Standardization of a protocol for shotgun proteomic analysis of saliva

Abstract Saliva contains numerous proteins and peptides, each of them carries a number of biological functions that are very important in maintaining the oral cavity health and also yields information about both local and systemic diseases. Currently, proteomic analysis is the basis for large-scale identification of these proteins and discovery of new biomarkers for distinct diseases. Objective This study compared methodologies to extract salivary proteins for proteomic analysis. Material and Methods Saliva samples were collected from 10 healthy volunteers. In the first test, the necessity for using an albumin and IgG depletion column was evaluated, employing pooled samples from the 10 volunteers. In the second test, the analysis of the pooled samples was compared with individual analysis of one sample. Salivary proteins were extracted and processed for analysis by LC-ESI-MS/MS. Results In the first test, we identified only 35 proteins using the albumin and IgG depletion column, while we identified 248 proteins without using the column. In the second test, the pooled sample identified 212 proteins, such as carbonic anhydrase 6, cystatin isoforms, histatins 1 and 3, lysozyme C, mucin 7, protein S100A8 and S100A9, and statherin, while individual analysis identified 239 proteins, among which are carbonic anhydrase 6, cystatin isoforms, histatin 1 and 3, lactotransferrin, lyzozyme C, mucin 7, protein S100A8 and S100A9, serotransferrin, and statherin. Conclusions The standardization of protocol for salivary proteomic analysis was satisfactory, since the identification detected typical salivary proteins, among others. The results indicate that using the column for depletion of albumin and IgG is not necessary and that performing individual analysis of saliva samples is possible.


Introduction
Saliva is a biological fluid composed of more than 99% water and less than 1% protein, electrolytes and other low-molecular-weight components. It originates mainly from three pairs of major salivary glands (parotid, submandibular and sublingual glands), as well as from 300 to 400 minor salivary glands present in the oral cavity. Saliva plays a key role in lubrication, chewing, swallowing and digestion. It protects the oral tissues and also provides biomarkers for local and systemic diseases 17 . Therefore, saliva contains more Human saliva is a biological fluid with enormous diagnostic potential. Because saliva can be noninvasively collected, it provides an attractive alternative for blood, serum or plasma 13 .
In the human saliva were identified 1166 proteins, and high portions of these proteins were found in serum. Currently, progress in salivary diagnostics has demonstrated that these contents can be very informative for detection of oral and systematic diseases 20 .
Proteomics, a new field of research centered on identification, quantitation, and characterization of proteins and their interplay, is largely based on the robustness, sensitivity, speed, and throughput of mass spectrometric procedures 6 . Currently, mass spectrometry is the basic technology for large-scale identification of these salivary proteins, and proteomic analysis of saliva has distinct advantages over blood, especially for proteins of low abundance 17,18 . One of the main challenges in proteomic analysis is the fact that highly abundant proteins can impair the identification of low-abundance proteins, considering the equipment dynamic range. In the case of saliva, albumin and immunoglobulin G (IgG), they are very abundant, and some authors have recommended using columns for depletion of these proteins during the extraction procedure 7,8 . Saliva functions are not only restricted to process food for digestion, considering that it contains a large number of proteins, which play important roles in the regulation of the immune defense and endocrine system and in the maintenance of mucosal tissue and dental health 1 .
Saliva may contain locally expressed proteins and other substances called biomarkers, which can be used as diseases' indicators, be closely related to an individual's health condition and change greatly when diseases occur. In general, most studies view saliva wrongly as a homogeneous body fluid. It is also not stable, but constantly in change, and its composition is affected among other things by sampling methodology, environment, periodicity, oral hygiene, psychological status and general health 6,13,20 .
Considering the importance of saliva in the oral cavity homeostasis, as well as its great potential as a diagnostic fluid, the aim of this study was to standardize a protocol to extract salivary proteins for further proteomic analysis. In the first test, we evaluated the need for using an albumin and IgG column to deplete these proteins during protein extraction. In the second test, we compared analysis of samples pooled from 10 volunteers with samples from individual analysis.

Material and methods
Ethical aspects and human subjects

Saliva collection
The volunteers were asked to rest for 15 min before collecting saliva, sitting upright. They were asked not to speak or eat before beginning to collect saliva.
First, they rinsed their mouths with 5 mL of drinking deionized water, then they were asked to swallow saliva for 5 min. After this period, the volunteers spit out all the saliva accumulated in the mouth in a plastic tube immersed in ice for 10 min (unstimulated flow).
The saliva samples were immediately centrifuged at 14,000 g for 15 min at 4°C to remove all debris, such as insoluble material, cell debris and food debris.
The supernatant from each sample was collected and frozen at -80°C until analysis. These procedures were based on previous studies 6,18 .

Preparation of the saliva samples
The experiments were performed into two phases.

Results
In the first test, when the albumin and IgG depletion column was used, the total amount of protein recovered from the pooled samples after extraction was 8 μg, while only 35 salivary proteins were identified. Among them are proteins typically found in saliva, such as alpha-amylase 1 and 2B, cystatin isoforms, hemoglobin isoforms and mucin 7, among others (Table 1). When the depletion column was not used, the amount of protein recovered was much higher (48.0 μg) and 248 proteins were identified, among them many typical components of saliva such as alpha-amylase 1 and 2B, many cystatin isoforms, carbonic anhydrase 6, lactotransferrin, lysozyme C, mucin 7, proline-rich protein 4, protein S100A9, serotransferrin, statherin, several hemoglobin isoforms, among others (  was not used. For the pooled sample, the amount of protein recovered after extraction was 54.02 µg, which allowed the identification of 212 proteins, including alpha-amylase 1 and 2B, carbonic anhydrase 6, cystatin isoforms (B, C, D, S, SA, SN), histatin 1 and 3, lysozyme C, mucin 7, protein S100A8 and S100A9, statherin, several hemoglobin isoforms, among others (Table 3). In the analysis of the individual sample, 25.13 μg of total protein were obtained and 239 proteins were identified, among which are alphaamylase 1 and 2B, alpha-enolase, carbonic anhydrase 6, many cystatin isoforms (B, C-D, S, SA, SN), histatin 1 and 3, Ig alpha-2 chain C region, Ig a chain C region, lactotransferrin, lysozyme C, mucin 7, protein S1008 and S100A9, serotransferrin, statherin, among other proteins (Table 4).

Discussion
This study aimed at standardizing a protocol for proteomic analysis of saliva that is sensitive, easy to perform and of low cost, to be used in  (Table 1). This figure increased to 248 when the column was not used ( Table 2). We believe this occurred because, when using the albumin and IgG depletion column, there was also depletion of other proteins, since using the column increases one more process in the methodology. We also believe that many proteins could bind to albumin and IgGs, thus being depleted together. Among the identified proteins, in both situations, are those typically found in saliva.
By contrast, when the depletion column was used, classical salivary proteins such as 14-3-3 proteins, histatins, statherin, lactoperoxidase, lactotransferrin, lysozyme C, neutrophil defensins, protein S100A9, serotransferrin and some cystatin isoforms were not identified. Thus, contrary to what was observed in gel-based proteomics 7 , in shotgun proteomics the use of albumin and IgG depletion column impaired protein identification according to our workflow. Some studies, in spite of that, report advantages in using depletion columns when more than one workflow is In the second test, we compared analysis of pooled samples (from ten individuals) versus individual analysis, without using the depletion column. In the individual analysis 239 proteins were identified (Table 4), while 212 proteins were identified in the pooled sample (Table 3). One-hundred and twenty three proteins were common to both groups (data not shown), and among them are most of the proteins typically found in saliva. The proteins exclusively found in the individual sample or in the pooled sample are not typically reported in saliva, which might be related to   However, this technique is considered bounded by the collision energy required in protein fragmentation, which is insufficient for proteins greater than 50 KDa, and its application is restricted to the analysis of purified proteins 4,9,11 . In addition, top-down proteomics method requires the use of one or more forms of separation prior to mass spectrometry analysis 12 .   In this way, shotgun proteomics covers the largest variety of detectable components, regardless of their mass, due to the proteolytic digestion of large proteins almost always generates peptides that can disclose the presence of the protein in a complex mixture. Due to these reasons, the number of salivary components currently detectable by shotgun proteomics approaches is more than five times greater than that of components detected by any other platform 2,10 . Thus, in this study we employed shotgun proteomics.

Moreover
Based on the results of the two tests, the protocol for salivary shotgun proteomic analysis was satisfactory, since it allowed the identification of many proteins, including those typically found in saliva.
Moreover, it is easy to perform and cheaper than the methods previously described, since it does not require the use of depletion columns. Furthermore, it allows individual analysis of the samples, which is very important in quantitative proteomics. Thus, this protocol could be used in future studies involving shotgun proteomic analysis of saliva.