Discovery of potential pancreatic cholesterol esterase inhibitors using pharmacophore modelling, virtual screening, and optimization studies

Figures & data

Figure 1.  Chemical structures of the training set compounds together with their experimental K_i values in µM used for pharmacophore generation.

Table 1.  Summary of the generated common feature pharmacophore models.

Hypotheses	Molecular features^a	Ranking score	Direct hit^b	Partial hit^c	Max. fit
Hypo 1	HHHAA	57.887	111111	000000	5
Hypo 2	HHHAA	57.587	111111	000000	5
Hypo 3	HHHAA	56.563	111111	000000	5
Hypo 4	HHHAA	55.823	111111	000000	5
Hypo 5	HHHAA	55.542	111111	000000	5
Hypo 6	HHHAA	54.082	111111	000000	5
Hypo 7	HHHAA	53.880	111111	000000	5
Hypo 8	HHHAA	53.792	111111	000000	5
Hypo 9	HHHAA	52.673	111111	000000	5
Hypo 10	HHHAA	49.212	111111	000000	5

^aH, Hydrophobic; A, hydrogen bond acceptor.

^bDirect hit indicates whether (1) or not (0) a molecule in the training set mapped every feature in the hypothesis.

^cPartial hit indicates whether (1) or not (0) a particular molecule in the training set mapped all but one feature in the hypothesis.

Table 2.  Best fit values and absolute energy for the training set compounds respective to “Hypo 1” and “Hypo 2”.

Compound	Ki µM	Fit value		Absolute energy (kcal/mol)
		Hypo 1	Hypo 2	Hypo 1	Hypo 2
1	0.011	4.263	2.708	37.411	44.775
2	0.013	1.614	2.592	24.981	24.981
3	0.018	4.999	1.905	68.057	68.057
4	0.019	3.110	2.791	33.584	47.717
5	0.025	3.028	5.000	13.881	12.766
6	0.04	1.471	3.202	17.044	15.805

Figure 2.  (A) Three-dimensional spatial arrangement of the best pharmacophore hypothesis “Hypo 1”. (B) The distance constraints between the chemical features. Green colour represents hydrogen bond acceptor (HBA) and cyan colour represents hydrophobic (HY) features.

Table 3.  Statistical parameters and scores from decoy set validation for “Hypo 1”.

S. No.	Parameters	Results
1	Total molecules in database (D)	90
2	Total no. of actives in database (A)	35
3	Total hits (Ht)	41
4	Active hits (Ha)	33
5	% Yield of actives [(Ha/Ht) × 100]	80.49
6	% Ratio of actives [(Ha/A) × 100]	94.29
7	Enrichment factor (E) [(Ha × D)/(Ht × A)]	2.06
8	False negatives [A − Ha]	2
9	False positives [Ht − Ha]	8
10	Goodness of hit (GH)*	0.72

*[(H_a/4H_tA)(3A + H_t) × (1 – (H_t – H_a)/(D – A))]; GH score of 0.6–0.8 indicates a very good model.

Figure 3.  Compound 1 (A) and compound 6 (B) in the training set mapped on “Hypo 1”. Green colour represents hydrogen bond acceptor (HBA) and cyan colour represents hydrophobic (HY) features.

Figure 4.  Pharmacophore mapping of final database hit compounds on the best pharmacophore hypothesis “Hypo 1”, (A) SEW00846 represented in green colour, (B) NCI0040784 represented in violet colour, (C) GK03167 represented in blue colour, and (D) CD10645 represented in cyan colour. In the pharmacophore hypothesis, green represents hydrogen bond acceptor (HBA) and cyan represents hydrophobic (HY) features.

Figure 5.  Binding orientations of database hit compounds: (A) SEW00846, (B) NCI0040784, (C) GK03167, and (D) CD10645 are shown in cyan, red, blue, and magenta colours, respectively. Hydrogen bonds are shown in dotted lines.

Figure 6.  2D representation of the database hit compounds: (A) SEW00846, (B) NCI0040784, (C) GK03167, and (D) CD10645.

Table 4.  GOLD scores, AutoDock binding energies, and SYLVIA synthetic accessibility scores of top 10 optimized compounds.

Compound	Structure	Precursor	GOLD fitness score	AutoDock binding energy	SYLVIA score^a
OPT-SEW-1		SEW00846	70.566	−7.87	4.631
OPT-SEW-2		SEW00846	70.500	−9.70	5.064
OPT-SEW-3		SEW00846	64.473	−7.73	5.208
OPT-NCI-1		NCI0040784	58.110	−5.33	3.573
OPT-NCI-2		NCI0040784	54.520	−4.29	3.554
OPT-GK-1		GK03167	64.581	−5.48	4.555
OPT-GK-2		GK03167	64.397	−5.63	4.568
OPT-CD-1		CD10645	67.749	−9.77	5.107
OPT-CD-2		CD10645	66.979	−7.94	4.397
OPT-CD-3		CD10645	64.408	−7.17	4.471

^aThe calculated score for the synthetic accessibility. A number between 1 and 10 denoting the synthetic simplicity and the complexity of a particular chemical compound, respectively.

Figure 7.  Binding orientations of the optimized compounds in the active site of Cease: (A) OPT-SEW-5, (B) OPT-NCI-1, (C) OPT-GK-1, and (D) OPT-CD-2 are shown in blue, green, yellow, and orange colours, respectively. Hydrogen bonds are shown in dotted lines.

Figure 8.  2D representation of the final optimized compounds: (A) OPT-SEW-5, (B) OPT-NCI-1, (C) OPT-GK-1, and (D) OPT-CD-2.