Controlling the fate of cancer cells with SF3B1 modulators
Posted 16th August 2019 by Jane Williams
Andrew Cook has worked in the pharmaceutical industry for over 20 years. Throughout his career, he has endeavoured to do things differently, a little out of the ordinary, leading him to his work at H3 Biomedicine on splice modulators. “Part of the reason it attracted me” he says, “is it’s so different. It’s a natural product-based drug discovery effort, and I had never done that before”.
“With SF3B1, we’re trying to control the fate of cancer cells by changing the expression of key proteins like Mcl-1 that control their survival. If we can modulate them by changing the way their RNA is expressed to an inactive form, there may be ways we can tackle diseases not possible by conventional techniques.”
What are the overall aims of H3? You’ve mentioned cancer – is there a specific type of cancer H3 is looking at?
H3 is unusual for a biotech in that we are not platform based and we’ll take on any type of cancer. As we are wholly owned by a Japanese pharmaceutical company called Eisai, the remit we have been given is not only to look at the more conventional targets, but also smaller niches where bigger pharma companies haven’t focused. We try to find anything that will tackle cancer, mainly small molecules, but other modalities too, including ADCs.
What is your research on currently?
SF3B1 is a key component in the proper processing of pre-messenger RNA. This messenger RNA is translated in every cell to produce the proteins required for normal function. Hotspot mutations in SF3B1 or other proteins that process mRNA are associated with certain types of cancer.
Through work conducted at Eisai, we had access to SF3B1 modulators and realised that the use of these compounds on cancer cells containing SF3B1 hotspot mutations might allow us to control the fate of these tumours. In effect, you stop those cancer cells being able to process messenger RNA and cause them to die while leaving normal cells untouched. The idea is that you have a very different way of trying to tackle a disease like cancer which is orthogonal to the more conventional approaches currently in the clinic.
Do you have any measure of how common that mutation is?
My presentation at the Medicinal Chemistry & Protein Degradation Summit is focused mainly on bladder and uveal melanoma. They have specific SF3B1 hot spot mutations which are in the 5 to 20 percent range.
Eisai had worked on SF3B1 splice modulators in the early 2000s, developing a clinical compound before actually discovering the target. We were able to use that work in combination with phenotypic screening of these hotspot mutation cell lines to enable the discovery of H3B-8800.
I’m going to show some of the structures and workings of splicing modulators. I’ll talk about discovery but also give a bit of background on how these molecules work. The initial work was phenotypic but now, with access to protein crystal structures, you can develop hypotheses about how the molecules work in the splicesome complex.
What are your thoughts on the introduction of AI and machine learning into the medicinal chemistry process?
It’s a tool that you should be using. It gives you another way of approaching drug discovery and allows access to other people’s results and learning outside of your current experience.
I heard someone from Roche recently say at a conference that the best software can now do as well as a good medicinal chemist. In three years’ time, they are hoping compound design will become a synergy between people and computers; in seven years’ time, who knows what computers will be able to do. But there’s always things that people can do that the software can’t.
I’m a big supporter of AI and people should embrace it, not fear it. The job will change, but it will help productivity. About 20 years ago, most of the drug targets were easier, but a lot of those have been done. Currently the industry is moving towards more difficult targets and approaches. That means we are going to have to use every tool available to be successful.
Why are people fearful of it?
It comes down to fear of being replaced. It’s very personal. But it’s going to help you make decisions and there will be some decisions it’s better at making than you. Not all this technology and decision making is going to work, and it will probably work out a little bit differently to what we currently expect. But it’s going to be interesting to see how it develops.
I’ve been in the industry long enough to realize there are lot of different techniques and technologies that have been introduced over time. Some successfully and some not. In each case the industry has had to learn where and how best to use the technology and AI is currently in the early stages of implementation.
In about 5 – 10 years time, I believe we will develop a series of tools and approaches that can expand the way we can tackle disease and that will be for the benefit of everyone.
Meet Andrew and other experts at the Medicinal Chemistry & Protein Degradation Summit. Download the agenda today.
Leave a Reply