NMR Characterization of Natural and Synthetic Anthocyanidins: Establishing a Foundation for Designing Functional O -Methylated Analogs

Abstract

Objective

As methylation affects the biological activities of secondary metabolites, novel methylated anthocyanins hold potential as new functional materials. While accurate structural determination is essential for the development of these novel derivatives, discrepancies in the nuclear magnetic resonance (NMR) assignments of anthocyanidins are observed in the literature. This study aims to achieve accurate NMR signal assignments for major natural anthocyanidins and a non-natural analog, then systematically elucidate the effects of O-methylation to establish reliable foundational data for future structural elucidation.

Methods

Five natural anthocyanidins (cyanidin, peonidin, delphinidin, petunidin, and malvidin) were isolated and purified from bilberry (Vaccinium myrtillus L.) extract following acid hydrolysis. Additionally, a non-natural analog, 3,7,3′,4′-tetra-O-methylcyanidin, was synthesized from quercetin. Comprehensive structural analysis of these compounds was conducted using ¹H-, ¹³C- and two-dimensional NMR spectroscopy.

Results

Signal assignments were achieved for all compounds, and previous literature discrepancies regarding the H-6 and H-8 signals were resolved. In ¹H-NMR, O-methylation generally induced downfield shifts for aromatic protons on the substituted rings. In ¹³C-NMR, ipso and para carbons typically exhibited downfield shifts, whereas ortho methine carbons shifted upfield. While 3,7,3′,4′-tetra-O-methylcyanidin exhibited trends similar to natural anthocyanidins, O-methylation on the A- and C-rings demonstrated slightly different shift changes.

Conclusion

In this study, we systematically summarized the fundamental trends of ¹H- and ¹³C-NMR chemical shifts for anthocyanidins. Furthermore, effects on the spectra due to O-methylation of hydroxyl groups including those on the A- and C-rings were confirmed. These findings provide an essential spectroscopic basis for identifying novel O-methylated anthocyanidins.

Keywords

NMR anthocyanin

Introduction

Modification of natural secondary metabolites by a relatively simple functional group, such as a methyl group, has been reported to significantly enhance or alter their biological activity. For instance, chrysin and quercetin, upon methylation to their derivatives 5,7-dimethoxyflavone and tamarixetin (4′-mono-O-methylquercetin), respectively, exhibit enhanced anticancer activity.^1,2 Similarly, methylated catechins, methylated derivatives of epigallocatechin gallate (EGCG) found in green tea, are known to possess much stronger anti-allergic activity than EGCG.^3,4 Numerous studies have shown that methyl modification contributes substantially to the biological activity of compounds. This large difference in activity, even within the same core structure, is mainly attributed to the position and number of methyl groups, which are thought to enhance compound stability, bioavailability, and binding affinity for target enzymes.⁵

Among the representative secondary metabolites, anthocyanins have garnered significant attention owing to their diverse biological activities, including potent antioxidant capacity,⁶ anti-inflammatory effects,⁷ and improvement of visual function.⁸ Among anthocyanins, O-methylated derivatives such as malvidin, peonidin, and petunidin are known. These compounds are characterized by structures in which one or two hydroxyl groups on the B-ring are methylated. This methylation has been reported to enhance the absorption of anthocyanins,⁹ improve their delivery rate to adipocytes, and consequently heighten their anti-obesity effects.¹⁰ Although several other naturally methylated anthocyanins besides peonidin, petunidin, and malvidin have been documented,^11-15 methylation predominantly occurs at the hydroxyl groups of the B-ring.

Therefore, the organic synthesis of novel, non-natural O-methylated anthocyanins to extend methylation to the hydroxyl groups on the A- and C-rings and increase the total number of methyl groups is anticipated to enhance known biological activities or lead to the discovery of new ones. This strategy has significant potential for the identification of new drug discovery seeds and functional materials. However, in developing synthetic methods to achieve structural diversity in O-methylated anthocyanins, structural determination is essential. Nuclear magnetic resonance (NMR) spectroscopy is indispensable for obtaining detailed structural information. Knowledge of the NMR chemical shift changes associated with the introduction of methyl groups to the hydroxyl moieties provides critical reference data for the structural determination of novel compounds. However, many reports on natural anthocyanins have been published, resulting in assignment discrepancies in the literature (Table SM 1). While the majority of the literature assigns the upfield signal between δ 6.6–7.0 to H-6 and the downfield signal to H-8, some reports^16-19 provide reversed assignments for these protons. Additionally, the chemical shifts for delphinidin reported by Kraus and Geraskin²⁰ exhibited an overall upfield shift compared to other literature. Therefore, precise and accurate assignment of NMR data is imperative.

In this study, we isolated and purified five major natural anthocyanidins from bilberries (Vaccinium myrtillus L.) after acid hydrolysis: cyanidin, peonidin, delphinidin, petunidin, and malvidin. Additionally, we synthesized 3,7,3′,4′-tetra-O-methylcyanidin to elucidate the effects of O-methylation on chemical shifts. The synthesis of this compound has been previously reported,²¹ and it features methoxy groups at positions not present in the five natural anthocyanidins mentioned above, serving as a model for comprehensively evaluating the effects of O-methylation across the A-, B-, and C-rings. We then conducted ¹H-NMR, ¹³C-NMR, and two-dimensional (2D)-NMR measurements, achieving complete and accurate signal assignments for all six anthocyanidins for the first time. Subsequently, we analyzed the resulting chemical shift changes. Through this analysis, we systematically clarified the relationship between the structures of poly-O-methylated anthocyanidins and their NMR spectra, thereby establishing foundational data essential for future exploration and identification of novel anthocyanidins.

Materials and Methods

Chemical Reagents

All chemicals were obtained from commercial sources and used without further purification. Methanol-d₄ (CD₃OD) was purchased from Thermo Fisher Scientific (Waltham, MA, USA). Trifluoroacetic acid-d₁ (TFA-d₁, for NMR) was obtained from Kanto Chemical (Tokyo, Japan). Quercetin (≥95% for liquid chromatography) was purchased from Shaanxi Jiahe Phytochem. (Shaanxi, China). Bilberry extract (BILBERON^®) was manufactured by Tokiwa Phytochemical (Sakura, Chiba, Japan). Zinc powder (for organic synthesis, average particle size of 6–9 µm) was purchased from FUJIFILM Wako Pure Chemical (Osaka, Japan). Boron tribromide (17% in dichloromethane) was purchased from Tokyo Chemical Industry (Tokyo, Japan). Dimethyl sulfate (first grade), potassium carbonate (reagent grade), sodium sulfate (reagent grade), acetone (reagent grade), methanol (reagent grade), acetonitrile (reagent grade), dehydrated methanol (for organic synthesis), dehydrated tetrahydrofuran (THF, for organic synthesis), 10% hydrogen chloride in methanol, chloroform (reagent grade), trifluoroacetic acid (TFA, reagent grade), and diethyl ether (reagent grade) were all obtained from Kanto Chemical (Tokyo, Japan). Purified water (Japanese Pharmacopoeia grade) was purchased from Kyoei Pharmaceutical (Kashiwa, Japan). For open-column chromatography, Wakogel^® C-200 (FUJIFILM Wako Pure Chemical, Osaka, Japan) and DIAION™ HP-20 (Mitsubishi Chemical, Tokyo, Japan) were used as stationary phases.

Preparation of Natural Anthocyanidin

Natural anthocyanidins were purified from bilberry extract after acid hydrolysis. Bilberry extract (100 g) was hydrolyzed by adding water containing 5% TFA (5.3 L) and heating at 97 °C for 1 h. After hydrolysis, the solution was applied to the DAION™ HP20 column. After washing the column with water containing 0.1% TFA, elution was performed with 50% aqueous methanol solution containing 0.1% TFA, yielding crude anthocyanin fractions. Each fraction was concentrated under reduced pressure and freeze dried. The dried crude fractions of each anthocyanidin were further purified using the procedure described below.

Preparation of Cyanidin Chloride and Malvidin Chloride

The dried material was fractionated on an octadecylsilyl (ODS) column (COSMOSIL 15C18-AR-II (250 × 50 mm inner diameter (i.d.); Nacalai Tesque, Kyoto, Japan)) using a 40% aqueous methanol solution containing 0.5% TFA. The fractionated solution was concentrated under reduced pressure and freeze dried to obtain cyanidin as a TFA salt. The dried material was dissolved in 10% hydrochloric acid in methanol, concentrated, and recrystallized to obtain cyanidin as a chloride salt (120 mg). Malvidin was prepared using the same procedure as that used for cyanidin and was obtained as its chloride salt (54 mg).

Preparation of Peonidin Chloride

The dried material was fractionated on an ODS column (Kilo Pack ODS SM50E (300 × 80 mm i.d.; Yamazen, Osaka, Japan)) using a 50% aqueous methanol solution containing 0.5% TFA. The fractionation solution was concentrated under reduced pressure and fractionated on an ODS column (Hydrosphere C18 (100 × 30 mm i.d.; YMC, Kyoto, Japan)) using an 18% aqueous acetonitrile solution containing 0.5% TFA. Subsequent operations were performed following the same procedure as that for cyanidin; peonidin was obtained as its chloride salt (23 mg).

Preparation of Delphinidin Chloride

The dried material was fractionated on an ODS column (Delta-Pak C18 (300 × 47 mm i.d., 15 µm; Waters, Milford, MA, USA)) using 30% aqueous methanol solution containing 0.5% TFA. Subsequent operations were performed following the same procedure as that for cyanidin, and delphinidin was obtained as its chloride salt (385 mg).

Preparation of Petunidin Chloride

The dried material was fractionated on an ODS column (Kilo Pack ODS SM50E) using a linear gradient of 40%–50% aqueous methanol containing 0.5% TFA. Subsequent operations were performed following the same procedure as that for cyanidin, and petunidin was obtained as its chloride salt (97 mg).

Preparation of 3,7,3′,4′-tetra-O-Methylcyanidin

The 3,7,3′,4′-tetra-O-methylcyanidin sample was synthesized with reference to a reported procedure²¹ via 3,7,3′,4′-tetra-O-methylquercetin, which was prepared from 3,5,7,3′,4′-penta-O-methylquercetin.²²

Synthesis of 3,5,7,3′,4′-penta-O-Methylquercetin

To a solution of quercetin (1.5 g, 0.50 mmol) in acetone (34 mL), potassium carbonate (6.9 g, 5.0 mmol) and dimethyl sulfate (3.8 mL, 4.0 mmol) were added. The mixture was refluxed for 7 h. After cooling, the reaction was quenched with water (39 mL). The resulting precipitate was collected by filtration and then recrystallized from methanol to afford 3,5,7,3′,4′-penta-O-methylquercetin (1.6 g, 87%).

Synthesis of 3,7,3′,4′-tetra-O-Methylquercetin

To a suspension of 3,5,7,3′,4′-penta-O-methylquercetin (1.00 g) in acetonitrile (25 mL), a 17% solution of boron bromide (BBr₃) in dichloromethane (5.4 mL) was added dropwise at 0 °C. The mixture was stirred at 0 °C for 1.5 h. The reaction was quenched with methanol (30 mL), and the solution was concentrated under reduced pressure to obtain the crude product. The crude product was recrystallized from methanol to afford 3,7,3′,4′-tetra-O-methylquercetin (0.78 g, 81%).

Synthesis of 3,7,3′,4′-tetra-O-Methylcyanidin

To a suspension of 3,7,3′,4′-tetra-O-methylquercetin (0.40 g, 0.11 mmol) and Zn powder (4.0 g, 61 mmol) in anhydrous methanol (2.7 mL) and anhydrous THF (4.0 mL), 10% hydrogen chloride–methanol solution (35 mL) was added at 0 °C. After stirring for 30 min at 0 °C, Zn was removed by filtration. The filtrate was stirred in air at room temperature for 20 h. The reaction solution was then added to chloroform containing 0.1% TFA (50 mL) and washed with water containing 0.1% TFA. The organic layer was dried over sodium sulfate and concentrated under reduced pressure to obtain the crude product. The residue was purified by silica gel chromatography (80% chloroform/hexane with 0.1% TFA→chloroform with 0.1% TFA→4% methanol/chloroform with 0.1% TFA). The fractionated solution was concentrated under reduced pressure, and the residue was crystallized from diethyl ether to afford 3,7,3′,4′-tetra-O-methylcyanidin as a TFA salt solid (0.074 g, 19%).

Nuclear Magnetic Resonance

NMR spectra were acquired using a JEOL JNM-ECZ600 NMR spectrometer (operating at 600 MHz for ¹H and 150 MHz for ¹³C) or a JEOL JNM-ECZ500 NMR spectrometer (operating at 500 MHz for ¹H and 125 MHz for ¹³C). Data were processed and analyzed using Delta NMR Software (version 6.2; JEOL, Tokyo, Japan). Unless otherwise noted, the samples were dissolved in CD₃OD for two reasons. First, it was utilized in the previous literature^18-20 whose NMR assignments required scrutiny. Second, the spectra were expected to be simplified through the elimination of hydroxyl proton signals via proton–deuterium exchange. When 3,7,3′,4′-tetra-O-methylcyanidin was measured in CD₃OD alone, signals attributed to the hemiacetal form predominated (Figure SM 68 Top). Consequently, the measurement was performed in CD₃OD containing 10% TFA-d₁ to ensure acidic conditions, thereby shifting the equilibrium toward the flavylium cation form (Figure SM 68 Bottom). Chemical shifts (δ) are reported in parts per million (ppm) relative to the residual solvent peak of CD₃OD (δ_H 3.31 and δ_C 49.00) as an internal standard. The following data are reported: chemical shift, multiplicity (s = singlet, d = doublet, dd = double doublet, br = broad), coupling constant (J) in Hertz (Hz), and integration. The assignments were confirmed by 2D NMR experiments (¹H–¹H correlation spectroscopy (COSY), ¹H–¹³C heteronuclear single quantum coherence (HSQC), and ¹H–¹³C heteronuclear multiple bond correlation (HMBC)). Chemical structures were drawn using ChemDraw software (version 18.0.0.231 (4029); PerkinElmer, Waltham, MA, USA).

Results

The ¹H-NMR, ¹³C-NMR, COSY, HSQC, and HMBC spectra were acquired for five natural anthocyanidins—cyanidin, peonidin, delphinidin, petunidin, and malvidin—and one synthetic analog, 3,7,3′,4′-tetra-O-methylcyanidin (Figure SM1–67). The resulting signal assignments are summarized in Tables 1 and 2 and in Figure 1. The singlet signal observed downfield in the ¹H-NMR spectrum was assigned to H-4, and C-4 was identified based on HSQC correlations. This characteristic deshielding reflects the high electrophilicity of the C-ring, induced by the trivalent cationic oxygen atom at position 1. Furthermore, this assignment aligns with previously reported NMR data for anthocyanidins, where H-4 chemical shifts typically range from δ 8.5 to 9.0,^18-20,23,24 making it a distinctive peak for the flavylium skeleton.

Table 1.

¹H-NMR Data for Anthocyanidins

	Cyanidin	Peonidin	Delphinidin	Petunidin	Malvidin	3,7,3′,4′-tetra-O-Methylcyanidin
4	8.55 (s)	8.58 (s)	8.52 (s)	8.57 (s)	8.62 (s)	8.86 (s)
6	6.61 (d, 1.4)	6.62 (d, 2.0)	6.60 (br d, 2.0)	6.62 (d, 2.0)	6.64 (d, 1.7)	6.74 (s)
8	6.85 (br s)	6.90 (d, 1.0)	6.82 (br d, 0.9)	6.88 (br s)	6.95 (d, 1.0)	7.14–7.21 (m)
2′	8.09 (d, 2.1)	8.13 (d, 2.1)	7.80 (s)	7.92 (d, 2.0)	8.03 (s)	8.06 (br s)
3′
5′	7.00 (d, 8.3)	7.03 (d, 9.0)				7.14–7.21 (m)
6′	8.20 (dd, 8.6, 2.1)	8.31 (dd, 9.0, 2.1)	7.80 (s)	7.87 (d, 2.0)	8.03 (s)	8.41 (br d, 8.3)
3-OCH₃						4.23 (s)
7-OCH₃						4.02 (s)
3′-OCH₃		3.99 (s)			4.00 (s)	3.97 (s)
4′-OCH₃						4.00 (s)
5′-OCH₃				3.98 (s)	4.00 (s)

The spectrum for delphinidin was recorded at 500 MHz, while all others were recorded at 600 MHz. The solvent was CD₃OD for all compounds except for 3,7,3′,4′-tetra-O-methylcyanidin, which was analyzed in CD₃OD containing 10% trifluoroacetic acid-d₁ (TFA-d₁). s = singlet; d = doublet; dd = double doublet; br = broad; coupling constant (J) in Hertz (Hz).

Table 2.

¹³C-NMR Data for Anthocyanidins

	Cyanidin	Peonidin	Delphinidin	Petunidin	Malvidin	3,7,3′,4′-tetra-O-Methylcyanidin
2	162.1	161.8	162.3	161.9	161.7	164.4
3	146.6	146.5	146.7	146.7	146.7	149.6
4	134.1	134.4	133.7	134.1	134.6	131.6
4a	113.6	113.8	113.5	113.7	114.0	114.6
5	158.1	158.2	158.0	158.1	158.2	158.2
6	103.2	103.2	103.1	103.2	103.2	103.6
7	169.0	169.2	168.8	169.0	169.3	170.4
8	94.8	95.0	94.8	94.9	95.1	92.8
8a	157.0	157.1	157.0	157.1	157.3	157.1
1′	122.0	121.8	120.7	120.7	120.7	122.4
2′	118.0	114.2	112.0	113.9	110.1	114.1
3′	147.5	149.6	147.6	147.5	149.8	151.2
4′	155.3	156.2	144.2	144.7	n.d.	158.0
5′	117.3	117.5	147.6	149.9	149.8	113.1
6′	127.2	128.5	112.0	108.0	110.1	129.6
3-OCH₃						58.5
7-OCH₃						57.6
3′-OCH₃		56.7			57.1	56.8
4′-OCH₃						57.0
5′-OCH₃				56.9	57.1

n.d. = not detected. The spectrum for delphinidin was recorded at 125 MHz, while all others were recorded at 150 MHz. The solvent was CD₃OD for all compounds except for 3,7,3′,4′-tetra-O-methylcyanidin, which was analyzed in CD₃OD containing 10% TFA-d₁.

Figure 1.

HSQC and HMBC correlations of anthocyanidins are indicated by bold line and arrows, respectively

In cyanidin, the B-ring protons displayed an ABX pattern. The signal at δ 8.20 (dd, J = 8.6, 2.1 Hz) was assigned to H-6′, which couples with the protons at the ortho and meta positions. In addition, the signals at δ 7.00 (d, J = 8.3 Hz) and δ 8.09 (d, J = 2.1 Hz) were ascribed to H-5′ and H-2′, respectively, based on their COSY correlations with H-6′. C-2, C-3, and C-1′ through C-6′ were identified based on HSQC and HMBC correlations involving H-4 and B-ring protons (Figure 1).

However, assigning H-6 and H-8 based on chemical shift differences and the HMBC spectra was difficult. This is because of their similar chemical environments, as oxygen atoms are bonded to the carbons ortho to both protons, and the fact that the pathways from H-4 to C-8a and C-5 both consist of four bonds. Previous studies on coumarins and quinolines reported ⁵J long-range coupling between H-4 and H-8 in the COSY spectra, which was absent between H-4 and H-6.²⁵ Similar long-range coupling has been reported for 3-deoxyanthocyanidin.²⁶ In the present study, we observed a COSY correlation between H-4 and the doublet signal δ 6.61 (J = 1.4 Hz), whereas no such correlation was observed for a broad singlet at δ 6.85. Based on these observations, we assigned the correlating signal (δ 6.63) as H-6 and the non-correlating signal (δ 6.87) as H-8. Subsequent analysis of the HSQC and HMBC correlations allowed the complete assignment of C-4a, C-5, C-6, C-7, C-8, and C-8a. This assignment approach for the A-ring was applied to all other anthocyanidins in this study.

In peonidin, the B-ring protons (H-2′, H-5′, and H-6′) were assigned analogously to those of cyanidin. A carbon signal at δ 156.2 showed HMBC correlations with all three of these protons and was consequently assigned to C-3′. In the ¹H- and ¹³C-NMR spectra, signals attributed to a methoxy group appeared at δ 3.99 and δ 56.7, respectively. A strong HMBC correlation between these protons of the methoxy group and C-3′ confirmed the position of the methoxy group at C-3′.

For delphinidin, the aromatic proton signal at δ 7.81 exhibits an integration of 2, consistent with the symmetrical B-ring protons H-2′ and H-6′. HSQC and HMBC correlations using these protons enabled the identification of C-2, C-3, and C-1′ through C-6′.

In the case of petunidin, signals attributed to a methoxy group were observed at δ 3.98 and δ 56.9 in the ¹H- and ¹³C-NMR spectra, respectively. The asymmetry of the B-ring allowed the exclusion of the C-4′ position for this methoxy group. Consequently, the carbon signal at δ 147.5, which exhibited an HMBC correlation with the methoxy protons, was assigned to C-3′. Furthermore, the signal at δ 7.92 (d, J = 2.0 Hz), showing an HMBC correlation with C-3′, was assigned to H-2′, while the signal at δ 7.87 (d, J = 2.0 Hz), displaying meta-coupling with H-2′, was assigned to H-6′. The remaining B-ring carbons were assigned by analyzing the HSQC and HMBC correlations (Table 1).

Malvidin was analyzed based on the symmetry of its B-ring, as for delphinidin. H-2′, H-6′, C-2, C-3, C-2′, and C-6′ were assigned using the HSQC and HMBC correlations. Among the signals showing HMBC correlations with H-2′ and H-6′, the downfield signal at δ 149.8 was assigned to C-3′ and C-5′, while the upfield signal at δ 110.1 was assigned to C-1′ and C-6′. Methoxy group signals appeared at δ 4.00 and δ 57.1 in the ¹H- and ¹³C-NMR spectra, and HMBC correlations confirmed that these methoxy groups are bonded to C-3′ and C-5′.

No distinct signal for C-4′ was observed in the 1D spectrum. Compared with the NMR data for malvidin-3-O-glucoside reported by Jordheim et al,²⁷ the signals for the A- and C-rings showed deviations of ∼2–3 ppm, while the signals for the B-ring (C-2′, C-3′, C-5′, C-6′) agreed well. In the HMBC spectrum, a cross-peak with H-2′/H-6′ was observed around δ 146, which is consistent with the reference data for C-4′. This suggests that the C-4′ signal is indeed located at this position (Figure SM 51).

For 3,7,3′,4′-tetra-O-methylcyanidin, an ABX pattern was observed for the B-ring protons. While the ortho-coupling constant was 8 Hz, splitting owing to meta coupling was not clearly resolved; the signal assigned to H-2′ (δ 8.06) appeared as a broad singlet, and the signal for H-6′ (δ 8.41) appeared as a broad doublet. H-2′, H-5′, and H-6′ were assigned using COSY correlations in the same manner as cyanidin, and C-2, C-3, and C-1′–C-6′ were assigned via HSQC and HMBC correlations. Furthermore, the four methoxy proton signals (δ 3.97, δ 4.00, δ 4.02, δ 4.23) exhibited strong HMBC correlations with C-3′, C-4′, C-7, and C-3, respectively, thereby confirming the locations of the methoxy groups.

Discussion

In the ¹H-NMR spectra, it was observed that O-methylation tended to cause a downfield shift in the signals of the aromatic protons on the substituted ring (Figure SM 69, 70). Based on a recent study,²⁸ it is inferred that while the inductive effects of hydroxyl and methoxy groups are comparable, the methoxy group possesses a weaker resonance effect, thereby reducing its electron-donating capacity. In the case of anthocyanidins, it is likewise suggested that the methoxy groups decreases the electron density of the aromatic ring, leading to the observed downfield shift.

Focusing on the B-ring, the H-6′ (para position of the 3′-methoxy group) signal of peonidin shifted downfield 0.11 ppm compared with that of cyanidin. Similarly, in petunidin, the H-2′ (para position of 5′-methoxy group) signal and H-6′ (ortho position of 5′-methoxy group) signal were shifted downfield 0.11 ppm and 0.06 ppm, respectively, relative to delphinidin. Furthermore, malvidin exhibited downfield shifts in H-6′ (para position of 3′-methoxy group) of 0.16 ppm and H-2′ (ortho position of 3′-methoxy group) of 0.11 ppm compared with petunidin. It has been reported that in ortho-substituted anisoles, steric hindrance can cause the O-CH3 bond to be out of the plane of the aromatic ring. This structural change leads to a decrease in electron-donating ability due to stereoelectronic effects, typically resulting in a smaller shift at the ortho positions than at the para positions,²⁹ and our observations appear to be consistent with this phenomenon. However, in 3,7,3′,4′-tetra-O-methylcyanidin, H-5′ exhibited a downfield shift of approximately 0.2 ppm despite being at the ortho position, representing a departure from the trend observed in the other anthocyanidins.

Regarding the O-methylation of the A- and C-rings, the signals for H-4, H-6, and H-8 all exhibited downfield shifts; notably, H-4 and H-8 showed substantial shifts of approximately 0.3 ppm. The downfield shift of H-4 is likely attributable to stereoelectronic effects resulting from the twisting of the 3-methoxy group, which is induced by steric hindrance from the adjacent B-ring. In contrast, H-8 exhibited a significantly larger shift compared with H-6, despite the ortho positions of the 7-methoxy group being unsubstituted. A similar trend was observed when comparing the chemical shifts of the corresponding flavonols, isorhamnetin³⁰ and retusin.³¹ Consequently, this phenomenon appears to be characteristic of the flavonoids possessing a hydroxyl group at C-5 and a methoxy group at C-7.

Turning to the ¹³C-NMR spectra, the observed changes were more complex than those found in the ¹H-NMR analysis. Initial examination of the B-ring revealed that the signals of ipso carbons generally exhibited a downfield shift of approximately 2 ppm, reflecting the difference in electron-donating ability between hydroxyl and methoxy groups. For the ortho carbons, the signals of methine carbons tended to shift upfield by 3.7–4.4 ppm, whereas quaternary carbons shifted downfield by 0.5–1.6 ppm. While the signals of para carbons consistently shifted downfield by 1.3–2.2 ppm, chemical shifts of the meta carbons remained largely unchanged. However, in a departure from the previously observed trends, the meta carbon (C-6′) in 3,7,3′,4′-tetra-O-methylcyanidin shifted downfield by 1.1 ppm compared with peonidin.

Similar to the B-ring, the signals of the ipso carbon in both the A- and C-rings shifted downfield by O-methylation, though the magnitude varied by position (3.1 ppm at C-3 and 1.2 ppm at C-7). The signals of the ortho carbon showed the same trend as the B-ring for the C-ring, whereas in the A-ring, despite both being methine carbons, C-6 showed almost no shift while C-8 shifted upfield compared with peonidin. Furthermore, the methoxy carbon at the 3-position exhibited a downfield shift of approximately 1.5 ppm compared with those at the 7, 3′, 4′, and 5′ positions. In anisole derivatives and methoxyflavones, the signals of methoxy carbon typically appear around 55–56 ppm. However, it has been reported that when both ortho positions are substituted, the signal can shift toward 60 ppm due to stereoelectronic effects.^32-34 In this study, despite being mono-ortho-substituted, the 3-methoxy carbon exhibited a downfield shift, which can be attributed to its non-planar orientation. This loss of planarity is presumably driven by steric hindrance from the adjacent B-ring.

This study has provided fundamental insights into the chemical shift variations resulting from the O-methylation of hydroxyl groups in both ¹H- and ¹³C-NMR. Nevertheless, data regarding the methylation of the 3-,7- and 4′-hydroxyl groups remain limited. For a more comprehensive discussion, it is essential to acquire NMR data for each mono-O-methylated anthocyanidin (at the 3, 5, 7, 3′ and 4′ positions), including the 5-methoxy derivative which was not investigated in this study. Such data would enable a systematic investigation into the positional effects of O-methylation at each site. Furthermore, confirming whether these trends persist in other solvents, such as dimethyl sulfoxide-d₆ in addition to CD₃OD, would establish a more robust spectroscopic foundation.

Conclusion

In this study, we conducted an NMR spectroscopy analysis of five major natural anthocyanidins (cyanidin, peonidin, delphinidin, petunidin, and malvidin) as well as a synthetic non-natural analog, 3,7,3′,4′-tetra-O-methylcyanidin. Through this analysis, we systematically established the fundamental trends of ¹H- and ¹³C-NMR chemical shifts within the anthocyanidin skeleton. Furthermore, spectral effects due to O-methylation of hydroxyl groups including those on the A- and C-rings were confirmed. However, regarding O-methylation at the 3, 7, and 4′ positions, the investigation was limited to the results obtained from 3,7,3′,4′-tetra-O-methylcyanidin, and the 5-mono-O-methylated derivative was not examined in this study. In addition to the present findings, further data acquisition for mono-O-methylated anthocyanidins and the investigation of solvent-dependent trends are expected to establish a more valuable informative foundation for the structural elucidation of newly synthesized or isolated O-methylated anthocyanidins.

Supplemental Material

Supplemental material - NMR Characterization of Natural and Synthetic Anthocyanidins: Establishing a Foundation for Designing Functional O-Methylated Analogs

Supplemental material for NMR Characterization of Natural and Synthetic Anthocyanidins: Establishing a Foundation for Designing Functional O-Methylated Analogs by Shunsei Shiozawa, Yukiko Kobayashi, Tadatoshi Yamashita and Jinwei Yang in Natural Product Communications.

Footnotes

Acknowledgements

We would like to thank Dr. Kumi Yoshida, Dr. Ryosuke Sugiyama, and Dr. Tsutomu Ishikawa for their valuable advice and technical assistance with this research, and Editage () for English language editing.

ORCID iD

Jinwei Yang

Author Contributions

Conceptualization, S.S., Y.K. and J.Y.; methodology, S.S., Y.K., T.Y. and J.Y.; investigation, S.S. and T.Y.; writing—original draft preparation, S.S. and T.Y.; writing—review and editing, S.S., Y.K. and J.Y.; visualization, S.S.; supervision, J.Y.; project administration, Y.K. and J.Y. All authors have read and agreed to the published version of the manuscript.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The author declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: All authors are employees of Tokiwa Phytochemical Co., Ltd., which manufactures the bilberry extract (BILBERON^®) used in this study.

Data Availability Statement

The data supporting the findings of this study are available within the article and its supplemental material.*

Supplemental Material

Supplemental material for this article is available online.

References

Walle

Kawamori

Wen

Tsuji

Walle

. Cancer chemopreventive properties of orally bioavailable flavonoids–methylated versus unmethylated flavones. Biochem Pharmacol. 2007;73:1288-1296. doi:10.1016/j.bcp.2006.12.028.

Sak

Lust

Kase

Jaal

. Cytotoxic action of methylquercetins in human lung adenocarcinoma cells. Oncol Lett. 2018;15:1973-1978. doi:10.3892/ol.2017.7466.

Sano

Suzuki

Miyase

Yoshino

Maeda-Yamamoto

. Novel antiallergic catechin derivatives isolated from oolong tea. J Agric Food Chem. 1999;47:1906-1910. doi:10.1021/jf981114l.

Suzuki

Yoshino

Maeda-Yamamoto

Miyase

Sano

. Inhibitory effects of tea catechins and O-methylated derivatives of (−)-epigallocatechin-3-O-gallate on mouse type IV allergy. J Agric Food Chem. 2000;48:5649-5653. doi:10.1021/jf000313d.

Maeda-Yamamoto

Ema

Shibuichi

. In vitro and in vivo anti-allergic effects of ‘Benifuuki’ green tea containing O-methylated catechin and ginger extract enhancement. Cytotechnology. 2007;55:135-142. doi:10.1007/s10616-007-9112-1.

Kähkönen

Heinonen

. Antioxidant activity of anthocyanins and their aglycons. J Agric Food Chem. 2003;51:628-633. doi:10.1021/jf025551i.

Zhu

Ling

Guo

, et al. Anti-inflammatory effect of purified dietary anthocyanin in adults with hypercholesterolemia: a randomized controlled trial. Nutr Metab Cardiovasc Dis. 2013;23:843-849. doi:10.1016/j.numecd.2012.06.005.

Nomi

Iwasaki-Kurashige

Matsumoto

. Therapeutic effects of anthocyanins for vision and eye health. Molecules. 2019;24:3311. doi:10.3390/molecules24183311.

Ichiyanagi

Shida

Rahman

Hatano

Konishi

. Bioavailability and tissue distribution of anthocyanins in bilberry (Vaccinium myrtillus L.) extract in rats. J Agric Food Chem. 2006;54:6578-6587. doi:10.1021/jf0602370.

10.

Skates

Overall

DeZego

, et al. Berries containing anthocyanins with enhanced methylation profiles are more effective at ameliorating high fat diet-induced metabolic damage. Food Chem Toxicol. 2018;111:445-453. doi:10.1016/j.fct.2017.11.032.

11.

Karrer

Widmer

. Über Primelnfarbstoffe. Pflanzenfarbstoffe V. Helv Chim Acta. 1927;10:758-763. doi:10.1002/hlca.19270100191.

12.

Harborne

. Flavonoids of the Primulaceae: anthocyanidins of the Primulaceae. Nature. 1958;181:26-27. doi:10.1038/181026b0.

13.

Harborne

. Plant polyphenols: 5. Occurrence of azalein and related pigments in flowers of Plumbago and Rhododendron Species. Arch Biochem Biophys. 1962;96:171-178. doi:10.1016/0003-9861(62)90467-8.

14.

Harborne

. Comparative biochemistry of the flavonoids-IV.: correlations between chemistry, pollen morphology and systematics in the family Plumbaginaceae. Phytochemistry. 1967;6:1415-1428. doi:10.1016/S0031-9422(00)82884-8.

15.

Momose

Abe

Yoshitama

. 5-O-Methylcyanidin 3-glucoside from leaves of Egeria densa. Phytochemistry. 1977;16:1321-1322. doi:10.1016/s0031-9422(00)94397-8.

16.

Hillebrand

Schwarz

Winterhalter

. Characterization of anthocyanins and pyranoanthocyanins from blood orange [Citrus sinensis (L.) Osbeck] juice. J Agric Food Chem. 2004;52(24):7331-7338. doi:10.1021/jf0487957.

17.

Ichiyanagi

Kashiwada

Ikeshiro

, et al. Complete assignment of bilberry (Vaccinium myrtillus L.) anthocyanins separated by capillary zone electrophoresis. Chem Pharm Bull. 2004;52(2):226-229. doi:10.1248/cpb.52.226.

18.

Zhang

Guo

Zhang

, et al. Separation, purification and identification of antioxidant compositions in black rice. Agric Sci China. 2006;5:431-440. doi:10.1016/S1671-2927(06)60073-410.1016/S1671-2927(06)60073-4.

19.

Franz

Serebnitskaya

Gudial

Wallis

. Structure assignment and H/D-exchange behavior of several glycosylated polyphenols. Arkivoc. 2014;5:94-122. doi:10.3998/ark.5550190.p008.583.

20.

Kraus

Geraskin

. Synthetic anthocyanidins from natural benzopyrans. Nat Prod Commun. 2016;11:1649-1650. doi:10.1177/1934578X1601101106.

21.

Kimura

Oyama

Murata

Wakamiya

Yoshida

. Synthesis of 8-Aryl-O-methylcyanidins and their usage for dye-sensitized solar cell devices. Int J Mol Sci. 2017;18:427. doi:10.3390/ijms18020427.

22.

Kimura

Kato

Oyama

Kondo

Yoshida

. Efficient preparation of various O-methylquercetins by selective demethylation. Nat Prod Commun. 2016;11:957-961. doi:10.1177/1934578X1601100722.

23.

Nomura

Ikuta

Fukai

Uzawa

. ¹³C NMR spectra of cyanidin and chrysanthemin. Heterocycles. 1985;23:121-126. doi:10.3987/r-1985-01-0121.

24.

Guo

Liu

Zhang

. Combination of supercritical fluid extraction with counter-current chromatography to isolate anthocyanidins from the petals of Chaenomeles sinensis based on mathematical calculations. J Sep Sci. 2013;36:3517-3526. doi:10.1002/jssc.201300873.

25.

Abraham

Fisher

Loftus

. Introduction to Nuclear Magnetic Resonance Spectroscopy. 1st ed. Wiley; 1988.

26.

Kouda-Bonafos

Nacro

Ancian

. Total assignment of ¹H and ¹³C NMR chemical shifts of a natural anthocyanidin, apigeninidin, using two-dimensional COLOC and HMBC techniques. Magn Reson Chem. 1996;34:389-392. doi:10.1002/(SICI)1097-458X(199605)34:5<389::AID-OMR888>3.0.CO;2-9.

27.

Jordheim

Fossen

Andersen

ØM

. Characterization of hemi-acetal forms of anthocyanidin 3-O-β-glycopyranosides. J Agric Food Chem. 2006;54:9340-9346. doi:10.1021/jf0619636.

28.

Comas-Vilà

Salvador

. Capturing electronic substituent effect with effective atomic orbitals. Phys Chem Chem Phys. 2025;27(20):10482-10491. doi:10.1039/D5CP01299F.

29.

Lambert

Olsen

Jaroszewski

. Stereoelectronic effects on ¹H nuclear magnetic resonance chemical shifts in methoxybenzenes. J Org Chem. 2006;71(25):9449-9457. doi:10.1021/jo061757x.

30.

Kulić

Steiner

VJN

Butterer

. NMR chemical shifts of common flavonoids. Planta Med. 2026;92(1):58-80. doi:10.1055/a-2706-7473.

31.

Juvale

Stefan

Wiese

. Synthesis and biological evaluation of flavones and benzoflavones as inhibitors of BCRP/ABCG2. Eur J Med Chem. 2013;67:115-126. doi:10.1016/j.ejmech.2013.06.035.

32.

Makriyannis

Knittel

. The conformational analysis of aromatic methoxyl groups from carbon-13 chemical shifts and spin-lattice relaxation times. Tetrahedron Lett. 1979;20(30):2753-2756. doi:10.1016/S0040-4039(01)86406-6.

33.

Calvert

Cambie

Davis

. ¹³C NMR spectra of polymethoxy- and methylenedioxyflavonols. Org Magn Reson. 1979;12(10):583-586. doi:10.1002/mrc.1270121008.

34.

Makriyannis

Fesik

. Methoxy group conformations of phenyl methyl ethers in solution. J Am Chem Soc. 1982;104(23):6462-6463. doi:10.1021/ja00387a058.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

4.98 MB

0.00 MB