At a Glance:
Name: Nancy Ng
Position: Director of Biomarkers Discovery and Development program, Protana, since 2003
Background: Manager, Drug Development program, Hemosol, 1997-2003; Postdoc, department of neurobiology, Stanford University, 1990-1993; PhD, department of clinical biochemistry, University of Toronto, 1990
Last week, Nancy Ng gave a talk at the Biomarker Pipeline conference in Boston on protein biomarker discovery from early to late stages. ProteoMonitor caught up with Ng after her talk to find out about her background and work as director of the biomarkers program at Protana.
How did you end up as the director of biomarkers at Protana?
My connection with Protana is quite an interesting story. I did my undergraduate degree in biochemistry, and for my graduate studies, I trained in classical protein studies. We were looking at the structure-and-function relationship of this family of blood-antifreeze proteins from fish, and looking at how completely different structures of the same protein were still able to perform the same function, which was basically to bind ice crystals and inhibit their growth. That was a very fascinating project. And the outcome of all of that was that I was trained in all the classical methods of protein chemistry. My supervisor ran a protein analysis facility, and so we really had the luxury of doing as much immunoanalysis and protein sequencing by Edmund degredation as we needed. So even back then, during my graduate work, I was sequencing proteins. That was at the University of Toronto in the department of clinical biochemistry.
So while I was doing all these protein chemistry techniques, I also had to learn all about clinical biochemistry, which is the study of human disease and the chemistry of the diagnostic tests that are used to diagnose. So I had to learn all of that at the same time. It seemed at the time to be quite disconnected — I never dreamed that I would be working in biomarkers and now using all that training that I got back then.
My encounter with mass spectrometry-based proteomics happened in 1999 when I was working for another company. At that time I was leading a drug discovery program. I had purified a very low-abundance protein. I worked with [a company called] Ocata Proteomics to analyze the protein by mass spectrometry. I don’t know if you’re aware, but Ocata actually became MDS Proteomics, which became Protana last year. And I was actually the first paying customer of Ocata. So it’s really interesting that I ended up working for Protana.
What did you purchase from Ocata?
Well, I had purified this protein, and there was this cell-based assay activity related to it, and we wanted to find out what it was. At that time, there weren’t really too many mass- spectrometry services available. We later found out that this protein was a novel protein, and they had to do some de novo sequencing on it. Working with them actually made quite an impression on me, and that’s one of the reasons I later joined them. It’s quite interesting how things work out.
What became of that protein?
Well, I’m no longer with that company — Hemosol — and I can’t talk about it.
What made you decide to make the move from Hemosol to Protana?
Well, Protana wanted to set up the biomarker program, so I joined in August of 2003. At that time my only experience in proteomics was based on my work with the Ocata scientists in 1999. And I had kind of kept an eye on them, and I kind of knew that they had published some really amazing work with the PATH maps. I was really excited to be working with a company that was conducting such cutting-edge research. The caliber of the people are really amazing. So that’s really one of the main reasons that I joined. I’m still amazed at the caliber of the people that they have here.
When I joined, they were switching over, starting to really refocus the company away from drug discovery over to biomarker discovery. Because of the nature mass spectrometry — it’s a semi-quantitative method — and what we were trying to do was obtain quantitative information from the technology, so basically we had to rethink and create new tools and apply statistical design so that we were able to obtain meaningful differential analysis data from this platform. And I think we’ve done so quite successfully. And we’ve done this in a really short period of time. So for both the mass spectrometry platform and the bioinformatics platform, we had to create new tools and create new ways of doing analysis such that we can obtain reproducible, stable platforms that are going to give us data that’s statistically significant.
When I first joined, there were just four of us at the time, and then gradually as the company was reorganizing, they decided that biomarkers was one of the main areas that they wanted to focus on, so now we have about 24 people.
What was one of the first things that you worked on?
At that time they had already established a collaboration with MDS Pharma Services, and they had actually just finished conducting a rat study of renal toxicity, and this was proof concept of really going from discovery through to developing an assay. So the study was done, we had the samples, and we basically had to create the tools and develop the platform to conduct that study.
That’s completed, and right now Pharma Services has selected about three candidate biomarkers. They’re still in the process of developing assays for all three candidates.
Does your company also do validation of biomarkers?
We always verify the data if we can using antibodies, and also if we can by looking at the raw MS data. With the assay development, that’s done in conjunction. The next step is to validate the assay using the samples that we used for discovery, and then move on to looking at human samples.
Has your company done any work on theranostics — biomarkers that are also potential therapeutic targets?
Well, what we’re focusing on mainly is to apply biomarkers for drug development, from pre-clinical to clinical. We also have some academic collaborations looking at biomarkers that can [have] potential prognostic and diagnostic use. Those are going to be much longer-term projects, because that’s going to be for physicians’ use, as opposed to use in drug development.
There are really two main streams. The level of validation is different depending on if the biomarker is going to be applied for drug development, in which they’re called “exploratory biomarkers”, versus use as either a surrogate endpoint or as prognostics or diagnostics. Because those would require actual clinical trials before they could be approved as a prognostics or diagnostics. So those are much longer-term projects. That’s not to say that any biomarker used during the drug development phase can’t eventually be used [as a diagnostic biomarker], but the [exploratory] biomarker can be used earlier on.
These [exploratory] biomarkers are used to select out which [drugs] are more efficacious? Which ones are safer? And to help drug developers to make more informed decisions, and hopefully, because of those informed decisions you’re going to get less attrition as it goes down the drug development pipeline. If [drug companies] could make better decisions earlier, that’s going to help them to save money.
What do you think of proteins being used as biomarkers? Is that something that’s harder to get through the FDA? In the FDA’s current Pharmacogenomics Guidance, protein biomarkers are not really addressed. Is this something that concerns your company?
Well, I think that as far as protein biomarkers go, the genomic biomarkers are actually something that are fairly new compared to protein biomarkers, which have been in use for many many decades. I think that actually genomic biomarkers, while their platform is quite stable, have not been in use for as long as protein biomarkers. For example, many of the clinical tests — a lot of enzyme assays, albumin itself is a biomarker — were protein biomarkers. If you look back at a clinical chemistry textbook from 30 years ago, you’ll actually see plasma protein profiles that they used to run standardly, and you’ll see different types of diseases will have different profiles. So even back then, they were using protein biomarkers to help analyze disease — not necessarily as a one-assay diagnostic, but to help physicians in their differential diagnostics, to help them come up with what’s wrong with the patient.
So with protein biomarkers, I think it’s still going to require the same rigorous attention. If you think about it, the protein biomarkers that they used to use — they spent decades studying it before it actually became in use. But now we’re trying to speed it up. So that’s why the FDA really needs to make sure that the same rigor and the same standards are applied to make sure this is a valid biomarker before it’s in widespread use, whether it’s a screening test, or if it’s a diagnostic. Because the implication of a diagnostic is that a physician may base whether he does surgery or not on the results of that test. So there’s profound responsibility that we have to take as to how valid it is.
I think if it’s screening, that’s a different application from determining which of ten candidates for drug development you should push through. It’s a different purpose. So that’s why exploratory biomarkers are really in a different class from a biomarker that’s going to be used as a surrogate endpoint, or as a diagnostic. You’re talking about people, versus whether or not a drug should move on in the drug development pipeline.
Do you find the field of protein biomarkers to be very competitive?
There’s a number of companies out there, and I think we’re all facing the same challenges. We all want to use biomarker discovery to help drug development. But I think because it’s still regarded as a new technology, people are kind of still waiting to see how it bears out. I think if we can get some successful biomarkers discovered through the use of proteomics-based discovery, then move on to become successfully validated biomarkers, I think people are waiting for that, and that’s just starting to emerge.
What are some of your company’s success stories, in terms of biomarkers?
We’re conducting some collaborations which I can’t talk about right now, but they’re more cancer focused. The two animal studies that we’ve conducted on the drug toxicity biomarkers and the biomarkers of disease progressions — we’re quite pleased with the results of those studies. I’m quite pleased with the work that we’ve done up until now.
I just wanted to mention that last week Protana received the Frost and Sullivan 2005 award for strategic leadership, and one of the main points they pointed out was that we had refocused our business model moving away from drug discovery towards biomarker discovery. That was very satisfying.
What kind of goals does your company have for the future?
I think we want to continue to optimize our platform to improve our throughput, and that can involve developing new technology. Throughput is pretty important and we want to use cutting-edge technology to help us get there. And then the other stream is to begin focusing more on validation. That’s a huge challenge. And I think what we need to do is to begin some pre-validation processes using mass spectrometry. Once we’ve done that, that’s going to help the validation stage a little bit better.
I’m very optimistic about the application of proteomics to biomarkers, but I think it’s just the beginning. I believe that protein biomarkers will remain essential, though there are other types of biomarkers as well.
Are you looking in the future to take any of these biomarkers that your company has discovered and develop them into therapeutics?
There’s always that possibility. But that’s not a short-term goal. We really need to establish our business as a service for biomarker discovery.