Bioguided Identification of Polymethoxyflavones as Novel Vascular CaV1.2 Channel Blockers from Citrus Peel

Abstract

The huge amount of citrus peel produced worldwide represents an economic burden for society. However, this agricultural by-product is a rich source of natural molecules, potentially endowed with interesting pharmacological activities. In this regard, we decided to investigate if the polymethoxyflavones contained in citrus peel waste could be exploited as novel vasorelaxant agents. A hydroalcoholic blond orange (Citrus sinensis) peel extract, obtained by ultrasonication, was partitioned in dichloromethane. Column chromatography allowed for the isolation of four polymethoxyflavones, namely, scutellarein tetramethyl ether, nobiletin, tangeretin, and sinensetin, identified by nuclear magnetic resonance (NMR) spectroscopy and UPLC-HRMS/MS and confirmed by multivariate curve resolution of NMR fractional spectra. The four molecules showed interesting in vitro vasorelaxant activity, at least, in part, due to the blockade of smooth muscle Ca_V1.2 channels. Molecular modeling and docking analysis elucidated the binding mode of the polymethoxyflavones at the homology model of the rat Ca_V1.2c subunit and provided the structural basis to rationalise the highest activity of scutellarein tetramethyl ether in the set and the dramatic effect of the additional methoxy group occurring in nobiletin and sinensetin. In conclusion, citrus peel can be considered a freely available, valuable source of vasoactive compounds worthy of pharmaceutical and/or nutraceutical exploitation.

Keywords: CaV1.2 channel blockers; docking studies; multivariate curve resolution; polymethoxyflavones.

Figure 1

Citrus ethanolic extract, as well as n-hexane and dichloromethane phases, are effective against phenylephrine-induced but not KCl-induced contraction. Rings stimulated either (A) by 25 mM or (C) 60 mM KCl (in the absence of a functional endothelium) or (B,D) by 0.3–0.6 µM phenylephrine [phe; endothelium-intact (+endothelium) or -deprived (−endothelium)] were challenged with cumulative concentrations of the citrus ethanolic extract (EE), n-hexane (E), or dichloromethane (DCM) phases. The effect of the solvent dimethyl sulfoxide (DMSO) is also shown. In the ordinate scale, relaxation is reported as a percentage of the initial tension induced by either KCl or phenylephrine. ((A) KCl 860 ± 322 mg, n = 3; (B) phe + endo 238 ± 136 mg, n = 4, and phe−endo 667 ± 75 mg, n = 3, p = 0.0046; (C) KCl DCM 640 ± 392 mg, n = 5, E 740 ± 237 mg, n = 5, and DMSO 648 ± 364 mg, n = 5, p = 0.8744; (D) phe + endo DCM 738 ± 364 mg, n = 5, E 1186 ± 555 mg, n = 5, and DMSO 760 ± 439 mg, n = 6, p = 0.2467; phe − endo DCM 1092 ± 217 mg, n = 5, E 880 ± 718 mg, n = 5, and DMSO 1069 ± 467 mg, n = 8, p = 0.7568; KCl, phe + endo, and phe − endo for DMSO p = 0.2182; for DCM p = 0.1199; for E p = 0.4370]. Data points represent the mean ± SD (n = 3–8). (B) * p < 0.05 vs. −endothelium, (C) * p < 0.05 vs. DMSO, and (D) * p < 0.05 vs. DMSO and # p < 0.05 vs. E, two-way ANOVA and the Bonferroni multiple comparison test.

Figure 2

Citrus M1 to M4 fractions antagonized KCl- and phenylephrine-induced contraction. Rings stimulated either (A) by 60 mM KCl (in the absence of a functional endothelium) or (B,C) by 0.3/0.6 µM phenylephrine (phe), (B) endothelium-intact or (C) -deprived, were challenged with cumulative concentrations of the citrus fractions M1–M4. The effect of solvent (DMSO) is also shown. In the ordinate scale, relaxation is reported as a percentage of the initial tension induced by either KCl or phenylephrine. [(A) KCl DMSO 648 ± 364 mg, n = 5, M1 602 ± 231 mg, n = 5, M2 1630 mg, n = 2, M3 1822 ± 436 mg, n = 5, M4 507 ± 276 mg, n = 3; (B) phe + endo DMSO 760 ± 439 mg, n = 5, M1 956 ± 499 mg, n = 5, M2 698 ± 383 mg, n = 5, M3 1244 ± 418 mg, n = 5, M4 815, n = 2; (C) phe-endo DMSO 1092 ± 217 mg, n = 8, M1 862 ± 335 mg, n = 5, M2 1080 ± 636 mg, n = 3, M3 1623 ± 278 mg, n = 6, M4 977 ± 757, n = 3; p = 0.1033; KCl, phe + endo, and phe-endo for DMSO see legend to Figure 1; KCl, phe + endo, and phe-endo for M1 p = 0.3291; for M3 p = 0.0815]. Data points represent the mean ± SD (n = 3–5). * p < 0.05 vs. DMSO, two-way ANOVA, and the Bonferroni multiple comparison test. M2 (A) and M4 (B) were excluded from the statistical analysis because they included only two replicates.

Figure 3

UPLC-PDA-MS chromatograms of M3 fraction and single PMF standards.

Figure 4

Composition profiles and pure spectra obtained by MCR-ALS analysis on the spectral matrix: nobiletin (NOB), scutellarein TME (SCU), sinensetin (SIN), and tangeretin (TAN).

Figure 5

Inhibition of KCl- and phenylephrine-induced contraction by scutellarein TME, tangeretin, nobiletin, and sinensetin. Rings stimulated by either 60 mM KCl (in the absence of a functional endothelium) or by 0.3 µM phenylephrine [phe; endothelium-intact (+endo) or -deprived (−endo)] were challenged with cumulative concentrations of (A) scutellarein TME, (B) tangeretin, (C) nobiletin, and (D) sinensetin. In the ordinate scale, relaxation is reported as a percentage of the initial tension induced by KCl or phenylephrine. [(A) KCl 1292 ± 460 mg, n = 5, phe + endo 954 ± 456 mg, n = 5, phe − endo 1584 ± 645 mg, n = 5; p = 0.2098; (B) KCl 1556 ± 448 mg, n = 5, phe + endo 1267 ± 590 mg, n = 6, phe − endo 2138 ± 1145 mg, n = 5; p = 0.2105; (C) KCl 2178 ± 618 mg, n = 5, phe + endo 1395 ± 504 mg, n = 8, phe − endo 1680 ± 382 mg, n = 5; p = 0.0515; (D) KCl 2082 ± 560 mg, n = 5, phe + endo 1016 ± 406 mg, n = 5, phe − endo 1898 ± 905 mg, n = 5; p = 0.0542]. Data points represent the mean ± SD (n = 4–8).

Figure 6

M3 fraction, scutellarein TME, tangeretin, nobiletin, and sinensetin inhibited I_Ba1.2 in single tail artery myocytes. Concentration-dependent effect of (A) M3 fraction and (C) vehicle (DMSO), scutellarein TME, tangeretin, nobiletin, and sinensetin on I_Ba1.2. On the ordinate scale, the current amplitude is reported as a percentage of the value recorded just before the addition of the first drug concentration. Data points represent the mean ± SD (n = 4–8). * p < 0.05 vs. DMSO, two-way ANOVA, and the Bonferroni multiple comparison test. (B) Traces (representative of 5 similar experiments) of I_Ba1.2, elicited by 250-ms clamp pulses to 10 mV from a V_h of −50 mV, measured in the absence (control) or presence of various concentrations of scutellarein TME. The effect of 10 µM nifedipine is also shown. (D) Effect of flavonoids on I_Ba1.2 kinetics of single tail artery myocytes. Time constants for activation (τ_act) and inactivation (τ_inact) were measured in the absence or presence of 100 µM tangeretin (tan), scutellarein TME (scu), nobiletin (nob), or sinensetin (sin). Lines and bars represent the mean ± SD (n = 5). * p < 0.05 vs. control (C), Student’s t-test for paired data.

Figure 7

The best fit at the level of protein backbone for the representative docking poses of tangeretin (colored in slate blue) and scutellarein TME (colored in sienna). A stick representation is used for heavy atoms of the ligand and protein side chains within 5 Å of the ligand (colored in light blue and tan for tangeretin and scutellarein TME complexes, respectively. Protein backbone atoms are represented as ribbons colored according to the side chains, using half-transparency in correspondence with the ligand. Hydrogen, nitrogen, oxygen, and sulfur atoms are painted white, blue, red, and yellow, respectively. A green wire representation is adopted for H-bonds.