Is the bioconformation of 5-deoxy-5-fluoro-D-xylulose affected by intramolecular hydrogen bonds?

Abstract

5-Deoxy-5-fluoro-D-xylulose (DFX) binds to the xylulokinase enzyme and, as a free ligand, it has preferential conformations governed by intramolecular interactions, such as hydrogen bonds and hyperconjugative interactions. The role of intramolecular hydrogen bonds on the bioconformation of DFX has not been studied yet, despite the relevance of this topic to explain the mode of interaction between the ligand and enzyme and, therefore, the action mechanism of this molecule. DFX presents several conformations in the gas phase and implicit water, as determined by theoretical calculations, but the main optimized geometries do not match the bioactive conformation nor the most stable docked structure. This indicates that even expected strong interactions, such as hydrogen bonds, are overcome by the enzyme induced fit of DFX. The natural consequence of this finding in three-dimensional quantitative structure–activity relationship (3D-QSAR) analysis is that the use of conformations obtained in a receptor-free environment can cause misinterpretation of the chemical and biological results.

---

Introduction

5-Deoxy-5-fluoro-D-xylulose (DFX) binds to the xylulokinase enzyme, which plays an important role in metabolic disease, since it catalyzes the ATP-dependent phosphorylation of D-xylulose to produce xylulose 5-phosphate, a key regulator of glucose metabolism and lipogenesis.¹ Molecular conformation is a property ruling the bioactivity of drug-like compounds, such as DFX, since the enzyme fitting depends on the three-dimensional arrangement of the small molecule. Nevertheless, the preferential conformations of an enzyme-free molecule are dictated by intramolecular interactions, such as steric and electrostatic effects, hyperconjugation (e.g. in the anomeric and gauche effects) and hydrogen bonds.

It has been found that hydrophobic and hyperconjugative interactions are not strong enough to keep the bioconformation of organofluorine anesthetics as in the gas phase or solution.^2–4 However, hydrogen bonds are known as being stronger forces than other noncovalent interactions,⁵ and usually ascribed as the responsible phenomenon for the architecture of supramolecular systems,^6,7 ligand–enzyme interaction⁸ and conformation of small molecules.^9,10 Despite the seeming lack of hydrogen bond involving fluorine as proton acceptor in solution,^10–13 other nucleophilic sites in a molecule, such as oxygen and nitrogen atoms, can induce conformational changes due to intramolecular hydrogen bonds.¹⁴

This work reports the theoretical conformational analysis of DFX (Fig. 1) in the gas phase, implicit and explicit water solution, and also docked inside the active site of xylulokinase (Protein Data Bank code 4BC5), in order to find out the role of intramolecular hydrogen bonds on the bioactive conformation of a small molecule containing fluorine and oxygen (as hydroxyl and carbonyl groups) as proton donors and acceptors. Ultimately, the outcomes of this study can aid the design of congeneric drugs. In addition, the suitability of three-dimensional quantitative structure–activity relationship (3D-QSAR) techniques in drug design is evaluated, since such molecular modeling approaches require geometry optimization and 3D-alignment to obtain molecular descriptors for further correlation with bioactivity values; if the optimized enzyme-free conformation of DFX diverges from its bioconformation, such as in enflurane,³ isoflurane⁴ and 2,4-dichlorophenoxyacetic acid,¹⁵ the use of 3D molecular descriptors can be of limited utility.

Fig. 1 Chemical structure of 5-deoxy-5-fluoro-D-xylulose (DFX) and its corresponding bioactive conformation (Protein Data Bank code 5FX).

Computational details

5-Deoxy-5-fluoro-D-xylulose (DFX) has 7 torsional angles and, therefore, the statistical Monte Carlo sampling was used to identify the possible energy minima for this compound. A total of 83 conformers were found and, after optimization at the ωB97X-D/6-311++g(d,p) level^16,17 that includes empirical dispersion corrections, 65 different geometries converged to energy minima (ESI†). From these, 15 conformers with non-zero population in the gas phase and 20 conformers in implicit water as solvent (using the polarizable continuum model¹⁸) were obtained. In addition, an explicit water molecule was placed surrounding the carbonyl oxygen of DFX in each of the 20 conformers obtained in implicit solution and the geometries were then optimized. Frequency calculations at this DFT level were employed to guarantee that the minima obtained were not saddle points and also to obtain the standard free energy values. Natural bond orbital (NBO)¹⁹ analysis was carried out to obtain the energies from hyperconjugative interactions, while QTAIM²⁰ analysis was performed to characterize hydrogen bonds. The docking calculations were performed in order to understand the physical–chemical impact of the interaction between the molecular target and the studied conformations. The crystal structure of DFX inside the xylulokinase active site was obtained from the Protein Data Bank (PDB codes 5FX for the ligand and 4BC5 for the enzyme) and used for the docking procedure and alignment of the optimized structures. The calculation of the docking energies of the rigid conformation inside the xylulokinase active site was performed using the software Molegro Virtual Docker (MVD).²¹

Results and discussion

Intramolecular hydrogen bond can drive the conformational isomerism of molecules and, therefore, it has been analyzed against conformational entropy and enzyme–ligand interaction (mostly due to intermolecular hydrogen bond) in order to find whether the enzyme-free conformation of DFX persists in a biological environment. First, the conformational isomerism of DFX was computationally studied in the gas phase (g) and implicit water solution (w) and, subsequently, the factors governing the conformer stabilization were evaluated using NBO and QTAIM analyses. From a diverse set of conformers, 15 geometries in the gas phase and 20 in implicit water solution have non-zero Gibbs population (≥1%). The relative energies using an explicit water molecule as solvent differ from those obtained in the implicit solvation model (9w against 1w as the most stable conformations – Table 1), revealing the important role of specific intermolecular solute–solvent interactions on the conformational isomerism of DFX. This kind of interaction is certainly related to conformational changes induced by the enzyme active site, as further discussed. Thus, it is anticipated that the main conformation in an enzyme-free environment can be different from the bioconformation.

Table 1 Relative energies (kcal mol⁻¹), conformer Gibbs populations (%) and dihedral angles (degrees) for the conformations of DFX in the gas phase and implicit water (data for the explicit water is given in parenthesis), optimized at the ωB97X-D/6-311++g(d,p) level^a

Gas	Grel^0	%	Erel	EL	ENL	HOC(H2)C(=O)	OCC=O	O=CCO(H)	C(=O)COH	(H)OCCO(H)	HOCC(F)	OCCF
a Grel^0 = relative standard Gibbs free energy; Erel = conformational relative energy; EL = Lewis-type energy; ENL = non-Lewis-type energy.
1g	0.0	23	0.0	9.6	9.6	−16.8	12.2	9.4	−13.4	57.9	76.2	173.4
2g	0.2	16	1.3	10.2	8.9	9.6	−4.8	−6.3	2.1	−166.4	−55.3	62.2
3g	0.4	11	1.0	5.9	4.9	4.0	−1.7	174.7	84.0	52.3	−51.0	56.8
4g	0.5	10	0.1	7.5	7.4	69.8	159.2	−11.1	6.2	−171.3	−53.5	60.1
5g	0.8	6	1.6	1.6	0.0	−1.2	0.6	176.7	164.2	−48.8	−56.1	57.7
6g	1.0	4	1.5	12.3	10.8	−16.6	10.8	101.6	5.1	74.9	−46.9	60.0
7g	1.1	4	1.5	8.7	7.2	4.1	−1.7	177.6	−172.9	48.6	81.7	−73.5
8g	1.3	3	1.6	6.2	4.6	−2.9	1.5	172.6	153.8	−42.2	40.7	−54.2
9g	1.3	3	1.6	15.9	14.3	−18.0	12.1	10.1	−13.5	66.7	64.8	65.2
10g	1.3	3	2.4	3.9	1.5	0.9	−3.0	−159.3	83.1	51.9	−145.1	178.8
11g	1.5	2	2.4	3.1	0.7	1.7	−3.6	−151.5	81.9	53.4	−148.7	179.0
12g	1.6	2	0.5	9.9	8.4	−55.5	156.9	−110.7	48.1	70.9	−51.5	57.8
13g	1.7	1	3.3	4.6	1.3	5.4	−1.8	177.7	−133.4	50.9	73.6	175.7
14g	1.8	1	2.1	16.1	14.0	2.7	−2.9	−171.9	83.1	−163.5	76.0	−176.5
15g	1.8	1	2.2	3.0	0.8	−9.0	6.2	−105.9	82.2	72.6	−166.3	176.9
Water
1w	0.0 (3.3)	34 (0)	0.1	9.6	9.5	−18.9	10.7	7.2	−10.5	64.1	67.6	63.2
2w	0.7 (1.4)	10 (7)	0.1	7.3	7.2	−19.1	11.0	5.8	−8.2	59.7	72.9	174.9
3w	0.7 (1.7)	10 (4)	0.7	5.0	4.3	2.0	−0.7	176.9	84.5	54.4	−53.2	59.8
4w	0.9 (2.3)	7 (2)	1.7	9.2	7.5	6.9	−3.2	−7.0	−165.7	176.9	−59.2	65.7
5w	0.9 (2.3)	7 (2)	0.6	7.4	6.8	−18.3	9.7	0.9	4.5	71.1	−59.2	62.3
6w	1.2 (2.1)	4 (2)	0.8	6.2	5.4	−0.8	0.6	−176.3	−170.5	50.3	78.6	−69.1
7w	1.3 (4.9)	4 (0)	1.3	1.7	0.4	2.1	−1.0	−176.6	−125.7	75.7	−48.9	60.6
8w	1.4 (5.1)	3 (0)	1.4	6.5	5.1	−16.8	9.4	−0.5	7.6	68.1	−74.0	177.4
9w	1.5 (0.0)	3 (77)	1.5	4.2	2.7	1.2	−0.3	179.0	83.5	53.5	−71.0	178.3
10w	1.6 (5.5)	2 (0)	0.7	13.0	12.3	6.4	−3.7	−2.0	2.6	−170.3	−161.0	−62.6
11w	1.7 (3.1)	2 (0)	0.5	10.1	9.6	68.2	159.7	−11.4	10.5	−168.6	−57.7	64.8
12w	1.7 (4.4)	2 (0)	1.6	3.3	1.7	−80.6	159.0	−7.0	13.9	70.1	−60.2	62.3
13w	1.7 (5.5)	2 (0)	1.3	14.2	12.9	0.5	1.0	−177.9	−102.8	−75.8	83.2	−66.1
14w	1.8 (1.7)	2 (4)	1.3	4.4	3.1	2.0	−0.7	176.8	84.5	54.4	−53.2	59.8
15w	1.9 (4.6)	1 (0)	1.5	5.5	4.0	−1.1	1.6	165.8	156.6	−45.2	47.1	−59.3
16w	2.0 (3.6)	1 (0)	0.0	14.1	14.1	68.8	153.4	−5.6	6.9	169.9	−66.6	−59.1
17w	2.0 (5.0)	1 (0)	1.9	1.9	0.0	−1.2	1.8	165.8	157.9	−42.3	−79.8	64.8
18w	2.1 (4.6)	1 (0)	2.0	3.1	1.1	2.3	−0.4	171.8	−90.5	66.2	−62.2	63.6
19w	2.4 (6.1)	1 (0)	0.4	21.4	21.0	−89.7	178.2	−6.1	5.0	−156.0	−43.0	−63.0
20w	2.5 (5.0)	1 (0)	2.0	3.0	1.1	19.35	−11.80	−35.74	39.39	−55.17	172.70	−71.94
5FX						175.0	4.4	167.1	−149.1	52.2	120.1	65.8

---

According to NBO analysis (Table 1 and Fig. 2), the most populated conformer in the gas phase 1g (23%) exhibits a network of hydrogen bonds involving hydroxyl groups and the carbonyl oxygen of DFX . According to QTAIM analysis, a single bond path between C=O and O–H was observed in 1g and characterized as a weak hydrogen bond, since both ∇²ρ(r) and H(r) are positive²² (Table 2). Hydrogen bonds also appear in the other conformers and can include the fluorine atom. However, other interactions lead to the greater stabilization of 1g relative to the remaining conformers; the full electronic energy of a system can be decomposed into non-Lewis (electron delocalization) and Lewis-type (steric and electrostatic effects) interactions using NBO analysis and, according to Table 1, conformer 1g combines high stabilization due to hyperconjugation and low repulsive effects. Table 3 shows that, in addition to hydrogen bond , the gauche effect arisen from the gauche arrangement between electronegative substituents allows important antiperiplanar hyperconjugative interactions (especially ).

Fig. 2 Crystal structure and the optimized geometries of DFX in the gas phase and implicit water. The corresponding hydrogen bonds obtained by NBO analysis (, in kcal mol⁻¹) are shown.

Table 2 QTAIM parameters^a used to characterize hydrogen bonds in DFX

Gas	ρ(r)	∇^2ρ(r)	H(r)
a Electron density – ρ(r), Laplacian of the electron density – ∇^2ρ(r), and total energy density at the bond critical point – H(r).
1g	+0.025374	+0.107403	+0.002732
2g a/b	+0.022415/+0.026278	+0.101106/+0.109487	+0.003063/+0.002569
3g	+0.023248	+0.101158	+0.002874
4g a/b	+0.026865/+0.027738	+0.110218/+0.107869	+0.002475/+0.001740
5g a/b	+0.023737/+0.020795	+0.101876/+0.095271	+0.002773/+0.003071
6g a/b	+0.025111	+0.106653	+0.002777
7g a/b	+0.023391/+0.017242	+0.101689/+0.066264	+0.002877/+0.000979
8g a/b	+0.023979/+0.023399	+0.102157/+0.093128	+0.002725/+0.002101
9g	+0.025217	+0.106525	+0.002722
10g	+0.023744	+0.101565	+0.002746
11g	+0.023546	+0.101048	+0.002759
12g a/b	+0.029000/+0.025460	+0.100256/+0.098074	+0.000656/+0.001851
13g	+0.023590	+0.102021	+0.002854
14g a/b/c	+0.023293/+0.021372/+0.019610	+0.100444/+0.074582/+0.075621	+0.002747/+0.001342/+0.001009
15g a/b	+0.021967/+0.016586	+0.098853/+0.058660	+0.002986/+0.001238
Water
1w	+0.023990	+0.105896	+0.002748
2w	+0.025530	+0.107701	+0.002702
3w	+0.024078	+0.101946	+0.002712
4w	+0.025306	+0.106788	+0.002707
5w a/b	+0.023050/+0.025495	+0.100900/+0.107370	+0.002897/+0.002668
6w a/b	+0.024152/+0.014591	+0.102167/+0.056479	+0.002706/+0.000971
7w	+0.024449	+0.102938	+0.002677
8w	+0.024754	+0.105767	+0.002796
9w	+0.024544	+0.103050	+0.002656
10w a/b	+0.023425/+0.026389	+0.101164/+0.109546	+0.002807/+0.002544
11w a/b	+0.026365/+0.030194	+0.108854/+0.115734	+0.002536/+0.001356
12w	+0.025323	+0.106211	+0.002691
13w a/b	+0.024118/+0.022593	+0.101867/+0.081293	+0.002657/+0.001642
14w	+0.024192	+0.102243	+0.002696
15w a/b	+0.024560/+0.022415	+0.103014/+0.089973	+0.002641/+0.002163
16w a/b	+0.024189/+0.023866	+0.102220/+0.099989	+0.002695/+0.002483
17w a/b	+0.027762/+0.031069	+0.112484/+0.119232	+0.002350/+0.001245
18w	+0.024320	+0.102572	+0.002690
19w	+0.024322	+0.102578	+0.002690
20w	+0.029483/+0.035355	+0.119041/+0.124195	+0.002233/−0.000144

---

Table 3 Hyperconjugative interactions between antiperiplanar donor and acceptor orbitals (kcal mol⁻¹) for the conformers of DFX

Gas	1g	2g	3g	4g	5g	6g	7g	8g	9g	10g	11g	12g	13g	14g	15g
	—	—	—	—	2.2	0.7	—	2.5	—	—	—	0.6	0.6	—	—
	4.7	0.9	4.2	0.6	—	4.8	4.3	—	4.8	4.8	4.8	5.6	5.6	1.1	4.5
	—	1.7	—	2.0	—	—	—	—	—	—	—	—	—	1.4	—
	—	—	—	—	2.5	0.6	—	2.2	—	—	—	—	—	—	—
	4.5	1.1	4.6	0.9	—	5.0	4.8	—	4.8	4.8	4.9	5.4	5.4	1.0	4.9
	—	1.7	—	1.6	—	—	—	—	—	—	—	—	—	1.8	—
	—	2.7	2.7	2.7	2.5	2.7	0.7	—	2.6	—	—	2.8	2.8	—	—
	0.8	1.4	0.9	1.3	1.0	1.2	6.4	5.2	1.6	0.7	0.7	1.1	1.1	—	0.8
	1.7	—	—	—	—	—	—	—	—	1.9	1.9	—	—	2.1	2.0
	—	5.2	5.3	5.3	5.0	5.3	1.2	0.6	5.6	—	—	5.5	5.5	0.7	—
	0.6	0.8	0.6	0.8	0.7	0.8	5.4	5.1	1.0	0.5	0.5	0.7	0.7	—	0.6
	1.9	—	—	—	—	—	—	—	—	1.9	1.9	—	—	1.7	1.8

---

Implicit water	1w	2w	3w	4w	5w	6w	7w	8w	9w	10w	11w	12w	13w	14w	15w	16w	17w	18w	19w	20w
	—	—	—	0.6	—	—	—	0.5	—	—	—	0.5	—	—	2.4	2.2	—	5.6	2.4	—
	4.9	4.8	4.2	4.8	0.9	4.5	4.4	4.8	4.3	4.8	0.8	4.8	—	4.4	—	—	0.7	4.9	—	0.9
	—	—	—	—	1.6	—	—	—	—	—	1.8	—	1.9	—	—	—	1.9	—	—	1.9
	—	—	—	0.5	—	—	—	—	—	—	—	—	—	—	2.2	2.4	—	—	2.2	—
	4.8	4.6	4.9	5.2	1.1	4.9	4.7	5.1	4.9	4.6	1	5.2	0.6	4.5	—	—	1.0	5.3	—	1.7
	—	—	—	—	1.7	—	—	—	—	—	1.7	—	1.6	—	—	—	1.6	—	—	1.1
	2.7	—	2.8	2.8	2.9	0.5	—	—	—	—	2.8	2.9	—	2.9	—	2.7	—	2.7	—	—
	1.5	0.7	1.1	1.3	1.6	6.5	0.7	0.7	0.7	0.7	1.6	1.3	6.5	1.4	5.8	1.4	6.3	1.4	5.8	6.6
	—	1.8	—	—	—	—	1.9	1.8	1.8	1.8	—	—	—	—	—	—	—	—	—	—
	5.6	—	5.5	5.5	5.4	1.0	—	—	—	—	5.5	5.5	0.9	5.2	0.8	5.3	1.3	5.5	0.8	0.8
	0.9	0.6	0.7	0.8	0.9	5.8	0.6	0.6	0.6	0.6	1.0	0.8	5.6	1.1	5.4	1.0	1.4	0.9	5.4	5.4
	—	1.9	—	—	—	—	1.8	1.9	1.9	1.9	—	—	—	—	—	—	—	—	—	—

---

A conformer with high energy in the gas phase became the most stable geometry in implicit water (1w) with 34% population, and it differs from 1g only by the O–C–C–F dihedral angle. Indeed, the six most stable conformations in the gas phase appear within the top ten preferential conformations in solution, such as conformer 2w (with a geometry similar to 1g) as the second most stable (10%). Again, in addition to intramolecular hydrogen bond, a balance between steric effect and hypercojugative interactions takes place and governs the conformational equilibrium of DFX in an enzyme-free environment (Tables 1–3 and Fig. 2). Surprisingly, 1g and 1w do not match the crystal structure of DFX, which is the ligand bioconformation attached to the active site of xylulokinase.¹

In order to compare the optimized geometries with the experimental crystal structure, only the O–C–C=O, O=C–C–O(H), (H)O–C–C–O(H) and O–C–C–F dihedral angles were taken into account, since the coordinates of the hydroxyl hydrogens are not experimentally accurate. The bioconformation is reasonably consistent only with 3g (11% in the gas) and 3w, 7w and 18w (summing up 15% in implicit water). Consequently, the intramolecular forces ruling the optimized geometries are certainly not the same as those governing the bioactive conformation of DFX. In order to check the possible intermolecular interactions that overcome the intramolecular ones as driving forces of the bioconformation of DFX, docking studies were performed including all optimized geometries inside the bioactive site of xylulokinase.

The docking results were validated after finding the highest intermolecular (ligand–protein) interaction for the bioconformation of DFX (PDB code 5FX) in the active site of xylulokinase (PDB code 4BC5), which is −78.8 kcal mol⁻¹. Conformers 3g and 3w were the most stable optimized structures in the biological target (Table 4), with intermolecular interaction energies of −76.6 and −76.0 kcal mol⁻¹, respectively, more than 2 kcal mol⁻¹ less stable than 5FX. Nevertheless, 3g and 3w, together with 7w and 18w, exhibited high stabilization in the active site of xylulokinase in comparison to most of the optimized geometries, since they have more structural similarity to 5FX. The preferential enzyme-free conformations 1g and 1w interact attractively with the active site of xylulokinase (−31.1 and −39.8 kcal mol⁻¹, respectively), but they are far less stable than the bioconformation in the protein, revealing the important role of enzyme fit over intramolecular interactions for the conformer stabilization in a biological medium. Actually, many other enzyme-free conformations exhibit more favoring ligand–enzyme interaction than 1g and 1w, as a special consequence of more effective hydrogen bonds between ligand and amino acid residues in the enzyme active site. For instance, Fig. 3 compares the main hydrogen bond interactions experienced by the bioconformation 5FX and the optimized conformers 1g and 1w, in order to show the probable cause of conformer stabilization in the biological environment.

Table 4 Intermolecular (ligand–protein) interactions (kcal mol⁻¹) for the main optimized conformers of DFX, obtained by docking studies^a

Conformer	Eligand–protein	Conformer	Eligand–protein
a The interaction energy for the remaining optimized conformers were all less stabilizing than −78.8 kcal mol^−1 (the Eligand–protein for 5FX).
1g	−31.1	1w	−39.8
2g	9.1	2w	−33.7
3g	−76.3	3w	−76.0
4g	8.2	4w	−35.6
5g	−48.0	5w	9.4
6g	−33.2	6w	−72.7
7g	−74.4	7w	−60.7
8g	−10.8	8w	−39.2
9g	−39.2	9w	−60.7
10g	−55.0	10w	−4.7
11g	−53.9	11w	9.3
12g	−64.0	12w	−33.6
13g	−67.8	13w	−42.0
14g	−31.1	14w	−75.4
15g	−45.9	15w	−10.0
		16w	−47.7
		17w	−2.8
		18w	−73.2
		19w	−73.2
		20w	0.6

---

Fig. 3 Conformers 5FX, 1g and 1w docked inside the active site of xylulokinase. Hydrogen bonds with amino acid residues are shown, predominantly in 5FX.

The stronger effect of intermolecular ligand–protein interactions over intramolecular hydrogen bonds and hyperconjugation as driving force of the conformer stabilization in a biological medium suggests that conformations obtained in an enzyme-free environment should be used with caution in conformation-dependent QSAR methods. Most 3D-QSAR methods use conformational screening and structure alignment rules that do not consider the biological target. Consequently, the chemical–biological interpretation obtained from molecular descriptors derived from optimized conformations (those not consistent with the bioconformation) can be inappropriate and inaccurate.

Conclusions

The electrostatic gauche effect in the free diphenhydramine cation stabilizes a conformation within the O–C–C–N⁺ fragment that is similar to its bioconformation.²³ Non-covalent interactions involving fluorine have shown to modulate conformations and reactivity, thus playing an important role to physical organic chemistry and chemical biology.²⁴ For instance, a fluorine substituent can lead to a change in the preferred molecular conformation, which can have important consequences in a lead-optimization program.²⁵ In addition, the C–F and C–NH₃⁺ bonds in 3-fluoro-GABA accommodate a stabilizing charge–dipole interaction, thus opening up prospects for further conformational studies of GABA with various GABA receptor types.²⁶ However, the most stable conformation of free DFX, which is subjected to a network of cooperative intramolecular hydrogen bonds, in addition to hyperconjugative interactions and low steric/electrostatic effects, does not correspond to the ligand conformation, thus revealing the enzyme induced-fit rather than non-covalent intramolecular interactions as the modulator for the bioactive conformation for this molecule. Consequently, QSAR studies based on geometries optimized in an enzyme-free environment can be of limited utility.

Acknowledgements

Authors thank FAPEMIG (grant number APQ-00383-15), CAPES and CNPq for the financial support of this research, as well as by the scholarships (to M. C. G and J. M. S.) and fellowships (to E. F. F. C., T. C. R. and M. P. F.).

References

1. R. D. Bunker, E. M. M. Bulloch, J. M. J. Dickson, K. M. Loomes and E. N. Baker, J. Biol. Chem., 2013, 288, 1643.
2. M. P. Freitas, M. Bühl, D. O'Hagan, R. A. Cormanich and C. F. Tormena, J. Phys. Chem. A, 2012, 116, 1677.
3. L. A. F. Andrade, J. M. Silla, S. L. Stephens, K. Marat, E. F. F. da Cunha, T. C. Ramalho, J. van Wijngaarden and M. P. Freitas, J. Phys. Chem. A, 2015, 119, 10735.
4. M. C. Guimarães, M. H. Duarte, J. M. Silla and M. P. Freitas, Beilstein J. Org. Chem., 2016, 12, 760–768.
5. G. Gilli and P. Gilli, The Nature of the Hydrogen Bond, Oxford University Press, Oxford, 2009.
6. N. H. Evans and P. D. Beer, Angew. Chem., Int. Ed. Engl., 2014, 53, 11716.
7. N. Busschaert, C. Caltagirone, W. von Rossom and P. A. Gale, Chem. Rev., 2015, 115, 8038.
8. H. Kubinyi, Hydrogen Bonding: The Last Mistery in Drug Design?, in Pharmacokinetic Optimization in Drug Research: Biological, Physicochemical, and Computational Strategies, ed. B. Testa, H. van de Waterbeemd, G. Folkers and R. Guy, Verlag Helvetica Chimica Acta, Zürich, 2001.
9. N. S. Nudelman and C. E. S. Alvaro, J. Phys. Org. Chem., 2011, 24, 1067.
10. J. M. Silla, R. A. Cormanich, R. Rittner and M. P. Freitas, J. Phys. Chem. A, 2013, 117, 1659.
11. L. A. F. Andrade, J. M. Silla, C. J. Duarte, R. Rittner and M. P. Freitas, Org. Biomol. Chem., 2013, 11, 6766.
12. R. A. Cormanich, M. P. Freitas, C. F. Tormena and R. Rittner, RSC Adv., 2012, 2, 4169.
13. R. A. Cormanich, R. Rittner, M. P. Freitas and M. Bühl, Phys. Chem. Chem. Phys., 2014, 16, 19212.
14. A. Lämmermann, I. Szatmári, F. Fülöp and E. Kleinpeter, J. Phys. Chem. A, 2009, 113, 6197.
15. M. P. Freitas and T. C. Ramalho, Cienc. Agrotecnol., 2013, 37, 485.
16. J.-D. Chai and M. Head-Gordon, Phys. Chem. Chem. Phys., 2008, 10, 6615.
17. K. Raghavachari, J. S. Binkley, R. Seeger and J. A. Pople, J. Chem. Phys., 1980, 72, 650.
18. C. Amovilli, V. Barone, R. Cammi, E. Cancès, M. Cossi, B. Menucci, C. S. Pomelli and J. Tomasi, Adv. Quantum Chem., 1998, 32, 227.
19. F. Weinhold and C. R. Landis, Discovering Chemistry with Natural Bond Orbitals, Wiley, Hoboken, 2012.
20. R. F. W. Bader, Atoms in Molecules: A Quantum Theory, Oxford University Press, Oxford, 1994.
21. R. Thomsen and M. H. Christensen, J. Med. Chem., 2006, 49, 3315.
22. S. J. Grabowski, Chem. Rev., 2011, 111, 2597.
23. F. M. P. de Rezende, L. A. F. Andrade and M. P. Freitas, J. Mol. Struct., 2015, 1093, 8.
24. Y. P. Rey, C. Thiehoff and R. Gilmour, Isr. J. Chem., 2016 DOI:10.1002/ijch.201600038.
25. H.-J. Böhm, D. Banner, S. Bendels, M. Kansy, B. Kuhn, K. Müller, U. Obst-Sander and M. Stahl, ChemBioChem, 2004, 5, 637.
26. M. D. Clift, H. Ji, G. P. Deniau, D. O'Hagan and R. B. Silverman, Biochemistry, 2007, 46, 13819.

---

Footnote

† Electronic supplementary information (ESI) available: Tables with energies and geometric parameters, and standard coordinates. See DOI: 10.1039/c6ra23423b

---