Nick Nelson:
(silence) Okay. Good afternoon, everyone. For our audience, I am Nick Nelson. I'm the senior vice president and chief business officer at Nautilus. Here with me today, I'll be interviewing Dr. Josh LaBaer. Josh is currently the executive director of the Biodesign Institute at ASU, and he's also recently just served as the president of the US Human Proteome Organization. So Josh, welcome.
Dr. Josh LaBaer:
Well, thanks. Thanks for having me.
Nick Nelson:
Appreciate your time. We're happy to have a chance to learn from you and hear your insights on the proteomics market and on Nautilus.
Dr. Josh LaBaer:
Great. Yeah.
Nick Nelson:
So if I could, I'd like to ask first if we could just kick it off with you sharing a little bit of background on yourself, your role at the Biodesign Institute, perhaps your areas of research today.
Dr. Josh LaBaer:
Sure, sure. Very brief background, I'm a MD PhD. I went to medical school at UCF, where I also did my PhD. I trained in Boston for many years. I did the full clinical training. I did an internship in residency at the Brigham Women's Hospital and then a medical oncology fellowship at the Dana Farber. So I got board certified in internal medicine and medical oncology. But I've long time been focused on research. My research primarily is involved in diagnostics and early detection of disease. Fundamentally, the tools that I use in my research involve proteomics, especially proteomics driven from expressing proteins from cloned copies of genes. I serve as an editor for the Journal Reporting on Research. I was on the board of scientific advisors for the National Cancer Institute for a number of years, and I'm an advisor for the National Cancer Institute's clinical proteomics program, the CB Tech Program.
Nick Nelson:
Wonderful. Thank you. So, I mean, given your background, it's a perfect opportunity for us to learn a little bit about your experience with the technologies out in the field today. Share with us a little bit on what your lab's doing, what technologies you're interested in or familiar with.
Dr. Josh LaBaer:
Right. So back when I was at Harvard, I started the Harvard Institute of Proteomics, and the approach that we have always taken was proteomics from their perspective of taking cloned copies of genes and producing proteins in a variety of settings in high throughput. The notion that we take is that we want to understand how, by producing proteins in high throughput, we can both understand what they do, but also how they are altered in disease or how they can trigger the ability to identify biomarkers for early detection of disease.
Dr. Josh LaBaer:
So my lab has cloned tens of thousands of full-length copies of genes. We have largest collection of genes for the human I think anywhere on the planet, close to the full complement of genes now, around almost 18,000 full-length genes, as well as tens of thousands of genes from both microorganisms and model systems.
Dr. Josh LaBaer:
A key part of our technology has been to produce those proteins in a protein microarray setting. So we print the genes on the chip and then synthesize the proteins in situ on the chip, which we can then probe with serum from patients to look for responses to proteins that betray the presence of diseases. So in that context, we've identified a number of biomarkers for the early detection of breast cancer that have been now licensed out and are part of a CLIA-approved test for breast cancer, as well as biomarkers for detections of various inflammatory diseases, autoimmune diseases, and also infectious diseases.
Dr. Josh LaBaer:
So my lab has also been involved in a number of other high throughput projects. We had a $40 million project with an agency called BARDA to assess responses in the context of a nuclear explosion using a technology called qPCR, which we've now pivoted to use for detection of COVID. We now run a clinical testing lab for COVID-19 that has processed more than a half million samples in the state of Arizona. Then more recently, we have been developing some serological assays for detection of disease and became part of the SeroNet, which is NIH's response to COVID-19 for serology. We're one of four capacity building centers in the country to develop a high throughput test for detecting response there.
Nick Nelson:
Wonderful. Thank you. So, I mean, I think it's pretty clear your background and experience with proteomics is extensive. One of the themes here, obviously, given Nautilus's interest in building next gen proteomics systems and analysis, we're really interested in the pros and cons of the methods that are out there in the field. Can you talk to us about your experience with proteomics platforms today, maybe some of the benefits and some of the drawbacks?
Dr. Josh LaBaer:
Right. So typically, most proteomics are just focused on sort of measuring proteins and samples, falls into two general classes. One would be sort of antibody-based, and the other would be mass spectrometry-based. So the antibody-based technologies have some advantages. If you have good antibodies to specific proteins, you can readily detect the presence of those proteins in samples. It's a relatively fast assay if you have the reagents. The challenge there, of course, is that you pretty much are forced to make antibodies to specific protein species almost one by one, really, and there is a wide variation of quality in affinity reagents to specific proteins. A lot, a lot of antibodies out there are not very good.
Dr. Josh LaBaer:
I've been involved in a project to try to help develop better tools for making antibodies better because there is such a variation and because you never know when you get an antibody, when you buy one whether it's going to be good or not. It's very much a buyer beware situation. Of course, to have antibodies against proteins, you have to know what proteins you're looking for. You have to know it in advance. You have to sort of plan ahead as to what you're looking for.
Dr. Josh LaBaer:
The mass spectrometry approach, a little bit different. The benefits of mass spectrometry is you don't necessarily need to know a priori what you're looking for. You can apply a mass spec broadly to samples. You can put a whole sample, and you kind of look at what's there. The problem with mass spec, of course, is that it's not a particularly sensitive technique. It really looks at abundant species. Anything that's rare or uncommon it's going to have a hard time detecting, and, of course, some of the most important biomarkers that we want to look for fall into that category of rare proteins.
Dr. Josh LaBaer:
The other problem with mass spectrometry, of course, is that it's taking sort of bulk measurements of the sample. I mean, a metaphor one could use here, not perfect, of course, is if you had a room of 1,000 people and you knew that there was $1,000,000 distributed in that room, you wouldn't necessarily know whether there was $1,000 in the pocket of every person in that room or whether there was $500,000 in one person and several people with $100,000 and everybody else had $100 in their pockets. There's a big difference in terms of understanding what's going on in that room and the dynamic, depending on how that is distributed. Mass spec can't tell you that, and that's, I think, something we need to get at here.
Nick Nelson:
Yeah, that's very helpful. Thank you. Maybe before I launch into my next question, I should mention just for the audience that the Dr. LaBaer is a longtime advisor of Nautilus and our chief scientist, Dr. Parag Mallick. He's also followed Nautilus since its inception back in 2016, so just to kind of set the stage for folks. So Josh, I think you've been tracking progress for a while. I wanted to just for a moment here, in your words, your understanding of how the technology works and maybe kind of some of the more recent reactions you've had to our data and the method.
Dr. Josh LaBaer:
Right, right. So what I think is really special about the approach that sort of Parag conceived of is this idea that you're going to take proteins from a sample and, using their technology, affix individual molecules of proteins on a kind of landing pad. So now what you've got is, on some surface, individual proteins separated from each other that you can then probe using affinity reagents to those proteins. The affinity reagents you use can either be very specific to a specific protein, or they can be general agents that bind a specific motif in proteins. That becomes important later, because what you're ultimately going to do is take a picture of which proteins are recognized, remove the reagents, and then come in with the next round of reagents. You can repeat that cycle over and over again using different affinity reagents and essentially measuring the frequency that a particular motif or a particular protein appears in your sample.
Dr. Josh LaBaer:
This allows you to do a couple of things. On one end of the spectrum, you can, by using ... and, again, this involves a lot of the sort of machine learning approaches that they're using, as well as some of the sort of informatics that they've cleverly worked out in terms of how to mix and match different affinity reagents. You can look at a whole proteome and ask what's there, basically identifying specific proteins by the content of the motifs they contain, or you can look at post-translational modification of proteins, if you have antibodies to those, and ask which proteins have which post-translational modifications, or you can look at just specific sets of proteins based on the interests that you have.
Dr. Josh LaBaer:
So kind of a wide spectrum of applications there, but the notion here fundamentally is that you are digitizing the proteome, right? So this gets at, in the context of genomics, it was a big deal when we could get single-cell genomics or when we could get single-cell measurements of the RNA in cells. That allowed us, for example, to look at a cancer and ask, "What's the distribution of specific mutations in individual cells in that cancer or gene expression?" Well, now we can do the same thing with proteins, and that kind of opens up ... It revolutionizes, really, how we think about proteomics.
Nick Nelson:
When you think about that, and I know you track what's out there in the emerging technology space, I mean, there's some other companies and other methods that have been in development for some time, some starting to emerge with data, others still early. I mean, what are you aware of in the community for other methods that kind of compare to what Nautilus is building?
Dr. Josh LaBaer:
So let me first say I don't think anything really compares to what Nautilus is doing. There are some new technologies out there. There are some antibody technologies that use antibodies to specific motifs followed by mass spectrometry, right? So that's a quick way at assessing specific proteins and species. It's kind of a hybrid method between the antibody approach and the mass spec approach. Then there are other types of affinity reagent protein, things like SomaLogic or things like that, where they have various binding reagents that can look for specific proteins in samples.
Dr. Josh LaBaer:
The problem with that particular approach, I think, is that the characterization of their different SOMAmers, as they call them, is not particularly great. Some of them are okay. A lot of them, they've sort of made them against a particular molecule in the hope that it buys with good affinity and good avidity and all that stuff, and, of course, they're looking at subsets that they've been able to make their reagents against. Everything else, they can't do.
Dr. Josh LaBaer:
So I think those are okay technologies. All of them still have the bottom line limitation that you're looking in bulk. None of them are digitized technologies. None of them allow you to look at specific protein molecules and ask what's specific about that particular molecule?
Nick Nelson:
That's really helpful. Thank you. I mean, the other question I have for you as a follow-on, I know we've been really trying to build some progress here in a few areas in the company. Then more recently, at the end of last year, we shared some data on proteoform detection. You got an early peek at that. So I wanted to kind of ask for the group here, what were your reactions to some of the data that you've seen coming out of Nautilus?
Dr. Josh LaBaer:
Yeah, right. So, obviously, I'm very excited. I'm excited on a couple levels. So first of all, I think, step-by-step, one of the things that Nautilus has done is sort of developed technologies at a lot of different levels for what they're trying to accomplish. I think each of those solutions is itself of great value. The ability to be able to lock down an individual protein on a surface apart from all other proteins is huge, and I think that in and of itself is an ability that will become useful in other platforms. The ability to get very strong signals in detecting an individual protein, again, huge, huge approach there.
Dr. Josh LaBaer:
But, of course, I think the thing that's especially exciting now is that they are realizing that they don't necessarily have to look at whole proteomes. They can look at specific modifications, specific proteomes, specific proteoforms and ask questions about the abundance of those individual species in specific samples. Again, to use kind of a silly metaphor here, if the traditional approach to proteomics would be kind of like looking at a store full of shoppers and being able to say that there are many people in that store buying books and some of them are buying books on pregnancy and there are many people in that store buying vitamins, some of them buying folate, that's one thing. But when you can look individually and say, "In these individual shopping carts, there are both books on pregnancy and bottles of folate," now you can say, "Aha. That's a shopper that's probably pregnant," right? It's a different kind of way of thinking, and it gets down to the importance of digitization.
Dr. Josh LaBaer:
So for me, this is now a whole new chapter, right? It allows us to say, "In these individual samples, I can tell you that these are these very specific proteoforms that have a phosphate on this amino acid and a phosphate on that amino acid, as opposed to just some proteins that have phosphates on this amino acid and some proteins in here that probably have it on that one." Now we can say they're on the same molecule, and it's a game changer.
Nick Nelson:
That's exciting. I'm interested to hear, now with this capability, what should we do with it? I mean, where do you point it? What are the disease areas or applications that make sense to use it?
Dr. Josh LaBaer:
Right, right. So I'm obviously biased. I'm a cancer doctor by training. So for me, one of the most important next steps for this kind of approach would be to identify biomarkers that either help me with the early detection of cancer, in which case I can save lives, or help me understand the characterization of specific cancers so that I can both better prognosticate outcomes and maybe better target my therapy, because I can say, "This subtype of cancer that we're looking at here would best respond to this class of drugs or this type of therapy."
Dr. Josh LaBaer:
So that's where I would go. Of course, the same could be said for autoimmune diseases. The same could be said for a whole series of other clinical conditions or even for infectious diseases. So lots of different sort of diagnostic possibilities, and the ability to have a tool that can sort of broad look at specimens will help me identify those biomarkers. Of course, those will be opportunities for Nautilus to get intellectual property on those biomarkers so that they essentially could own their own diagnostic tests to look for specific diseases.
Nick Nelson:
That's an interesting thought, because in watching and observing the market evolve, particularly the genomics market, there's just a tremendous amount of clinical application now. Proteomics hasn't followed suit. What do you think the barriers have been that have held the proteomics technologies back from really entering the clinic in a meaningful way?
Dr. Josh LaBaer:
Right, right. Well, I mean, I think a key issue here for sort of most of these technologies are technologies that work in laboratories, particularly things like mass spectrometry, which although it's a science is also an art, and it requires very careful practitioners. So it's not easy for clinicians to imagine ordering tests to kind of getting mass spectrometry reads. I think that's, in my view, one of the things I think Nautilus ought to be thinking about in the future, is Nautilus has developed a technology that can measure individual modifications on specific proteins. That's going to lead them to develop unique diagnostic tests for specific conditions, but they are the ones who have the technology that could be used to actually read those diagnostics, right?
Dr. Josh LaBaer:
So I think that they should be building boxes that they can place in clinics, that they can place in CVSes, that they can place in doctors' offices or hospitals that can actually read the very biomarkers, the very diagnostic tests that they will have discovered, because they've got the tool to read those quickly. I think there's a whole market there in the clinics that I think could be very important for them.
Nick Nelson:
I mean, when you think about it, there's been certainly a rush in terms of new diagnostic applications, particularly in liquid biopsy. So this is an area that a lot of us follow closely, and you look at the application of things like cell-free DNA for cancer as a biomarker.
Dr. Josh LaBaer:
Right.
Nick Nelson:
What do you think the pros and cons are in looking at nucleic acids versus proteins in a liquid biopsy setting?
Dr. Josh LaBaer:
Right. Well, cell-free DNA is certainly interesting. In the context of cancer, it is really not readily detectable until it's already too late, until the cancer is really advanced. By and large, DNA tells you what could happen, whereas proteins tell you what is happening, right? Proteins get modified in the course of disease. They are, in fact, the drivers that make disease disease. I would argue that all disease is a result of protein malfunction, and virtually almost every therapeutic either alters proteins or is itself a protein. So to me, I'm very much a protein chauvinist, and I think that's where we really need to go for active diagnostics and prognostics. So that's where I think we should be focusing right now, is what's happening at the protein level and what's happening at the modification of the protein level? What's happening in the post-translational protein level?
Nick Nelson:
I've got another question kind of on that same thread. Then we're starting to get a feel now for the diagnostic potential and the biomarker discovery potential, PTMs and proteoforms. That's taking shape. Another area is clearly there's a need to just unlock more data in the protein, right?
Dr. Josh LaBaer:
Right, right.
Nick Nelson:
I mean, there's just a tremendous appetite for that in the market today. So we've built this roadmap within Nautilus, and I think you've seen it, where we're starting to make headway over the course of the year. The first major milestone is we want to unlock, say, 2,500 proteins per run, and then we want to move into 10,000 proteins per run and then ultimately the whole proteome.
Dr. Josh LaBaer:
Right.
Nick Nelson:
So as you think about that, I want to hear, in your view, does 2,500 proteins in a run unlock something new for you? What would you do with it?
Dr. Josh LaBaer:
Well, I mean, first of all, I think the ability to look at five proteins, maybe, [inaudible 00:21:01] species is already a big deal to me. 2,500 is fantastic. Obviously, as Nautilus has the ability to look at larger and larger protein species and then, in parallel, invoke machine learning or deep learning techniques so they can actually understand what those patterns mean, that becomes sort of the facial recognition of disease, right? I mean, this is all of a sudden now, they can start looking at both biology and medicine in a way that we've never been able to look at before, right? Because now they're coupling this ability to do deep learning with all of these potential data points in a much broader way. So I think it's huge, and I think the beauty of what they've got is that they can, in the short term, do specific diagnostics on specific proteins and specific proteoforms, use that to help bootstrap the development of better reagents to keep looking at more and more proteins, and broaden their capacity to look at the entire proteome.
Nick Nelson:
That's great. Thank you for the feedback. I think I actually covered most of the questions I had.
Dr. Josh LaBaer:
Great.
Nick Nelson:
Were there any other topics that you wanted to bring up or [crosstalk 00:22:21]?
Dr. Josh LaBaer:
No, I think this is a really exciting opportunity. I mean, I think it really is the evolution of proteomics, right? The next step beyond looking at bulk, starting to look much more broadly by digitizing the information and then coupling that with sort of deep learning approaches to really imagine what you're looking at.
Nick Nelson:
Great. Well, thank you again for your time. We so appreciate it. Then, of course, we'll keep you posted on progress.
Dr. Josh LaBaer:
Great. Can't wait to see what happens next.
Nick Nelson:
All right. Very good. Take care.
Dr. Josh LaBaer:
All right. Thanks.
Nick Nelson:
All right. Bye-bye.
