NEW YORK (GenomeWeb) – It's been several years since GenomeWeb last checked in with researchers in the niche field of tear fluid proteomics.
When looking for people to talk to, one name kept coming up: Roger Beuerman, a researcher at the Singapore Eye Institute.
"He has done the most work on tear proteomics of anyone in the world," Richard Semba, an ophthalmologist at Johns Hopkins University who studies the eye proteome, told GenomeWeb.
Five years ago, GenomeWeb reported on a Beuerman-led study to validate protein biomarkers for dry eye syndrome. It's the disease that provided him entry to the world of the tear proteome, and now he's leading a growing number of ophthalmologists studying inflammation in the eye and diseases like age-related macular degeneration.
In addition to major research and clinical studies studies, Beuerman holds patents on biomarkers and consults with ophthalmology groups around the world. He's now focused on getting the technology into doctors' offices, putting his research to clinical use.
Beuerman spoke last week with GenomeWeb from Singapore. Below is an edited transcript of that interview.
Why look at tears in the first place?
Tears are not just some watery exudate. They're an extracellular fluid. If you told someone doing proteomics of breast cancer, "Here, I can show how you can get pure extracellular fluid from around a tumor," they'd be very happy, because that contains signaling molecules and molecules being expressed by cancerous cells and immune cells next to that area. That's the same thing we're finding in the eye, that the tears tell us an awful lot about the disease process.
Proteomics is an attractive area, particularly for eye diseases, as it turns out. One of the reasons is you don't have to jab needles into people and draw blood.
The blood proteome picks up signals from wherever. Tears are a dilute solution and don't have the same amount of protein as blood, but at the same time, it's close to the disease. As an extracellular fluid, it's taking its parts from a lot of sources. We can see changes associated with the retina, the lacrimal gland, and we've now begun to use the tears to look at systemic disease as well. We can even see changes associated with diabetes.
What was the catalyst in launching the push for diagnostics in this field?
In 2009 we published an article in the Journal of Proteome Research where we found a bunch of potential biomarkers for dry eye in humans. That was one of the momentous events that developed more interest in the field.
A lot of the field has revolved around the condition [of dry eye] and it has become recognized as one of those problems that really haunts the elderly and women and is something that tends to make your life miserable. It makes your vision unstable as you get older and it hurts and scratches. The diagnostic methods are not particularly good and there needed to be quantifiable diagnostic methods that can be very easily understood and communicated between clinics around the world.
What are some other diseases that tear proteomics could help with?
Retinal disease like age-related macular degeneration (AMD) is one. We're turning out now a broader clinical trial in AMD. If it works out, we'll be able to set up a screening tent in a shopping center, take tears from people, look at them for biomarkers, and, if need be, tell people, "You should go see a doctor about this."
Keratoconus — when the cornea becomes misshapen and forms a triangular shape — had some of the same problems as dry eye. There were a lot of tests, but different clinics disagreed with outcomes and there were absolutely no quantifiable results where you could say, "This is a stage three patient and here are the results that show it's definitely stage three."
With diabetes there are terrible side complications in the eye. One of the biggest problems worldwide now is the sight-threatening aspects of diabetes, diabetic retinopathy. If not for tear proteomics, you'd have to carry out advanced studies to look inside the retina. If you can obtain information from something as accessible as tear fluid, you can easily do screening.
What have been the biggest changes driving your work?
One of the reasons the field started expanding was not because we were all that much smarter, but because the instruments became better than they had been in the past. With the instruments used to publish the 2009 paper, we could find 300-400 proteins. But the method required so much protein, that you couldn't do it on an individual, per-patient basis. You had to collect either across a group of patients or collect tears from one person in a serial fashion.
With the advent of the new mass spectrometers starting in the mid-2000s, you could begin to look at tear proteins and do the proteome from a few microliters of tears.
Now we can now do a proteomic analysis on 1 microliter of tears and publish papers on a tear proteome of 1,500 proteins. That's changed things a lot.
We're also able to carry out bioinformatics on samples. That was something new, as the analysis is different from genomics. People were aware of genomics, but proteomics at that time was a little scary. At that point we were talking about having maybe 40,000 total human genes and maybe 200,000 to 300,000 proteins.
That's come down to 22,000 genes and maybe 40,000 proteins. The available proteins are pretty tractable and it's not as scary as people once thought. That kind of set the stage to do something serious with proteomics.
What unique challenges are there in sample collection and preparation and analysis?
Sample collection isn't so bad. Preparation is tedious.
We do it a number of different ways, often using fire-polished microliter capillaries. In some situations you use 1 microliter tubes, other times 5 to 10 microliter tubes. You put these along the tear meniscus, and by capillary action you allow the tears to run up into the tube. Then we dilute them five to 10 times. Because of the mass spectrometry technology, we like to run a third or a half of the study samples all at once just to maintain conditions. Everything is very reproducible, but it's the optimal way to do it. We run the sample through a high-pressure nano-liquid chromatography system.
You can run through 30 samples in a week without too much trouble, sometimes you can do more.
What kind of mass spec instrumentation have you been using?
We've been using the Sciex mass spectrometers for some time now. The TripleTOF 5600 is very good for discovery, and we've been using that quite a bite over the years now. We can also do clinical trial samples with that, which is an advantage.
Another very nice mass spec is the Thermo Fisher Scientific Orbitrap and that also works extremely well. Recently that has been modified so it is more useful for clinical trials.
When doing a clinical trial that you know what you are looking for we use multiple reaction monitoring. There you've limited the number of proteins you're looking at to between five and 50. You're looking at just those, not the whole field of the proteome. That's very accurate. The Sciex machine is good for that.
What's the next big thing that will happen in this field?
Right now, mass spec is sitting right on the threshold where it can duplicate the sensitivity of western blot. That's something everybody would like to do. Because of the limits of sensitivity, we do not detect some growth factors well and we don't do cytokines well. We can do some cytokines, but not all of them. If we can bring the sensitivity down to the level needed for cytokines and such so it's comparable to the sensitivity of western blots, that will be the next huge leap because then we'll be able to quantify a huge array of proteins.
What are you working on right now?
With AMD we have found some biomarkers that look like they really can tell us about the staging of the disease. We think that could be useful for large-scale screening. A lot of people are walking around and don't know they have the beginning of this disease. It progresses slowly, and once they realize they're overtaken by symptoms, it's really hard to treat.
We're even looking at contact lens wearers. Some people wear contacts well and some have trouble. It's long been thought that something in the tear fluid causes that. Before you start to wear contacts, we could take a tear sample and say whether you're likely to have trouble wearing contacts.
When and how will tear fluid proteomics be put into routine clinical practice?
We've published a number of papers that have made [clinical diagnostics] a lot more tractable. Medicine requires a lot of validation and we have just now validated [tear fluid proteomics] over and over. To get it out into practice, you have to bring it down to some device that a nurse can use.
There's one company, Advanced Tear Diagnostics, that's beginning to use a couple of biomarkers and we're working with them to add a lot more, so you can do quantitative analysis in the clinic. I'm in the process of licensing biomarkers to them; we have certain understandings and certain types of papers that have been signed.
We'll refine this and say, "Proteomics biomarkers X, Y, and Z will tell us about this condition. We can measure those." One big challenge is to boil this down to a device that can fit in a doctor's office. We're working on that as well.
If everything goes well, this will be a device that can be run easily in a small or large clinic, which can be used to rapidly quantify anywhere between four and eight biomarkers in the tears. We could potentially have different panels, depending on the type of disease.
Point-of-care tear fluid proteomics. Will that be antibody based?
It's a little different. It will allow us to quantify things down with a sensitivity that will allow us to look at growth factors and cytokines. It's probably better not say any more.