Eric Springman, Ph.D., Celtaxsys CSO

neutrophil celtaxsys

CDD Spotlight Highlights

“It’s just amazing how science comes and goes in waves. Some of that is related to the tools that you have at your disposal, but when I started out in discovery science, particularly the innate immune system – neutrophils, and this is where I dwell right now, so it’s pretty fascinating to me how it’s coming around. Neutrophils were viewed as pretty unsophisticated dumb cells in that they were, as we called them, terminally differentiated, which had a lot of implications to it. They just basically couldn’t think anymore as cells think, they were just preprogrammed to do something and then die. And that really has turned around on its head to where even the cells that we have thought of for a long time now, for the last 20 years as primitive, are actually not at all, they’re very sophisticated cells.”

Eric Springman

Dr. Eric Springman is the former Senior Director and founder of the Biology Department of Locus Pharmaceuticals where he led the Company’s initial IND. Prior to that, he was Acting Head of Arthritis Therapy at Roche Pharmaceuticals in Palo Alto, CA.and at Arris/Axys Pharmaceuticals in South San Francisco, CA, where he was co-leader of the partnered project resulting in Merck’s Phase III drug candidate, odanacati. In 2007, Dr. Springman was appointed by Governor Edward Rendell to the Pennsylvania Drug, Device & Cosmetics Board.

Interviewed by Barry Bunin, CEO Collaborative Drug Discovery, Inc.

Listen to the Full Audio

Audio MP3

Edited Interview Transcript

Barry Bunin
Eric, since we worked together before in the lab, I thought I would start with a personal question about your career as a true scientist. What was interesting and cool for each piece of work that you’ve collaborated on over the years?

Eric Springman
I think the thing that attracts scientists to science is that every day is different, and every project is different. It’s always sort of a self renewing job. Just thinking back, starting out early on at Arris (Arris Pharmaceuticals), we had a very interesting projects on cysteine proteases, and we had some interesting challenges with early compounds. We weren’t sure exactly how they were acting, and it was a major breakthrough to figure out how certain compounds in the series were acting, so that we could help lead the chemists to the right chemical series to go after. It was a very successful project, certainly for the Cathepsin K and then subsequently for other cysteine proteases in the field.  So that was, I would say, one of the more interesting projects still to this day that I worked on. There were so many interesting challenges to it. And I think every time, as a biochemist and biophysicist, working closely for many years with chemists, being really at the interface of biology and chemistry of drug discovery, I really liked that aspect of the job. It’s like translating biology for chemists and chemistry for biologists, and being somewhere in the middle in both cases. But it is always interesting to me to see the people at work and the creative solutions on both sides of the fence that people come up with to solve the problems that come up in drug discovery. I think nowhere is it more than when you’re trying to get an optimized lead candidate that everybody’s really under the gun to make progress in areas that are seemingly at times mutually exclusive, like potency and oral availability, I have just loved the many projects that I’ve worked on and how that’s come together to create optimized compounds that could even be considered as drug candidates.

Barry Bunin
That’s excellent. Just moving from Arris forward in time and in bringing this relevant to CDD. At your last company you used another informatics vendor, and currently you’re working with CDD, so the question is why is CDD better?

Eric Springman
Actually I guess I’m fortunate to have grown up in the sort of electronic era of biological databasing and electronic notebooking, and at every stop, at Arris, at Roche, at Locus (Locus Pharmaceuticals), I’ve used electronic databasing, from some pretty sophisticated stuff back in the 90s at Arris and then moving to Roche where it was, I would say, a put together a system because it had come from a number of different companies. I think there were six different databasing systems that we had in place and the one used on a regular basis was really a homegrown one. So it was very rigid. If you needed anything, if you needed to change a unit, you had to go to programmers. Then at Locus we started out as a computational drug design company and in the spirit of being cutting edge, we went with a totally electronic notebooking system and electronic databasing system, that I would say overall was very successful. But again, I think the two things that I’ve observed over the years is that to this day, I think electronic databasing has been very good at capturing chemical data but really challenging when it comes to capturing biological data. I think somewhat because of the variable nature of the way you do biology, but also I think it’s just hard to capture some of the things like a pharmacokinetic curves in an electronic format, for reasons I don’t really understand. But, and so here’s where I came from, in my last job for almost ten years we used an electronic notebooking and electronic databasing system that was from a very large vendor. First of all it was very expensive, which for my current company of seven people would be prohibitive. Secondly, we were fortunate in Locus in that we had people who were expert computer programmers. Otherwise, we would have been constantly going back to the vendor to reprogram things just to suit our needs. So it was, to me, while a workable solution, it was very cumbersome to customize for anything you wanted to do in your own shop. Which is not a limit when you’re at Roche and you have programmers, or you’re even at Locus where we had programmers who could deal with it, but not here at a small company. What attracted me to CDD was basically it’s a very flexible, off-the-shelf, very affordable solution that solves a great need that we had here when I came to Celtaxsys. Which is we had lots of data in Excel spreadsheets, none of it had structures associated with it, and so we had no way of really relating structure and activity to this point.

Barry Bunin
Great. So you can talk more about the ease of customization and the technology informatics aspects, but also maybe connect that to the end goal of your research at Celtaxsys and what is interesting and different about your approach and why that will matter if it is successful. So going from the detail of the data and the need for customization to the broader intellectual challenges if you will.

Eric Springman
Celtaxsys started as a platform discovery company looking at immune cell functions, such as migration. So we had amassed a bunch of data on compounds that were effective at modulating various aspects of immune cell function, but for anyone who is trying to select a compound or move a compound forward, you need to be able to compare things head-to-head. So looking through Excel data sheets and trying to compare two different compounds, you either have to upload it all into your head, or you have to come up with some sort of system for making decisions, and I think that’s one of the most important things that we’ve had here at Celtaxsys is just choosing out of the 35 leads that we started with, choosing two that were the most promising leads. We really couldn’t have done that without some data organization, and now going into the future, we want to optimize those leads, so we’re going to be generating more chemical data and more biochemical data associated with that. As you go forward obviously the decision tree becomes more complex. There are more parameters that you have to consider. So I think really the only effective way to capture and catalog that information and to search it is really through a databasing system. Again what I found with the CDD team is it’s a very flexible and nimble organization that is very responsive to the kinds of data that we have, and to be honest our cellular data is a functional cellular screen. So it’s pretty complex, and most databases cannot really even handle the types of complexity that we are trying to convey in a summary format.

Barry Bunin
Absolutely! I think also the web-based collaborative aspect is kind of interesting to talk about in the context of your work. Because you travel between San Francisco and Atlanta if my memory is right, and you may have other people that you’re collaborating with either in the company or externally, and so just a little bit about working in this geographically dispersed manner and just some of the communication between you and others…

Eric Springman
I’m glad you brought that up Barry. I kind of passed over that in my mind, but you are right on point there. One of the things that attracted me to CDD was that it supports a multi-centered, a virtual organization that we can have. Our people working out of our San Francisco office, out of our Atlanta office, or even when we’re on the road look up data securely online. And the other thing is we use consultants for some of our chemistry optimization work, and so flexibility to give them access to a secure online database is really priceless, because we can’t have them come in just to be able to look at compound information, and we don’t have to ship our database on CD. We don’t have e-mail things around if we don’t want to. So again I think that is a very big value in the web-based format of CDD. There are a lot of companies out there who are a few people trying to do the job of a 250 person business unit, so I could see that this is a pretty big need in the industry. I’m not aware that any of the other packages I ever used were amenable to that kind of usage.

Barry Bunin
I want to talk a little bit about innovation in terms of practical techniques in either molecular cell biology or other areas. What are some of the new practices that you found interesting and that you think will impact the future?

Eric Springman
I came in my career on the tail end of I would call the pharmacology era, where companies would have a team of chemists making interesting compounds and sticking them into animals to see what they did, or they were looking for things like physiologic readouts like blood pressure changes. Then we went through the whole era of very much target focused drug discovery, and it all started with: “here’s an interesting target, let’s find a compound that does something to it, and then we’ll figure out what it really does physiologically later.” That was kind of interesting, but  what is really interesting to me is to see that it’s coming back around to more of a “let’s figure out physiology first” through functional screenings, and then we have the tools now to associate the pharmacology with targets. It’s not a done deal. It’s challenging, but again, I think the outcomes from functional screening are becoming more interesting because it goes back to my enzymology roots. You’re always looking from a target-based perspective, you’re looking for the rate limiting step in some process or the key point of intervention, and I think when you’re doing functional screening, you inherently find those points that have a really strong effect in the first place. The second challenge that evolved during the target based era, right when we expanded to combinatorial chemistry and high throughput screening, was that a lot of the compounds weren’t really all that drug like. They might be effective on the target, but they still weren’t permeable. And when you’re looking at more biological screens, you know there are some built-in aspects. If your target is intracellular, you don’t get a readout unless you are cell permeable. So I think there are features to the compounds that are getting better, and obviously we, as an industry, have gotten a lot smarter about how to make compounds that at least have a higher probability of being taken up into biological systems. To me that’s still one of the great frontiers is connecting the preclinical or nonclinical biology to what really happens in the human patient population, so translational research, and the emergence of translational research, and we’re really not there yet. I think that’s a vast area where we can improve drug discovery, is just being better at predicting how our model systems, which are really the weak point, can translate into a human therapeutic. From a mechanistic standpoint, I think as an industry we’ve gotten wiser. Mechanism seems to translate better than just some modeled effect, so there’s some promise there, but again I think the human body is so complicated and it’s not the same as a dog, or a rat, or a pig, or a horse, or whatever you’re looking at, that our model systems could be better.

Barry Bunin
I like that you cover the full gamut from the target based work with the cathepsin K project to some of the phenotypic examples. Is there a fascinating commercial or scientific development or study, that you think will be interesting for folks in the broader research community to be aware of?

Eric Springman
Just as a disclaimer I’ve spent so much time over the past few weeks just digging into my own area and into the R&D development program we have going on right now, that I have to really give that some thought, so if we could come back to that maybe something will cue up in my mind. You know again being a scientist, I think even the most esoteric remote things seem pretty interesting. I’m talking so I can think about this.

You know, what goes around, comes around, and we sort of dwell in an area that’s re-emergent, and I’ll use this as my example. It’s just amazing how science comes and goes in waves. Some of that is related to the tools that you have at your disposal, but when I started out in discovery science, particularly the innate immune system- neutrophils, and this is where I dwell right now, so it’s pretty fascinating to me how it’s coming around. Neutrophils were just viewed as pretty unsophisticated dumb cells in that they were, as we called them, terminally differentiated, which had a lot of implications to it, that they just basically couldn’t think anymore as cells think, they were just preprogrammed to do something and then die. And that really has turned around on its head to where even the cells that we have thought of for a long time now, for the last 20 years as primitive, are actually not at all, they’re very sophisticated cells. I’ll just relate an example, when I was in my graduate work we were trying to identify MMPs –  matrix metalloproteinases, and we were looking at the mechanism of activation of the matrix metalloproteinases, but they hadn’t been sequenced yet. So we had very limited information. We were trying to get the sequence, and we had a discussion in our lab, and the discussion basically went like this: “You can’t get any mRNA from a neutrophil, because they don’t make any mRNA because they’re terminally differentiated.” That really sticks in my mind because now we know very differently. Neutrophils can really adapt and even direct a lot of the flow of traffic in there, and I think this is an emerging area, much like we’ve seen with the macrophage M1, M2 story.  I think we’re going to find, and we are finding, neutrophil subsets that parse out in either disease states or localizations, and part of it has to do with our ability to study them more accurately with the tools such as single cell imaging flow cytometry, which give you a really deep insight into all kinds of functional aspects of a single cell. So to me this has been a very interesting awakening that might be interesting to a general audience, I don’t really know. I live it, so it’s interesting to me. But the fact that these cells we thought for so long were basically the sort of pawns of the immune system, now are actually coming around to being more part of the royalty.

Can I interview you back for a second Barry?

Barry Bunin
Absolutely. I like it, you’re putting me on the spot now.

Eric Springman
And this is interesting conversation, because I think your whole name of Collaborative is interesting because it seems like in some ways we are inching our way to collaboration in industry. Usually there are barriers to collaboration. I know we collaborate with our peers within a company, and we collaborate with our chosen peers within academia, but we don’t usually collaborate between companies, although it’s a little bit different now because small companies like Celtaxsys, we like to find other companies that have orthogonal kinds of capabilities that we can partner up with. But just in general I think the idea of collaborative drug discovery is really fascinating, and I was just wondering, what’s your take on the status of scientists contributing collaboratively to drug discovery? What’s the status of that and where do you think it can actually go?

Barry Bunin
That’s a great question, and I think about it in terms of what is maximally possible in theory, and then what’s practical given the reality in terms of the intellectual property requirements. The answer from that analysis is that technically there is going to be huge synergies between various companies or other constituents working together, and you need to reconcile that with each group’s needs for their own business challenges, but I think it’s very doable. So one example from the Gates Foundation, there were recently some press releases about this, they are working with seven big pharmas and a number of tuberculosis screening labs in a collaborative project. So you can imagine the challenges and opportunities there are. Another quote that is kind of relevant, is from Draper Fisher Jurvetson in a talk they mention they’re funding all sorts of different companies and what they found is that there are cases where there are great technical synergies, even though the companies each have their own business challenges. So the grace and the elegance is in that we think of collaboration as anything. It can be collaboration between two people in the same company down the hallway or two completely different organizations. But what we’ve been figuring out is that when people have more control over being able to partition their data, keep it separate and secure, and to decide if when and whom to share with, actually the more collaboration paradoxically we facilitate. And the reason is that getting information captured even in a private vault for just one person, for example someone looking to QC their own data before they share with someone else, has a certain activation barrier associated with it. And then there is a second one, which is what everybody sort of struggles with, is how do I decide who I share with or collaborate with. Actually that activation barrier is not necessarily any higher than just capturing what’s the most interesting information, be it from your own experiments or someone else’s, unpublished or published. So I think we are just at the beginning, at the beachhead, and how do you give people the maximum freedom to collaborate with anyone else that could have a positive impact. I think we’ve done a good job of starting to solve the data problems there, and then from that is going to follow the social problems and the contractual problems, but I think there are examples where you know it’s water moving downhill. One example I’ll just cite with, NIH had an industry NIH roundtable on drug repositioning and repurposing, which you could use something that was safe and efficacious after phase 2 but maybe not being developed for commercial reasons or other reasons, but it might be a useful candidate for a rare neglected disease. So now with NCATS and TRND programs, there are a number of projects that are looking at getting into that group synergy. So I think the solution is going to be that it’s not going to happen overnight, but I think the solution is to let everybody decide for themselves if, when ,and with whom they want to share data with and to sort of gently encourage more collaboration while letting people have the control. I think that’s the maximum efficiency that you can have, going back to what I was saying at the beginning, pairing the theoretical ideal with what’s realistically possible. So that’s what we’re focused on, and then to make that happen requires a variety of software tools, requires a variety of processes and content that encourages us to get the efficiencies that we have seen outside of the drug discovery process. And it’s interesting because in our industry you need to collaborate, and you need to be able to have something that’s a hook for intellectual property that you can get a premium on, so that’s the challenge and that’s the opportunity. I hope that answers your question.

Eric Springman
Yeah, it does.

Barry Bunin
Well maybe that’s a good note to end it on. unless there is anything else. I especially thank you for interviewing me. It’s always fun talking with you.