Modelling Biology

Company to Watch – Cellarity

In the final article in the Company to Watch series on Cellarity, Big4Bio editor Marie Daghlian spoke with Chief Scientific Officer Laurens Kruidenier about what impressed him about the company, what he sees as the strength of the Cellarity platform and how it works, and why we should keep our eyes on the company.  by Marie Daghlian‍

‍

Marie: You just joined Cellarity. What’s your background, why did you decide to join Celllarity, and what impresses you about the company?

‍

‍Laurens: I really joined Cellarity for its disruptive platform. It's also a platform that has the potential to deliver groundbreaking new medicines for many patients. I'm a research biologist by training with a PhD in mucosal immunology. What I realized early in my career is that I just wanted to spend my time and energy on problems that really move the dial, finding solutions that break new ground. That was important to me throughout my career over the last 25 years, in academia and in larger and smaller biotech. But throughout that, in three different countries, I've always had a pioneering mindset and I've worked with that mindset across the full gamut of drug discovery and development. In my last role as CSO at Prometheus Biosciences, we were pioneering a way to improve clinical practice by bringing precision medicine out of oncology into the inflammation space. When I learned about Cellarity, I felt this is the ultimate pioneering approach because at Cellarity we really want to fundamentally change, not one aspect of the drug discovery and development process, but basically the whole paradigm.

‍

‍Marie: And did you hear about Cellarity before they approached you or when they approached you?[caption id="attachment_19447" align="alignright" width="200"]

‍

Laurens: I heard about them. With these things it's always about timing. I had just helped build Prometheus out to a publicly listed company with three phase 2 trials, ongoing right now. It’s exactly that kind of experience that I want to exceed at Cellarity—do it again, but at a much bigger scale.

‍

I joined Cellarity because I'm really impressed about its focus, a complete focus on the cell. I think across the industry, we're facing a major challenge and that is that we are seeing a growing unmet need clinically in many complex diseases. At the same time, we don't often know a lot about the underlying molecular pathology of these complex diseases. That's a big challenge because what we're trying to do in the industry is to address these diseases by making more and more selective drugs—small and large molecules that just hit one particular protein in your body--and we've seen successes there. It can work. But I think it's now time to really see a paradigm shift and apply a more target agnostic approach that will allow us to look for more and basically embrace the complexity so we can look for more, see more what's happening under the hood of that disease, and if we are able to address that, do more for more patients. For me, that was really something that took me by the hook when I looked at Cellarity.

‍

It's important to understand that now the time is ripe for this approach, I don't think we could have done this approach maybe five or 10 years ago. We've seen major advances in high resolution data, single cell technologies, machine learning, AI approaches. I’ve actually deployed all of these things in my previous role where we were trying to identify the right patient for the right treatment. Here, we're using that technology to embrace the complexity of the disease and really see it not as a biological problem, but as a computational one--we look at the complete cell, not just a single protein within it as the fundamental unit of the disease. That’s important. We can then understand computationally what the changes are that occur in the cell that makes the cell transition from a healthy to a disease state and visualize the transition from health to disease as a mathematical equation. If we can do that, we can model the actions of various pharmaceutical interventions on that and design novel therapies that address this other set of changes, not just that single bit of target. That turns discovering drugs into engineering drugs, which makes sense because most things we currently don't discover, we engineer them and why would that be any different with drugs? I think that's another thing that really impressed me about the approach. In extension to that, because this approach is target agnostic, you would expect it to work in basically any disease.

‍

When I joined, I was impressed by the breadth of the cell types and disease areas that Cellarity had already been able to tackle and validate, and the speed by which that had happened. Because if you can engineer drugs, you can basically really speed up your process. You don't have to build 10,000 molecules, test them and fine tune that they work, and then go from there. Really, you can jump straight to what you expect to work. Cellarity Maps is a key component to this.

‍

Basically, Cellarity Maps are a digital representation of process within the tissue. What we do in practice is to take a diseased tissue and then disperse all the cells in that tissue and collect transcriptional profiles. These are mRNA profiles for each cell individually in the tissue. And these inputs basically build the the Cellarity Map, which is not a theoretical model but a digital representation of actual data. For each of these cells, we looked under the hood at the transcriptional events that happen in that cell and catalog that. And that's what the map represents. The map reveals cells that make up a tissue, the diversity of the behaviors that each of these cells undergoes at that moment in time in that tissue and the relationships between them. Then we compare the Cellarity map of a healthy cell or a healthy tissue to a Cellarity map of a disease tissue and that allows us then to identify cellular dysfunctions, basically a set of changes that occurs as all these individual cells move from a healthy to a disease state. That is the basis of our whole platform where we use the Cellarity maps to then find ways to address this particular cellular dysfunction. We do this by using our Intervention Library, our own purpose-built resource that basically links pharmacological interventions to the images that we see that allows identification of the interventions most likely to address that complexity and then importantly—I’m a biologist remember—we test that. It's not an exercise. We find these molecules. And then we verify that in the lab, we have a big lab team here to make sure that, indeed, these predictions come true.

‍

‍Marie: For a few of the molecules, you've already found that they work in higher animal species and that you're moving forward into humans.

‍

‍Laurens: Correct. The breadth as well as the speed by which we can basically now do drug engineering is an impressive piece of it. It takes serendipity out of the whole process.

‍

‍Marie: There are other companies focused on single cell biology. What is unique about Cellarity versus the others? Or do you think they are all racing for the same solutions?

‍

‍Laurens: There's a couple of aspects to this question. Some companies are service providers and we are not. We are not using our platform to drive classic target or biomarker discovery. We're not using our platform to identify basic, let's say combinations of existing drugs that might help a certain indication. We're not using our technology for precision medicine basically to match existing drugs to the right patient subgroups. We're not using our technology to improve the current drug discovery process, make it more efficient, for example, by improving target identification or drug screening. All of that is not what we do. What we do entirely re-architects the way that we do drug discovery. We solve for a functional readout combined with a cell behavior to change, and not a proxy like traditional drug discovery does by finding a particular target. What this approach allows us to do uniquely versus the other companies is to understand biology at a more complex level and address this complexity and uncover previously undiscovered mechanisms and complex behavior that you will not have discovered with a traditional target centric approach. Basically, it's about modeling biology. It's about testing interventions and seeing medicines, then, as reprogramming devices. To my knowledge, Cellarity is the only company that uses this kind of technology in combination with machine learning approaches to address a particular problem like this.

‍

Marie: How do the various teams work together to make sure the data generated is high quality?

‍

‍Laurens: It's a really good question because for this approach to work, we really need to have the various disciplines work very closely together. We have biologists, medicinal chemists, data scientists, computational scientists. This platform can only work if we foster a deep multidisciplinary approach and that's exactly what we do. It's one of the reasons I joined this company because in large pharma companies that's hard to achieve. This is ingrained in every program we do. Our program teams are multidisciplinary and they work together to ensure all the right questions are asked, not just from a biological perspective or from a computational perspective: are we collecting the right kind of data from the right kinds of samples? Is this data helpful? It sounds like an obvious thing, but sometimes people just collect data for data’s sake, but you want to collect data helpful in addressing the particular question you ask as a team. I see this as a dialogue between the disciplines and we really foster a culture where we also learn from each other because we've all been trained in different areas and we speak different languages. You start to speak the same language. You start to understand each other better, and therefore you can formulate new questions and dive much deeper into scientific questions than if a different discipline would just ask the questions by themselves.

‍

‍Marie: Can you give an example of how the technology worked in a certain situation?

‍

‍Laurens: An example I would like to mention is our lead metabolic disease program, and it's a good example of how our plan can identify new biology that's totally unexpected because we look at this complete behavior. And you find things that you would not have identified through a conventional target specific approach. What we have identified in this program is a novel insulin sensitizer that does not appear to use the PPAR-γ pathway. And I'm mentioning the PPAR-γ pathway because it's infamous. This was the darling pathway in the industry for many years, and we had the glitazones on the market as a major class of oral anti-diabetics that have now been withdrawn for the most part because of unacceptable side effects, including cardiovascular side effects. Our platform allowed us to identify new biology that has the same outcome, because we optimize against outcomes behaviors, but that can avoid the PPAR-γ pathway and these kinds of problems. That's what we believe.

‍

What we did in this program is, we started with fat cells, and we asked ourselves the question: “What if we could reprogram white fat cells, these are fat cells that store fat, into brown fat cells, these are fat cells that burn fats? Wouldn't that be great because we can then have a significant impact on several metabolic diseases.” And this includes obesity, but also type two diabetes. So, we built a Cellarity Map based on mouse fat tissue. We have since also looked at other fat cells and other species, but we started with the mouse. We created a Cellarity Map to model the transition between these white storing fat cells to the brown burning fat cells, and then deployed our platform.

‍

And what we very quickly found through our machine learning algorithms was multiple interventions that were predicted to drive this biology. At that point it's only a prediction for intervention. We have the map and we say, “Hey, here's a bunch of molecules that can drive, really promote that transition from white cells to brown cells. That's the effect we wanted to see.” What was really encouraging, of course, is then when we took these predicted interventions and we tested them, in vitro and in vivo, we actually found really impressive improvements in insulin sensitivity, glucose tolerance, and lipid profiles across multiple model systems in vitro and in vivo. What we also found is that these molecules did not process PPAR-γ pathway activity, the liability I talked about earlier. This is, in my mind, quite remarkable because it shows us that our in silico predictions are coming through. We have a really high hit rate. We don't need to search tens of thousands of molecules to find the one that works. We can hone straight in on the ones that have the desired outcome. And we got that pretty fast so it really highlights the ability of this platform to uncover new biology, and then work at a speed that allows us to uncover pharmacological interventions that will allow us to address this biology at ultimate speed and ultimately in the human.

‍

‍Marie: Can you tell me how long the process took and is this one of the molecules that you're hoping to push into clinical studies?

‍

‍Laurens: Let's say based on my experience, pretty fast, and yes, it is a lead molecule.

‍

‍Marie: What's the future look like? What are the near future plans?

‍

‍Laurens: I think the future looks very rosy. Of course, the reason I joined is to progress our pipeline, and the future will be to push our lead programs into the clinic. That's number one. Number two, we want to do more with the platform, we want to expand our pipeline. We want to prove that this platform works in many different kinds of cell types and disease indications. So, the plan will be to start new programs, new diseases, and importantly, as a platform company, we're not forgetting the platform itself, you know, the goose that that lays the golden egg. I see a lot of potential in how we can enhance this platform by adding new data types, new analysis capabilities. So, our goal really is to build Cellarity into the next-gen biotech that leads the way in a new age of discovery and development. It’s the reason I joined. And it comes back to the theme in your first question, the pioneering mentality here is definitely what I'm attracted to.

‍

‍Marie: These are good reasons why we should keep our eyes on the company. I have a feeling you're probably going to spend a bit of time talking to potential partners to broaden the platform’s reach.

‍

‍Laurens: Of course. It will be a key part of our strategy. We cannot run 10 phase 3 trials.

‍

‍Marie: Thank you for your time.

‍

‍Laurens: Thank you.

‍

‍This interview has been edited for clarity and readability.