Role of sialidase in glycoprotein utilization by Tannerella forsythia

Abstract

The major bacterial pathogens associated with periodontitis include Tannerella forsythia. We previously discovered that sialic acid stimulates biofilm growth of T. forsythia, and that sialidase activity is key to utilization of sialoconjugate sugars and is involved in host-pathogen interactions in vitro. The aim of this work was to assess the influence of the NanH sialidase on initial biofilm adhesion and growth in experiments where the only source of sialic acid was sialoglycoproteins or human oral secretions. After showing that T. forsythia can utilize sialoglycoproteins for biofilm growth, we showed that growth and initial adhesion with sialylated mucin and fetuin were inhibited two- to threefold by the sialidase inhibitor oseltamivir. A similar reduction (three- to fourfold) was observed with a nanH mutant compared with the wild-type. Importantly, these data were replicated using clinically relevant serum and saliva samples as substrates. In addition, the ability of the nanH mutant to form biofilms on glycoprotein-coated surfaces could be restored by the addition of purified NanH, which we show is able to cleave sialic acid from the model glycoprotein fetuin and, much less efficiently, 9-O-acetylated bovine submaxillary mucin. These data show for the first time that glycoprotein-associated sialic acid is likely to be a key in vivo nutrient source for T. forsythia when growing in a biofilm, and suggest that sialidase inhibitors might be useful adjuncts in periodontal therapy.

Fig. 1.

Growth of T. forsythia biofilm on glycoprotein surfaces. (a) Biofilm growth was assessed for the wild-type strain in the absence (black bars) or presence of oseltamivir (grey bars). Biofilms were set up in triplicate and wells were supplemented with 6 mM sialic acid (sialic) or 6 mM sialyl-lactose (sia-lac), or coated with 6 mM mucin, fetuin and asialofetuin (as-fet), 2 µg human serum ml⁻¹ (serum) or saliva, with and without the addition of 10 mM oseltamivir to the TSB medium, as indicated at the time of inoculation. Glycoproteins were coated on the 96-well plate overnight at 4 °C and washed before inoculation with T. forsythia to a final OD₆₀₀ of 0.05. Data are the mean number of cells with sds after harvesting from wells after 5 days. Differences between mean values in this experiment were evaluated by t test, with P<0.05 being taken as the level of significance (*statistically significant). (b) Biofilm growth was compared as above between the wild-type (wt) strain and the ΔnanH mutant strain of T. forsythia (ΔnanH).

Fig. 2.

Initial attachment of T. forsythia to glycoprotein-coated surfaces. (a) Adhesion to glycoproteins was assessed for the wild-type strain in the absence (black bars) or presence of oseltamivir (grey bars). Biofilms were set up in triplicate and wells were supplemented with 6 mM sialic acid (sialic), or were coated with 6 mM mucin, fetuin and asialofetuin (as-fet), 2 µg human serum ml⁻¹ (serum) or saliva, with and without the addition of 10 mM oseltamivir to the TSB medium, as indicated at the time of inoculation. Glycoproteins were coated on the 96-well plate overnight at 4 °C and washed before inoculation with T. forsythia to a final OD₆₀₀ of 0.05. Data are the mean number of cells with sds after harvesting from wells after 3 h. Differences between mean values in this experiment were evaluated by t test, with P<0.05 being taken as the level of significance (*statistically significant). (b) Adhesion to glycoproteins was compared as above between the wild-type (wt) strain and ΔnanH mutant strain of T. forsythia (ΔnanH).

Fig. 3.

Adhesion of the ΔnanH mutant strain of T. forsythia to glycoprotein surfaces without (black bars) and with purified NanH (rNanH) (grey bars). Biofilms were set up in triplicate and wells were supplemented with 6 mM sialic acid (sialic) or 6 mM sialyl-lactose (sia-lac) to the TSB medium, or were coated with 6 mM mucin, fetuin and asialofetuin (as-fet), 2 µg human serum ml⁻¹ (serum) or saliva, as indicated at the time of inoculation. Glycoproteins were coated on the 96-well plate overnight at 4 °C and washed before inoculation with T. forsythia to a final OD₆₀₀ of 0.05. Data are the number of cells with sds after harvesting from wells after 3 h. Differences between mean values in this experiment were evaluated by t test, with P<0.05 being taken as the level of significance (*statistically significant).

Fig. 4.

Release of free sialic acid from glycoprotein by recombinant NanH (rNanH). Six millimolar fetuin [0.48 µg sialic acid ml⁻¹, fetuin (0.48)], 0.002 mM mucin [0.48 µg sialic acid ml⁻¹, mucin (0.48)] or 6 mM mucin [19 µg sialic acid ml⁻¹, mucin (19)] was incubated in the presence (white bars) and absence (grey bars) of 5 mM of the sialidase inhibitor siastatin B (+sias) at 37 °C for 3 h. A₅₄₉ was measured after colour development (*statistically significant at P<0.05).

Fig. 5.

Removal of lectin-binding moieties by rNanH. Fetuin at 10 µg ml⁻¹ was incubated with increasing concentrations of rNanH (0, 0.78 and 1.5 mg ml⁻¹) for 3 h before running on an SDS-PAGE gel, blotting onto a nitrocellulose membrane, probing with 5 µg ml⁻¹ biotinylated SNA lectin, and visualizing with streptavidin–HRP and luminescent substrate, before incubation with X-ray film. Clostridium tetani sialidase (Cl-sia) (50 U) was used as a positive control, with asialofetuin (as-fet) and 1.5 mg rNanH ml⁻¹ as negative controls.