Bisphosphonate inhibition of a Plasmodium farnesyl diphosphate synthase and a general method for predicting cell-based activity from enzyme data

Abstract

We screened 26 bisphosphonates against a farnesyl diphosphate synthase from Plasmodium vivax, finding a poor correlation between enzyme and cell growth inhibition (R(2) = 0.06). To better predict cell activity data, we then used a combinatorial descriptor search in which pIC(50)(cell) = a pIC(50)(enzyme) + bB + cC + d, where B and C are descriptors (such as SlogP), and a-d are coefficients. R(2) increased from 0.01 to 0.74 (for a leave-two-out test set of 26 predictions). The method was then further validated using data for nine other systems, including bacterial, viral, and mammalian cell systems. On average, experimental/predicted cell pIC(50) correlations increased from R(2) = 0.28 (for an enzyme-only test set) to 0.70 (for enzyme plus two descriptor test set predictions), while predictions based on scrambled cell activity had no predictive value (R(2) = 0.13). These results are of interest since they represent a general way to predict cell from enzyme inhibition data, with in three cases, R(2) values increasing from approximately 0.02 to 0.72.

Figure 1

Structures of the 26 compounds investigated in cell (Plasmodium falciparum) and enzyme (Plasmodium vivax FPPS) assays.

Figure 2

Correlation plots for cell (Plasmodium falciparum) and enzyme (Plasmodium vivax FPPS) assays and predicted cell activities from the training and test set data, obtained by using the combinatorial descriptor search method. (a) Plot showing correlation between cell pIC₅₀ (= −log₁₀(IC₅₀)) and enzyme pIC₅₀ values. (b) Best correlation between predicted cell pIC₅₀ (enzyme plus two molecular descriptors) and experimental pIC₅₀ values: training set results. (c) Test set pIC₅₀ predictions (leave-two-out analysis) plotted against the experimental values. The R² value increases from ∼0 to 0.74, when adding the two molecular descriptors to the enzyme data. The colored circles (A—C) indicate bisphosphonates with side chains having different relative hydrophobicities and potencies in the cell assay (low, intermediate or high, respectively), as discussed in the text.

Figure 3

Correlation plots for the cell and enzyme assays, and predicted cell activities from the training and test set data in Dictyostelium discoideum (a—c), Leishmania donovani (d—f), and Streptococcus pneumoniae (g—i). (a) Plot showing the correlation between cell (D. discoideum) pIC₅₀ and enzyme (FPPS) pIC₅₀ values. (b) Correlation between predicted cell pIC₅₀ values (from the best combination of enzyme plus two molecular descriptors) for the training set, with the experimental pIC₅₀. (c) Correlation between test set pIC₅₀ predictions obtained from a leave-two-out analysis. (d) Plot showing the correlation between cell (L. donovani) pIC₅₀ and enzyme (FPPS) pIC₅₀ values. (e) Correlation between predicted cell pIC₅₀ values (from the best combination of enzyme plus two molecular descriptors) for the training set, with the experimental pIC₅₀. (f) Correlation between test set pIC₅₀ predictions obtained from a leave-two-out analysis. (g) Plot showing the correlation between cell (S. pneumoniae) pIC₅₀ and enzyme (UPPS) pIC₅₀ values. (h) Correlation between predicted cell pIC₅₀ values (from the best combination of enzyme plus two molecular descriptors) for the training set, with the experimental pIC₅₀. (i) Correlation between test set pIC₅₀ predictions obtained from a leave-two-out analysis.