The University of Cape Town Research and Ethics Committee approved this study (ethics approval number REC REF: 283/2004).

The ECL™ Western Blotting Detection Kit was from Amersham Biosciences (Amersham UK). Nitrocellulose membrane and dialysis tubing were from Kimix (chemical and laboratory suppliers, SA). Polyclonal rabbit anti-MUC5B (Lum 5B-2), goat anti-MUC7, goat anti-rabbit and rabbit anti-goat horse radish peroxidise (HRPO) linked secondary antibodies were from Santa Cruz Biotechnology, Inc (Santa Cruz, California). The CD4⁺ CEM SS cells were from AIDS Research and Reference Reagent Programme (Germantown, USA). RPMI 1640, L-Glutamine and heat-inactivated fetal bovine serum were from Gibco (Massachusetts, USA). IL-2 and p24 antigen kit were from Roche Diagnostics (Germany) and Vironostika HIV-1 Antigen kit Biomérieux (France) respectively. Sepharose CL-4B and reagent solvents such as guanidinium chloride (GuHCl), phenylmethylsulfonylfluoride (PMSF), caesium chloride (CsCl), ethylenediaminetetra-acetic acid disodium salt (Na₂-EDTA), N-ethylmaleimide (NEM), and 3-((3-cholamidopropyl)-dimethyl-ammonio)-1-propane-sulfonate (CHAPS) were from Sigma (UK). Trypan Blue Dye solution was from Merck (Germany).

Saliva was collected from healthy symptom free 'normal' female and male volunteers (who declared a risk-free lifestyle) from the laboratory who abstained from eating and drinking for at least 2 h prior to collection. Whole saliva was stimulated by chewing on parafilm and collected into 6 M GuHCl containing a spectrum of protease inhibitors (10 mM EDTA, 5 mM NEM, 1 mM PMSF) and 0.1% CHAPS pH 6.5. Samples were collected on ice and then stored at -20°C.

Crude salivary mucus was solubilised by overnight mixing at 4°C on a revolving rotor. The insoluble debris was then separated from the soluble mucus by centrifugation at 4400 g for 10 min at 4°C.

Aliquots of the supernatant (20 ml) were chromatographed on a Sepharose CL-4B gel filtration column equilibrated and eluted with 4 M GuHCl containing 10 mM EDTA, 5 mM NEM and 0.05% CHAPS at pH 6.5 and at a flow rate of 48 ml/h at room temperature. Following PAS and protein (A₂₈₀) assays, the void (V_o) and included volume (V_i) materials were pooled separately, dialysed against three changes of distilled water at 4°C and freeze-dried.

For the HIV inhibition assay, crude saliva was collected into 0.1 M Tris-HCl, 2% (w/v) EDTA and 5 mM PMSF pH 7.5 (saliva:buffer ratio 1:2). After solubilisation by overnight mixing at 4°C on a revolving rotor, the insoluble debris was separated by centrifugation at 10 000 g for 10 min at 4°C. The supernatant was then dialysed over three changes of distilled water at 4°C and freeze-dried.

The mucin that eluted in the included volume (V_i) of the column was subjected to caesium chloride isopycnic density-gradient ultra-centrifugation, twice for 48 h at a 105 000 g and 4°C in a Beckman L45 ultra-centrifuge 28. Briefly, V_i samples dissolved in 4 M GuHCl containing a cocktail of proteolytic inhibitors (as described above) were adjusted to a density of 1.39 to 1.40 g/ml with caesium chloride prior to centrifugation. Mucin rich fractions were pooled, dialysed against three changes of distilled water at 4°C and freeze-dried.

SDS-PAGE was carried out according to the method of Laemmli 29 in a buffer containing 0.2% SDS using 4% (w/v) stacking gels and 10% separating gels 3031, in a Hoefer Mighty Small mini-electrophoresis system. Samples were prepared in reducing gel loading buffer containing 2% SDS, 10% glycerol, 0.01% bromophenol blue and 5% mercaptoethanol and boiled for 2 min prior to loading. After electrophoresis gels were stained for carbohydrate with PAS 32 and for protein with Coomassie Brilliant Blue G-250.

After gel electrophoresis mucins were transferred to a nitrocellulose membrane (Nitrocellulose, 0.22 Micron) using a semi-dry electroblotting unit at 0.8 mA/cm². The transfer buffer used contained 192 mM glycine, 25 mM Tris, 1.3 mM SDS and 20% (v/v) methanol. After electro-blotting non-specific binding was blocked by incubating the membranes overnight in 5% (m/v) low fat milk powder in TBS, 0.05% Tween-20 (TBST) at 4°C. The membranes were then washed with TBST for 3 × 5 min and incubated for 2 h with rabbit anti-MUC5B and goat anti-MUC7 polyclonal antibodies diluted in 5% (m/v) low fat milk powder in TBST at 1:500 and 1:100 dilutions respectively. The membranes were washed for 3 × 5 min with TBST and incubated for 1 h with HRPO linked secondary antibodies (goat anti-rabbit and rabbit anti-goat) diluted in 5% (m/v) low fat milk powder in TBST at 1:5000 dilution. After another TBST wash (3 × 5 min) bands that interacted with the antibodies were detected by exposing the membrane to ECL detection.

The amino acid contents of purified MUC5B and MUC7 mucins were analysed using a high pressure liquid chromatography (HPLC) system. The analysis procedure was similar to that of Klapper 33 and Cohen and Strydom 34. The samples were vacuum-dried and placed in a hydrolysis vessel containing some constant boiling HCl and 1% (v/v) phenol. The vessel was cleaned with nitrogen gas and sealed under vacuum. The samples were then hydrolysed in the gas phase at 110°C for 24 h. Following hydrolysis, the vials were cooled and vacuum dried to remove the residual HCl. The dried samples were re-dissolved in citrate buffer pH 2.2 and injected into a HPLC column from Waters Associates, Medford, MA., packed with a cation exchange resin (sulfonated polystyrene cross-linked with divinylbenzene) and eluted with a series of buffers ranging from a low (0.25 M trisodium citrate, pH 3.05) to high (0.25 M sodium nitrate, pH 9.5) pH. Detection was carried out using post column derivatization with o-phthalaldehyde (OPA), a fluorescent reagent that reacts with all the amino acids except proline. The relative ratios of the individual amino acids for each sample was determined and compared to each other. For proline detection samples were treated with sodium hypochlorite prior to post column derivatization with o-phthalaldehyde (OPA).

The toxicity of crude saliva and purified salivary MUC5B and MUC7 mucins to the CD4⁺ CEM SS cells was determined by a toxicity assay. Briefly, 500 μl of the CD4⁺ CEM SS cells in RPMI complete containing 10% Fetal Calf Serum, 1% Penicillin/Streptomycin antibiotic, 10 μmol Fungin and 50 μmol 2-mercaptoethanol (final concentration 2.5 × 10⁶ cells/ml) was incubated with 250 μl of IL-2 and 250 μl (0.9 mg) of crude saliva or purified salivary MUC5B and MUC7 mucins in a CO₂ incubator for 24 h. As controls CEM SS cells with IL-2 only and IL-2 without CEM SS cells (blank) were used. After spinning at 1400rpm for 5 min, cells were re-suspended in 500 μl of RPMI and live and dead cells were counted using Trypan blue exclusion criteria. The percentage of viable cells was calculated as live cells/total cells × 100.

The anti-HIV-1 activities of crude saliva and purified salivary MUC5B and MUC7 mucins were tested in an inhibition assay according to the method of Nagashunmugam et al. 3. Briefly, the crude saliva and purified salivary MUC5B and MUC7 mucins were dissolved in 0.25% PBS and (500 μl or 0.9 mg each) were mixed with 4 ml of the subtype D HIV-1 supernatant fluid (SNF) and incubated for 60 min at 37°C. As controls heat inactivated HIV-1 and HIV-1 plus media (RPMI 1640 with 10% fetal calf serum and IL-2) were used. The virus was first isolated from an AIDS patient by the Department of Medical Virology, Tygerberg Hospital, University of Stellenbosch, in February 1986, and it was fully characterised and sequenced subsequently. At the end of the incubation period the mixtures (virus plus mucins) and the control (virus plus media) were filtered through 0.45 μm pore size cellulose acetate filter (25 mm diameter), and both the unfiltered and filtered samples were incubated with CD4⁺ CEM SS cells at 37°C at a concentration of 0.5 × 10⁶ cells/ml for 30 min, 1 h and 3 h. Cells were then washed three times with PBS to remove free virus and cultured. Supernatant fluid was harvested on Day 4 and viral replication was measured by a qualitative p24 antigen assay. Endpoints were calculated by the Reed-Muench formula and the 50% tissue culture infective dose (TCID₅₀) was expressed as the highest dilution that produced a positive qualitative p24 antigen result. All samples were done in triplicate and the anti-HIV-1 activity of each mucin was tested in a serial tenfold dilution (10^-1 to 10^-4).

Sepharose CL-4B gel filtration demonstrated that saliva contains two species of carbohydrate-rich mucin macromolecules which are distinguishable according to their size. Of these two populations, the glycoprotein (MG1) was excluded in the void volume (V_o) 23 and the glycoprotein (MG2) in the included volume (V_i) of the gel filtration column 23 (Fig. 1). A sharp PAS positive peak was seen eluting in the void volume and a broader peak in the included volume of the column, which was alongside a protein positive peak. The protein peak that eluted in the void volume, coinciding with the PAS positive peak, was much smaller.

To check the purity of the materials eluted in the V_o and the V_i of the gel filtration column, fractions eluting under these peaks were pooled separately, dialysed extensively against distilled water, freeze-dried and subjected to 10% SDS-PAGE and stained for protein with Coomassie Brilliant Blue G-250 (Fig. 2a and 3a) and for carbohydrate with PAS (Fig. 2b and 3b).

The PAS positive void volume material gave a strong band sitting on the top of the stacking and running gels, very characteristic of the electrophoretic behaviour of large-sized mucins (Fig. 2b) 30. On the other hand the material that eluted in the V_i showed a prominent band of greater electrophoretic mobility and of relatively smaller size than V_o mucin, in that the mucins slightly entered the running gel (Fig. 3b). Also there were a range of fainter bands ranging in size from 66–30kDa (Fig. 3b, lanes 1–3 and 5 and 6). More contaminant protein associated with mucin was present in the V_i peak (Fig. 3a) than the V_o peak (Fig. 2a) suggesting the need for further purification of the material that eluted in the included volume of the column.

Material that eluted in the included volume of the gel filtration column was subjected to density-gradient ultra-centrifugation in caesium chloride. Mucin rich fractions from the first density gradient (Fig. 4a) were pooled and subjected to a second spin in caesium chloride (Fig. 4b), after which the mucin rich fractions were pooled, dialysed against distilled water and freeze-dried. SDS-PAGE analysis of these mucins showed less protein contaminant (Fig. 5, lane 2), except for low molecular weight material of 30kDa and under. As indicated by the arrows, the PAS stain of the same material gave two prominent bands in the region of 220kDa and 46–60kDa (Fig. 5, lane 3). Traces of material were also seen at 21.5kDa and under (Fig. 5, lane 3).

To determine the identity of the V_o and V_i mucins from the gel filtration column material eluting under these peaks was subjected to 10% SDS-PAGE and then transferred to nitrocellulose membrane and probed with rabbit anti-MUC5B and goat anti-MUC7 polyclonal antibodies respectively. The results showed that the V_o material sitting on the top of the gel is MUC5B (Fig. 6a, lanes 2 and 3) and the two bands slightly entered the running gel that interacts with the anti-MUC7 polyclonal antibody are the two glycoforms of MUC7 (Fig. 6b, lane 3). The two glycoforms of MUC7 in Figure 6b are indicated by arrows.

Following the amino acid analysis (Table 1) both MUC5B and MUC7 mucins were found to contain higher amounts of threonine, serine, glutamic acid, glycine and aspartic acid. However, the amount of threonine in MUC7 is very low. Of these signature amino acids of mucins, serine, threonine, and proline were found to comprise 26% and 28% of the MUC5B and MUC7 mucins respectively.

Prior to the HIV inhibition assay the toxicity of the crude saliva and purified salivary MUC5B and MUC7 mucins to the CD4⁺ CEM SS cells was determined by toxicity assay. Based on this assay (data not shown) no cell death or toxicity of these mucins to the CD4⁺ CEM SS cells was detected.

Subsequent to the toxicity assay the anti-HIV-1 activities of the crude saliva and purified salivary MUC5B and MUC7 mucins were determined by HIV inhibition assay. The result revealed that after a 30 min incubation period during which the mixtures three separate mixtures (HIV-1 plus crude saliva), (HIV-1 plus MUC5B) and (HIV-1 plus MUC7) were incubated with the CD4⁺ CEM SS cells, no HIV-1 infection of these cells was observed (Fig. 7). Simultaneously, to determine the effect of the incubation period between the CD4⁺ CEM SS cells and the mixtures described above on the infectivity of these cells, incubation was carried out at different time-points (1 h and 3 h). No difference in the rate of inhibition due to time differences was detected (Fig. 7). Serial tenfold dilution (10^-1 to 10^-4) of the mucins was also done to determine the anti-HIV-1 activity of these mucins at the highest dilution. All the mucins showed anti-HIV-1 activity down to the concentration of 0.09 μg or in the range of 900 μg to 0.09 μg (data not shown).

In the controls, where the virus was incubated with the media only, prior to addition to the CD4⁺ CEM SS cells at all time points (30 min, 1 h and 3 h), 100% of HIV-1 replication or infection of the CD4⁺ CEM SS cells was detected (Fig. 7). However no HIV-1 infection was revealed when heat inactivated HIV-1 was used (Fig. 7).

To determine whether the HIV-1 activity of the crude saliva and purified salivary MUC5B and MUC7 mucins is by aggregation or trapping of the virus, the mixtures (HIV-1 plus crude saliva), (HIV-1 plus purified salivary MUC5B), (HIV-1 plus purified salivary MUC7) and (HIV-1 plus media), at the end of the 60 min incubation period, were filtered through 0.45 μm pore size cellulose acetate filter (25 mm diameter) and the filtrates were incubated with the CD4⁺ CEM SS cells at all time-points (30 min, 1 h and 3 h). While 100% of viral replication or infection of the CD4⁺ CEM SS cells was detected with the filtrate of the control (HIV-1 plus media) (Fig. 8), no HIV-1 replication or infection of the CD4⁺ CEM SS cells was detected when the filtrates from the three mixtures were used (Fig. 8).