Is any protein target truly undruggable? 

Though most cancer targets have been considered conventionally “undruggable” (roughly 80 percent, according to Liron Bar-Peled, PhD, an investigator in Mass General’s Krantz Family Center for Cancer Research), that definition does not mean that solutions don’t exist—it means that the traditional approach of identifying a binding site and designing a molecule to fit it doesn’t work. 

“Undruggable” targets usually do not have a hydrophobic pocket structure that can bind small molecules, work through protein-protein interactions and the formation of protein complexes, have highly conserved active sites, or have intrinsically disordered or unknown tertiary structures. 

So, it’s not that these targets cannot be drugged but, rather, are challenging and require firing up a bit more of the human noggin (plus a bunch of computers) than it takes to throw small molecules at proteins and see what sticks. 

Michelle Chen - insilico medicine
Michelle Chen, PhD, Chief Business Officer at Insilico Medicine

“This is a problem that artificial intelligence (AI) is uniquely positioned to solve,” Michelle Chen, PhD, chief business officer at Insilico Medicine, told GEN. 

At the American Association for Cancer Research (AACR) Annual Meeting 2024, happening April 5–10 in San Diego, Insilico will show the latest fruits of their labor using AI in cancer drug discovery and development as the clinical-stage generative AI-driven drug discovery company will present five preclinical programs at poster presentations. 

“We want companies and fellow cancer researchers to see the remarkable value of AI in target and biomarker discovery and de novo drug design in developing new cancer drugs, driven by a team of cross-trained AI experts and drug developers,” said Chen. “We hope they will leave with assurance that Insilico’s AI platform can be used to create highly optimized, potent, and efficacious molecules that can serve as potentially best-in-class therapeutic options in treatment-resistant cancers and as promising candidates for partnering.” 

The inhibitors 

While Insilico has made its biggest splash with INS018_055, it’s lead program in fibrosis that received the FDA’s first Orphan Drug Designation for a drug discovered and designed using artificial intelligence AI, and has moved into Phase II clinical trials with patients, the AI company has been working on innovating therapeutics to treat cancer since its earliest days. 

“[Insilico Medicine has] particular knowledge in the cancer space, as that is where we first applied our earliest research in generative chemistry in 2016, applying generative adversarial networks (GANs) to generate novel small molecules against cancer,” said Chen. 

All five of Insilico’s announcements are for the discovery and development of novel “inhibitors,” which isn’t a coincidence. Many successful cancer drugs being used today turn out to be inhibitors against cancer targets, e.g., Gleevec (a tyrosine kinase inhibitor), Osimertinib (an EGFR inhibitor), Vemurafenib (a BRAF inhibitor), Bevacizumab (an anti-VEGF inhibitor), plus chemotherapy agents, many of which are also inhibitors. 

Insilico’s AI algorithms can help prioritize indications by analyzing information from a massive dataset that includes millions of omics samples (e.g. expression data, single cell sequencing, methylation data, proteomics data), pathway analysis, and text documents that include publications, clinical trials, patents, grants, and more. These data can be filtered based on novelty, safety, and tissue specificity. 

While Chen said that Insilico focuses on developing drugs against targets where there is a high unmet need and limited existing treatment options, these inhibitors were designed based on a scientific rationale, not a business decision. 

“To pick targets for cancer, we always look for a balance between novelty, confidence, and commercial tractability, as well as a maximum benefit to the patient,” said Chen. “We find targets that are moderately novel and have been mentioned in the literature. While these targets might be of interest to other companies, the programs are still in the early stages and demonstrate blockbuster potential and compelling mechanisms of action.” 

Chen also said that Insilico has been looking at treatment-resistant cancers. As more and more target therapies become available, tumors find a way to develop resistance to the current treatment, and how to overcome cancer treatment resistance has become important. That’s why they have developed ISM6331—a small-molecule inhibitor against TEAD, which is being presented at AACR. 

In vivo anti-tumor efficacy 

The five posters all demonstrate data that includes preclinical characterization that demonstrates either “in vivo anti-tumor efficacy” or “anti-cancer effects.” 

For ISM8001, a novel and selective FGFR2/FGFR3 dual inhibitor, this process has seemingly taken three or so months. Insilico announced in December that ISM8001 had been nominated for preclinical candidacy. At AACR, Insilico will give an update on their findings that show ISM8001 has strong anti-tumor activity in multiple cell line-derived xenograft mouse models, stopping tumor growth by 80 to 120 percent. 

Interestingly, at AACR 2023, Insilico presented posters for four candidates that showed preclinical promise, none of which are being presented at this year’s conference. However, that doesn’t mean that they’ve disappeared into the ether from which they came. 

 As an example, Insilico presented data for ISM3091 last year. It is a small-molecule inhibitor of USP1, which has become a synthetic lethal target for tumors with BRCA mutations and could be the best of its kind. On April 17, 2023, one day after Insilico presented preclinical data for ISM3091, the FDA cleared Insilico’s Investigational New Drug (IND) application for the candidate. Since then, Insilico and Exelixis have signed an exclusive license agreement that gives Exelixis the worldwide right to develop and market ISM3091. It is currently being tested in a first-in-human Phase I study with people who have advanced solid tumors that do not have homologous recombination. 

In addition to the $80 million upfront deal with Exelixis, Insilico made a deal worth up to $500 million with Menarini for a KAT6 cancer inhibitor to treat breast cancer (which was not presented at either AACR 2023 or 2024).  

Read more about the AI-generated drug development and discovery process in GEN’s in-depth interview with Insilico founder and CEO Zhavoronkov.  

Previous articleDigital In Vitro Transcription Models Could Aid mRNA Manufacturing
Next articleIambic Rhythm: AI Drug Developer Enters the Clinic, Targeting HER2 Cancers