Founder, president, chairman,
Name: Charles Witkowski
Position: Founder, president, chairman, and CEO of Protein Discovery
Background: Founder of Protein Discovery, 2001; project manager at Motorola, semiconductor product sector, 2000-2001; MBA, Vanderbilt University, 2003
Based in Knoxville, Tenn., Protein Discovery uses high-throughput mass spectrometry to develop products and services for molecular research, drug discovery and development, and diagnostics.
Earlier this month at the Dow Jones “Health Care Innovations 2006” conference, Charles Witkowski, its founder and president, made a presentation on clinical mass spectrometry addressing challenges affecting the routine adoption of mass specs for medical research and clinical practice.
ProteoMonitor caught up with him this week to get his thoughts on the subject and what lies ahead for the mass spec industry and other proteomic-based businesses for the future.
Why did you start Protein Discovery?
I guess I got my entrepreneurial roots when I started my first company when I was 17. It was a medical equipment company. I sold that company when I was 23 or 24. It was equipment for facial surgery. And I went back to get my MBA, and it was while I was getting my MBA that I worked for Motorola Semiconductor. But I knew that when I went back to get my MBA that the intention was to augment my skill set such that I could start a company that was in the space of electronics and molecular biology.
I’m guessing that you weren’t specifically looking at starting a company dealing with proteomics, or protein-based applications, or mass specs.
I was, specifically mass spectrometry. I felt that there was tremendous opportunity there, but more than that, my father was a scientist and also a physician, and so I grew up hearing about it. And as a business person, I just felt like we as a society were making tremendous progress in the area of life sciences. I believe in 100 years, people will look back at this as the age of life sciences with the sequencing of the human genome and a lot of the advances that are coming to market, coming to bear right now.
That was my general interest, to be involved in that. I saw, in terms of a detection platform, it seemed to me, and still seems, that there is no detection platform that rivals the performance or power that one can achieve with mass spectrometry. More specifically, I and probably everyone else in the industry recognized that there was a real need for sample preps to really enable mass spec to do the type of things that it should be able to do.
We were looking at developing products, sample preparation products for life sciences applications in mass spectrometry. That remains our business model.
The first product that we launched was just back in June this year, a specialty reagent for mass spectrometry called PPS Silent Surfactant. And it is an acid-cleavable surfactant for protein solubilization and cell-membrane disruption.
We actually launched a service based on some proprietary technology called tissue imaging mass spectrometry and we launched that service in November of last year and provide that service to pharmaceutical partners on a fee-for-service basis. The last product which we’ll be launching in about a month at HUPO is kind of our flagship product, if you will. That is called our MALDIplex. It is an instrument, and cartridge, and software package for the rapid parallel preparation of up to 96 samples for MALDI analysis.
You have an MBA. Do you have a science background?
I do not. I’m a business guy. I grew up around science and I have a passion for science, but my formal training is in business. I’m familiar with [scientific concepts]. I can remember going with my dad and spending nights at the morgue. But I did have to learn quite a bit about the field of proteomics in general when I got into it. But it’s something that comes naturally to me. Like I said, it’s fascinating to me.
How did you learn about mass specs? Even though your dad’s a doctor, they don’t use them.
No, they don’t. It was really when I was kind of investigating the area of what was happening in molecular biology, and whenever one starts to look pretty deep there, you see mass spectrometry everywhere.
Recently, you gave a talk about the use of mass specs as a routine aspect of medical research and clinical practice. What exactly do you mean by that?
It’s clearly already being used routinely in clinical research and drug discovery, but we do foresee that mass spectrometry is going to have a special place in areas of diagnostics. And one area that we’re particularly interested in is in anatomical pathology. I think mass spectrometry holds tremendous potential in that area for providing massive on-tissue multiplexing which is something that’s currently not achievable. Using techniques such as fluorescence you can see if you’re lucky maybe five, six analytes at a time, whereas with mass spectrometry, and tissue imaging and some of the reagents that we’re developing, we can look at dozens or hundreds of analytes simultaneously with special resolutions. And we think that that’s really a unique enabling application of mass spectrometry.
The focus of the talk was ‘What does it take to enable the extension of mass spec into the clinic?’ Because you’ve been hearing about it for years and people have been talking about SELDI and all this stuff. We really do believe that it is simplifying and making the whole sample preparation process automated and reproducible.
Much of the criticism is that so much of it isn’t reproducible. So we’re not there yet.
I think in a label-free approach, it is very, very difficult and I’m not sure we will get there. But with some of the chemistries we’re developing, other I-tags — these are tags that are photo-cleavibly linked to reporter molecules, i.e. antibodies, for example they can go down on a tissue — we can get down there. And in areas such as anatomical pathology, there’s a real power to using mass specs as opposed to fluorescence.
For clinical practice, what is realistic for the near term?
First of all, the number one thing that is required is a platform, start to finish from sample collection through data analysis, that is standardized and reproducible. And once that platform is in place, of course, we believe that it will be adopted by the marketplace much in the same way as other diagnostic platforms have in the past. We believe they will start off as homebrew tests that provide unique value to the actual patient, [and then] will eventually go through more timely and costly FDA approval process for in vitro diagnostics. And eventually, we believe [they] could be used in either reference labs and/or in hospitals by themselves. The key, though of course, is building that platform that provides a level of automation, ease of use, and reproducibility specifically, as well as sensitivity.
I think the real fork coming right now is that existing platforms do not provide the level of sensitivity and selectivity required for existing assays, so the field has [said], ‘OK, we can’t see the low-abundance stuff, let’s see if we can find anything in this high-abundance range.’ And thus far that’s been difficult. Combine that with a lack of selectivity or reproducibility and I think it’s probably unrealistic that that approach will make it to the clinic.
You hear from researchers and even the business side of the industry that technology has lagged and researchers are being held back from doing things they’d like to do. Would you agree with that and if so, what can be done about it?
I think we’re moving forward and we’re getting a grasp on some of the issues such as dynamic range and that sort of thing. But there are two approaches — the bottom up approach, of course, has a lot of nice benefits, but the problem is still technological in nature in that the throughput is so low that the number of samples that need to be analyzed to get to a statistically significant conclusion can take years and years at the current pace. There’s a lot of good work occurring there through the use of monolithic columns and microfluidics that can have tremendous impact. But we’re actually approaching it from the top down whereby we with our multiplex sample prep station can prepare up to 96 serum samples in parallel in about an hour for MALDI analysis, and once they’ve been prepared and the ion signals of interest have been validated across several hundred samples, then we can actually perform an identification step off our MALDI plate that permits widespread identification that’s also required.
You’re advocating a quantitative approach to proteomic research. Why hasn’t the industry taken up that approach?
We have heard from researchers that some of the techniques to do so are difficult to implement, difficult to automate — [I’m] talking about some of the labeling techniques — and they’re difficult to implement in a high-throughput manner. I think that really boils down to it. We still perceive a high barrier there. It hasn’t been adopted as quickly and widespread as one might think. But even so, with the quantitative approach, you still have to run enough samples, especially if you’re looking at biomarker discovery, that you can get high enough levels for it to make sense. And I think that’s where we’re seeing the shortfall of a lot of the bottom-up strategies that employ labels and that sort of thing.
Is your company dependent on the technology that’s being developed and built by mass spec or liquid chromatography companies?
No, it’s not. We are developing our own sample-prep front ends. We do have, currently, somewhat of a partnership with Bruker, and we talk a lot with the other mass spec companies. We feel that that part of it will continue to advance as necessary, but it’s really what you do with the sample upstream, we find, that has the biggest impact. Going all the way upstream to how it’s collected, of course, has a huge impact, the sample that is. So that’s really our focus, that front-end piece