Fig 1.
Characterization of enzyme activity.
(a) In order to test the effect of pH on CaN activity, CaN assay was performed at different pH. Our data suggested that pH 7 is optimum for CaN activity. To study the effect of bivalent metal ions on CaN activity, enzyme activity assays were performed at different MnCl2 (b) and MgCl2 (c) concentrations. The data indicated that bivalent manganese increased CaN activity almost two times (b) whereas magnesium had not effect (c). Note that (a), (b) and (c) were performed in absence of CM. (d) The concentration of CM which produces 50% activation (IC50 CM) of CaN was determined by plotting initial reaction velocities at different Ca2+ and CM concentrations. Our data indicated that CM produces half maximum CaN activity at 1:1 molecular ratio. (e) Enzyme Km was determined by plotting initial velocity at different substrate (RIIP) concentrations. The Km was 213 μM. (f) DMOS tolerance of the enzyme was assayed by performing the assay at different DMSO concentration ranging from, 0.125 to 4%. Our data indicate that at ≤0.5% DMSO did not alter enzyme activity. The assays were done in duplicate (except for experiment in Fig 1d) and the data were expressed as means and standard error. The fittings were done using Prism software.
Fig 2.
Miniaturization of the assay using phosphate standard.
(a) The effect of different volumes of standard phosphate solution was studied by adding it into 384 Perkin Elmer white proxy plates. The malachite green volume was kept at ratio of 1.25:1 (reaction volume: developing reagent volume). (b) The influence of the malachite green quantity was measured using 200pmol phosphate in 5 μL volume. (c) Z score was plotted against malachite green volume. Z score linearly decreases with malachite green volume. (d) A phosphate standard curve was generated using white plate fluorescence quenching versus absorbance assay in 96 well plates. Inset shows high sensitivity phosphate detection at lower concentrations. Our data indicate that the 5μL assay volume with 2.5μL developing reagent produces the optimum result using white proxy plates. In panels a and b, raw data was converted to OD and then % of quenching was calculated.
Fig 3.
White plate fluorescence assay was validated using CaN and a known inhibitor. (a) Complete enzyme assay was performed in 272 replicates (full reaction) and compared with 80 replicates of no enzyme control. Assay was done manually in 384 well plates. The assay yielded a Z score of 0.63. (b) Endothall, a known inhibitor of CaN, was used to validate the assay format. The assay was done in duplicate and the data point represents mean and standard error. IC50 was calculated by prism software, using Eq 5, was 12.63μM.
Fig 4.
The complete assay was automated and performed in a similar way mimicking all steps of the actual screening. 100% DMSO was used to replace the compound. (a) Two full 384 plates were used. Plate 1 and 2 yielded a Z score of 0.91 and 0.81; % CV 13.99 and 15.12 for the full reaction, respectively. (b) Z scores of 4 pilot plates screened in duplicate using this automated assay format.
Fig 5.
Approximately 1400 compounds were screened, in duplicate, against CaN using this assay format. 50% inhibition was set as the threshold to identify hits. Out of 16 hits identified, 4 were selected on the basis of Lipinski's rule and commercial availability. Two of the selected compounds produced a dose-dependent inhibition and at least more than 50% inhibition. The reaction was done in duplicate and the data point represents mean and standard error. The IC50 were determined using Prism software, using Eq 5, and the values were (a) 106.4 μM for LDN-0013905 (flunarizine hydrochloride), (b) 4.5 μM for LDN-0013906 (tri-fluoperazine dihydrochloride). The fact that tri-fluoperazine dihydrochloride (LDN-0013906) is a known inhibitor of CaN further validated our assay format.
Fig 6.
Chemical structure and formula of identified hits.
Chemical structures, masses and formulas of the inhibitors are listed in the tabular format.