November 2007
Three years ago, scientists supported by the National Institute of Dental and Craniofacial Research (NIDCR), part of the National Institutes of Health, began taking the first full inventory of the proteins that normally are produced in our salivary glands. Now, one of those scientists and his colleagues offer a first glimpse into how this new research tool can be applied to detect subtle changes in the protein content of a person’s saliva that may be linked to an oral or systemic disease. As reported in the November issue of the journal Arthritis and Rheumatism, the scientists detected 42 proteins and 16 peptides in saliva that clinically discriminated between people with the primary form of Sjőgren’s syndrome and healthy volunteers. These data far surpass previous efforts to identify protein biomarkers for primary Sjőgren’s syndrome, a chronic autoimmune condition of the salivary and tear glands that affects about two million Americans, mainly women. The scientists also identified 27 distinct gene transcripts that were over produced in the saliva of those with primary Sjogren’s syndrome. The Inside Scoop spoke with the study’s senior author Dr. David Wong, a scientist at the University of California at Los Angeles School of Dentistry. He offered his thoughts on the paper’s findings, its implications, and the recent research progress with salivary diagnostics.
Some people have asked why develop saliva as a diagnostic fluid? Isn't blood good enough?
One part of the answer is saliva can be readily collected. That’s a big plus. As most dentists and physicians will tell you, they treat patients everyday who wince at a needle poking their skin or mouth. Saliva also is easier than blood to store and, if needed, to ship. But there’s a second part to this answer that really is worth emphasizing.
What's that?
The salivary glands do not exist in isolation from the rest of the body. In other words, they are not just a series of glands that secrete fluid into the mouth. The salivary glands are connected to the rest of the body through blood vessels that nourish them and neural networks that innervate them. So, they contain information about health and disease and, with its ease of collection, saliva could serve as an ideal diagnostic fluid. That’s been the vision that has moved the concept forward through the years.
That movement was relatively slow during the 20th century. Where did things stand with salivary diagnostics just five years ago?
Well, salivary diagnostics was fairly well defined in terms of its physiology and biochemistry. The scientific gaps arose in defining its clinical utility. As I mentioned, we didn’t know whether or not saliva could become a mainstream diagnostic fluid. It’s still not a mainstream diagnostic fluid, but things have changed in a very good way over the last few years.
How so?
A real turning point was NIDCR’s support of a consortium of laboratories – which included my lab at UCLA - to inventory for the first time the proteins produced in the major salivary glands. Here at UCLA, we also assembled independently a companion inventory of over 3,000 gene transcripts produced in the saliva. With these two toolboxes – I call them diagnostic alphabets - we can begin to evaluate more methodically and systematically the potential of saliva as a diagnostic fluid for oral and systemic diseases.
You can cast a wide net.
Exactly. Before the arrival of these toolboxes, we basically were shooting in the dark. If we heard a research group had found a promising biomarker in blood or urine for a given disease, we looked in saliva and see if it was there, too. We were playing a “me-too” game. What was lacking was a solid scientific foundation to develop the full diagnostic potential of saliva. With these toolboxes now in hand, the game has changed. We can look de novo for signs of a given disease. In other words, we know healthy people have on average 1,166 distinct proteins in their saliva. That gives us 1,166 places to look for measurable changes in protein expression that might be indicative of disease.
And you can do so by comparing the molecular content of saliva from a healthy person with that of an individual known to have Sjogren's or another disease? Is this what you and your colleagues did in this paper?
That’s exactly right. In our pilot study, we collaborated with Dr. Arjan Vissink and his group in the Netherlands. Arjan has studied Sjőgren’s for many years, and he provided saliva samples from10 people diagnosed with primary Sjőgren’s. We compared the samples with those from 10 healthy volunteers. We cast our investigative net, measured the protein content in both groups, and pulled out 42 proteins and 16 peptides that were either over or under expressed in the Sjőgren’s group. So, these data really highlight the maturation of the salivary proteome as a diagnostic tool.
What about the other toolbox?
We opened the transcriptome toolbox and profiled the gene transcripts in the saliva. Using very stringent criteria, we found 27 gene transcripts that were significantly upregulated in the Sjőgren’s group. What’s really striking here is 19 of the 27 genes are involved various aspects of the inflammatory process.
So that tells you that you're in the right ballpark? Sjogren's syndrome is an autoimmune disease, and the inflammatory process is reflected in the saliva.
Correct. This suggests to me that saliva contains the constituents to reconstruct the disease process. It will be exciting to explore this possibility further and tease out more information of potential diagnostic value. Right now, a person faces a battery of expensive and time consuming tests to get a diagnosis of primary Sjőgren’s syndrome. Salivary diagnostics could make it simple, quick, and painless.
What is the sensitivity and specificity of the markers that you found? How tightly do they correlate with the disease?
That hasn’t been determined yet. Nor do we know which of these molecules might be associated with early disease and thus might be important for early detection. Keep in mind, our findings are only research leads at this point. Each protein and gene transcript must be validated in at least two follow up studies to be considered as a marker of disease.
Do you plan to follow up these leads in the near future?
Absolutely. We are in the midst of planning a larger validation study right now.
What do you have in mind?
We’ll need to have at least 50 people with primary Sjőgren’s and 50 healthy volunteers. That’s the general design. But we’ll need to do these independent validation studies at least twice. You do the first study to validate the markers and then, based on its findings, you build a prediction model. In the second validation study, you independently test the reliability of the prediction model. If the outcome is good, you can begin to envision an algorithm, possibly combine the most informative molecules into a panel of diagnostic markers, and proceed to clinical testing.
You've already ploughed through this validation process in your work with salivary markers for oral cancer.
That’s right. In December 2004, my lab published a paper that reported on four gene transcripts that were elevated in the saliva of patients with oral squamous cell carcinoma. Since then, my lab has worked extremely hard to take these markers through the validation process. I’m pleased to say that a few weeks ago, these salivary oral cancer RNA biomarkers were validated as markers for oral squamous cell carcinoma by the Early Detection Research Network, or EDRN. The EDRN is an initiative supported by the National Cancer Institute that develops and validates markers for the early detection of cancer.
Where do these gene transcripts come from?
They come from almost every conceivable source. The transcripts are in the ductal fluid secreted into the mouth as saliva. They’re obviously in the acinar and ductal cells of the salivary glands, as well as in gingival crevicular fluid. That leaves open the issue of transcripts reaching the salivary glands from a distant disease site, and we’re studying that question right now using animal models. Hopefully, these studies will fill in some of the blanks. But, as mentioned earlier, the larger point is you cannot draw a line between the mouth and the rest of the body. If a part of the body is unhealthy, it will be reflected elsewhere. The real blessing here is that this information seems to be reflected in a non-invasive fluid that we can readily collect in the mouth.
In the Sjogren's paper, you evaluated saliva collected from the parotid gland, oneof the major salivary glands, and whole saliva. Whole saliva is the total secretion from all of the major and minor salivary glands. You found that whole saliva had more diagnostic value. Was this surprising and why?
It was surprising. If you had asked me to beforehand, I would have said the parotid fluid would be more informative. That’s because most of the oral pathology that we are aware of with Sjőgren’s occurs in the parotid gland. It would seem like a logical place to look for signs of disease, particularly because whole saliva contains other glandular constituents as well as gingival crevicular fluid. So, it was an unexpected but welcome outcome.
Why "welcome'?
Let’s say a rheumatologist wants to take a saliva sample. He or she won’t need to ask a patient to sit down for an hour to have a parotid gland drained. The doctor can collect a whole saliva sample in a matter of seconds.
It sounds like this paper has ramped up your work load?
Well, it’s a good position to be in. I might add that about 5 percent of those with Sjőgren’s syndrome develop malignant lymphoma, and we’re looking into that, too. Now that we have these toolboxes, I think the field will move much faster to crack saliva’s diagnostic codes. But we still need that one big research victory. Imagine the day when there’s a news headline that saliva can be used for the early detection of a systemic disease. If the science is good and credible, you can almost begin to see how the table will turn. Now we knock on doors for help. The day that happens, they will be knocking our doors.