Opening Keynote Spotlights Novel Target in Genomically Unstable Tumors

BOSTON—When I first attended SLAS two years ago, I was struck by how much automation is embedded in virtually all areas that GEN covers, from wet-lab research through drug discovery all the way through to bioprocessing. So, it came as no surprise that the keynote which kicked off the first full day of talks at this year’s meeting in Boston, slated to run from Feb 7–11, placed biology squarely at center stage.

The opening keynote was delivered by Serena Silver, PhD, chief scientific officer from Accent Therapeutics. The bulk of her talk focused on the work on KIF18A protein as a target for precision oncology therapies but included mentions of the importance of automation particularly for performing complex high-throughput screening assays in pharma and biotech. Prior to joining Accent, Silver led the molecular pharmacology group at Novartis, and after that worked at Fulcrum Therapeutics leading the development of new assay modalities and complex disease models.

Now at Accent, her focus is on developing small molecule protein inhibitors in the context of oncology. One such protein is KIF18A, part of a family of motor proteins. By way of background from the keynote, several studies into KIF18A have highlighted its role in mitosis particularly in cells with a high degree of genomic instability due to things like defective DNA repair or recombination defects or chromosomal rearrangements.

“What’s really exciting and intriguing about KIF18A is that unlike many other kinesins…which are required in every cell of your body every time it wants to divide, KIF18A is really only required in cells where there’s a high degree of chromosomal instability” and that “need some extra help to get through mitosis,” Silver explained during her talk. “We and others became really excited about this target several years ago [when] a number of publications showed that in those cancer cells … when you inhibit or you knock out KIF, you get cell division stalling” and ultimately apoptosis. In contrast, normal cells or cancer cells that do not have chromosomal instability, cell division proceeds normally “even if you’ve knocked out KIF.”

Throughout the presentation, she shared examples where KIF18A was inhibited in normal cells and stable cancer cells that remained largely unaffected following treatment and contrasted them with the response to treatment in cells that were in a state of chromosomal instability, which were affected.

This selectivity presented an opportunity for Accent to develop KIF18A inhibitors that target cancers with high levels of chromosomally unstable cancer cells. To that end, the company has developed a small molecule drug dubbed ATX-295, which Silver described as “a highly selective and potent KIF18A inhibitor with good drug-like properties” that works by preventing KIF18A from moving along the mitotic spindle during cell division disrupting chromosome alignment in unstable cells. It accomplishes this by binding the interface of tubulin and the KIF18A protein.

This particular small molecule was initially identified by Amgen several years ago. Accent built on that foundation by solving its crystal structure and making adjustments to its scaffold. This included changes to attached groups such as performing a silicon atom replacement. Essentially, “we looked for ways that we could improve on it” to “actually reinforce interaction,” she explained.

Among other datasets, Silver shared results from testing the modified molecule both in ovarian cancer cell lines and xenograft models which showed that the inhibitor caused dose-dependent tumor regression in a chromosomally unstable model but had no effect on a chromosomally stable one. Additionally, she shared data from a large-scale screen performed using the Broad Institute’s PRISM platform on nine cell lines and 45 major tumor types. That screen confirmed that cells with whole genome duplications, used as a surrogate measurement for chromosomal instability, were more sensitive to Accent’s inhibitor. The screen also identified other potential indications for the drug, including head and neck, lung, and breast cancers which have high levels of genomic instability.

“The beauty of this is also not just the cell viability readout in these nine cell lines, it’s that it’s incorporated and integrated with all of the data that’s been generated over the years, including data from CRISPR [experiments], as well as expression data, sequencing data and other packets of information,” she said. It is “incredibly powerful to bring that all together [and] what’s also exciting to us is that … we could get all the raw data, put [it] into our own profiling systems and analysis pipelines as well.”

The company launched a Phase I/II clinical trial for ATX-295 last year and has begun dosing patients. The treatment is indicated for adult patients with advanced/metastatic platinum-resistant or refractory ovarian cancer.

Silver wrapped up her talk with a nod to AI’s potential in oncology particularly for learning from tumor images. As an example, Accent has “started to explore ways that we might use AI image analysis to identify chromosomal instability from tumor tissue” using solutions from Imagene, a developer of foundation models for precision medicine.