Amy Herr's research focuses on bioinstrumentation innovation to improve quantitative measurements in life sciences and translating that work to provide better clinical diagnostics. Amy is Professor of Bioengineering at UC Berkeley.

Transcript

Speaker 1:        Spectrum's next. 

Speaker 2:        Mm MM. 

Speaker 3:        Yeah. 

Speaker 1:        Welcome to spectrum the science and technology show on k a l x [00:00:30 ] Berkeley, a biweekly 30 minute program bringing you interviews featuring bay area scientists and technologists as well as a calendar of local events and news. 

Speaker 4:        Good afternoon. My name is Renee Rao and I'll be hosting today's show. Our guest this week is Amy, her associate professor of bioengineering at UC Berkeley. Amy is a teacher and a researcher. Her research focuses on bioinstrumentation innovation to improve quantitative measurements in life sciences [00:01:00 ] and how to translate that work to provide better clinical diagnoses. She is a pioneer in the new field of proteomics. Brad swift and I interview Amy, her. 

Speaker 5:        Amy, her. Thanks very much for coming on spectrum and welcome. Thank you. I'm very happy to be here. How did you become interested in bioengineering? So I am actually a trained mechanical engineer and I think what really peaked my interest in bioengineering was during graduate study in mechanical engineering. I realized that a lot of [00:01:30 ] the measurement and instrument challenges that exist that face engineering today really are in the life sciences. So this messy area where things are not necessarily tractable or well-described protein measurement is an area that I've been interested in for some time and I've been working on. And it's especially challenging from the perspective of designing instrument technology, measurement technology. What are protein biomarkers and what makes them elusive? Yeah. So protein biomarkers really is just sort of a catch [00:02:00 ] all phrase for indicators of disease state, um, indicators of living, organisms, response to treatment, just sort of indicators of what's going on in the organism at a particular time. 

Speaker 5:        So there's many different types of biomarkers. You may have heard quite a bit about this genomics revolution and our use and understanding of information that's coming from nucleic acids. And what we're really looking for in Dow is building on what we've learned from our understanding of nucleic acids. How can we try [00:02:30 ] to understand proteins, which are the effectors of function, if you will, in living organisms and really try to use that information from proteins to understand all of these questions surrounding disease. So who has a disease, who might respond to specific treatments, who might not respond to specific treatments? How you are responding to specific treatments and in our mind it's released the next phase of what genomics has laid the groundwork for an area that we call proteomics. Can you give us a quick run through [00:03:00 ] of how molecular diagnosis works now and what new things you are trying to detect and what new information we can get from those? 

Speaker 5:        I guess it has been striking to me as an instrument designer, innovator developer. If you take a look at our understanding of the role of proteins in disease right now, there's a treasure trove I would say, of information that's come out of basic discovery. So trying to understand what proteins are upregulated or downregulated or modified in [00:03:30 ] response to disease or treatment of disease. Right. So I would say there's definitely more effort that needs to be done in discovery, but we've done a lot of great work in discovery. A huge challenge and unmet need to use the engineering design terminology that exists right now is we have these potential indicators of disease or response to disease or prognosis, but very, very few of them have made it into a clinical setting into a diagnostic. Right now there are less than a hundred [00:04:00 ] different biomarkers that are being used for diagnostics. 

Speaker 5:        That includes nucleic acids of DNA, RNA and proteins as well, just metabolites as well, right? So very, very few of the known existing bio molecules are being used in any way as a diagnostic measurement. And so there's really a huge gap right now between all of these promising markers that have been identified and those that are currently being used to make a diagnosis. So one of the things that we're [00:04:30 ] trying to do is to just build a basic framework for measurements that will allow people to make many, many, many measurements of a particular biomarker of potential interest so that you can look at many, many different patients' samples, many, many different disease states. We won't be really data limited. So the technologies that we use right now for a lot of these protein biomarkers to see whether or not the promising ones actually answer a clinical question, they're really rate limiting. 

Speaker 5:        [00:05:00 ] They're really slow or they require a lot of material and in some cases this biospecimens these materials from patients are precious, hugely limited, right there, sparingly available. So we're just trying to think about ways that we can use these microfluidic architectures that require just tiny amounts of sample to run one measurement. How we can use those to scale up to make thousands of measurements. We're right now tens of measurements can be difficult and to make those measurements on, you know, a [00:05:30 ] microliter of sample from a patient as opposed to tens to hundreds of microliters. So that for us, this so-called biomarker validation question getting from yet this might work too. Okay, here are the clinical questions this marker can or cannot answer as the gap that we're trying to fill. Are you building these instruments? A major focus of my research group is looking at innovating new instrumentation, new technologies. 

Speaker 5:        So by understanding the underlying physical principles [00:06:00 ] of the types of transport that we use. So electrophoresis and diffusion and by understanding unmet clinical or life sciences needs. So questions or challenges that currently exist out in life sciences laboratories or in clinical laboratories. We're basically trying to bring those two aspects together to develop new tools. All of the new tools that we develop are developed really to meet an unmet need either in the clinical setting or the life sciences setting and they're built with an understanding these underlying principles, but they all [00:06:30 ] have to be validated. So when we make a measurement with a new tool, we have to have some confidence in how well our measurement reflects our current understanding of the systems. And we typically do that by using conventional gold standard measurement technologies where appropriate. I think recently we've just come into this really interesting and exciting gray zone where we can make measurements that there really are no existing tools to be able to validate whether our measurement makes sense or not. And so we've had to put some effort and careful thought [00:07:00 ] into how do we validate our measurements using maybe indirect approaches so that we can say with some confidence the limits and the benefits of the tools that we're introducing. 

Speaker 4:        You said earlier that a lot of your research comes from trying to meet the unmet needs of both the life sciences and the technological aspects. How do you go about picking which needs to meet? Do you find ones that you think, okay, well this is doable, or do you find ones that you think, maybe no one else can do this? I'm going to work on it? 

Speaker 5:        Right. [00:07:30 ] That's a great question. So as an engineer, as an engineering designer, one of the first things that we do is really try to understand the world around us and try to understand how people approach existing problems, how they define those problems, why they approach them in a particular way. But I think this is one of the most exciting aspects of the work that we do. It's certainly true that if you get this first stage, this identification and understanding of unmet needs wrong, you're going to go down the wrong path, but if you get it right, you can make a huge difference in terms [00:08:00 ] of how people are approaching either science or medicine and our work is really translational in that way. So we're engineers and we're passionate about making excellent measurements and as you say, measurements that are currently not possible are the measurements that we're really looking to impact. 

Speaker 5:        Measurements that are currently possible but needs significant improvement. We do focus on those as well, but when you can find a measurement that when you're talking to a biologist and explaining kind of what you can do and they look at you and say, oh my gosh, there's no [00:08:30 ] way I could do that right now, then you know you've hit upon something that's really important to at least consider further to fill a gap and unmet need that's out there at the present time. In many ways, I think it reminds many of us of why we chose to be engineers in the first place. I mean, certainly I can speak for myself and say I'm really excited about being able to make measurements that no one else can make. And understanding how those measurements, how good they are, how much more improvement they need, and maybe trying to understand the physics an