAM) — the bit of DNA that serves as a recognition site for the Cas nuclease to lock on to — was also mutated in NRF2 in the tumor cells, just enough so that the researchers could design a guide RNA that would take the CRISPR-Cas9 editing complex to the gene in the tumor cells, while sparing the gene in normal cells.
"We had kind of given up on trying to determine a vehicle that would only bring the CRISPR-Cas to a tumor cell. But now, we can deliver the CRISPR-Cas complex to all cells, and in our hands so far, this will only cleave the genome of the tumor cells," Kmiec said. "We're going to use the mutagenic programming that goes on in the tumor cell against itself."
Indeed, this important new finding has caused Kmiec and his colleagues to change tack, somewhat, starting with the lung cancer subtype that they were working on. Though they were originally studying adenocarcinoma, they've now switched to squamous cell carcinoma. Most lung cancer research is focused on adenocarcinoma, Kmiec explained, and there's a lot of targeted therapy research in adenocarcinoma that's focused on NRF2. The group wanted to see if the same could be done for squamous cell carcinoma.
"This doesn't mean that the technique would not be useful for adenocarcinoma as well," Kmiec said. "We also think this occurs with esophageal cancer, where these changes make CRISPR act only at the tumor cell and not the normal cells. We made a scientific decision to pursue [squamous cell carcinoma] because that puts it into a zone that is quite unique and should move things forward quickly."
However, this change in tactics did require researchers to start over and test the toxicology and efficacy of the technique in animal studies, as well as engage in new pre-IND talks with the FDA.
They published a study in Molecular Cancer Research earlier this month detailing their identification of the unique PAM that facilitates site-specific cleavage of NRF2 present only in tumor genomes.
They cataloged somatic NRF2 mutations in various cancer cases reported in The Cancer Genome Atlas and reported 214 cases of NRF2 mutations, predominately found within the protein's Neh2 domain, also known as its KEAP1-binding domain. These mutations disrupt KEAP1 binding and lead to constitutive expression of NRF2 in cancer cells. The most common NRF2 mutation that they observed, R34G, created the new PAM for Cas9 recognition, which consisted of two to six nucleotides juxtaposed to the CRISPR seed binding sequence, facilitating alignment and cleavage of the DNA.
In their study, the researchers also concluded that this technique could eventually be used in many different types of cancer where NRF2 keeps cells resistant to chemotherapy. "By destroying the NRF2 gene in a tumor cell population, the source of chemo-resistant cells is immaterial, because CRISPR-directed gene editing is active on tumor cells regardless of origin," they wrote. "A differential PAM site exists in tumors from esophageal, head and neck, uterine-endometrial, and bladder-urothelial cancers, which might provide a path to exploit tumor cell sequence variance as a method of selection."
Indeed, Kmiec said, the group's overarching goal is to prove that solid tumors can be treated with CRISPR.
"We're trying to develop more of a [treatment] paradigm here," he explained. "Could this be a general principle approach to other solid tumors, both in NRF2 and other genes?" Kmiec believes the answer could be yes, if a PAM site is created that is unique to the tumor cell of a lineage of normal cells.
"It solves a huge problem, and that is how you deliver the active molecule to a tumor cell as opposed to a normal cell," he continued. "How to use that CRISPR is unique — it has to get to the site, but once it gets there, if it acts only on the tumor DNA, then it is much more effective than trying to bang away at some receptor on the cell surface."
The next step, Kmiec said, is to determine the frequency within the population of the R34G mutation the researchers wrote about in their study. Toward that end, they are combing through available databases to look at sequence data from squamous cell carcinoma patients. Once that work is done, the group can begin thinking about studying this technique in human trials.
"We believe it's going to work," he said. "The proof-of-concept work is all done and we have to define a patient population. So that's kind of the next step in our focus right now."