The major components of human submandibular-sublingual saliva (HSMSL) are mucins, amylases, cystatins, proline-rich proteins and statherin. Structure-function studies of these molecules have been hampered by the small amounts of purified materials that can be isolated from human secretions. The present study describes an integrated purification protocol for the large-scale preparation of many of these molecules. To dissociate partially heterotypic complexes among salivary molecules, HSMSL was initially fractionated into four pools by gel filtration with 6 M-guanidine hydrochloride. Subsequent fractionation of these four pools by gel-filtration and ion-exchange chromatography resulted in the purification of high- and low-Mr mucins, neutral and acidic cystatins, acidic and basic proline-rich proteins and statherin. Many variants or isoforms of these salivary molecules have been identified and biochemically characterized. Biochemical studies indicated that the low-Mr mucin exists as two isoforms which vary in their sialic acid to fucose ratios. Three isoforms of acidic cystatin S were characterized which differ in their phosphate content. Two isoforms of a basic proline-rich peptide were identified; the smaller peptide was a truncated form missing the first seven amino acids.

Saliva has proven to be a discriminating element in forensic arenas, an effective indicator of acute diseases of salivary glands, and a promising probe for drug monitoring. With the advent of sensitive immunochemical assays, the compositional profile of human salivary secretions has been expanded considerably. Thus, the establishment of a range of "normal values" for a variety of "intrinsic" and "extrinsic" salivary components represented the initial step to use saliva as a diagnostic tool of oral health status. Unfortunately, numerous cross-sectional studies have shown a wide individual variation in the salivary composition of healthy populations, thus precluding its use as a diagnostic chair-side test for the screening of the most common chronic oral diseases (e.g., caries and periodontal disease). A possible explanation may arise from the wide functional versatility of salivary molecules. For instance, it has been recognized recently that in addition to its digestive properties, salivary amylase may modulate bacterial colonization, whereas histatins are not only antifungal but also bactericidal. Thus, low levels of already known antimicrobial salivary molecules (e.g., secretory IgA, lactoferrin, and lysozyme) could be compensated with higher concentrations of other molecules with antimicrobial activity, such as amylase and histatins. Consequently, for caries and periodontal diseases, longitudinal sialochemical studies may yield more insight than cross-sectional studies.

The goal of the present study was to begin characterizing the amylase-binding component(s) on the surface of Streptococcus gordonii G9B. Alkali extracts but not phenol-water extracts of this bacterium inhibited '25I-amylase binding to S. gordonii G9B. To identify the bacterial components involved in amylase binding, the alkali extract was subjected to affinity chromatography on amylase-Sepharose. Immunoblotting with a rabbit antiserum against S. gordonii G9B revealed that a 20-kDa streptococcal component was eluted from the amylase-Sepharose with 1% sodium dodecyl sulfate (SDS), 2 M KSCN, or 0.1 M sodium citrate buffer, pH 4.5. Subsequently, the 20-kDa component was prepared from alkali extracts by electroelution from preparative SDS electrophoresis or by gel filtration chromatography. This component was trypsin sensitive, and an antibody raised against it inhibited the binding of '25I-amylase to S. gordonii G9B. Indirect immunofluorescence microscopy and immunogold electron microscopy demonstrated that both bound amylase and the 20-kDa component were localized to the cell division septum on dividing cells or to polar zones on single cells. In addition, exponentially growing bacteria bound more 1251-amylase than stationary-phase cells did. Collectively, these results suggest that a 20-kDa amylase-binding component is present on the surface of the nascent streptococcal cell wall. * Corresponding author. with proteolytic enzymes abolishes amylase binding to S. gordonii, implicating a surface protein as the amylasebinding component (11, 12, 39). However, amylase binding could also be abrogated by treating the bacteria with periodate, suggesting a role for streptococcal carbohydrate in this interaction (39). Recently, several amylase-binding components have been identified on S. gordonii NCTC 7868 (Challis) (12). Little is known, however, about the chemical composition of these molecules or their distribution on the streptococcal surface. Thus, the goal of the present study was to characterize and localize the amylase-binding component(s) on S. gordonii G9B. MATERIALS AND METHODS Bacterial strains and culture conditions. The following amylase-binding strains were studied: S. gordonii G9B (37), Challis (NCTC 7868), Blackburn (NCTC 10231), and FAS4 (39) and S. crista CR311 (22). The non-amylase-binding strain S. sanguis ATCC 10556 was also used in control

Salivary proteins play an important role in the maintenance of the oral ecology. Previous studies have indicated that human submandibular-sublingual and parotid salivas can selectively suppress the in vitro infectivity of herpes simplex virus 1. The purpose of this study was to identify the salivary components in human submandibular-sublingual saliva that modulate in vitro infectivity. Assessment of the interaction of viral particles with salivary components was accomplished using an in vitro solid-phase assay. These experiments revealed that herpes simplex virus particles selectively interact with the members of the salivary proline-rich protein and cystatin families. Subsequent yield reduction assays demonstrated the ability of proline-rich proteins and salivary cystatins to inhibit the viral replication, with basic proline-rich peptides being more effective. Subsequent assays suggest that basic proline-rich peptides reduced the virus titer by interfering with penetration and/or cellular processing of virus within the target cell. Collectively, these results further suggest that salivary proteins have an important role in the host defense mechanism against recurrent herpesvirus infection.

The low-molecular-mass human salivary mucin has at least two isoforms, MG2a and MG2b, that differ primarily in their sialic acid and fucose content. In this study, we characterize further these isoforms, particularly their peptide moieties. Trypsin digests of MG2a and MG2b yielded high- and low-molecular-mass glycopeptides following gel filtration on Sephacryl S-300. The larger glycopeptides from MG2a and MG2b had similar amino acid compositions and identical N-terminal sequences, suggesting common structural features between their peptides. An oligonucleotide probe generated from the amino acid sequence of the smaller glycopeptide from MG2a was employed in Northern-blot analysis. This probe specifically hybridized to two mRNA species from human submandibular and sublingual glands. A cDNA clone selected from a human submandibular gland cDNA expression library with antibody generated against deglycosylated MG2a also hybridized to these two mRNA species. In both cases, the larger mRNA was polydisperse, and the hybridization signal was more intense in the sublingual gland. In addition, the N-terminal amino acid sequence of the larger glycopeptide was found to be part of one of the selected MG2 cDNA clones.