At A Glance
Name: Paul Utz
Position: Assistant professor of medicine, division of rheumatology and immunology, Stanford University, since 1999
Background: Medical residency, postdoctoral fellowship, Brigham and Women’s Hospital, 1996; MD, Stanford University School of Medicine, 1991; BA, biology, King’s College, Wilkes-Barre, PA, 1986
How did you end up combining arrays with autoimmune diseases?
I was born in Pennsylvania and went to medical school here at Stanford. While I was an undergraduate I went to a little school called Kings College which is in Wilkes-Barre, Penn. While I was there I had a really good pre-med advisor and he suggested that I do research. I was interested in immunology, so I spent two summers doing research in Buffalo, NY, at Roswell Park. I had a great experience there and ended up going to medical school at Stanford. The main reason I chose Stanford is that they’re so strong in immunology. Arguably, they have the best immunology program in the world. So while I was at Stanford, one of my med school roommates was working in Pat Brown’s lab, so I spent a lot of time in Pat Brown’s lab as a medical student at the time that Pat was developing the DNA microarrays. And I had talked with him back then about wouldn’t it be great if you could do this for proteins — couldn’t you spot antibodies or use it for protein-protein interaction?
And then I went on and did a residency at Boston at Brigham and Women’s Hopsital. I did internal medicine and rheumatology there. And I again chose the Brigham because they have such a strong immunology program. And the protein array stuff— you completely forget about that stuff during residency because it’s all patient care-related stuff. So I was actually working in a totally different field for four years, which is RNA splicing.
I met Pat again when he came out to give a talk at Brigham, just before I took the job at Stanford. I talked with him and said it would be great to do something with arrays. Then I took the job at Stanford in 1999. I got to Stanford and realized there was a tremendous opportunity to change direction in my research. I was much more interested in doing what’s called translational research, which is to do something that lets you go from bench to bedside.
Around that time — it was December of 1999 — Bill Robinson walked into my office. Bill is one of the rheum fellows. He said well, I’ve got this idea of doing protein microarrays and I’ve actually got a lot of preliminary data on it. I’ve been working on it for about a year and half with Brian Haab. He said would you be interesting in getting involved, and also are you interested in starting a company? So this sort of thing doesn’t happen at many places where a fellow walks in and says I’ve got a great idea, do you want to totally change the direction of your lab and would you like to start a company?
So I suggested that they start working on the lupus arrays because of the arrays that you would do, they would actually be the easiest because the antibodies of that particular disease are very high titer and very specific, and I have tons of reagents in my lab to do these studies. I’ve been collecting sera samples for five years, I have about 1,000 serum samples. So we set about doing that and that ended up being the first paper that we published, which was the Robinson paper in Nature Medicine in 2002. That’s pretty much how things got started. I think Stanford was the perfect place to do it because Stanford is a very technology-driven place and there are tons of really smart engineering people there with a lot of good ideas, and they’re just looking for people like me who are trained in medicine to be able to come up with good applications.
Had you started working on protein arrays or DNA arrays before you got the proposal to start a company?
No, I had not done it. Other than the original discussions with Pat back in 1991, and then thinking about it a little during my fellowship years, I hadn’t really thought about [protein arrays].
I think that the protein array stuff, Bill Robinson had been thinking about it since 1997. Really all I provided to him was give him the ideas of what the best disease would be to study. Bill is very interested in rheumatoid arthritis and multiple sclerosis, and the problem with those diseases is in RA there are only a few antigens known, and most of those were being discovered around the time he started his work. For MS, there are many antigens that have been hypothesized, but patients for the most part don’t make antibodies, so you can’t really study that disease.
I think a big reason I got into it was I had a huge interest in lupus. Bill had been developing the protein microarrays and we just thought, ‘This will work out really well.’
What are the advantages of using protein arrays to study autoimmune diseases such as lupus?
For lupus, there are probably 50 to 100 antigens that have shown to be targeted in that disease, and the linear epitope that is targeted is not known. The beauty of studying lupus with arrays is you can look at many antigens at one time, and many linear epitopes. The other thing is that in lupus and other connective tissue diseases like scleraderma and myocitis, no two patients are alike. Every one is different. We don’t know why that’s the case. We hypothesized that there are probably differences in their serum proteome, either in their antibodies or circulating cytokines, or some other circulating factor that might allow us to look at their serum and say, ‘OK, you have this bad form of lupus and you’re going to die really soon and we need to do something about it.’ Or, ‘You’re going to do just fine, and I don’t want to give you this toxic medication when you don’t need it.’
The other important aspect for the arrays is that for the DNA vaccine technology, there the critical thing that you have to know is what the antigen is to encode it into the plasmid. We had come up with the idea that you could use the array to identify the antigen that a patient was reacting against — figure out what the antigen was — then clone up the antigen, make a DNA plasmid encoding that antigen, and now you’ve got a therapeutic. So you could go from discovery to having a therapeutic in a month. And the other beauty is you could give patient-specific therapy, which we’re decades away from. But ultimately, you could say Patient A is responding to this, and we’re going to give that person A, and patient B is responding to this so we’re going to give that person B. So you would give them individualized antigens.
As it is now, if someone is diagnosed with lupus, would you know what antigens he or she is reacting to?
We could figure out some of them by using standard ELISAs. There’s a standard ELISA for about 10 to 15 different antigens that are main targets in systemic, rheumatic diseases. But we don’t have antigen-specific therapy developed, so patients are all getting very non-specific and toxic therapies. Things like high doses of steroids and chemotherapeutic agents.
How well does the DNA vaccine technology work?
In animal models, it works very well for a model of multiple sclerosis, called experimental autoimmune encephalitis. It also works in a model of diabetes called the non-obese diabetic mouse. And it works in a model of arthritis called the collagen-induced arthritis model. So we know that it works in at least three different diseases, and we’re testing it in lupus. But the lupus models are very long — between six months and two years, so it takes longer to get the results there. And we’ve taken it in to the clinic specifically for multiple sclerosis first. It’s already in the Phase I/Phase II trial with ten patients who have been enrolled so far. And the next disease target is juvenile diabetes. We’re getting all the materials together to file an IND to start treating patients.
Where are the arrays being manufactured?
Right now the company, called Bayhill Therapeutics, has licensed the array technology from Stanford and we have a collaborative agreement between Bill’s lab and my lab and Bayhill. So they’re manufactured at Stanford and then provided to Bayhill for studies. Then they Bayhill scientists probe the arrays and scan and analyze them. Bill Robinson and I are involved in helping them with the data analysis, etc.
The people in my lab are pretty jazzed by the stuff I do with the company. They also realize that if they make a big discovery and they think it’s something that can be taken to the clinic, that we’ll work with the office of patent and licensing to take it into the clinic. So I think they feel that the work they do actually means something and it’s not just playing around with mice or fruit flies. That’s really the whole goal of the work that we’re doing is to translate the discoveries into something that would be useful in the clinic.
My goal in my career would be to make a discovery at the bench, in general starting with material from the patients, and then ultimately develop something by industry that could go back to the clinic. So my dream would be to make a discovery at the bench, or some contribution to a discovery, and then see it developed and later be able to prescribe it in the clinic. It’s pretty frustrating with these autoimmune diseases — I happen to be on service this month — and I have some patients both in my clinic and in the hospital who have autoimmune diseases for which we have absolutely no treatment. Nothing will work. And it’s very frustrating and it’s be nice to develop something that could actually make a difference for these patients.
Would you say you’ve been successful so far, for example with the vaccine technology, in bringing discoveries from the bench to the clinic?
I think so. We always joke about shots on goal. So we look at this [vaccine] as our first shot on goal. It might go in and be a goal, or it might get bounced back by the goalie — you never know. I think if this succeeds for multiple sclerosis, it’s going to completely change the way we take care of all autoimmune patients. There are 90 autoimmune diseases, and of these 90 autoimmune diseases, half of them have no good treatment. If we could develop a good treatment for those patients and use this platform, it would be huge. So yes, I would say if this works, it’s more than a shot on goal, it’s more like winning the Superbowl or the Stanley cup.