A team of scientists from five US universities has completed what is being called the most comprehensive map of the human salivary proteome to date, a development that is expected to push along the long-term goal of creating new saliva-based diagnostic tests for a wide array of diseases.
In an article published in the March 25 online edition of the Journal of Proteome Research, the researchers reported that they identified 1,166 proteins in saliva collected from the parotid and submandibular/sublingual glands, including a number that are found in plasma, data that could add to the likelihood that saliva-based diagnostics could eventually replace blood-based tests for some diseases.
The proteins found are only for those from salivary gland secretions, or ductal saliva, not from whole saliva, which comprises proteins from bacterial waste products and dead cells from mucous membranes, among other things. But because the parotid and submandibular/sublingual glands produce 99 percent of all saliva, the proteins identified make up essentially a first draft of the whole salivary proteome, said David Wong, associate dean of research at the University of California, Los Angeles School of Dentistry, and an author on the JPR article.
“It sets the stage now to refine it: what are the post-translational modifications, where do these proteins come from, and, most importantly, how can we use them in a clinical sense?” he said of the research, which is the culmination of a four-year, $12 million project funded by the National Institute of Dental and Craniofacial Research.
The schools that participated in the project were the Scripps Research Institute and University of Rochester, which formed one team; the University of California, Los Angeles, and University of Southern California, which formed another group; and the University of California, San Francisco.
Because ductal saliva is believed to contain a high percentage of proteins found in blood, the hope is that whole saliva will someday become a non-invasive replacement for blood as a fluid for diagnostic tests, Wong said. While some saliva-based diagnostics are already in common use, such as HIV tests, blood still is viewed as the most sensitive and specific fluid for detecting most diseases.
The JPR study, however, “serves as an important foundation to use saliva as a clinical fluid,” Wong said. “Now we have … a diagnostic alphabet in saliva that we can now use systematically and compare healthy people’s saliva with people who have diseases.
“If we compare enough people, we can begin to generate a fingerprint, a diagnostic signature based on saliva, which can be easily obtained [because] it’s non-invasive, non-painful,” he added.
Building on their research, the authors of the JPR study as well as individuals involved in the Human Proteome Organization’s Plasma Proteome Project are preparing an article for submission that compares the salivary proteome with the blood proteome, he said.
Up to now, work on saliva-based protein biomarkers has been scattered, resulting in mixed findings, Wong said. Even now, saliva is “not scientifically respected” because it is open to the environment, which can muddy scientific results. Additionally, he said, saliva has negative cultural and social connotations, which has hindered its use as a fluid for research.
But interest in saliva as a research fluid is growing. Along with mapping the saliva proteome, a nine-year project to develop saliva-based biosensor technologies, such as lab-on-a-chip, to help detect diseases in their early stages, is being funded by the NIDCR, said Eleni Kousvelari, associate director for biotechnology and innovation at the agency.
The study “serves as an important foundation to use saliva as a clinical fluid. Now we have … a diagnostic alphabet in saliva that we can use systematically and compare healthy people’s saliva with people who have diseases.”
The first phase of the technology-focused project is finished, she said, and the work described in the JPR article “is very important for the second phase” of the project, she added.
Of the 1,166 proteins identified by Wong and his colleagues, 914 were produced by the parotid gland and 917 by the submandibular/sublingual gland with roughly 700 proteins produced by both glands.
While other groups, including some authors of the JPR article, have worked at creating a salivary proteome, the NIH effort is the largest such effort. Tim Griffin, a professor of biochemistry at the University of Minnesota, who has been studying salivary proteins that may be implicated in oral cancer, called the study “the most comprehensive, in-depth, and systematic” map of the salivary proteome he has seen. He did not belong to any of the teams that did the research.
“It’ll be interesting to see … how whole saliva compares to these purified gland secretions because whole saliva brings in other possible things like nasal secretions and different things that maybe they don’t have in this data set at this point,” Griffin said. “I think the trick for all of this — whether it’s whole saliva or the purified gland [secretions] — is that you’re going to get this baseline protein map, but … to use this for disease diagnosis, it’s more of [the] quantitative side of things — what changes in these saliva samples under certain diseases?”
The authors of the JPR article write that in addition to allowing for the evaluation of candidate biomarkers for oral diseases such as oral cancer and Streptococcus mutans titers, the creation of a comprehensive catalog of proteins found in parotid and submandibular/sublingual salivary glands can prove useful in the evaluation and monitoring of systemic conditions such as ovarian and breast cancers.
“Our data suggest that a number of other markers of systemic disease may also be present in saliva,” they say. “A recent example is the discovery of a panel of IFN-regulated proteomic and genomic biomarkers in whole saliva that are highly discriminatory of patients with primary Sjögren’s syndrome.”
Blood, Spit, and Tears
In total, Wong and his colleagues looked at the ductal saliva from 23 adult volunteers. A number of different separation and fractionation technologies were used by the three groups, including reverse-phase high pressure liquid chromatography, metal chelate affinity chromatography on Talon resin, 2D SDS-PAGE, ZOOM IEF, and strong cation-exchange chromatography on a polysulfethyl A column.
Mass spectrometry was done on ThermoFinnigan’s, now Thermo Fisher Scientific’s, LCQ Deca, LCQ DecaXP, and LTQ 2D ion trap instruments; the Thermo LTQ Orbitrap; Applied Biosystem’s QStar Pulsar XL with a nanoelectrospray interface; and ABI’s 4800 LC-MALDI TOF/TOF.
The three groups each used the European Bioinformatics Institute’s Human International Protein Index “to derive peptide and protein identifications,” although each group used a different version. For integration, the protein identifications for each group were standardized to IPI version 3.24.
The groups were able to find 173,811 peptide and 68,789 protein identifications, producing 11,592 distinct peptide sequences matching 2,153 protein accession numbers.
Integration resulted in 1,166 non-redundant clusters with each cluster containing at least one peptide distinct from other peptides used in other clusters. For each cluster containing isoforms, fragments, or homologs, one representative protein was chosen.
The UCSF researchers identified 197 parotid and 205 SM/SL proteins. The Scripps/Rochester team identified 855 parotid and 848 SM/SL proteins, and the UCLA/USC team identified 401 parotid and 309 SM/SL proteins. Proteins that were identified by all three teams totaled 152 from the parotid gland and 139 from the SM/SL gland.
The researchers also compared their findings with the plasma proteome and tear proteome. They downloaded a set of 889 high-confidence plasma protein identifications and 491 tear proteins. Those mapped them to IPI v.3.24, reducing the number of plasma proteins to 657 and tear proteins to 467.
Matching those results against their salivary proteome findings, they found 172 plasma proteins also in human saliva, “including the most abundant species, which is possible evidence of vascular leakage or the contribution of fluid from the interstitial compartment,” the authors say. They also found in human saliva 259 proteins produced by the lacrimal gland, which produces tears.
Among the proteins they found that were common to saliva, plasma, and tears were cystatins B, C, SA, S, and SN, zinc-alpha-glycoprotein, and prolactin-inducible protein.
The “brute force” strategy was a novel approach, said the University of Minnesota’s Griffin.
“It’s really kind of like throwing the kitchen sink at it in terms of the methods they used, and it makes sense,” he said. “I think that’s the novelty of it — the ability to have three different groups doing this in a large scale way, so that they do pretty much anything and everything they can to pull out as many proteins as possible.”
Also using immunoblots to confirm novel components and proteins that were identified by mass spectrometry, the authors saw “noticeable differences” in antigen reactivity among donors and saliva secretions “suggesting that the composition of the primary salivary proteome is gland-specific and may differ among individuals; formation of protein complexes is another source of sample-to-sample variation.”
They say further studies will be needed to explore the differences “in the context of oral health and disease, as it is likely that individual differences are biologically relevant.”
They call their study and its results “the initial edition of the secreted salivary protein catalogue,” and acknowledge that more sophisticated methods than were used for their work will be needed to target specific questions involving post-translational modifications, splice variants, protein quantification, and other issues.
A top-down approach, rather than the bottom-up methods they used, will be able to address polymorphic isoforms, the authors say.
“Although current top-down proteomic strategies are not amenable to high-throughput measurements and thus are used to answer highly focused questions, approaches such as top-down and de novo sequencing may become the methods of choice to address the complexity inherent in biological fluids such as human saliva,” they say.