Fig. 5

In vitro validation of in silico predictions on chondrocyte phenotype changes. A Concept of in silico identification of potential drug targets. B Secreted ALP activity, relative to DNA quantity, positively linearly correlates with Col10a1 gene expression during hypertrophic differentiation with and without Ihh treatment. Results of one representative experiment. Each point is the average of 3 replicates and bars denote standard deviation. C Effect of PKA or SMAD3 activation as measured in silico and in vitro in ATDC5 (N = 3 replicates, histograms show average fold change in ALP activity relative to control and bars are standard deviations, p-values are computed with one-tailed t-test and Welch’s correction) and human chondrocytes from OA donors (N= 4 donors with 3 replicates each, p-value is computed with one-tailed linear mixed effect model). In silico activation was performed by setting the variables to their maximum value (1.0), in vitro PKA (resp. SMAD3) activation was performed with Forskolin 1μM (resp. Activin 100ng/ml) for 24h. D Single and combinatorial drug screening in ATDC5 with selected conditions based on in silico predictions. Boxplot of the series of conditions across independent replicates (z-scores of ALP activity fold change) with control conditions in purple. Conditions significantly lower than the control (combined p-value < 0.05) have dark grey borders and dots (Wilcoxon rank-sum test with BH correction and combined probabilities over independent runs). For each condition, dots are the average of biological triplicates, summary statistics are represented by a horizontal line for the median of independent experimental repetitions and a box for the interquartile range. The whiskers extend to the most extreme data point that is not >1.5 times the length of the box away from the box. Blue labels indicate potent conditions predicted by the in silico model, gray labeled conditions are added to the experimental set-up for information. CM stands for ‘control medium’, medium1 has 0.02% of DMSO and medium 2 0.035%. * Indicates in silico predicted conditions without significant decrease of ALP activity in vitro