One of the lesser-known companies I’ve been advising for the past year is Forcyte Biotechnologies, a very early-stage biotech company spun out of UCLA that is focusing on “mechanobiology”. You’ve probably never heard of them yet. They’re frankly not led by a well-known team, or backed by prominent investors. But neither was Recursion when I joined, but despite being the underdogs, we went on to significantly shape the TechBio landscape. So I’ve got a soft spot for the not-so-well-known in the world, and perhaps a bit of an aversion to those startups launched with $100M out of the gate, with the opportunity handed to them on a silver platter. This post is for the little guys of the world, and I’ll focus today on Forcyte. Don’t worry, other little guys are coming 🙂
When first introduced to Forcyte, I scratched my head and wondered “what on earth is mechanobiology?” In short, it’s the study of the mechanics of cells. In their hands, it’s specifically the examination of “the contractile cell function underlying the physicality of biology and its universal roles in health and disease.” They basically measure, very systematically, how cells contract or relax under various stimulants - by cytokines, genetic perturbations, small molecules, etc. “Okay, that’s cute,” I initially thought, while simultaneously thinking “that seems way too narrow and specific a niche to build a platform-based drug discovery company upon.” But when I stepped back to think about contractility in cells, I started to realize how many therapeutic areas this could affect:
Fibrosis
As a condition that occurs when cellular tissues thicken, harden or scar due to a buildup of collagen and other extracellular matrix components, it’s natural that cellular contractile function could be a useful lens under which to study fibrotic indications and potential therapeutics here.
Cardiovascular
Contractile dysfunction is heavily involved in cardiomyopathies. Decreased contractility could be an indication of stiffening or thickening of cardiac tissue.
Ocular
The ciliary muscles in our eyes contract and relax, changing the thickness of the lens which ultimately enables the eyes to focus on objects at various distances. It’s reasonable that the study of mechanobiology, in particular cellular contractile function, could aid our ability to discover new treatments for various ocular conditions.
Respiratory
Our lungs obviously expand and contract, but this is passive and a result of the contraction/relaxation of our diaphragm muscles. But the bronchial smooth muscle tissue of our airways can also contract, causing breathing difficulties, as occurs for patients with asthma or COPD.
Now, some of these indications with relevance to cellular contractile function have easily accessible treatments (such as albuterol, a bronchodilator used to relax the airway muscles in asthma and COPD), many do not. Other companies are bringing forward drug candidates whose functions effectively center around cellular contractility, such as Pliant’s phase 2 integrin inhibitor Bexotegrast, which prevents force transmission from fibroblasts to ECM to unlock TGFβ). Or Rhopressa, a ROCK inhibitor approved for glaucoma. While “mechanomics” still seems to me to be far from a drug discovery panacea, I now better see how it can provide relevant data to drug discovery across a wide range of therapeutic areas.
So what does Forcyte Bio actually do? They have developed a proprietary platform called the FLECS platform, wherein they manufacture their own custom microtiter plates coated with an elastomeric film containing an array of micropatterns (plus signs and crosses) to which single cells can adhere and contract under stimulation, inducing a change to the micropatterns. Relaxed cells generally leave the micropattern shape and size the same as those without adherent cells, while contracting cells shrink the micropatterns. Leveraging custom computer vision software, they are able to examine multiple cell types and measure the relative contraction induced by various perturbations (secreted factors, small molecules, CRISPR knockouts, etc.).
I frankly love this kind of stuff, because they are rooted in radical empiricism with a functionally interpretable readout while applying technologies (both physical and digital) to generate scaled discovery data - at a scale that has not been possible in the past due to technological limitations. At this point, they aren’t predicting which perturbations will relax cells contracted by a stimulant (such as TGFβ), but one can envision a future in which their “atlas of mechanobiology” has grown sufficiently broad and deep that they can learn how to predict cellular contractility of a perturbation under hundreds of conditions from real data generated under only a dozen conditions.
One of the things that struck me here almost immediately is how while this platform could be used as a straight-up phenotypic drug discovery system (screening large-scale small molecule libraries against various contraction models - they already have some compelling in vivo data on some of their earliest discoveries), it could also be used as the starting point for novel target discovery in these relevant therapeutic areas. Over the past year, the team at Forcyte has explored RNAi perturbations on their platform, and later integrated CRISPR knockout systems in a suitable manner for their FLECS plates, examining hundreds of genes to identify which knockouts significantly reduce the contractile effect induced by various cytokines. With only a few dozen genes examined, they have already identified a couple of pronounced potential genetic drivers of cellular contractile function. I can’t wait to see what they find once they examine thousands of genes, let alone the whole genome.
Some might look at them and think, “Where’s the generative AI? Where are the LLMs? It’s 2024, isn’t it?” While there’s potential for machine learning to be applied to Forcyte’s data in the future, I think their team is doing a fantastic job applying sound applied data science and engineering to unlock novel discoveries coming from data uniquely available to them at this scale. That is precisely where they should focus, as any generative AI algorithm or LLM is limited by the data available to it - and Forcyte is generating a valuable data modality that I hope proves useful in our efforts to bring better medicines to patients faster.