Docking Analysis in Research for Novel Enzyme Inhibitors
Edit

Approaches for increasing effectiveness of docking analysis in Prediction of inhibitory potency of small molecules, emphasizing in selection of appropriate enzyme 3D-structure and in calculation of probability of binding factors, based on docking analysis to the target site and to the whole enzyme.

Docking enzyme inhibitors whole enzyme docking box 3D- enzyme structure inhibition prediction Probability of Binding Factor PF Prediction Factor PF

1. Enzymes as Drug-Targets

Enzymes constitute a major class of drug targets with enzyme inhibitors dominating among approved drugs for a great number of pathologic disorders, among which inflammation, hypertension, hypercholesterolemia, diabetes, viral and microbial infections etc. Despite the numerus approved agents, research in this field continues for the development of safer drugs with less side effects, ability to act against mutated stains or targeting to novel enzymes with pharmaceutical interest.

The vast majority of enzyme inhibitors are small molecules which bind to the active site (competitive inhibitors) or other sites of the enzymes (allosteric sites).

2. Docking Analysis

Docking analysis is one of the tools used to estimate the probability of small molecules to form stable complexes with the enzymes, binding to the active or allosteric sites thus leading to enzyme inhibition. The change in Free Energy (ΔG) during complex formation is a measure of the stability of the complex. A negative ΔG with high absolute value is indicative of a stable complex.

In Docking analysis, a computational program is used to estimate the change in Free Energy (sometimes characterized as Estimated Binding Energy, Eest) for different orientations of binding of the small molecule at a determined target site of the enzyme which usually is the active site or an allosteric site if there is one. Usually, a descent inhibition is observed if the calculated Eest is lower than -5.5 Kcal/mole. The lower Est the higher inhibition. Since the inhibition potency is measured by the concentration of the inhibitor which reduces enzyme activity to 50% (IC50), the lower Est corresponds to lower expected IC50 value for the potential inhibitor.

Application of Docking Analysis presupposes crystallographic studies of the enzyme at a pure form or in complex with the substrate or a known inhibitor and the existence of 3D structures of the enzyme at the 3D-structure Protein Databases. Such structures can be found through several links such as https://www.ncbi.nlm.nih.gov/Structure/index.shtml, https://www.rcsb.org/docs/search-and-browse/basic-search, https://www.ebi.ac.uk/pdbe/entry/pdb/1A5Y

3. Approaches for Increasing Effectiveness of Docking Analysis in Prediction of the Inhibitory Potency of Small Molecules

3.1. Selection of Appropriate Enzyme 3D-Structure

Docking-based prediction of inhibitory action is not always accurate. The accuracy of such predictions is a goal for the scientists who are working on computational programs or the ones applying these programs as a tool to facilitate their research. Interestingly, docking analysis using different 3D-structures of the same enzyme usually gives different results. This is mainly because of the flexibility of the enzymes and their capacity to change conformation when interacting with their substrate or inhibitors. Docking analysis using an enzyme structure which was crystallized in a complex with a competitive inhibitor is ideally used to predict binding at the active site because of the ability of the enzymes to slightly change adapting a comformation "favoring"  binding to the substrate or inhibitor. A more accurate prediction is obtained when the complexed inhibitor at the crystallized enzyme form (initial ligand) resembles the studied compounds in magnitude or even in structure [1]. In an analogue way, docking analysis using an enzyme structure which was crystallized in  complex with an allosteric inhibitor is preferred for the prediction of the capacity of small molecules to act as allosteric inhibitors of the enzyme. The use of the same crystal structure for all the compounds of a group is preferred to produce comparable results if the compounds are of the same magnitude. The use of multiple available structures and the calculation of an average binding energy, Eest, has been proposed by some scientists [2][3][4]. The specific characteristics of the docking target (coordinates of the docking center and dimensions of the docking box) are also important for effective docking prediction and are usually adjusted to the magnitude and specific characteristics of the group of studied compounds [5].

3.2. Calculation of Probability of Binding Factors, based on Docking Analysis to the Whole Enzyme

If the docking box is expanded to include the whole enzyme, several sites of probable binding are revealed. This may imply the presence of allosteric site(s) or may indicate the presence of strong binding site(s) with no effect on the catalytic activity of the enzyme. This will antagonize binding at the target site, reducing the potency of predicted inhibitors.

Docking to the whole enzyme in parallel with docking targeted to the active site or allosteric site has been proposed for better prediction results [6]. In this case, in compounds which strongly bind at the active site, the probability of acting as potent competitive inhibitors is reduced if there are other sites of stronger binding potential (lower Estimated Binding Energy in docking results). In such cases, the experimental IC50 value tends to be much higher than the predicted one [7].

Better prediction results are obtained if a Probability of binding Factor is calculated taking into account the binding at other sites with Estimated binding Energy (Eest, ΔG) lower than that of binding at the target site.

Such a factor was calculated for prediction of inhibitory potency of competitive and allosteric inhibitors against PTP1b with good prediction results leading to correlation between the Prediction Factor and the in vitro calculated ΙC50 described by an asymmetrical sigmoidal curve with r2 = 0.9692 [6].

A Probability of binding Factor, PF, based on docking to the target site and to the whole enzyme, also taking into account the frequency of binding, was calculated for the prediction of competitive inhibitors against DPP4, leading to a linear regression between PF and in vitro calculated logΙC50 with R2 = 0.9881, p = 0.00005347 [7]. This is a great improvement compared to the correlation of the Eest calculated by docking to the target site box (Eestts) with logIC50 which exported much poorer linear regression results with R2 = 0.5932 and p = 0.07316.

The results in this study [7] indicated that the determined IC50 values of the competitive inhibitors followed well (R2 = 0.9881, p = 0.00005347) the following equation:

logIC50 = 0.6704824 * PF + 6.451999

where: PF = Eestts − { Σ[(Eestx - Eestt) * (vx/100)] * 10},

Eestts: Estimated Binding Energy to the target site when the docking is applied to the target site box.

Eestt: Estimated Binding Energy to the target site when docking is applied to the whole enzyme.

Eestx: Estimated Binding Energy to the site x, lower than Eestt, when the docking is applied to the whole enzyme.

vx: frequency % of the docking pose with Estimated binding Energy Eestx

Figure 1. Correlation of log IC50 of competitive inhibitors with the Probability Factor (PF) [7].

References

  1. Eleftheriou, P.; Petrou, A.; Geronikaki, A.; Liaras, K.; Dirnali, S.; Anna, M. Prediction of enzyme inhibition and mode of inhibitory action based on calculation of distances between hydrogen bond donor/acceptor groups of the molecule and docking analysis: An application on the discovery of novel effective PTP1B inhibitors.. SAR QSAR Environ. Res. . 2015, 26, 557-576.
  2. Bello, M.; Martínez-Archundia, M.; Correa-Basurto, J. Automated. Automated docking for novel drug discovery. Expert Opin. Drug Discov. 20013, 8, 821-834.
  3. Knegtel, R.M.; Kuntz, I.D.; Oshiro, C.M. Molecular docking to ensembles of protein structures. J. Mol. Biol. . 1997, 266, 424-440.
  4. Osterberg, F.; Morris, G.M.; Sanner, M.F.; Olson, A.J.; Goodsell, D.S. Automated docking to multiple target structures: Incorporation of protein mobility and structural water heterogeneity in AutoDock. Proteins. 2002, 46, 34-40.
  5. Tsolaki E.; Eleftheriou P,; Kartsev V.; Geronikaki A.; Saxena A.K. Application of Docking Analysis in the Prediction and Biological Evaluation of the Lipoxygenase Inhibitory Action of Thiazolyl Derivatives of Mycophenolic Acid. Molecules. 2018, 23(7), 1621.
  6. Ganou, C.A.; Eleftheriou, P.T.; Theodosis-Nobelos, P.; Fesatidou, M.; Geronikaki, A.A.; Lialiaris, T.; Rekka, E.A. Docking Analysis Targeted to the Whole Enzyme: An Application to the Prediction of Inhibition of PTP1B by Thiomorpholine and Thiazolyl Derivatives. SAR QSAR Environ. Res. . 2018, 29, 133-149.
  7. Amanatidou, D.; Eleftheriou, P.; Petrou, A.; Geronikaki, A.; Lialiaris, T. Τhiazolidine-4-one Derivatives with Variable Modes of Inhibitory Action Against DPP4, a Drug Target with Multiple Activities and Established Role in Diabetes Mellitus Type II. Pharmaceuticals. 2025, 18(1), 52.
More
Information
Contributor MDPI registered users' name will be linked to their SciProfiles pages. To register with us, please refer to https://encyclopedia.pub/register :
View Times: 670
Revisions: 3 times (View History)
Update Date: 22 Jan 2025
Video Production Service