PMC

Molecular mass and pI distribution of the entire human salivary proteome. (A) The distribution was skewed to relatively low-molecular-mass components. (B) In contrast, the salivary proteins had a relatively broad range of pIs.

Human parotid and SM/SL salivas share many of the same protein components. The protein composition of each ductal fluid is presented as a Venn diagram. Fifty-seven percent of proteins were found in both parotid and submandibular/sublingual (SM/SL) salivas, whereas 27% were unique to the gland(s) of origin.

Human parotid and SM/SL gland secretions share many protein components with human tears and plasma. (A) Areas of overlap depict the number of proteins that were identified in both tears and parotid (top) or SM/SL saliva (bottom). (B) Similarly, parotid (top) and SM/SL gland secretions (bottom) contained a substantial number of proteins that are also found in plasma.

Venn diagrams illustrating the number of parotid (left) and SM/SL (right) proteins that were identified at each research site. Areas of overlap depict the number of proteins that were contributed by more than one group, with the core proteome—proteins that were identified by all three groups—shown in the center. Areas where the circles diverge show the number of protein identifications that were site-specific. Differences are likely attributable to the different methodologies employed by each research team.

Array view of the minimally overlapping human salivary proteome. Data were grouped according to the research team that identified each protein, the salivary gland where it was produced, and percent sequence coverage. The extent of coverage positively correlated with the number of research groups that identified a particular protein. The core proteome included one SM/SL-specific protein. Blue, protein identified; yellow, protein not identified; red, high-sequence coverage; yellow, intermediate-sequence coverage; and green, low-sequence coverage.

Relative allocation of the proteins identified in parotid and SM/SL saliva according to their gene ontology annotations. Components of both fluids were similarly distributed with regard to cellular locations (A), molecular functions (B), and biological processes (C). A high percentage of salivary proteins are predicted to be extracellular components or to reside within organelles (A). Salivary proteins were commonly involved in binding and catalysis (B). Parotid and SM/SL proteins had the highest distribution in metabolic and regulatory pathways (C).

Sequence coverage and cellular origin of salivary proteins identified by single or multiple groups. (A) Histogram of the sequence coverage for proteins that were identified in parotid and SM/SL salivas. The shaded areas depict the number of proteins that were reported by one (gray), two (dark gray), or all three (black) research groups. (B) Comparison of the cellular location of salivary proteins identified by one, two, or all three research groups with a reference data set containing all human IPI database entries. Ratios of protein associations with the extracellular, plasma membrane, cytoplasm, or nuclear subcellular compartments were calculated using Ingenuity software. In some cases, this information was not known.

Detection of novel salivary proteins and those with low sequence coverage. An immunoblot approach was used to confirm the presence in saliva of a portion of the proteins that were identified by using MS approaches. Most of the analyses shown employed parotid and SM/SL saliva samples collected as the ductal secretions from 5 females (lanes 1–5) and 4 males (lanes 6–9). As shown in panel F, whole saliva samples were collected from 10 females (lanes 1, 4–12) and 2 males (lanes 2–3). Proteins were selected for validation based on the number of research groups that reported the identification and the extent of sequence coverage. In general, emphasis was placed on the analysis of novel components and proteins with the least MS evidence of their presence in ductal saliva. (A) Samples were separated on 4–12% BIS-TRIS gradient gels under reducing conditions, and proteins were visualized by staining with Coomassie brilliant blue. Immunoblots detected the following: (B) CEA (IPI00027486, 6% sequence coverage, 1 group reporting), (C) kallikrein-1 (IPI00304808, 89% sequence coverage, 2 groups reporting), (D) apoE (IPI00021842, 21% sequence coverage, 1 group reporting), (E) TIMP-1 (SM/SL only, IPI00032292, 67% sequence coverage, 2 groups reporting), (F) HLA-G alpha chain, a nonclassical class I histocompatibility antigen (whole saliva, IPI00015988; a single peptide was sequenced by 1 research group), (G) prostatic acid phosphatase (IPI00396434, 8% sequence coverage, 1 group reporting), and (H) calpain-1 catalytic subunit (IPI00011285, 15% sequence coverage, 1 group reporting). E, apoE; AII-E, a complex of apoAII and apoE; AII-E-AII, AII-E with an additional apoAII; E-β-amyloid, a complex of apoE and β-amyloid.