“We’re using CDD Vault as a way of storing our chemical structures, as a way of searching them, as a way of storing all of our assay data. We’re also using it for doing calculations, for analyzing our data, for creating reports and communicating the data to others, and then also, very importantly, for working securely with collaborators. So it’s been instrumental for all of those purposes.”
President, CEO and a co-founder of Oligomerix
Dr. Moe has 20+ years industrial experience having held senior management positions in product development working on international teams in both early and late stage diagnostic, biotechnology and biopharmaceutical companies including Gene-Trak/Amoco Technology Ventures/Vysis, bioMerieux, and Mosaic Technologies. Prior to founding Oligomerix, he was Director of Product Development at Pyrosequencing, Senior Molecular Biologist at Spire Biomedical, and Director of Product Development at Q-RNA, Inc.
Dr. Moe received his Ph.D. degree in Molecular Biology/Molecular Biophysics from Wesleyan University, did his postdoctoral studies at Vanderbilt University in the Center for Molecular Toxicology where he was jointly appointed in the Biochemistry Department in the Medical School, and the Chemistry Department in the College of Arts and Sciences, and has an MBA degree with a concentration in entrepreneurial studies from Boston University.
Interviewed by Shirley Louise-May, Collaborative Drug Discovery, Inc.
I’m here with James Moe, President and CEO and a co-founder of Oligomerix. Oligomerix is a small biotech company primarily interested in addressing tauopathies. Tau is a protein in the central nervous system (CNS) that has been implicated in neurodegeneration if there are problems with the pathways downstream or upstream of tau. James, could you introduce us to the nature of Oligomerix’s research?
Sure, I’m James Moe, President and CEO of Oligomerix. We’re a company that’s focused on tau as a therapeutic target of neurodegeneration for Alzheimer’s disease (AD) but the work that we’re doing also applies to numerous other diseases. In fact, there’s about twenty different tauopathies that are known to date. Alzheimer’s disease (AD) is a dementia that is characterized by the pathological hallmarks, plaques and tangles composed primarily of aggregated amyloid beta or tau protein, respectively.
When we embarked on this project we chose to go after tau as a target based on the reproducible pattern of the development and spread of tau pathology during the course of AD, and because it was a differentiated approach from most other programs that were targeting the amyloid pathway. We also decided to further differentiate ourselves by targeting the oligomeric aggregates of tau rather than the larger and more inert fibril aggregates of tau. Now, after having embarked on this path about eleven years ago or so, most of the drug candidates developed against the amyloid target have failed in late stage clinical trials. So increasingly, there’s a lot of interest in Tau as a target and we have a pretty decent head start there. We’ve always believed that it’s an important target for AD, so we’re trying to discover drugs that prevent Tau from aggregating and developing its neurotoxic function.
Very Nice. Now that you’ve told me about the nature of your research, I’d like to hear more details of your approach, like what is the end goal, and what are either your modest or the more audacious goals of your programs?
At Oligomerix we’re developing proximity-based screening assays for identifying inhibitors that prevent tau from self-associating at the beginning of the aggregation process. Our end goal would be to identify compounds that we could take into clinical studies – that is the ultimate end goal – and that would require us to bring a partner on board and to get to IND enabling studies. Right now we are in the candidate lead optimization stage for the project.
So how did you come up with the active compounds that you’ve found so far? Up until this point have you only used the biology as a guide? Have you been screening library compounds without medicinal chemistry support?
No, we’ve used both supports. On the biology side, initially, when we decided to embark on a program to screen against tau as a target, we wanted a solid foundation for the research. So we decided not to use the fibrillation type of assays that people were using because they are not physiological, that was our departure from the norm. Secondly, most other researchers in the field didn’t use full length tau for technical reasons. Third, most everybody else was using tau sequences with mutations that are not present in AD. Finally, they were using aggregation enhancers or facilitators in their assays. We decided, since we’re trying to prevent tau-tau interaction by development of a proximity assay, that we needed to work with full length tau. Because – importantly – if we prevent tau from aggregating, we would want to have normal tau that could still be functional or would be able to undergo normal clearance. In choosing this path we were able to discover, for the first time, tau’s ability to self-truncate once it forms these types of aggregates. Other people missed that because they were using already truncated tau.
Initially, our medicinal chemistry was internal and basic – looking through our hits and then designing new structures. When our medicinal chemist left, at the time we thought it would be a very temporary solution, we decided to do our chemistry through a CRO. We still work with a third party who does all our medicinal chemistry, but have now brought on a consultant medicinal chemist and intend to bring this function in-house.
Ultimately, our end goal is to enable a large pharma to more rapidly bring candidates into the clinic. We, as a small entity, would not be able to do that on our own or in a timely manner, to bring the funding along to do that, or to have the expertise for the full ride. However, we would like to enable our strategic partner to do that to enable our global common audacious goal of finding effective therapeutics for AD.
OK. At Oligomerix, what are the means to the end (however you interpret that question)?
So the means are – my interpretation – how do you get there? So that’s the money question, that’s how I interpret it. Well, we started this project with two scientists plus myself and got Oligomerix started by writing an NIH grant – an SBIR. That’s how we got the company off the ground. NIH has been a very good partner. We’ve raised our own money, too. Actually, half of the money we’ve raised comes from NIH grants and the other half from investors. Most of our investors are individual investors although we have one institutional investor right now, Wheatley Partners. Ultimately, it’s the investment dollars that come in that enable this program. So that’s what I would interpret as the means. But also, to be in this business, you need to have novel technology. In terms of Oligomerix, the majority of the technology we are using we’ve developed inside the company. So, in other words, we created it, we’re not in-licensing our novel methods from an academic institution. I think that sets us apart as a small company in this field.
In the area of tauopathies, what new techniques or information have you found particularly interesting?
Well, as an example of one of the interesting things, because we were working with full length tau, we found that tau was self-truncating, and the question is ‘What is the purpose of that function?’ Specifically, tau has an intrinsic proteolytic function, so that when it forms certain types of aggregates, it self-truncates. So, the real question is whether the function is pathological or protective? In trying to understand why this function may have developed in tau, we noted that tau is a protein that stabilizes microtubules. So why would it have an intrinsic proteolytic function? Well, it also turns out that tau is involved in the stress response system. Work in other labs has shown that if you treat cells in culture with oxidative conditions, tau goes into the nucleus and somehow reduces the number of double strand breaks in the DNA. I speculate that this novel function may be involved in that process and that maybe, this function is going awry in the rest of the neuron in AD. This is consistent with our finding that expression of some of these truncated fragments of tau is toxic. From a purely scientific perspective, tau is a fascinating protein to be working on as its normal and disease-related functions are still being unraveled inside and outside of neurons.
So can I just ask a clarification – is tau’s proteolytic function something that you believe should be only functional in the nucleus and in a disease state it’s happening outside the nucleus?
Yes. However, there may be other normal functions for tau proteolytic activity, as well both inside and outside of cells that we have not considered.
OK, so how does Oligomerix use CDD Vault? And why did you choose CDD Vault?
One of the reasons why we chose CDD Vault over other packages that we were looking at was that we felt that CDD fit well, given that we’re a smaller entity, and we didn’t want to choose packages where we thought that big pharma was going to be able to muscle away the attention of the provider. So we thought that they fit well with our size and they had all the features that we needed. Indeed, their customer service has been exemplary!
How are we using it? Well we’re using CDD Vault as a way of storing our chemical structures, as a way of searching them, as a way of storing all of our assay data. We’re also using it for doing calculations, for analyzing our data, for creating reports and communicating the data to others, and then also, very importantly, for working securely with collaborators. So it’s been instrumental for all of those purposes. And when we get into a partnership, we are looking forward to both ourselves and our partnering company being able to access some of the same data and contributing data into the same database, CDD Vault is very functional for that. Another primary concern is having a database where we have better security over our molecules.
And you believe that you have better security with a cloud-hosted solution like CDD Vault than you might have had in-house?
I think so, yes.
And in securely exchanging data with partners?
So, I have to ask if you’ve ever gone the DropBox option?
We need more powerful drug discovery capabilities than merely a file exchange folder provides. However, we now use DropBox for standard documents as well.
But it has no functionality?
Yeah, basically. We are also using Merrill database for data sharing and have a lot of other types of data that we put in the Merrill database. We just keep it open all the time because it’s pretty cost effective, given the amount of data we have relative to say a larger company. You know, those are the two kinds of cloud-based type hosted solutions we use at Oligomerix; CDD Vault and Merrill. Then there’s still a lot we’re doing internally. We are looking for more ways to leverage using the cloud, but at the same time we want to be smart about doing it, so we’re sort of being careful in how we implement it and only use providers with trusted track records like CDD, which has been securely hosting sensitive data for a long-time.
So, how long you’ve been using CDD Vault? If you want to, can you enumerate some of the things that you like about it and some of the things that you hope will either be coming in future upgrades or things that you think would, if they were addressed, make the product even better.
Yes, I can easily think of two or three things we like about it. I think the search functionalities have been very useful. We’ve been happy with those and also the reports that we create. Also, being able to easily update protocols. It’s been very good too, as we develop new assays, then, to design them so that we can use them to calculate IC50’s and derived parameters like that. I also really like that the physico-chemical properties are present so you can, if you’re making a report, check off the ones you want and that sort of thing.
How long have you been using CDD Vault?
Let’s see, I’m trying to think exactly. We are in our third year now using CDD Vault.
Right, and you do mostly compound-centric searches or assay-centric searches? I mean, what kind of searches do you most commonly use?
We have many of both compound- and assay-centric searches. One thing we haven’t done, is, put a lot of the imaging data into CDD vault, but that’s something I’d like to do in the future. We do have that kind of data, you know, but right now people maintain that separately. One issue that I already talked to Charlie Weatherall about is in the graphing of curves. When you have outliers on your plots, you can sequester the outliers but it doesn’t automatically change the axis of the plot. We haven’t yet figured out how to change the axis automatically but it seems there must be a way. So that part, I think, is one little area that could be improved, however the good news is that you can manually remove outliers. But these are little things that could be addressed, as we don’t get that many outliers. I was telling one of my technicians, ‘just don’t have outliers and we won’t have to worry about it!’
Sure, I bet he loved that. All right, can you also describe the reports that you generate? Are they primarily for internal use?
Yes, internally or with our CRO, we selectively and securely share our CDD Vault also with our CRO. But we create our reports with password protection, so, a little different than what others do, the good news is with CDD Vault you can provide different levels of access privileges to different projects. So when we create an Excel file then, if it has structures, we save it as a password-protected file and then when we send it to our CRO’s that way. Then, in a separate email we send them the password. That, we feel, gives us some level of security that somebody won’t intercept both messages and get sensitive information. We just want to be cognizant that we’re protecting our structural information more than anything.
All right. So why don’t you describe a memorable interaction you’ve recently had with another brilliant scientist?
Well, when I was at the Alzheimer’s International conference this year, there was a lot of talk about Tau PET ligands. As some people who work in the field know, over the last few years there has been a lot of advancement in amyloid research because of amyloid imaging capability. Now, finally, tau PET ligands are available and are being validated. What was fascinating was, when you look at amyloid and tau imaging patterns in the same brains, you might have both amyloid pathology and tau pathology. And if you look at where the two pathologies are overlapping you see regions of hypo-connectivity. So, in other words, because of this hypo-connectivity or concentrated protein aggregation, there is the potential for loss of function in those regions. Nobody really has demonstrated that, but when I was talking to the Chief Scientist of Advent and he pointed those regions out to me, I thought that was very fascinating because it implies that there’s a lot of neuronal loss occurring in the places where the two pathologies are occurring in the same region. And I think that’s saying something about the disease mechanism, that maybe both pathologies are necessary for synaptic loss. I think it’s fascinating because for so many years, researchers have been struggling to understand what the relationship between these two pathologies. So I think we’re finally starting to get at that becoming clearer with this new imaging data. What I found was amazing, sitting in that hall that day, seeing a lot of good presentations of tau imaging data, that this is all new – it’s new to everybody in this room. These are the world’s experts on tau imaging and to hear them, sitting there, seeing it, at the same time with everyone else, it felt like, wow, it’s almost like you’re looking up the road and you know these are the tools that are going to help us. They’re going to help us find treatments that actually can cure or slow down or halt or arrest neurodegeneration and that’s really exciting. What a treat, as a scientist, to be sitting in that room.
What technologies do you use for your research? And why?
When we embarked upon this project, we developed a novel proximity screening assay where we were able to screen large chemical libraries and then used cellular assays as secondary screens. Then, to demonstrate that our screening assays are selecting compounds that will work in vivo, we conduct mouse studies. Yet, the problem with doing mouse studies is that there isn’t any mouse model that really recapitulates Alzheimer’s disease and so you’re only looking at a mechanistic piece of the disease. In our case, because tau was our target, we were looking at the tauopathy models. Now, nine out of ten of those models have frontal temporal dementia mutations that are completely irrelevant to Alzheimer’s disease so that’s one of the shortcomings. Being a mechanistic piece of the disease and not the whole disease, that’s another shortcoming. But at the same time, if you can show in one of these models that you’re engaging your target then it also becomes a way of convincing the other stakeholders that you’ve developed some valuable starting point for potential therapeutics that could go into clinical development. So the in vivo studies are really a way of building confidence in potential partners, a proof of a mechanism, proof of validation, that your screening assays are informative.
So how do you work collaboratively within your group and you can tell me about any collaborations with others, what works well, what can be better?
Well, AD projects are so complicated that you have to work collaboratively because no one person has enough skill to do all of the different aspects of it. We have had a number of academic collaborations, though we’ve started to limit those and now we’re mainly collaborating with CROs. One issue with working with academic organizations is that they don’t want to work for fee for service and our investors don’t want to have to negotiate licenses for everything that we’re developing. So that’s why we started to work more with CROs. We are anticipating very collaborative efforts with our partnering company and we intend to be doing most of the biology and some of the chemistry. Hopefully, the right partner will be doing the majority of the chemistry or the chemistry will be done through their CRO network. It will be very appealing, if we’re working with a major pharma and they’re doing the ADME, the P.K. and long term toxicity studies, for us to be able to search through each other’s databases around any jointly developed IP. So the way that I envision doing that is that we both use CDD Vault with bigger projects. So that’s definitely something that I’m sold on and that we’re looking forward to that in the future.
Is there any aspect of collaboration that you go outside of CDD Vault for or you hope that they put into CDD Vault in the future?
There’s not anything that I can think of right now. But I certainly could ask my group to see if anybody else has any ideas from working in our CDD Vault because they may have suggestions, or some may become apparent when we get involved in a larger pharma collaboration. There may be some features that they’re used to with their own databases that we don’t even know exists, for instance. But right now there’s nothing that comes to mind. So, its meeting our needs at this stage very well.
Can I ask, have you had anything customized for you to date? Have you said ‘if we could have it done this way’ and CDD Vault has risen to the occasion?
I don’t think we necessarily had things customized but I think Charlie (Weatherall) has helped us out a lot in terms of determining how to best set up our CDD Vault, how to import our data and do our calculations automatically off of the raw data given the way that we’re running our experiments. That requires a certain amount of back and forth to get it just right. So I think he’s been very receptive to anything that we needed to do, and that’s been good. We’re not heavily into computational modeling approaches at this time however this is something we may put more emphasis on in the future. We have gotten to the point where we have standardized a lot of our methods and now we’re enforcing best practices so that people don’t change the methods, in other words, we have S.O.P.’s that are signed off on and you have to use those when you run the assay which helps when you are looking at these parameters across many different compounds. So once you develop and fully validate an assay you don’t have to worry about it changing over time. This is just a question, but would there be value for companies like us who didn’t need to have a Merrill database, because they can get everything in one database with CDD Vault with versioning and an audit trail, I mean that’s something to think about but, for future upgrades.
So more of an IP-like module?
Yes, a module where you could put in all of your documents, all of your SOPs, you know, all of your reports, all of your collaborators reports, things like that. I mean, in the Merrill database we have a write-up around our experiment and we have notebook references, and we have conclusions, you know, so they don’t take up a lot of physical space, in the sense that these aren’t very big files. Some of the imaging files probably are pretty big, but in general they are not very big files. So it’s something that if CDD Vault were to build in that functionality then a company like us wouldn’t need to have a Merrill database.
How far would an Electronic Laboratory Notebook go towards that?
Yes, that’s what I think will be the solution. Right now our notebooks are physical notebooks but they’re all scanned, for instance. So, we are evaluating electronic laboratory notebooks at this time, but we haven’t decided to adopt one yet.
You know that in 2016, CDD is developing a fully functional electronic notebook?
Yes, we’d like to think that we will certainly get to evaluate it. I think we know we have to migrate to an electronic format because at some point when we start to work with the large pharmas, it’s going to be easier for them to interface with us. We don’t want a situation where we have apples and oranges when we’re trying to compare things and creating resistance to collaboration just by virtue of the way they are, or we are handling information, so that’s important. I have seen also, that even the scanned notebooks have been extremely helpful because, you know, it’s a lot easier to access information electronically than going to the filing drawer, pulling out the notebook and finding the page. So we’re modernizing in those ways and would be interested to see the CDD Vault ELN that when it gets rolled out.
All right. So, I have a few questions about the process of who in your company uses CDD, who logs on, about how often they log on? Who does the searches? Do you have people in different functions using CDD Vault in different ways? Like, you have your biologists who are calculating curves, you have your medicinal chemists doing their SAR mining… What are the diverse roles or ways that people in your company are using CDD?
Our biologists perform screening assays and import the data to CDD Vault and our medicinal chemist and scientists search and analyze the data. We also have a regular meeting that cycles through all of the functional areas. When we have those, we use CDD Vault to generate reports that are used to communicate the project status to the whole team. So we’re kind of using it both for internal reporting and then also for a little bit with the external reporting as well.
And then how do they securely get the information to you?
They send us encrypted files with the SMILES codes and then we incorporate it into our CDD Vault. It’s working out really well. We are working with Allen Reitz from Fox-Chase Chemical Diversity Center, he’s a highly respected medicinal chemist.
Are you using the new visualization module in CDD Vault (CDD Vision)?
Yes. We have started to use it to make these four dimensional plots and they’ve actually been very good for communicating our chemical space. In fact we had a poster at the Alzheimer’s Association international conference where we included a colorful plot in the poster directly from CDD Vault. I think it was a really good communication tool and some of the people coming up were big pharma med chemist-type guys and they were very interested in the plots. So it was a useful way to communicate to them what our series looks like across multi-parameters that are related to being drug-like. We’d like to expand off of those early analyses because, of course, we were very novice at using it. I saw today in Charlie’s presentation that there are multiple ways that we can enhance our use of it given how powerful the new Visualization module really is. Now the modeling feature is very interesting. We did go through that and do one model just to see how it worked. I can’t give you an answer as to whether it was predictive. I can’t tell you whether the model was effective at finding hits because, I think we built such a limited model for the first time that we tried it out, that whether the model was functional or not wasn’t that important. It was more like a proof-of-concept for the experience of building the model. But what I would like to do is to try to build a model to use as a way of separating real trends from outlier singletons. We have many singletons from our assays that are interesting and we’ve done work around three of them. We’ve focused mostly on one series because when you have a series you have more confidence in the underlying signal. When we performed a large screen of a hundred thousand compounds, some of our series had quite a large number of hits in them. And so if you’re getting thirty or forty hits, all of the same structural chemotype, it easier to focus there. But we have also found some success from the limited work that we’ve done on the singletons. So the question is, instead of trying to do synthesis around many singletons, could we do some virtual screening using a CDD built model and would some of those singletons come up more interesting than others and maybe help us to prioritize them. So I’d like to see if that works.
Another thing is, now that we have a lead series, we’re in discussion with some pharmas about whether we could go into their database and pull out similar structures that we can then test. So one of the first things that I would love to do is to run a validated CDD model on those structures to see if any hits come up. Then the real proof in the pudding is, as you know, do any of them work in the assay? So that’s really where you see the potential uses of the model. What would make the model valid is its ability to be predictive. But it doesn’t mean that it has to predict everything. If it’s predicting a subset and it’s pointing you in a direction, that would make it useful even if it’s not predicting everything perfectly. Because we use multiple approaches to solving these hard questions in searching for drug-like chemical series for Alzheimer’s disease. So again, I think around the singletons, which tend to get ignored anyway, that would be one area that I’d like to try to use the modeling more. Of course, you know my background is in molecular biophysics, which makes me not afraid of, or hesitant to use the modeling functionality.
Have your research efforts benefitted from your use of CDD Vault?
Absolutely, our research efforts have benefited using CDD Vault. It has been a very useful and enabling experience and I recommend it highly to potential users. Also, I am looking forward to seeing some of the other modules you are planning to roll out including the electronic laboratory notebook. Also, I want to take this time to thank you and Barry for the opportunity for me and Oligomerix to participate in the Spotlight!
This blog is authored by members of the CDD Vault community. CDD Vault is a hosted drug discovery informatics platform that securely manages both private and external biological and chemical data. It provides core functionality including chemical registration, structure activity relationship, chemical inventory, and electronic lab notebook capabilities!
CDD Vault: Drug Discovery Informatics your whole project team will embrace!