Synthesis of a Pyridone-Based Phthalimide Fleximer and Its Characterization and Supramolecular Property Evaluation.

In this study, a novel pyridone-based phthalimide fleximer, that is, ethyl 5-cyano-6-(3-(1,3-dioxoisoindolin-2-yl)propoxy)-4-(3-methoxyphenyl)-2-methylnicotinate, was synthesized, and its structure was established by the single-crystal X-ray diffraction method. The supramolecular self-assembly of the titled compound through noncovalent interactions was then investigated thoroughly. The titled compound crystallized with two symmetry-independent molecules (A and B, Z' = 2). In agreement with experimental observations, our density functional theory calculations also showed that the titled compound has a flexible motif and can occur in various conformations, including molecules A and B. The investigation of the supramolecular framework revealed that the molecules are notably bound by the nonclassical C-H···O and C-H···N hydrogen bonds and C-H···π interactions. Hirshfeld surface analysis was carried out to quantify the various intermolecular interactions. The dual anti-inflammatory activity of the tilted compound was also explored by molecular docking in the active sites of 5-LOX and COX-2 receptors, which revealed good binding affinities of -9.0 and -8.6 kcal/mol, respectively.

Introduction

In crystal engineering and supramolecular chemistry, noncovalent interactions play a pivotal part in making three-dimensional (3D) structures. Although there was a greater interest in developing robust and novel molecules with strong covalent bonds between the atoms, the study of non-covalent interactions is now a distinct field and currently trending. Molecules can interact with each other in various ways. Thus, understanding intermolecular interactions is important in crystal engineering and supramolecular chemistry.[1,2] It also plays an important role in other areas such as biochemistry, protein folding, drug design, and material science.[1] Moreover, understanding molecular recognition in biological systems requires an understanding of noncovalent interactions.[3−5] The self-assembly of molecules through noncovalent interactions has resulted in remarkable supramolecular architectures with a wide range of applications.[6,7] For instance, the host–guest interaction of cyclodextrins with drugs allows cyclodextrins to encapsulate drug molecules and deliver them to specific targets.

Furthermore, supramolecules have been used for developing sensors, catalysis, metal extraction, and data storage and processing. Generally, noncovalent interactions refer to hydrogen bonds;[8−11] halogen bonds;[12,13] ionic bonds;[14−16] and weak interactions such as van der Waals, hydrophobic,[17,18] aromatic π–stacking interactions, and so forth.[19,20] These powerful driving forces lead to diverse interactions between molecules such as hydrogen bonds of the types O–H···O, N–H···O, C–H···N, and C–H···O and other weaker interactions of the types I···I, O···I, N···Cl, C···H, C···C, and so forth.[21] Furthermore, C–H···lp (lone pair), C–H···π, and π···π interactions also act as cohesive forces in the supramolecular assembly of molecules.[3,22] Understanding these forces is also important for solid-state materials with desirable properties in supramolecular chemistry.[3,23]

Both 2-pyridone and phthalimide are N-heterocycle derivatives that have a diverse set of biological properties. N-Heterocycles are, in fact, one of the most common nuclei found in the structure of many FDA-approved drugs and bioactive natural products.[24] For instance, the structural motif of several 2-pyridone derivatives has potent biological actions in antitumor,[25] antiproliferative,[26] antibacterial,[27] SARS-CoV-2 main protease inhibitor,[28] antihepatitis B virus,[29] and analgesic activities.[30] The phthalimide motif is also one of the prominent structures often incorporated in the preparation of biologically active compounds. The phthalimide subunit is present in many biologically active compounds as androgen receptor antagonists[31] and has antimicrobial,[32] anticonvulsant,[33] antitumor,[34] hypoglycaemic,[35] anxiolytic,[36] and anti-HIV-1 activities.[37]

Moreover, several 2-pyridone and phthalimide derivatives have good anti-inflammatory activity,[38−40] and they have been reported as a feature of various new nonsteroidal anti-inflammatory drugs.[41−43] For instance, compounds 1 and 2 (Scheme ) are highly selective and potent inhibitors of COX-2. Compound 1 is a 2-pyridone derivative that has been reported to selectively inhibit the activity of the COX-2 enzyme at a concentration of 1.95 μM, which is comparable to the drug rofecoxib.[44] However, compound 2 is a phthalimide-based derivative with high inhibition activity (IC50 = 0.18 μM) against COX-1/COX-2 enzymes.[45] Moreover, it is highly selective toward COX-2 (SI = 668), comparable to the selectivity of the drug celecoxib. Furthermore, compound 2 has a greater anti-inflammatory activity (ED50 = 54.0 mg/kg) than diclofenac (ED50 = 114 mg/kg).

Scheme 1

Representative Examples of the Selective COX-2 Inhibitors (1 and 2) and the Designed Compound Based on the Molecular Hybrids of 1 and 2

Encouraged by these findings, we have explored the development of molecular hybrids of 2-pyridone and phthalimide motifs to design and synthesize novel COX-2 and 5-LOX inhibitors. The design strategy also involved the incorporation of flexibility by adding a propyl linker to connect 2-pyridone and phthalimide pharmacophores. This flexible linker will give greater degrees of freedom to the titled compound to orient itself in the best possible conformation with the least energy and high binding affinity.

Therefore, herein, we report the synthesis and investigation of intramolecular interactions in the supramolecular self-assembly of a dihydropyrimidinone-based phthalimide fleximer, that is, ethyl 5-cyano-6-(3-(1,3-dioxoisoindolin-2-yl)propoxy)-4-(3-methoxyphenyl)-2-methylnicotinate. The titled compound crystallizes with two symmetry-independent molecules in the asymmetric unit, that is, Z′ = 2. Hirshfeld surfaces were calculated for both the symmetry-independent molecules to study the intermolecular contacts between the molecules. The energy framework of the molecules was calculated to understand the nature of intermolecular forces in the self-assembly of the molecules. The relative stability of the structures of molecules (A and B) and their optimized conformers is studied using density functional theory (DFT). A molecular docking study was also performed to understand the various noncovalent interactions between the titled compound and the residues in the active sites of 5-LOX and COX-2 proteins.

Results and Discussion

X-ray Crystallography Investigation

The titled compound ethyl 5-cyano-6-(3-(1,3-dioxoisoindolin-2-yl)propoxy)-4-(3-methoxyphenyl)-2-methylnicotinate crystallizes with two symmetry-independent molecules (A and B) in the asymmetric unit, that is, Z′ = 2 (Figure ). Crystallographic details of the titled compound are shown in Table . It crystallizes in the triclinic crystal system of the P1̅ space group with four molecules per unit cell. The carbon skeletons of molecules A and B are virtually superimposable. Although the molecules are structurally the same, there is a large difference between the two molecules in the orientation of atoms in space. These differences can be visualized from the overlay of the two structures, which have a root-mean-square deviation (rmsd) value of 4.14 Å (Figure ). Moreover, there are slight differences at the crystal packing level where each of the two molecules, A and B, forms its two-dimensional (2D) hydrogen-bonded network.

Figure 1

showing an asymmetric unit of compounds A and B with 30% ellipsoid probability. Hydrogens are omitted for clarity.

Table 1

Crystal Data and Structure Refinement Parameters of the Titled Compound

CCDC number	2062952
empirical formula	C28H25N3O6
formula weight	499.51
temperature (K)	296
crystal system	triclinic
space group	P1̅
a (Å)	8.296(8)
b (Å)	12.579(12)
c (Å)	23.36(2)
α (deg)	87.172(10)
β (deg)	83.908(10)
γ (deg)	89.644(11)
volume (Å3)	2421(4)
Z	4
ρ (g/cm3)	1.370
μ (mm–1)	0.098
F(000)	1048.0
crystal size (mm3)	0.22 × 0.18 × 0.14
radiation	Mo Kα (λ = 0.71073)
2Θ range for data collection (deg)	0.878–26.158
reflections collected	30 582
independent reflections	9376
data/restraints/parameters	9376/3/673
goodness-of-fit on F2	1.028
final R indexes [I ≥ 2σ (I)]	R1 = 0.1183, wR2 = 0.3168
final R indexes [all data]	R1 = 0.1186, wR2 = 0.3935
largest diff. peak/hole/e Å–3	0.60/–0.63

Figure 2

Overlay diagram of molecules A (green) and B (gray).

The C–N bond lengths, that is, C11–N2 and C12–N2 (1.313 and 1.343 Å) and C40–N5 and C39–N5 (1.340 and 1.318 Å), are in good agreement with the values of other 2-pyridone derivatives.[46−48] The pyridone ring of both molecules A and B (N2, C8, C9, C11, C12, and C14 in molecule A and N5, C36, C37, C39, C40, and C42 in molecule B) is almost planar as the deviation of these atoms from the mean plane of the ring is very low. Moreover, the sum of the angles about those atoms is almost 360°. Furthermore, the atoms attached directly to the pyridone ring of both the molecules, that is, O4, C10, C7, C15, and C13 in A and O10, C38, C35, C43, and C41 in B, are almost coplanar with the pyridone ring. Interestingly, the atoms C18 and C19 in A and C46 and C47 in B of the propyl chain are also almost coplanar with the pyridone ring with just a slight deviation from the plane by 0.035 and 0.209 Å, respectively, in A and 0.022 and 0.175 Å, respectively, in B. The torsional angles O4–C11–C9–C8 in A and O10–C39–C37–C36 in B are −173.47° and 172.65°, respectively. Likewise, the two-ring systems of the phthalimide ring in both molecules A and B (N3 and C21–C28 in A and N6 and C49–C56 in B) are planar. The phthalimide and pyridone rings of A and B are linked through a propyl chain C18–C19–C20 and C46–C47–C48. The dihedral angle between the planes of the phthalimide ring and the pyridone ring of A and B is 75.49° and 74.70°, respectively.

Supramolecular Framework

The titled molecule has no strong hydrogen bond donors. However, it has good hydrogen bond acceptors such as N and O. Because of this, classical hydrogen bonds are not observed other than nonclassical hydrogen bond interactions with weak C–H donors. Nevertheless, the C–H···O and C–H···N interactions alone played a crucial role in forming an extensive supramolecular network. Indeed, the two symmetry-independent molecules (A and B) exhibited distinct supramolecular arrangements (Table ). Subsequent levels of architectures were studied to compare conformers A and B. They are distinguished using graph set descriptor R(n), where superscript “a” refers to the number of acceptors, subscript “d” is the number of donors, and “n” is the number of atoms involved in the pattern. The C3–H3A···O2 interactions in molecule A linked the two neighboring homomolecular (--AA--) into a centrosymmetric dimeric structure with R22(20) graph set motifs. These dimers are further interconnected by C23–H23···π3 interactions that propagated into chains (Figures and 5a). The pyridone nitrogen N2 of molecule A does not involve hydrogen bond interactions with any H-bond donor, unlike N5 of molecule B. In molecule B, the C52–H52···N5 and C31–H31A···O8 interactions result in the formation of a centrosymmetric dimer of R22(24) and R22(20) graph set motifs, respectively, forming homomolecular (--BB--) chains (Figure ). Molecules A and B are arranged in layers of alternate zigzag or corrugated patterns. The layers of A and B molecules are linked together by a trifurcated interaction of O12, that is, C23–H23···O12, C18–H18A···O12, and C56 = O12···C21 (Figure ). The C23–H23···O12 and C18–H18A···O12 interactions formed a graph set of R22(11) between molecules A and B. Furthermore, the A and B molecules are interconnected by interactions C47–H47B···O5 and C47–H47A···C25 (Figure ). Moreover, the interchain connections of the centrosymmetric dimers of molecules A and B result in infinite chains due to the extensive C–H···π interactions. In molecules A and B, a similar pattern of C–H···π interactions involving phthalimide, pyridone, and phenyl rings was observed (Figures and 5). For molecule A, the C–H···π interactions with the benzene ring of phthalimide are C23–H23···Cg, C20–H20A···Cg, and C20–H20B···Cg with distances of 3.865, 3.145, and 3.407 Å, respectively. While C24–H24···Cg and C25–H25···Cg interactions with a distance of 3.392 and 3.546 Å, respectively, involved the pyridone ring, C16–H16A···Cg and C16–H16B···Cg with distances of 3.465 and 3.613 Å, respectively, involved the phenyl ring (where Cg is the centroid). Similarly, in molecule B, the C–H···π interactions with the benzene ring of phthalimide are C51–H51···Cg, C48–H48A···Cg, and C48–H48B···Cg with distances of 3.840, 3.152, and 3.376 Å, respectively. In contrast, those involving the pyridone rings are C52–H52···Cg and C53–H53···Cg interactions with a distance of 3.308 and 3.438 Å, respectively (Figure ). The C44–H44A···Cg and C44–H44B···Cg interactions involving the phenyl ring in B have the distances of 3.635 and 3.461 Å, respectively.

Table 2

Hydrogen Bonds and Other Intermolecular Interactions of the Titled Compounda

intermolecular interactions	D–H [Å]	D–H···A [Å]	D···A [Å]	D–H···A [deg]
C47–H47B···O5(a)	0.970	2.649	3.378(9)	132.16
C23–H23···O12(a)	0.930	2.565	3.365(9)	144.44
C31–H31A···O8(b)	0.961	2.519	3.350(1)	144.80
C33–H33···C9(c)	0.930	2.874	3.694(1)	147.77
C45–H45C···O1(b)	0.960	2.694	3.588(1)	155.17
C19–H19A···O11(c)	1.209	2.601	3.507	130.35
C41–H41B···O3(d)	0.960	2.698	3.532(9)	145.55
C52–H52···N5(b)	0.930	2.741	3.586(1)	151.53
C46–H46B···O6(e)	0.970	2.469	3.340(8)	149.35
C47–H47A···C25(f)	0.971	2.861	3.572(1)	130.82
C56–O12···C21(f)	1.178(8)	3.199(9)	4.371(1)	174.95
C18–H18A···O12(f)	0.970	2.451	3.312(9)	147.77
C51–H51···O6(c)	0.929	2.539	3.344(9)	145.03
C3–H3A···O2(e)	0.959	2.699	3.632	164.44
Intramolecular Interactions
C48–H48B···O11	0.970	2.874	3.351	111.37
C13–H13A···O2	0.960	2.425	3.091	126.25
C20–H20A···O4	0.970	2.462	2.845	103.19
C18–H18B···N2	0.971	2.630	2.711	84.26
C20–H20A···N1	0.970	3.613	4.562	166.25
C17–H17B···π (C1, C2, C4, C5, C6, C7)	0.970	3.207
C17–H17C···π (C1, C2, C4, C5, C6, C7)	0.970	3.785
O3···π (C1, C2, C4, C5, C6, C7)		3.613
C45–H45A···π (C1, C2, C4, C5, C6, C7)	0.960	3.978
C45–H45B···π (C1, C2, C4, C5, C6, C7)	0.960	3.046
O9···π (C29, C30, C32, C33, C34, C35)		3.577

Symmetry code: (a) x, y, z; (b) −x, 2 – y, −z; (c) x, 1 + y, z; (d) −1 + x, y, z; (e) −x, 1 – y, 1 – z; (f) 1 – x, 1 – y, 1 – z.

Figure 4

Intermolecular connections in molecules A and B form chains parallel to the crystallographic c-direction.

Figure 5

C–H···π interactions in molecules (a) A and (b) B.

Figure 3

(a) Packing arrangements of A (orange) and B (blue) and (b) intermolecular interactions between A and B and formation of the R22(11) graph set.

Topology of the Crystal Structures of A and B

The topological parameter of the crystal structure is done by using ToposPro software.[49] Geometrical and topological approaches are the two types of ToposPro methods. The first type is implemented in DiAn and IsoCryst and includes routine geometrical computations (distances, angles, and rms planes) and crystal structure visualization. The second type consists of various methods for examining the connectivity features of the entire crystal space. The programs AutoCN and ADS contain the majority of topological operations. The program AutoCN calculates the adjacency matrix, which gives information about connectivity.

We observed no classical hydrogen bond and specific bonds in the adjacency matrix and IsoCryst in crystal packing, as shown in Figure . The topological analysis of the molecular packing reveals the 14-coordinated underlying of gpu-x 3D topological type (Figure ). This type of topology was found for 13 681 molecular structures. This standard representation analysis was done when complex molecules were considered as nodes, and van der Waals bonds were considered edges. The gpu-x net contains only one node that means both molecules A and B play a topologically similar role in the structure.

Figure 6

Crystal structure visualization in IsoCryst.

Figure 7

Original and underlying nets for the crystal structure. The corresponding underlying nets have a uninodal 14-coordinated gpu-x topology (ZA = C28H25N3O6).

Hirshfeld Surface

Hirshfeld surface analysis is an effective tool for studying and quantifying several intermolecular interactions in supramolecular compounds. These studies corroborate the intermolecular interactions in the crystal structure discussed in the previous section.

Hirshfeld’s surface of each of the two symmetry-independent molecules, A and B, is slightly dissimilar as molecules A and B have slight variation in their interatomic contacts and the relative percentage contribution of those contacts. Typically, red, white, and blue colors on the Hirshfeld surface projected over dnorm correspond to interatomic interactions that are short, equal to, and greater than the van der Waals interatomic distance, respectively. Accordingly, the small red spots on the Hirshfeld surfaces of molecules A and B mapped over a dnorm range from −0.5152 to 1.1464 Å indicate the short interactions involving nonclassical hydrogen bonds, that is, C–H···O and C–H···N interactions (Figure ). The 2D fingerprint plots depicting H···H, H···O/O···H, C···H/H···C, and H···N/N···H interactions are presented in Figure . In Figure , the H···H, O···H, and N···H interactions appear as pairs of distinct spikes where the upper (di < de) and lower (di > de) spike represents the donor and acceptor characters of the atoms, respectively. A pair of distinctly pointed spikes at (di, de) ≈ (1.3, 1.1 Å) and (di, de) ≈ (1.45, 1.15 Å) represents the O···H/H···O and N···H/H···N interactions, respectively, in molecule A, whereas (di, de) ≈ (1.3, 1.1 Å) and (di, de) ≈ (1.5, 1.1 Å) represent the O···H/H···O and N···H/H···N interactions, respectively, in molecule B.

Figure 8

Hirshfeld surface of the dnorm property and 2D fingerprint plots of A and B.

Hirshfeld surfaces mapped over the shape-index in a range from −1.0 to 1.0 Å for molecules A and B have an intense red π-hole, which indicates the possible C–H···π interactions in the supramolecular assembly (Figure ). The curvedness of molecules A and B does not indicate the possible presence of π···π interactions as the flat regions on the Hirshfeld surface mapped over curvedness are unapparent. The C–H···π interactions are also evident from the 2D fingerprint plot as the C–H···π interactions decompose into C···H contacts and appear as a pair of characteristic wings (Figure ). The C···H contacts or C–H···π interactions of molecules A and B contribute about 17.6 and 19.0% of the total Hirshfeld surfaces, respectively. In molecules A and B, H···H interactions (de + di ≈ 2.2 Å) contribute to the Hirshfeld surface, amounting to 45.1 and 44%, respectively (Figure ). The other significant contributors to the Hirshfeld surface are O···H, C···H, and N···H interactions (de + di ≈ 2.4 Å, de + di ≈ 2.75 Å, and de + di ≈ 2.6 Å, respectively) (Figure ).

Figure 9

Shape index and curvedness of A and B.

Figure 10

Percentage contribution of various intermolecular interactions of A and B to the Hirshfeld surface.

The Hirshfeld surface also gave information about the global shape, that is, globularity (G) and asphericity (Ω). Globularity (G)[50] is defined as the ratio of the surface area of a sphere to the surface area of a Hirshfeld’s surface, having the same volume of the sphere (Ssphere/SHS). The value of G is unity for a sphere, and the value becomes lesser than unity as the molecular surface is more organized. Asphericity (Ω)[51] is a measure of the anisotropy of the molecule. The asphericity of isotropic and prolate objects is assumed to be 0 and 1, respectively. Globularity (G) values of molecules A (0.663) and B (0.685) are lesser than unity. This indicates that molecule B has a higher deviation from the spherical surface. Furthermore, the asphericity (Ω) value of molecules A (0.238) and B (0.265) indicates that molecule B has a higher deviation from isotropy. Furthermore, the void domain calculations revealed that the unit cell of the titled crystal compound has a void volume of 265.24 Å3 and a void surface area of 926.24 Å2, which is 10.9% of the unit cell volume (Figure S3). This indicates that the crystal molecules are closely packed, and no large cavity is present.

The enrichment ratio (ER) is calculated from the interatomic contacts between pairs of interacting atoms (X, Y) derived from the Hirshfeld surface analysis.[52] This quantity allows prediction of the tendency of two atoms, X and Y, to form intermolecular interactions. Generally, the chemical elements (X, Y) with a greater ER value than unity indicate a high propensity to form intermolecular contacts. In comparison, pairs with lower ER values than unity tend to avoid intermolecular contacts. Therefore, the ER is useful to highlight the favorable contacts that are important driving forces in the supramolecular assembly of molecules. Table shows that H atoms in both molecules A and B generate more than 69% of the total molecular surface, while the contributions of N are the lowest, being only about 6.1–6.4% in both the molecules.

Table 3

ERs of Molecules A and B

A					B
actual contacts (%)					actual contacts (%)
atoms	H	C	N	O	atoms	H	C	N	O
H	45.1				H	44		contacts	(%)
C	17.6	1.5			C	19.0	1.3
N	9.7	0.9	0.3		N	9.9	0.9	0.3
O	21.7	2.1	1.0	0.1	O	21.1	2.3	0.9	0.1
surface %	69.6	11.8	6.1	12.5	surface %	69.0	12.4	6.4	12.3

Moreover, the ER value of O···H (EOH = 1.24, 1.24) and N···H (ENH = 1.10, 1.12) is greater than unity in both molecules A and B. This indicates that O···H and N···H interactions are favorable and act as an important contributor to the stability of the crystallized molecules. Although the H···H contacts are the most abundant interactions (44–45.1%) in both molecules, the ER of the H···H interaction (EHH = 0.93, 0.92) is slightly impoverished and disfavored. The C···H interaction in molecule A is slightly disfavored and only slightly enriched (ECH = 1.07). However, the C···H interactions in molecule B are favorable (ECH = 1.11). This corroborates with the slightly lowered energy between molecules exhibiting C–H···π interactions in B, as discussed later. Moreover, the ERs of C···C, C···N, C···O, N···N, N···O, and O···O interactions are impoverished with low ER values and are disfavored.

Energy Framework of Molecules A and B

Energy framework calculations were performed in CrystalExplorer 17.5 using the B3LYP/6-31G (d,p) functional basis set to better understand the nature of different intermolecular interaction energies in the crystal packing of molecules A and B. The interaction energies for packing molecules A and B were calculated by generating a cluster of molecules within a radius of 3.8 Å around the selected molecule. The interaction energies are calculated with the scale factors of k_disp = 0.871, k_ele = 1.057, k_pol = 0.74, and k_rep = 0.618. Tables S5 and S6 summarizes the net interaction energies and their component energies, such as electrostatic (Eele), dispersion (Edisp), polarization (Epol), and repulsion energies (Erep). The net interaction energies of molecules A and B are −160.8 and −161.9 kJ/mol, respectively. Among these energies, dispersion forces play a dominant role over other forces. They serve as the main pillar in stabilizing the crystal packing, while the polarization forces have the least contribution. The total dispersion interaction energies of molecules A and B are −191.1 and −190.3 kJ/mol, respectively, while the polarization forces are −18.0 and −19.0 kJ/mol, respectively. The higher role of dispersion forces in the supramolecular assembly of the titled compound is predominantly due to the extensive contribution of H···H and C···H interactions to a great degree, which are about 44–45.1 and 17.6–19%, respectively. Consequently, the interaction energies between the atoms that exhibited the C–H···π interactions in molecules A and B are the lowest. For instance, in A, the energy between the molecules having the C–H···π interactions (symmetry operation: x, y, z; 1 + x, y, z and 1 – x, 1 – y, 1 – z) is about −99.2 kJ/mol. In B (symmetry operation: x, y, z; 1 + x, y, z and 1 – x, 2 – y, 1 – z), this energy is about −100 kJ/mol, which is slightly lower than that in molecule A. Moreover, in molecule A, the graph set R22(20) due to the C3–H3A···O2 interactions corresponds to an energy of −24.7 kJ/mol (Edis = −31.4, Eelec = −9.4, Epol = −2.2 Erep = 22.9). In molecule B, the C52–H52···N5 and C31–H31A···O8 interactions that formed graph sets of R22(24) and R22(20) corresponds to an energy of −42.5 kJ/mol (Edis = −56.5, Eelec = −10.1, Epol = −3.4 Erep = 32.2) and −20.2 kJ/mol (Edis = −25.3, Eelec = −3.3, Epol = −3.4, Erep = 23.5), respectively. The energy framework diagram of the Coulomb, dispersion, and total energies is given in Figure .

Figure 11

Depiction of the energy framework diagram for the Coulomb, dispersion, and total energies of A and B for a cluster of molecules with a radius of around 3.8 Å from the selected molecule. The intermolecular interaction energies between the pairs of molecules are shown as tubes. The radius of the tube is proportional to the relative strength of the intermolecular interaction.

DFT Calculations

Interestingly, two conformers of the titled compound are observed in the crystal structure. The overlay diagram of these two conformers (A and B) shows that they have entirely different structures (Figure a). Therefore, to explore the other possible conformers of the titled compound and their stability, eight conformers of the titled compound were first generated using the MMFF94 force field and the systematic rotor search method as implemented in the Avogadro program. The crystal structure conformers (A and B) and the eight other conformers obtained using the systematic rotor search method were fully optimized at the M06-2X/6-31+G(d,p) level of DFT in the gas phase.[53] The Gaussian 09 software suite[54] was used for all DFT calculations. It is found that after geometry optimization, the crystal conformers A and B have the same zero-point energy (ZPE)-corrected total energies (−1696.585818 hartree), that is, they are equally stable. The ZPE-corrected total energies (expressed in kcal/mol) of different conformers relative to the optimized crystal conformer A/B are shown in Figure . It is, in general, believed that the motif of the crystal structure can change appreciably after complete geometry optimization. However, it is found that crystal conformers remain almost unchanged even after complete optimization at the M06-2X/6-31+G(d,p) level, and the rmsd is ∼0.01. The relative energies of the various conformers obtained by DFT calculations show that conformers A/B, C, and G are almost equally stable, with the difference between their energies being only 0.48 kcal/mol, which is less than the thermal energy. Conformers D and F are more stable than conformer A by 3.46 and 2.55 kcal/mol, respectively (Figure ). The energy difference between the most and least stable conformers is ∼6.5 kcal/mol (Figure ). This indicates that all the conformers can occur in the crystal structure. The overlay diagram of conformers shows that structures of conformers C and G are almost similar to those of the crystal conformers A and B, respectively (Figure b). It is further evident from the overlay diagram that conformer F is almost a mirror image of conformer D, and their enhanced stability relative to the optimized crystal geometry is due to the more pronounced π···π interaction (Figure c). A close examination of the overlay diagram of conformers shows that the titled compound has a very flexible motif; therefore; this molecule can crystalize in various conformations.

Figure 12

Overlay diagrams of (a) optimized crystal conformers [A (red) and B (pink)]; (b) crystal conformers [A (red) and B (pink)], conformer C (blue), and conformer G (purple); (c) crystal conformers [A (red) and B (pink)], conformer D (green), and conformer F (cyan); and (d) crystal conformers [A (red), B (pink)], conformers E (black), H (yellow), I (orange), and J (gray).

Figure 13

Plot of relative energies of different conformers.

Molecular Docking

The combined inhibition of the COX-2 and 5-LOX pathways has been a well-known pharmaceutical technique for developing a more effective anti-inflammatory medication with fewer and less severe side effects. It can be accomplished either by administering a combination of COX-2 and 5-LOX inhibitors or solely by using a single compound with dual action.[55,56] However, the preparation of a single compound with dual activity has attracted considerable attention due to its lesser side effects.[56] To estimate the dual anti-inflammatory activity of the titled compound, a molecular docking technique was employed to investigate its binding affinity in the active sites of 5-LOX and COX-2.

The molecular docking studies with COX-2 and 5-LOX enzymes yielded almost identical binding affinities and binding modes for conformers A and B (the overlay diagram of A and B in the active site is shown in Figures S4 and S5, respectively). Conformers A and B exhibited good binding affinity toward 5-LOX and COX-2 enzymes. Their binding affinities with COX-2 are −8.8 and −9.0 kcal/mol, respectively, while both A and B have the same binding affinity of −9.1 kcal/mol with the 5-LOX enzyme. The binding affinities of conformers A and B are comparable with the binding affinities of the native ligands of 5-LOX (−9.7 kcal/mol) and COX-2 (−9.9 kcal/mol), respectively. Conformers A and B occupied the active site of 5-LOX with two hydrogen bond interactions, one between the phthalimide carbonyl oxygen and residue Arg68 and the other between oxygen from the methoxy group and Arg101 (Figure ). The hydrogen bond distances with Arg68 and Arg101 residues are 2.373 and 2.139 Å, respectively, while the diaphysial axis-metacarpal head angles (DHAs) are 126.073° and 140.866°, respectively. Moreover, the titled compound is stabilized in the active site by the π···cation interaction between the pyridone ring and residue Arg101. The additional stability to the titled compound was reinforced by π···alkyl interactions with residues Leu66, Val110, His130, and Val107.

Figure 14

Binding mode of the titled compound in the active site of 5-LOX and COX-2.

In the active site of COX-2, one of the phthalimide oxygens acts as a bifurcated acceptor and forms two hydrogen bond interactions with residue Arg120 with a distance of 2.510 and 2.20 Å and DHA bond angles of 124.522 and 128.213°, respectively (Figure ). The fused phthalimide rings and the phenyl ring formed π···σ interactions with residues Val116, Leu93, and Leu352. Furthermore, the hydrophobic interaction with residues Ala527, Val349, Val523, and Tyr355 lowered the energy of the titled compound in the active site of COX-2 and gave it additional stability.

Conclusions

A novel pyrimidinone-based phthalimide fleximer, that is, ethyl 5-cyano-6-(3-(1,3-dioxoisoindolin-2-yl)propoxy)-4-(3-methoxyphenyl)-2-methylnicotinate, was synthesized, and single-crystal X-ray diffraction (SCXRD) was used to determine its crystal structure. The titled compound crystallizes in the P1̅ space group with two symmetry-independent molecules in the asymmetric unit. According to our investigation of the X-ray results of the crystal structures, the molecules are held together in the supramolecular framework by weak nonclassical hydrogen bond interactions C–H···O, C–H···N, and C–H···π. Analysis of the Hirshfeld surface reveals that H···H contacts are the most abundant interactions in the crystal packing. The ER derived from the Hirshfeld surface revealed that O···H and N···H interactions are the most favored and an important contributor to the stability of molecules in the supramolecular self-assembly. However, the interatomic contacts with impoverished ERs such as C···C, C···N, C···O, N···N, N···O, and O···O interactions are less significant and disfavored. The energy framework calculations indicate that dispersion forces play a dominant role over other forces in the crystal packing, while the polarization forces have the least contribution. The molecular docking study reveals an excellent binding affinity of the titled compound toward 5-LOX (−9.0 kcal/mol) and COX-2 (−8.9 kcal/mol) receptors. The DFT calculations reveal that the titled compound has a flexible motif and can crystallize in various conformers, including molecules A and B, observed in the crystal structure. Both conformers A and B have similar extended structures found in the crystal structure study but differ in the torsion angle of the methylene linker as observed in Hirshfeld analysis, docking, and DFT studies. The plane of the rings of both arms in both conformers A and B showed similarity in crystal analysis and surface interactions, controlled by intermolecular interactions. The flexible nature of the compound increases the binding affinity by providing better availability at different binding sites of the 5-LOX and COX-2 receptors. Overall, the study has provided a better understanding of the intermolecular interactions in the supramolecular assembly of the pyridone-based phthalimide fleximer.

Experimental Section

Synthetic Procedure (Scheme )

Compound 9 was synthesized according to the method reported in our previous paper.[48] In a 100 mL round-bottom flask, 2-pyridone derivative 9 (0.0068 mmol) was dissolved in dimethylformamide (DMF), and potassium carbonate (0.0072 mmol) was added and stirred for 20 min. After that, 2-(3-bromopropyl)isoindoline-1,3-dione (11) (0.0068 mol) was added and stirred for 12 h. The progress of the reaction was monitored using thin-layer chromatography (30% EtOAc in hexane). After completion of the reaction, DMF was concentrated under reduced pressure using a rotary evaporator, and the mixture was extracted with EtOAc (50 × 3 mL). The combined organic layers were then dried with anhydrous Na2SO4 and filtered. The crude mixture was purified by SiO2 flash chromatography with 15% EtOAc/hexane to obtain compound 11 (A and B) (Scheme ).

Scheme 2

Synthesis of the Pyridone-Based Phthalimide Fleximer

(a) Piperidine, EtOH, RT, 10 min; (b) I2 (10 mol %), EtOH, reflux 4 h; (c) DDQ, MW 2 min; (d) K2CO3, DMF, RT, 12 h.

Ethyl-5-cyano-4-(3-methoxyphenyl)-2-methyl-6-oxo-1,6-dihydropyridine-3-carboxylate (9)

Yield = 81%. H NMR (300 MHz, CDCl3): δ 0.84–0.89 (3H, t, CH3, J = 7.2 Hz); 2.62 (3H, s, CH3); 3.83 (3H, s, CH3); 3.92–3.99 (2H, q, CH2, J = 7.2 Hz); 6.89–7.03 (2H, m, Ar–H); 7.27 (1H, s, Ar–H); 7.36–7.41 (1H, t, Ar–H, J = 8.1 Hz); 13.67 (1H, s, NH); C NMR (75 MHz, CDCl3): δ 14.2, 19.6, 55.8, 61.8, 109.0, 110.5, 113.5, 115.3, 115.8, 121.2, 129.6, 133.5, 151.4, 160.5, 161.5, 165.0, 169.4.

Ethyl-5-cyano-6-(3-(1,3-dioxoisoindolin-2-yl)propoxy)-4-(3-methoxyphenyl)-2-methylnicotinate (11)

Yield = 62%. Melting point 109–111 °C. H NMR (300 MHz, DMSO): δ 1.29–1.31 (3H, t, CH3, J = 7.2); 2.18–2.22 (2H, m, CH2, J = 6.3); 2.55 (3H, s, CH3); 3.83 (3H, s, CH3); 4.06–4.08 (2H, t, CH2, J = 6.6); 4.22–4.25 (2H, q, CH2, J = 6.9); 4.65–4.68 (2H, t, CH2, J = 7.8); 6.95–6.99 (1H, d, CH, J = 1.8); 7.08–7.11 (1H, d, CH,J = 7.8); 7.36 (1H, s, CH); 7.45–7.48 (1H, t, CH, J = 3.1); 7.78–7.82 (2H, dd, CH, J = 3.5); 7.88–7.92 (2H, dd, CH, J = 3.5). C NMR (75 MHz, DMSO): δ 14.1, 22.0, 26.7, 38.8, 60.9, 65.3, 91.6, 109.5, 114.6, 123.7, 125.5, 129.3, 129.5, 132.2, 134.8, 139.4, 154.5, 164.7, 169.7, 166.0, 167.9. IR (neat) (cm–1) ν: 1087.85 (C–C); 1157.29 (C–O–C); 1558.48 (C=C); 1705 (NC=O); 1785 (C=OOCH2CH3); 2226 (C≡N); 2927 (C–H). Solubility: ethyl acetate, methanol, ethanol, and acetone.

Single-Crystal X-ray Diffraction

The crystals of the titled compound formed under slow evaporation of ethyl acetate were isolated and subjected to single-crystal X-ray diffraction using an Oxford Diffraction Xcalibur CCD using monochromated Mo Kα radiation (λ = 0.71073 Å). SHELXS-97 was used for solving the structures, and refinement was done based on F2 by employing a complete a full-matrix least-squares technique.[57] Mercury software (version 3.1)[58] was used to study and generate the packing of crystals. Table summarizes the title compound’s crystal data and structure refinement details.

Hirshfeld Surface Analysis

The 2D fingerprint plots and Hirshfeld surface were calculated using a CIF file of the titled compound in the Crystal Explorer 17.5 program.[59] The Hirshfeld isosurface is based on “di” and “de” distances, where “di” is the distance to the nearest nucleus internal to the surface and the “de” is the distance from the point to the nearest nucleus external to the surface. The normalized contact distance dnorm can be calculated using di and de as followswhere rivdw and revdw are the van der Waals radii of the atoms.

The dnorm property mapped over a range from −0.5152 to 1.1464 Å is used to interpret and quantify intermolecular interactions. The curvedness and shape index properties were mapped over a range from −0.5152 to 1.1464 Å. The energy framework calculation was performed using the B3LYP/6-31G (d, p) functional basis set to quantify the dispersion, electrostatic, polarization, and repulsion energies by generating a cluster of molecules with a radius of around 3.8 Å from the selected molecule.

DFT Calculation

The Gaussian 09 software suite was used for all DFT calculations with the hybrid meta-GGA functional M06-2X[53] and Pople’s basis set 6-31+G(d,p)[54] to examine the stability of conformers A and B and their different conformers. The M06-2X functional is a reliable function developed by Truhlar’s group for studying the relative stability of different conformers, noncovalent interactions, rate constants, and thermochemistry of various molecular systems.[60−62]

AutoDock Vina[63] was used to carry out molecular docking calculations. The X-ray crystal structures of 5-LOX and COX-2 were retrieved from the Protein Data Bank (RCSB) (PDB id: 6NCF and 5KIR, respectively). Before docking, the protein was prepared by removing cofactors, water molecules, and the cocrystallized native ligand. Subsequently, polar hydrogens and Kollman charges were added to the proteins in AutoDock tools software, and the file was saved in the pdbqt format. The CIF file of the titled compound was converted to PDB format in Mercury software, and it was used for docking without further minimizing the energy of the structure. The native cocrystallized ligand was used for assigning the grid parameters required for docking. The grid parameters employed for docking in 5-LOX were centered at x = 11.277, y = −21.891, z = −18.408; 20 × 20×20, while for docking in COX-2, they were centered at x = 23.287, y = 0.587, and z = 34.435; 20 × 20 × 20. The exhaustiveness parameter was set to eight modes. Lastly, Discovery Studio and PyMOL were used to visualize and analyze the docking poses.