Fig 2: Molecular interactions at the surface of the biosensor biochip. Schematic illustrating the functionalisation of the gold-coated biochip with firstly anti-IgG antibodies (red), and secondly with anti-MMP-8 capture antibodies (blue), followed by detection by successive addition of sample (containing MMP-8) and anti-MMP-8 detection antibody (green). The binding of the anti-MMP-8 detection antibody to MMP-8 bound to the immobilised complex results in a perturbation of mass at the biochip surface sufficient to generate a phase change (?f) in the surface acoustic wave detected by the biosensor control box. In later experiments, the anti-MMP-8 capture antibody was bound directly to the biochip, omitting the anti-IgG antibody step. A full description of the prototype biosensor structure, microelectronics and detection system has been recently published28.

Fig 3: Analytical performance of the biosensor for the assay of salivary MMP-8 in periodontal disease. Analysis of salivary MMP-8 in healthy volunteers (N = 56), patients with gingivitis (N = 47) and patients with periodontitis (n = 65) using ELISA (a) and the biosensor assay (b). Data are means of duplicate measurements and are presented as box and whisker plots: boxes represent median (line) and interquartile ranges and whiskers the minimum and maximum range. Data was analysed using Kruskal-Wallis one-way ANOVA with Mann-Whitney U post hoc test, *P < 0.001. Correlation of salivary MMP-8 analysis of clinical samples (n = 168) using ELISA and biosensor assays (c). Data were analysed using Spearman’s Rank correlation, Rho = 0.795, P < 0.001.

Fig 4: Response of the biosensor to hrMMP-8. Biosensor phase change response to increasing concentrations of hrMMP-8 (0–1000 ng/ml). Overlaid plots of raw data from a single experiment showing phase shifts recorded as a function of time after the addition of the detection antibody at t = 0 (a). Graph illustrating the relationship of relative phase change between samples containing MMP-8 and controls (?f) (b). Phase change measurements were made after final buffer wash (see Methods). Data are derived from 3 independent experiments each comprising parallel duplicate measurements on separate microchips and are presented as means ± SD (n = 6).

Fig 5: Comparative biosensor analysis of salivary MMP-8 before and after treatment for periodontitis. Analysis of salivary MMP-8 in a sub-group of patients (n = 20) before and after 6 months after non-surgical treatment for periodontitis. (a) Data presented as ‘box and whisker plots’: boxes represent median (line) and interquartile ranges and whiskers the minimum and maximum range. Data were analysed using a paired samples t-test *P = 0.026. (b) Figure showing the changes in levels on MMP-8 before and after treatment in saliva samples from individual patients.