Two N-(2-phenyl­eth­yl)nitro­aniline derivatives as precursors for slow and sustained nitric oxide release agents

Two new precursors to sustained NO-releasing materials have been characterized. The structures are compared with reported analogues, revealing significant differences in mol­ecular conformation, inter­molecular inter­actions, and packing that result from modest changes in functional groups. The structures are also discussed in terms of potential NO-release capability.

Abstract

Notwithstanding its simple structure, the chemistry of nitric oxide (NO) is com­plex. As a radical, NO is highly reactive. NO also has profound effects on the cardiovascular system. In order to regulate NO levels, direct therapeutic inter­ventions include the development of numerous NO donors. Most of these donors release NO in a single high-concentration burst, which is deleterious. N-Nitro­sated secondary amines release NO in a slow, sustained, and rate-tunable manner. Two new precursors to sustained NO-releasing materials have been characterized. N-[2-(3,4-Di­meth­oxy­phen­yl)eth­yl]-2,4-di­nitro­aniline, C₁₆H₁₇N₃O₆, (I), crystallizes with one independent mol­ecule in the asymmetric unit. The adjacent amine and nitro groups form an intra­molecular N—H⋯O hydrogen bond. The anti conformation about the phenyl­ethyl-to-aniline C—N bond leads to the planes of the arene and aniline rings being approximately perpendicular. Mol­ecules are linked into dimers by weak inter­molecular N—H⋯O hydrogen bonds such that each amine H atom participates in a three-center inter­action with two nitro O atoms. The dimers pack so that the arene rings of adjacent mol­ecules are not parallel and π–π inter­actions do not appear to be favored. N-(4-Methyl­sulfonyl-2-nitro­phen­yl)-l-phenyl­alanine, C₁₆H₁₆N₂O₆S, (II), with an optically active center, also crystallizes with one unique mol­ecule in the asymmetric unit. The l enanti­omer was established via the configuration of the starting material and was confirmed by refinement of the Flack parameter. As in (I), there is an intra­molecular N—H⋯O hydrogen bond between adjacent amine and nitro groups. The conformation of the mol­ecule is such that the arene rings display a dihedral angle of ca 60°. Unlike (I), mol­ecules are not linked via inter­molecular N—H⋯O hydrogen bonds. Rather, the carb­oxy­lic acid H atom forms a classic, approximately linear, O—H⋯O hydrogen bond with a sulfone O atom. Pairs of mol­ecules related by twofold rotation axes are linked into dimers by two such inter­actions. The packing pattern features a zigzag arrangement of the arene rings without apparent π–π inter­actions. These structures are compared with reported analogues, revealing significant differences in mol­ecular conformation, inter­molecular inter­actions, and packing that result from modest changes in functional groups. The structures are discussed in terms of potential NO-release capability.

Introduction  

Nitro­gen monoxide, commonly known as nitric oxide (NO), is produced in vivo from l-arginine (Miller & Megson, 2007 ▸). The conversion of l-arginine to NO and citrulline is catalyzed by nitric oxide synthases along with a series of cofactors. Notwithstanding its simple structure, the chemistry of NO is complex. As a radical, NO is highly reactive and it is consumed quickly within a radius of 100 µm. Therefore NO, which has profound effects on the cardiovascular system, must be produced slowly in small fluxes at the inner walls of blood vessels by endothelial cells (Zhou et al., 2006 ▸).

Since the discovery of the vaso-relaxing effects of NO (Cauwels, 2007 ▸), its physiological impact on the cardiovascular system has been the most studied. Endothelial cells, which line the lumen of the blood vessels, produce NO from l-arginine. The conversion is catalyzed by endothelial nitric oxide synthase (eNOS). The lypophilicity, small size, and chemical lability of NO allows for easy passage through cell membranes, without the use of channels or receptors, to neighboring cells (Miller et al., 2004 ▸). The effects of NO on the underlying vascular smooth muscle cells (VSMC) include vasodilation and inhibition of VSMC proliferation, as well as its migration to the endothelial layer (Jeremy et al., 1999 ▸).

The salutary roles of NO in the functioning of the cardiovascular, nervous, respiratory, and immune systems, in addition to its protecting functions of the heart, brain, and kidneys, have been well documented (Jobgen et al., 2006 ▸). On the other hand, high NO concentrations cause apoptosis and other complications, including septic and hemorrhagic shock, multiple sclerosis, neurodegenerative diseases, rheumatoid arthritis, ulcerative colitis, and cancer (Jeremy et al., 1999 ▸; Lundberg et al., 2015 ▸). A wide variety of diseases and disorders have been ascribed to NO malfunctions. These maladies include endothelial dysfunction, hypertension, cardiovascular disease, asthma, pulmonary hypertension, erectile dysfunction, pre-eclampsia, and insulin resistance (Giles, 2006 ▸). The overall effects on the cardiovascular system can lead to further endothelial injury and atherosclerosis (Stasch et al., 2011 ▸).

In order to better regulate NO levels, several recommendations have been made, which include a polyphenol-rich diet (Anter et al., 2004 ▸), moderate alcohol consumption (Lucas et al., 2005 ▸), and regular exercise (Hambrecht et al., 2003 ▸). Direct therapeutic inter­ventions include the development of numerous NO donors. Most of these donors release NO in a single high-concentration burst, which is deleterious (Cai et al., 2005 ▸). We have reported a series of N-nitro­sated secondary amines that release NO in a slow, sustained, and rate-tunable manner (Wang et al., 2009 ▸; Yu et al., 2011 ▸; Curtis et al., 2013 ▸, 2014 ▸; Lagoda et al., 2014 ▸). Furthermore, the released NO has been shown to inhibit the proliferation of human aortic smooth muscle cells, a contributing factor to the progression of atherosclerosis (Yu et al., 2011 ▸; Curtis et al., 2013 ▸).

We report herein the syntheses and X-ray crystal structures of two secondary amines, the precursors to NO donors. These secondary amines, namely N-[2-(3,4-di­meth­oxy­phen­yl)eth­yl]-2,4-di­nitro­aniline, (I), and N-(4-methyl­sulfonyl-2-nitro­phen­yl)-l-phenyl­alanine, (II), differ in their hydro­philic and lipophilic balances (HLBs). We hypothesize that the HLB of an NO donor will play a crucial role in the overall NO release rates of these compounds in an aqueous medium (phosphate-buffered saline solution). Compound (I) was prepared by the reaction of 3,4-di­meth­oxy­phenethyl­amine and an activated aromatic monofluoride, i.e. 2,4-di­nitro­fluoro­benzene. The syn­thesis of (II) was accomplished by the reaction of l-phenyl­alanine with 4-methyl­sulfonyl-2-nitro­fluoro­benzene.

Experimental  

Synthesis and crystallization  

For the synthesis of (I), sodium bicarbonate (0.373 g, 4.44 mmol), 3,4-di­meth­oxy­phenethyl­amine (97%) (0.539 g, 2.88 mmol), and 2,4-di­nitro­fluoro­benzene (0.536 g, 2.88 mmol) were weighed separately in glass vials and transferred to a 50 ml round-bottomed flask equipped with a magnetic stirrer bar. The vials were rinsed with N,N-di­methyl­acetamide (DMAC, 10 ml) and the rinses transferred to the reaction vessel. The reaction mixture was allowed to stir at room temperature for 1 h, after which time the mixture was poured onto a saturated sodium chloride solution (100 ml) to precipitate the crude product. The crude product was extracted with di­chloro­methane (75 ml) and washed twice with deionized water (100 ml). The organic layer was dried over anhydrous magnesium sulfate, followed by gravity filtration. The solvent was evaporated under reduced pressure using a rotary evaporator. The crude product, an orange solid, was recrystallized from a di­chloro­methane–diethyl ether solution (1:1 v/v) as crystals suitable for X-ray diffraction (yield 55%; m.p. 379–380 K). ¹H NMR (300 MHz, CDCl₃): δ 9.10 (d, 1H), 8.57 (s, 1H), 8.24 (m, 1H), 6.84 (m, 4H), 3.89 (s, 3H), 3.86 (s, 3H), 3.64 (q, 2H), 3.02 (t, 2H); ¹³C NMR (300 MHz, CDCl₃): δ 149.29, 148.13, 148.12, 136.01, 130.29, 129.66, 124.54, 120.71, 113.85, 111.68, 111.57, 55.88, 45.10, 34.50; IR (NaCl, ν, cm⁻¹): 3357, 3107, 2934, 1621, 1590, 1517, 1465, 1423, 1335, 1305, 1263, 1238, 1196, 1144, 1084, 1027; MS (m/z) (% base peak): 347 (8), 281 (5), 207 (18), 166 (22), 151 (100), 137 (8), 107 (11), 77 (16).

Compound (II) was prepared by dissolving 4-methyl­sul­fonyl-2-nitro­fluoro­benzene (0.565 g, 2.58 mmol) in tetra­hydro­furan (THF, 15 ml) in a 100 ml round-bottomed flask equipped with a magnetic stirrer bar in the presence of sodium bicarbonate (2.185 g, 24.6 mmol). The reaction mixture turned yellow upon the addition of sodium bicarbonate. l-Phenyl­alanine (0.406 g, 2.46 mmol) was dissolved in deionized water (25 ml) in a 50 ml beaker and the solution was subsequently transferred into the reaction vessel using a funnel. Water (5 ml) was used to wash the funnel. The reaction vessel was fitted with an air condenser and the mixture was allowed to stir at room temperature for 12 h, after which time the THF was removed under reduced pressure using a rotatory evaporator. The resulting aqueous solution was washed with diethyl ether (50 ml) to remove excess starting material. The crude product was precipitated via acidification of the aqueous solution (pH ca 2) using 12 M hydro­chloric acid. The precipitated crude product was extracted using ethyl acetate (50 ml), which was collected and washed with deionized water (50 ml) three times. The organic layer containing the crude product was dried over anhydrous magnesium sulfate, followed by gravity filtration. The solvent was removed using a rotary evaporator to yield a bright yellow solid. Crystals suitable for X-ray diffraction were obtained by recrystallization from a di­chloro­methane–hexane solution (4:1 v/v) (yield 58%; m.p. 443–445 K). ¹H NMR (300 MHz, DMSO-d ₆): δ 8.58 (d, 1H), 8.50 (d, 1H), 7.92 (dd, 1H), 7.27–7.15 (overlapping peaks, 6H), 4.94 (q, 1H), 3.24 (d, 2H), 3.20 (s, 3H); ¹³C NMR (300 MHz, DMSO-d ₆): δ 172.25, 146.61, 136.40, 134.13, 130.82, 129.84, 128.83, 127.75, 127.40, 127.35, 116.67, 56.33, 44.09, 37.04; IR (solid, ν, cm⁻¹): 3379, 3176, 3090, 1733, 1611, 1518, 1367, 1299, 1145; ESI–MS (m/z) calculated for C₁₆H₁₆N₂O₆S [M − H]⁻ 363.38, found 363.03.

Refinement  

Crystal data, data collection and structure refinement details are summarized in Table 1 ▸. H atoms on C atoms were included in calculated positions as riding atoms, with C—H bond lengths of 0.95 (ar­yl), 0.98 (meth­yl), 0.99 (methyl­ene) and 1.00 Å (methine). H atoms on N and O atoms were located from difference Fourier syntheses and were refined isotropically without any restraints (see the N—H and O—H distances in Tables 2 ▸ and 3 ▸). For (II), the Flack x parameter was determined (Table 1 ▸), and the probability P2 = 1.000, G = 0.97 (8), and a Hooft y parameter of 0.02 (3) were determined from an analysis of the Bijvoet differences (Hooft et al., 2008 ▸). We acknowledge that the crystal of (II) was longer than optimal, but the needles were difficult to cut cleanly and the refinement results do not appear to have been adversely affected.

Table 1. Experimental details.

	(I)	(II)
Crystal data
Chemical formula	C16H17N3O6	C16H16N2O6S
M r	347.33	364.38
Crystal system, space group	Monoclinic, P21/c	Monoclinic, C2
Temperature (K)	173	173
a, b, c (Å)	13.0087 (12), 7.2092 (6), 17.0024 (16)	21.260 (3), 5.7264 (6), 14.2762 (16)
β (°)	101.297 (7)	103.456 (7)
V (Å3)	1563.6 (3)	1690.3 (4)
Z	4	4
Radiation type	Mo Kα	Mo Kα
μ (mm−1)	0.12	0.23
Crystal size (mm)	0.30 × 0.23 × 0.11	0.80 × 0.14 × 0.04
Data collection
Diffractometer	Rigaku XtaLAB mini	Rigaku XtaLAB mini
Absorption correction	Multi-scan (REQAB; Rigaku, 1998 ▸)	Multi-scan (REQAB; Rigaku, 1998 ▸)
T min, T max	0.828, 0.987	0.815, 0.991
No. of measured, independent and observed reflections	16236, 3578, 2913 [F 2 > 2.0σ(F 2)]	10478, 4851, 4201 [I > 2σ(I)]
R int	0.030	0.029
(sin θ/λ)max (Å−1)	0.649	0.704
Refinement
R[F 2 > 2σ(F 2)], wR(F 2), S	0.039, 0.102, 1.04	0.039, 0.092, 1.03
No. of reflections	3578	4851
No. of parameters	232	235
No. of restraints	0	1
H-atom treatment	H atoms treated by a mixture of independent and constrained refinement	H atoms treated by a mixture of independent and constrained refinement
Δρmax, Δρmin (e Å−3)	0.25, −0.21	0.30, −0.29
Absolute structure	–	Flack x determined using 1620 quotients [(I +) − (I −)]/[(I +) + (I −)] (Parsons et al., 2013 ▸)
Absolute structure parameter	–	0.01 (4)

Computer programs: CrystalClear-SM Expert (Rigaku, 2011 ▸), SHELXS97 (Sheldrick, 2008 ▸), SHELXL97 (Sheldrick, 2008 ▸), SHELXL2014 (Sheldrick, 2015 ▸), CrystalMaker (Palmer, 2013 ▸) and CrystalStructure (Rigaku, 2010 ▸).

Table 2. Hydrogen-bond geometry (Å, °) for (I) .

D—H⋯A	D—H	H⋯A	D⋯A	D—H⋯A
N1—H1⋯O1	0.844 (18)	2.029 (18)	2.6495 (15)	129.8 (15)
N1—H1⋯O1i	0.844 (18)	2.822 (18)	3.644 (1)	165.3 (15)

Symmetry code: (i) .

Table 3. Hydrogen-bond geometry (Å, °) for (II) .

D—H⋯A	D—H	H⋯A	D⋯A	D—H⋯A
O6—H61⋯O4i	0.83 (4)	1.90 (4)	2.720 (3)	168 (4)
N1—H1⋯O1	0.87 (3)	1.98 (4)	2.642 (3)	132 (3)

Symmetry code: (i) .

Results and discussion  

N-[2-(3,4-Di­meth­oxy­phen­yl)eth­yl]-2,4-di­nitro­aniline, (I), crys­tallizes in the monoclinic space group P2₁/c with one independent mol­ecule in the asymmetric unit (Fig. 1 ▸). The adjacent amine and nitro groups form an intra­molecular N—H⋯O hydrogen bond (Table 2 ▸) consistent with previous observations in mol­ecules of this type (Wade et al., 2013 ▸; Payne et al., 2010 ▸; Gangopadhyay & Radhakrishnan, 2000 ▸; Clegg et al., 1994 ▸; Panunto et al., 1987 ▸). Aniline (I) is chemically very similar to 2,4-di­nitro-N-(2-phenyl­eth­yl)aniline, (III), whose structure we reported recently (Wade et al., 2013 ▸), with (I) differing only in the presence of the meth­oxy groups in the 3- and 4-positions on the arene ring of the phenyl­ethyl group. In spite of the similarity of the two mol­ecules, there are some notable differences in their conformations. Most importantly, the dihedral angle between the planes of the two six-membered arene rings in (I) is ca 106°, while the rings in (III) are nearly parallel (dihedral angle ca 2°). This difference is attributable primarily to the C1—N1—C7—C8 torsion angle, which is 177.32 (12)° in (I) and −85.5 (2)° in (III). Other less drastic differences include rotations about the N1—C1 [C7—N1—C1—C6 torsion angle = −10.9 (2)° in (I) versus −0.9 (2)° in (III)] and C4—N3 [torsion angle O3—N3—C4—C3 = −12.6 (2)° in (I) versus −1.2 (2)° in (III)] bonds.

Figure 1.

The mol­ecular structure of (I), showing the atom-labeling scheme and 70% probability displacement ellipsoids for the non-H atoms. The intra­molecular hydrogen bond is shown as a dashed line.

As was the case in (III), neighboring mol­ecules of (I) are related across centers of inversion so as to permit an inter­molecular N—H⋯O hydrogen bond between the amine group on one mol­ecule and the nitro group on the other mol­ecule (Fig. 2 ▸ a). The amine H1 atom thus participates in a three-center hydrogen bond with two nitro O1 atoms, and each O1 atom serves as an acceptor for both intra- and inter­molecular hydrogen bonds (Table 2 ▸). The inter­molecular H1⋯O1ⁱ distance of 2.822 (18) Å [symmetry code: (i) −x + 1, −y + 1, −z + 1] is significantly longer than the distance of 2.381 (19) Å observed in (III). It is also beyond the typical range for this type of amine–nitro N—H⋯O inter­action (Panunto et al., 1987 ▸), making it a weak hydrogen bond at best. The nonparallel intra­molecular arrangement of the arene rings and the overall bent shape of the mol­ecules of (I), among other factors, lead to a packing pattern (Fig. 2 ▸ b) in which π–π stacking inter­actions appear to be precluded.

Figure 2.

(a) Mol­ecules of (I) are linked into dimers across centers of inversion by weak amine–nitro N—H⋯O hydrogen bonds (dashed lines). The view is onto the (010) plane. [Symmetry code: (#) −x + 1, −y + 1, −z + 1.] (b) An alternative view of the same set of mol­ecules as in part (a), showing the nonparallel arrangement of the arene rings. Displacement ellipsoids are drawn at the 70% probability level.

Enanti­omeric N-(4-methyl­sulfonyl-2-nitro­phen­yl)-l-phenyl­alanine, (II), crystallizes in the noncentrosymmetric monoclinic space group C2 with one mol­ecule in the asymmetric unit (Fig. 3 ▸). The absolute configuration was assigned based on the known configuration of the l-phenyl­alanine starting material and established through refinement of the Flack parameter (Parsons et al., 2013 ▸). As in (I), the amine H atom forms an intra­molecular N—H⋯O hydrogen bond with the adjacent nitro group (Table 3 ▸). The C—O bond lengths of the carb­oxy­lic acid group are consistent with the placement of the acidic H atom on atom O6 [C9—O6 = 1.324 (3) Å] and with atom O5 being part of a free carbonyl group [C9—O5 = 1.191 (3) Å]. l-Phenyl­alanine (II) differs from 4-methyl­sulfonyl-2-nitro-N-(2-phenyl­eth­yl)aniline, (IV) (Wade et al., 2013 ▸), only in the presence of the carb­oxy­lic acid group on the phenyl­ethyl group.

Figure 3.

The mol­ecular structure of (II), showing the atom-labeling scheme and 70% probability displacement ellipsoids for the non-H atoms. The intra­molecular hydrogen bond is shown as a dashed line.

As was the case with (I), the presence of this one additional group substanti­ally changes the conformation of the mol­ecule. The dihedral angle between the two arene rings is ca 62° in (II), but ca 107° in (IV). The difference is attributable primarily to the torsion angles C1—N1—C8—C10 [149.3 (2)° in (II) versus −175.70 (7)° in (IV)] and N1—C8—C10—C11 [−178.12 (18)° in (II) versus −61.20 (8)° in (IV)]. In addition, there are smaller but noticeable differences in the positioning of the nitro [O1—N2—C2—C1 torsion angle −20.2 (3)° in (II) versus −4.4 (1)° in (IV)] and methyl­sulfonyl [C7—S1—C4—C5 torsion angle 81.4 (2)° in (II) versus 110.98 (6)° in (IV)] groups.

Unlike (I), the mol­ecules of (II) are not joined through inter­molecular amine–nitro N—H⋯O hydrogen bonds. The mol­ecules of (II) are positioned with the amine and nitro groups of adjacent mol­ecules directed towards each other (Fig. 4 ▸ a), but the closest inter­molecular distance, i.e. H1⋯O2(−x + , y + , −z + 1) of ca 3.1 Å, is too great even for a three-center hydrogen bond. Instead, mol­ecules related by twofold rotation along the b axis are linked into dimers by two inter­molecular O—H⋯O hydrogen bonds between the carb­oxy­lic acid and sulfone groups (Table 3 ▸). Similar O—H⋯O inter­molecular inter­actions between hy­droxy and sulfone groups have been observed in 4,4′-sulfonyl­diphenol (Glidewell & Ferguson, 1996 ▸). These inter­actions and the chirality of the structure of (II) preclude the familiar inversion-related hydrogen bonding between carb­oxy­lic acid groups seen in many classic structures. The overall packing is such that the arene rings form a zigzag pattern (Fig. 4 ▸ b) that is devoid of π–π inter­actions between nearby parallel rings.

Figure 4.

(a) Mol­ecules of (II) are linked into dimers across twofold axes of rotation by carb­oxylic acid–sulfone O—H⋯O hydrogen bonds (dashed lines). The view is onto the (010) plane. [Symmetry code: (#) −x + 2, y + 1, −z + 1.] (b) An alternative view of the packing showing the zigzag arrangement of the arene rings. Displacement ellipsoids are drawn at the 70% probability level.

Recent unpublished work in our laboratory has indicated that mol­ecules with more hydro­philic character have lower NO-release rates in aqueous solution than similar mol­ecules with greater lipophilic character. We believe that mol­ecules with higher HLB (more hydro­philic) form more organized micelles that restrict NO release, while mol­ecules with lower HLB (more lipophilic) form less organized micelles that are less restrictive of NO release (Israelachvili, 2011 ▸). The presence of the carb­oxy­lic acid group on (II) would give it a greater hydro­philic balance than the previously reported 4-methyl­sulfonyl-2-nitro-N-(2-phenyl­eth­yl)aniline (Wade et al., 2013 ▸). On this basis, we would expect (II) to show a lower NO-release rate. By contrast, the addition of the two meth­oxy groups on (I) might be expected to lower the HLB relative to 2,4-di­nitro-N-(2-phenyl­eth­yl)aniline (Wade et al., 2013 ▸); the melting point of (I) is 379 K, while that of the parent without the meth­oxy groups is 425 K, suggesting weaker inter­molecular attractions in (I), leading to a higher NO-release rate. Ongoing experiments are underway to test these hypotheses and develop a better understanding of the relationship between structure and NO-release behavior.

Supplementary Material

CCDC references: 1472725, 1472724