Q&A with Lorenz Studer, M.D.

Cenk Sumen (CS): How did you get into science and who are the people that most influenced you on this path?

Lorenz Studer (LS): It’s a long story. I got interested in brain research; I really wanted to find out what can go right and what can go wrong in the brain. I studied many areas of neuroscience until I got into the area of fetal transplantation. This was back in Europe and considered a crazy idea- that you could take nerve cells, transplant them, and they would actually function in a new brain! That blew my mind at the time. How you could get circuits to work, can you actually manipulate that, how does it affect the person, could you use it to study diseases? These were early days. We actually went from the idea to a clinical trial back in Switzerland when I was a medical student. Maybe the science wasn’t ready at the time. It taught me that if you push very hard and you have the right people around, you can actually go all the way toward translation. But that was exactly the time in my career that I saw clearly that wasn’t going to be the long road, for many ethical reasons and practical reasons, it’s not going to be a routine approach.

Actually I was thinking about gene therapy at the time; that would be a cool way to fundamentally fix a gene that’s the cause of a disease. That was the time that progenitor cells came up in a few studies. Some of the first isolation of neural stem cells. Then I decided to really go into progenitor biology because there is so much that is unknown. That was the first time that I moved to the States. Ron McKay was my postdoctoral mentor, one of the discoverers of neural stem cells and the Nestin gene. It made me think differently, beyond the simple idea of isolating cells and putting them into the brain, really understanding the developmental biology. I started thinking about how the embryo does all those steps. It was also a lot of frustration. It turns out that actually brain stem cells can be influenced to make neurons or glia, but you can’t influence them much as to what kind of neurons they make, because that decision is made earlier. So I started working on embryonic stem cells, since that is the earliest one that you can get. Suddenly things started to lock in! All the signals that we talk about, such as Hedgehog, or FGF signals, they were actually the same things as in the embryo. That was for me the next big insight, how to use these signals on pluripotent stem cells. It really changed my focus toward working on pluripotent stem cells in my own lab at Sloan-Kettering.

Another big influence was Jim Rothman, a very famous person who described v-SNAREs and t-SNAREs in vesicular fusion, who hired me. As a scientist he was very impressive even before he won the Nobel Prize. His intellect and vision were great. I remember that he asked me “what will you do ten years down the road, what will you be known for?” There were two things: one was curing Parkinson’s disease. The other was understanding the signals that make all the different cell types of the brain. The funny things is now we are still doing more or less those two. We didn’t cure Parkinson’s disease obviously, but again that is the journey. We are at the stage that we got clinical approval for actually using pluripotent stem cell derived dopaminergic neurons in Parkinson’s patients. On the other side, with regard to making many lineages we can make about 50 different neuronal lineages reliably, we know exactly the logic of developmental biology. There are still many other mysteries left, many interesting questions such as timing. We can now move toward translation.

CS: Thanks for sharing that. This is relentless work. My next question is on the neural crest, a fascinating lineage of cells in evolution. There are cranial and dorsal ones. How committed are they to those pathways, if you influence them in vitro? If you transplant them do they stay in those?

LS: They are definitely quite a plastic cell lineage, some people even call them pluripotent since they make more than one germ layer. Even if they come from ectoderm they can make mesoderm, although they can’t usually make endoderm. They are an unusually plastic somatic lineage. But there are still some restrictions. In the past it was actually thought that they are even more plastic, before our work and those of other labs in the pluripotent field. The first step was: how do you make neural crest? It’s kind of funny since one of the reasons we got into neural crest was that they came up as contaminations, they didn’t properly differentiate, they would come crawling out! If we wanted to make CNS we had to understand neural crest development. But my heart is very close to neural crest. I think they are one of the most fascinating lineages.

Coming back to your question, the front to the back as you said there is the cranial, the vagal neural crest, the trunk neural crest, and the sacral neural crest. They go from the head region all the way to the tail. There seems to be actually some restriction or at least preference in those lineages. It was a bit surprising, as shown by Nicole Le Douarin, the transplantation studies, the chick quail studies. There was maybe some difference described between Hox+ and Hox- neural crest. But even within the Hox+ neural crest nobody actually saw different behaviors. For example, we are very interested in the neural crest that populates the gut, the enteric nervous system. We already published a Nature paper in 2016 on making the vagal neural crest, which is probably the biggest proportion of the enteric nervous system. Now we have newer work in making the sacral neural crest, which is fascinating because these cells have real therapeutic potential. That’s another translational effort that we are pushing to treat children with Hirschsprung Disease. These are children where the enteric neural crest is never really properly entering or migrating along the gut, and they get all sorts of complications. It’s very fascinating from that perspective. All the movement, all the fates they have to do. Those neural crest cells not only make glial cells and neurons, but they have the capacity to make more than 40 different neurotransmitters, so it’s an extremely complex system, nearly as complex as the brain, a brain in your gut, and exactly how all this is specified is quite fascinating.

We are doing single-cell studies, chromatin related studies to try to figure out the mechanism for what makes them so unique at that stage. We would like to understand their development but we’d also like to use these cells, which is the part that I always felt passionate about. I like the very basic developmental questions but I also like the fact that we understand the system well enough that we actually dare to use it. Not only just for the sake of knowledge, which is obviously a very good reason already, but also in the sense of a construction thing. There is this argument that if you can build something you can understand it. And the final proof that you can build it is that you’re brave enough to use it. That’s one way to think about it.

CS: These are clearly very useful cells therapeutically. Are there any markers that would push them toward commitment to the four lineages? How plastic are those commitments?

LS: We haven’t really done extreme switches. We’ve done some transplantation studies. The different lineages have preferences: for example, you can inject cranial neural crest into gut, they do survive, they can make some neurons but they don’t really have the same capacity for migration and repopulation of the gut. They never change their Hox code. This is like an address code; for example, from two to five it’s cranial, it goes down all the way to thirteen which is sacral. Basically a nice hierarchy that is determined early in development. That seems to be largely maintained and locked in. The wrong Hox code in an inappropriate region can carry out some functions but they don’t have all the necessary properties. The way you make them is to simply give them the same signals they would get in development. You create a zip code system. Instead of just giving them a transcription factor you recreate the system that happens in development, to tell them which position to be, and if you tell them which position to be they will do it automatically, with morphogens such as FGF, and Wnt signals which are very important for the initial step, and retinoid signals for vagal neural crest, and tissue pattern related signals for the most posterior structures. This is like a language you can learn—when do you use which signal at which time of development? Depending on how you count there are seven or eight main pathways, and you can use them again and again in a sequential way, like playing a musical instrument very precisely. It follows exactly the system of development. The difference between mouse and human is that you have to play much faster on the mouse; in humans it goes much slower. It’s the exact same piece in a mouse, you just play it faster.

CS: Faster tempo in the mouse! That’s very fascinating. Do you have any comments as to the role of technologies in speeding up time to clinic? I know this is something you are very passionate about. Things like automation, manufacturing systems, AI, machine learning, anything that you feel is especially relevant that we need to push into to enable translation from bench to bedside?

LS: We need to have a better way to QC products. Right now we just have a few release markers. It’s an iterative process. As you mentioned, AI and machine learning should help us make that in a much more comprehensive way. Right now we might still miss something even if we have seven release markers, that probably doesn’t describe the product under all conditions. That’s something that we’d obviously like to have. Not only for understanding the final product but also for optimization. We now have a very good imaging system where we can follow cells at every moment in time. There are many molecular markers that you can feed into the system. This is something that we are starting to play with. It’s not ready yet for GMP manufacturing. On the research side, we can see how we can use that to optimize the processes. If we start with a new cell line, you have to go through all the steps again; you need to define everything. But if you have an understanding of the process in a more holistic way, I think that will help quite a lot.

With regard to automation, I think that is clearly also a big point in manufacturing. Currently it is used more for disease modeling. We use patient-specific cells or engineered cells to capture disease for drug discovery. The bottleneck there is that for complex diseases you need thousands and thousands of patients before you can say something, to have subgrouping with predictive markers. For stem cells you probably need less if you do it within the right lineage. You don’t look for the whole body; you look for what is relevant for the disease. But on the other hand, you still have hundreds of lines. You can’t do that without some kind of automation. There are a lot of challenges where machine learning can be helpful. Can you find processes that are robust enough that they can work across hundreds of lines? Then you can do things massively parallel.

We have also developed pooled approaches where you can pool hundreds of lines, reverse engineer them, and see what happens to each line within a pool. For large-scale manufacturing where automation is going to be particularly important, it’s about achieving scale with no human intervention using bioreactors. One area where this could become important is if you want to do it patient specific, GMP manufacturing for each patient in parallel. Right now for each line you need a GMP room. You make the product and you qualify it. It’s very expensive obviously. If you can do it in small machines, you can do GMP manufacturing in a desktop-like system. This could be further improved, but it raises all kinds of regulatory questions. It’s not an easy path to get clearance from the FDA. We hope it’s going to be easier in the future. Imagine having to do that for hundreds of lines! Obviously we can’t do three-year animal studies on each of the lines. There has to be a better way from the regulatory perspective, to convince that we have a very safe product. I think for the whole field that is one of the main challenges.

Another challenge is finding the right CRO that can do this technology. The whole field will move forward as more people get to this stage. One of the main practical bottlenecks is finding good CROs to manufacture pluripotent-based therapies, carry out testing in the brain with complex surgery. There is still quite a big gap on how to handle that.

CS: Quality is often overlooked, and hopefully the regulators will catch up. Do you have any advice for young scientists that you’d like to share with us?

LS: The main idea is to follow your passion. You don’t need to think necessarily that you need to know your passion before you get in. When I initially started off doing Parkinson’s research, I thought we already knew a lot and I never thought I’d eventually be so passionate about it. Once you get in, you need to go to a good place to do research, to learn the craft, get the feel and be open to being passionate, then you need to think hard where you want to make a contribution, and then follow that dream. It’s not running after the fancy new technologies necessarily, but figuring out where you can make a contribution. Think where you can go deep and not just wide. To fundamentally change something you need enough focus on it. Create your own niche where you can really make a difference. Don’t be afraid if people tell you “this is a crazy idea!” Crazy ideas a few years later can suddenly become exciting. And a few years after that people can think it’s a bad idea. You shouldn’t be too much influenced by fashion in science. You should be open to new technology, every technology if possible should be integrated into your own focus, but not necessarily just running after technology itself, except if you are a technology development person, that’s another thing.

If you really want to make a biological difference, I think it’s important to have this focus. Personally for me at least, a big motivator was also to have a chance to make a difference downstream. The translational piece was a big point for me. It’s different for different people. It’s interesting to be able to get to a stage where you can actually apply your knowledge. To have people work at that interface is very, very important.

Dr. Sumen is the chief technology officer at Stemson Therapeutics and a member of the ASGCT Communications Committee.