Solubilizing Effect of Caffeine on Precipitate of Risperidone Oral Solution and Tea Beverage

Abstract

The rates of risperidone in precipitates from 5 risperidone oral solutions and tea beverages (green, black, and oolong teas) against risperidone used were measured by quantitative (q)NMR, indicating that as the amount of caffeine in the beverages increased, the rate decreased significantly. A solution of caffeine (0, 5, 10, and 30 μmol) in tartaric acid buffer (pH 3.0) was added to a solution of risperidone and (−)-epigallocatechin-3-O-gallate (EGCg) (each 3.0 μmol) in tartaric acid buffer (pH 3.0). Judging from the amounts of risperidone and EGCg in the precipitates, it was considered that risperidone formed a 1 : 1 complex with EGCg. As the amount of caffeine increased, the amounts of risperidone and EGCg in precipitates decreased significantly, suggesting that caffeine has a solubilizing effect on precipitates of the 1 : 1 complex of risperidone and EGCg. Furthermore, the amount of risperidone in precipitates formed when compounds with a xanthine derivative, such as diprophylline, theobromine, and theophylline, were added was less than when they were not added, indicating that such compounds with a xanthine derivative have the same solubilizing effect as caffeine on precipitate of the 1 : 1 complex of risperidone and EGCg.

INTRODUCTION

Risperidone, a second-generation atypical antipsychotic, is widely used to treat schizophrenia, and due to the nature of the disease, liquid oral preparations are generally sold. However, it is well-known that mixing risperidone oral solutions with tea beverages such as green or black tea can cause compatibility changes,¹⁾ and all manufacturers of such solutions have issued warnings. Regardless of this, in elderly care facilities and other places where there are no pharmacists, risperidone oral solution is sometimes mixed with tea beverages and administered to patients. Each manufacturer has published data on compatibility changes with each beverage. Here, data on compatibility changes reported by each manufacturer were obtained by measuring the amount of risperidone remaining in the mixture when risperidone oral liquid was mixed with various tea beverages. As shown in Table 1, the data on compatibility changes varied markedly among the 5 manufacturers of risperidone oral solution.

Table 1. Rate of Risperidone Remaining in the Mixture of Risperidone Oral Solutions and Tea Beverages

	Rate of risperidone in mixture (%)b)
	Aa)	B	C	D	E
Black tea	60.1	98.9	No data	102.3	21.2
Oolong tea	69.3	100.0	No data	100.2	22.8
Green tea	82.2	99.6	No data	97.7	86.9

a) A–E are manufacturers of risperidone oral solutions. b) Values (%), as announced by the manufacturers A–E, were measured rates of risperidone in the mixture of the tea beverages (100 mL) and risperidone oral solution (3 mg/3 mL).

We investigated the amount of risperidone in the precipitate on mixing risperidone with tea beverages using quantitative (q)NMR. Using qNMR, we previously reported that when risperidone is taken with catechin-rich beverages, the efficacy of risperidone might be reduced, mainly because the 2,3-cis gallated catechins (−)-epigallocatechin-3-O-gallate (EGCg) and (−)-epicatechin-3-O-gallate (ECg) form complexes with risperidone and precipitate²⁾ (Fig. 1). qNMR data were published as an international standard in December 2022 and included in the 18th revised Japanese Pharmacopoeia,³⁾ which has been used in various fields in recent years. In this study, we investigated compatibility changes of risperidone oral solution with tea beverages and their causes using qNMR.

Fig. 1. Risperidone and (–)-Epigallocatechin-3-O-gallate (EGCg)

MATERIALS AND METHODS

Materials

Risperidone was purchased from LKT Laboratories Inc. (St. Paul, MN, U.S.A.). EGCg was purchased from Sigma-Aldrich (St. Louis, MO, U.S.A.). Caffeine, theophylline, diprophylline, theobromine, l-tryptophan, nicotinamide, and 2-naphthol were purchased from FUJIFILM Wako Pure Chemical Corporation (Osaka, Japan). The 5 manufacturers of risperidone oral solutions (1 mg/mL) were: Kyowa Pharmaceutical Industry Co., Ltd., Osaka, Japan; Takata Pharmaceutical Co., Ltd., Saitama, Japan; Towa Pharmaceutical Co., Ltd., Osaka, Japan; Dojin Iyaku-Kako Co., Ltd., Fukushima, Japan; and Janssen Pharmaceutical K.K., Tokyo, Japan. Green and black teas were from Oi Ocha, ITO EN, Ltd., Tokyo, Japan; Kirin Gogo-no-Kocha Oishii Muto, Kirin Holdings Co., Ltd., Tokyo, Japan; and oolong tea was from Suntory Oolong Tea OTPP, Suntory Holdings Limited, Osaka, Japan.

NMR Experiments

¹H-NMR spectra were recorded at 30°C on JEOL JNM-ECZ400R (Tokyo, Japan) operating at 400 MHz using a 5-mm φ sample tube. In general, ¹H-NMR experiments were performed with 32 K data points covering a spectral width of 10,000 Hz with an approximately 3.7 s pulse interval and 16 scan times. Deuterated dimethyl sulfoxide (DMSO-d₆) (Eurosotop, Saint-Aubin, France) was used as a measurement solvent. Chemical shift values are expressed in ppm downfield using sodium 3-(trimethylsilyl)-1-propane-1,1,2,2,3,3-d₆-sulfonate (DSS-d₆; FUJIFILM Wako Pure Chemical Corporation).

Quantitative ¹H-NMR (qNMR) was performed using the following optimized parameters: probe temperature, 30°C; spinning, on; number of scans, 60 times; spectral width, 20 ppm; pulse interval, 60 s; pulse angle, 90°.

Relaxation time (T₁) in DMSO-d₆ (sample concentration, 5.0 μmol) was measured using the Inversion Recovery method with the following optimized parameters: probe temperature, 30°C; spinning, off; number of scans, 16 times; 90°pulse angle, 5.9 μs; and relaxation delay used for measurement, 10.0, 2.0, 1.5, 1.0, 0.8, 0.6, 0.4, and 0.2 s.

Preparation of Precipitate Made from Risperidone Oral Solutions and Tea Beverages

Five risperidone oral solutions containing risperidone (1 mg/mL) were each poured into 3 kinds of tea beverages (green, black, and oolong teas, each 30 mL). As a result, turbidity and precipitation occurred. The mixtures were stirred at room temperature for 40 s and centrifuged at 3000 rpm for 15 min to obtain a supernatant and precipitate. The precipitate was dried under reduced pressure for 1 d to obtain a solid, which was dissolved in DMSO-d₆ (600 μL) containing DSS-d₆ (5.0 μmol) and subjected to qNMR measurements under the above conditions.

Preparation of Precipitate Made from Risperidone, EGCg, and Caffeine

The risperidone oral solutions contained risperidone (1 mg/mL), tartaric acid as a solubilizer, benzoic acid as a preservative, and sodium hydroxide as a pH adjuster, with the pH kept at 2.0–4.0. Tartaric acid buffer (pH 3.0) was used as a solubilizer and pH adjuster for risperidone and tea catechins. A solution of caffeine (0, 5, 10, and 30 μmol) in tartaric acid buffer (pH 3.0, 0.5 mL) was added to a solution of risperidone (1.23 mg, 3.0 μmol) and EGCg (1.38 mg, 3.0 μmol) in tartaric acid buffer (pH 3.0, 1.0 mL). Furthermore, a solution of risperidone (1.23 mg, 3.0 μmol) in tartaric acid buffer (pH 3.0, 0.50 mL) was added to a solution of EGCg (1.38 mg, 3.0 μmol) and caffeine (0, 5, 10, and 30 μmol) in tartaric acid buffer (pH 3.0, 1.0 mL). As a result, the mixture separated into a supernatant liquid and precipitate. The precipitate was dried under reduced pressure for 1 d to obtain a solid, which was dissolved in DMSO-d₆ (600 μL) containing DSS-d₆ (5.0 μmol) and subjected to qNMR measurements under the above conditions.

Changes in Chemical Shifts of Proton Signal of the 1 : 1 Complex of EGCg and Risperidone by Adding Caffeine

Solution of caffeine (0, 2.0, 4.0, 6.0, 8.0, 10.0, and 12.0 μmol) in tartaric acid buffer (pH 3.0, 0.3 mL D₂O) was added to a solution of risperidone (0.41 mg, 1.0 μmol) and EGCg (0.46 mg, 1.0 μmol) in tartaric acid buffer (pH 3.0, 0.5 mL D₂O). ¹H-NMR measurements of the mixtures were performed.

Preparation of Precipitate Made from Risperidone, EGCg, and Various Compounds

A solution of risperidone (1.23 mg, 3.0 μmol) in tartaric acid buffer (pH 3.0, 0.5 mL) was added to a solution of various compounds with or without a xanthine derivative (5.0 μmol) and EGCg (1.38 mg, 3.0 μmol) in tartaric acid buffer (pH 3.0, 1.0 mL). As a result, it was separated into a supernatant liquid and precipitate. The precipitate was dried under reduced pressure for 1 d to obtain a solid, which was dissolved in DMSO-d₆ (600 μL) containing DSS-d₆ (5.0 μmol) and subjected to qNMR measurements under the above conditions.

The above experiments were performed 4 times, and the average value of 4 integrated values of corresponding signals obtained by qNMR measurement was calculated.

RESULTS AND DISCUSSION

Quantification of Precipitate Made from Risperidone Oral Solutions and Tea Beverages

The 5 risperidone oral solutions were each added to 3 kinds of tea beverages (green, black, and oolong teas) to obtain the resulting precipitates, which were measured by qNMR (Fig. 2).

Fig. 2. qNMR Spectra of Precipitate from a Mixture of Risperidone Oral Solution and Green Tea (a), Risperidone Alone (b), and EGCg Alone (c)

The precipitate, risperidone (1.23 mg, 3.0 μmol), and EGCg (1.38 mg, 3.0 μmol) were dissolved in DMSO-d₆ (600 μL) containing DSS-d₆ (5.0 μmol) and subjected to qNMR measurement.

We previously reported that regarding the amount of tea catechins in a precipitate obtained by mixing risperidone and catechin mixture: amounts of 2,3-cis gallated catechin (EGCg) were much larger than those of other green tea catechins.²⁾ The catechin mixture was a fraction of extracted catechins included in green tea. Therefore, in this study, amounts of EGCg and risperidone in the precipitate were measured by qNMR. Proton signals in ¹H-NMR spectra for H_a, H_b, and H_c bonding to the benzene ring of risperidone and H_2′,6′ of EGCg [6.432 ppm, singlet, 2H] of 2,3-cis gallated catechin EGCg (Fig. 2) do not overlap with other signals. Characteristic coupling between H_b and F was observed in the proton signal for H_b of risperidone [7.299 ppm, double-double-doublet, 1H]. Therefore, integrated values of the proton signal for H_b of risperidone and H_2′,6 of EGCg were used for quantification by qNMR, and the rates of risperidone in the precipitates obtained from the 5 risperidone oral solutions and 3 kinds of tea beverages (green, black, and oolong teas) against risperidone used were evaluated (Fig. 3). A trimethyl group of DSS-d₆ [0 ppm, singlet, 9H of trimethyl group (CH₃)₃−] was used not only as a reference substance but also as an internal standard substance.

Fig. 3. The Rate of Risperidone in Precipitates from a Mixture of Risperidone Oral Solution (A–E) and Tea Beverages (Green, Black, and Oolong Teas) against Risperidone Used

(a) The rate of risperidone in precipitates against the amount of risperidone in the oral solution (1 mg/1 mL). (b) (A–E) are manufacturers of risperidone oral solutions.

As shown in Fig. 3, the rates of risperidone in the precipitates against risperidone contained in the oral solution (1 mg) were highest in black tea, followed by green and oolong teas, regardless of the manufacturer. According to the manufacturer’s published values, the amount of caffeine per 100 mL was 20, 13, and 10 mg in oolong, green, and black teas, respectively. It was shown that the higher the caffeine content, the lower the rate of risperidone in the precipitate. We considered that these rates of risperidone were related to the amount of caffeine, one of the main components of tea. We previously reported that a solution of risperidone and EGCg in tartaric acid buffer (pH 3.0) afforded a precipitate that was a complex of risperidone and EGCg.²⁾ It has been suggested that caffeine forms precipitates through “creaming-down” with tea polyphenols such as catechins, theaflavins, and thearubigins.⁴⁾ Also, we previously reported that caffeine formed complexes with EGCg in water via intermolecular hydrophobic interactions.^5–8) Thus, it was suggested that water-soluble caffeine bound to the complex of risperidone and EGCg through some kind of intermolecular interaction, thereby having a solubilizing effect on the precipitate of the complex. Previously, Zoglio et al. reported that caffeine enhanced the dissolution rate of ergotamine tartrate by a factor of 3 at gastric pH through interactions between ergotamine tartrate and caffeine.⁹⁾

Quantification of Precipitate Made from Risperidone, EGCg, and Caffeine

Solutions of 0, 5, 10, and 30 μmol of caffeine (Fig. 4) in tartaric acid buffer (pH 3.0) were added to solutions of equimolecular amounts of risperidone and EGCg (each 3 μmol) in tartaric acid buffer (pH 3.0) to yield precipitates. Furthermore, solutions of risperidone (3 μmol) in tartaric acid buffer (pH 3.0) were added to solutions of EGCg (3 μmol) and 0, 5, 10, and 30 μmol of caffeine in tartaric acid buffer (pH 3.0, 1.0 mL) to afford precipitates.

Fig. 4. Caffeine and Xanthine Derivative

The amounts of caffeine, risperidone, and EGCg in the precipitates obtained were measured by qNMR (Fig. 5). Integrated values of the proton signal for a methyl group of N₇–CH₃ of caffeine [3.910 ppm, singlet, 3H], H_b of risperidone, and H_2′,6′ of EGCg, which do not overlap with other signals, were used for quantification by qNMR. The amounts of risperidone, EGCg, and caffeine in precipitates made from the mixtures were evaluated. As the amount of caffeine increased, the amount of precipitate that formed decreased (Fig. 6). Judging from the amounts of risperidone and EGCg in the precipitates, it was considered that risperidone formed a 1 : 1 complex with EGCg (Figs. 7 and 8). Based on our previous studies,^5–8) risperidone was taken into the hydrophobic space formed by 3 aromatic A, B, B′ rings of EGCg to form a strong 1 : 1 complex of risperidone and EGCg. The solubility of the 1 : 1 complex in tartaric acid buffer rapidly decreased compared to that of risperidone and EGCg alone, and the 1 : 1 complex precipitated from tartaric acid buffer due to its high hydrophobicity. Ikeda et al. reported that risperidone formed a 1 : 1 complex with EGCg through 2 intermolecular hydrogen bonds between 2 nitrogen atoms of risperidone and 2 hydroxyl groups of B′ ring of EGCg.¹⁰⁾

Fig. 5. qNMR Spectra of Precipitate from a Mixture of Risperidone, EGCg, and Caffeine (a), Caffeine (b), Risperidone (c), and EGCg Alone (d)

The precipitate, caffeine (1.94 mg, 10 μmol), risperidone (1.23 mg, 3.0 μmol), and EGCg (1.38 mg, 3.0 μmol) were dissolved in DMSO-d₆ (600 μL) containing DSS-d₆ (5.0 μmol) and subjected to qNMR measurement.

Fig. 6. Precipitate That Developed When Caffeine Was Added to a Mixture of Risperidone and EGCg

A solution of caffeine (0, 5, 10, and 30 μmol) in tartaric acid buffer (pH 3.0, 0.5 mL) was added to a solution of risperidone (1.23 mg, 3.0 μmol) and EGCg (1.38 mg, 3.0 μmol) in tartaric acid buffer (pH 3.0, 1.0 mL).

Fig. 7. Amount of EGCg (a), Risperidone (b), and Caffeine (c) in Precipitate When Caffeine Was Added to Mixture of Risperidone and EGCg

A solution of caffeine (0, 5, 10, and 30 μmol) in tartaric acid buffer (pH 3.0, 0.5 mL) was added to a solution of risperidone (1.23 mg, 3.0 μmol) and EGCg (1.38 mg, 3.0 μmol) in tartaric acid buffer (pH 3.0, 1.0 mL). The precipitate was dissolved in DMSO-d₆ (600 μL) containing DSS-d₆ (5.0 μmol), and the amounts of risperidone, EGCg, and caffeine in the precipitate were measured by qNMR.

Fig. 8. Amount of EGCg (a), Risperidone (b), and Caffeine (c) in Precipitate When Risperidone Was Added to a Mixture of EGCg and Caffeine

A solution of risperidone (1.23 mg, 3.0 μmol) in tartaric acid buffer (pH 3.0, 0.5 mL) was added to a solution of EGCg (1.38 mg, 3.0 μmol) and caffeine (0, 5, 10, and 30 μmol) in tartaric acid buffer (pH 3.0, 1.0 mL). The precipitate was dissolved in DMSO-d₆ (600 μL) containing DSS-d₆ (5.0 μmol), and the amounts of risperidone, EGCg, and caffeine in the precipitate were measured by qNMR.

From the quantification by qNMR, as the amount of caffeine increased, the amounts of risperidone and EGCg in precipitates decreased significantly, suggesting that caffeine has a solubilizing effect on precipitates of the 1 : 1 complex of risperidone and EGCg (Figs. 7 and 8). Since a precipitate of risperidone alone did not dissolve even when an equimolar or double molar amounts of caffeine were added to the precipitate of risperidone without EGCg, it is considered that caffeine does not have a solubilizing effect on risperidone alone, but has a solubilizing effect on the 1 : 1 complex of EGCg and risperidone.

Changes in chemical shifts of proton signals of risperidone and EGCg moieties in the 1 : 1 complex of EGCg and risperidone by adding solutions containing regular amounts of caffeine in tartaric acid buffer (pH 3.0) were investigated. As shown in Figs. 9 and 10, a notable upfield shift of the singlet (2H) for H_2″,6″ of the B′ ring of the EGCg moiety was observed, and no other significant shift of proton signals was observed. The upfield shift of singlet (2H) for H_2″,6″ of the B′ ring was thought to be caused by magnetic anisotropic shielding by the ring current from the xanthine skeleton of caffeine. Therefore, it was thought that a π–π interaction between the B′ ring of the EGCg moiety and caffeine was formed. As a result, the interaction with water-soluble caffeine resulted in the dissolution of the 1 : 1 complex.

Fig. 9. ¹H-NMR Spectra of 1 : 1 Complex of EGCg and Risperidone by Adding Regular Amounts of Caffeine

A solution of caffeine (0, 2, 4, 6, 8, 10, and 12 μmol) in tartaric acid buffer (pH 3.0, 0.3 mL D₂O) was added to a solution of 1 : 1 complex of risperidone (0.41 mg, 1.0 μmol) and EGCg (0.46 mg, 1.0 μmol) in tartaric acid buffer (pH 3.0, 0.5 mL D₂O). (a) Caffeine (0 μmol), (b) caffeine (2 μmol), (c) caffeine (4 μmol), (d) caffeine (6 μmol), (e) caffeine (8 μmol), (f) caffeine (10 μmol), and (g) caffeine (12 μmol).

Fig. 10. Changes in Chemical Shifts of Proton Signal of Singlet for H_2″,6″ of the B′ Ring of EGCg Moiety in the 1 : 1 Complex of Risperidone and EGCg by Adding Caffeine

A solution of caffeine (0, 2, 4, 6, 8, 10, and 12 μmol) in tartaric acid buffer (pH 3.0, 0.3 mL D₂O) was added to a solution of 1 : 1 complex of risperidone (1.0 μmol) and EGCg (1.0 μmol) in tartaric acid buffer (pH 3.0, 0.5 mL D₂O). EGCg H_2″,6″ (Δ ppm) is expressed as the difference between 6.994 ppm (at caffeine 0 μmol) and chemical shift values of H_2″,6″ proton signal of EGCg (at caffeine 0, 2, 4, 6, 8, 10, and 12 μmol).

Quantification of Precipitate Made from Risperidone, EGCg, and Xanthine Compounds

We investigated whether compounds other than caffeine also had a solubilizing effect on precipitates of the complex of risperidone and EGCg. We focused on the xanthine derivative of caffeine and investigated the solubilizing effect of compounds with or without a xanthine derivative (Fig. 4). The solubilizing effect of pharmaceuticals with a xanthine derivative for the treatment of bronchial asthma, that is, diprophylline, theobromine, and theophylline was investigated.¹¹⁾ To compare with the solubilizing effect of xanthine compounds, that of non-xanthine compounds, tryptophan, nicotinamide, and cyclo(l-Pro-l-Ala)¹²⁾, was also investigated. The amount of risperidone and EGCg in the precipitate obtained from a mixture of risperidone and EGCg in tartaric acid buffer (pH 3.0) was compared with the amount of risperidone and EGCg in precipitate obtained when risperidone was added to a mixture of compounds with or without a xanthine derivative and EGCg in tartaric acid buffer (pH 3.0) (Figs. 11 and 12).

Fig. 11. Amount of EGCg and Risperidone in Precipitate When Risperidone Was Added to Mixture of Compounds with a Xanthine Derivative and EGCg

A solution of risperidone (1.23 mg, 3.0 μmol) in tartaric acid buffer (pH 3.0, 0.5 mL) was added to a solution of compounds with a xanthine derivative (5.0 μmol) and EGCg (1.38 mg, 3.0 μmol) in tartaric acid buffer (pH 3.0, 1.0 mL). The precipitate was dissolved in DMSO-d₆ (600 μL) containing DSS-d₆ (5.0 μmol) and subjected to qNMR measurements.

Fig. 12. Amount of EGCg and Risperidone in Precipitate When Risperidone Was Added to Mixture of Compounds without a Xanthine Derivative and EGCg

A solution of risperidone (1.23 mg, 3.0 μmol) in tartaric acid buffer (pH 3.0, 0.5 mL) was added to a solution of compounds without a xanthine derivative (5.0 μmol) and EGCg (1.38 mg, 3.0 μmol) in tartaric acid buffer (pH 3.0, 1.0 mL). The precipitate was dissolved in DMSO-d₆ (600 μL) containing DSS-d₆ (5.0 μmol) and subjected to qNMR measurements.

The amounts of risperidone and EGCg in precipitates produced when compounds with a xanthine derivative, such as diprophylline, theobromine, and theophylline, were added were lower than the amounts of those in precipitates produced when they were not added. This indicates that such compounds with a xanthine derivative have a solubilizing effect on the precipitate. In contrast, the amounts of risperidone and EGCg in precipitates produced when compounds without a xanthine derivative such as tryptophan, nicotinamide, and cyclo(l-Pro-l-Ala), were added were higher than the amounts of those in precipitates produced when they were not added. This indicates that these compounds without a xanthine derivative did not have a solubilizing effect on the precipitate. Therefore, the solubilizing effect of risperidone in a precipitate might be due to the presence of a xanthine derivative.

CONCLUSION

It can be concluded that when risperidone oral solution is taken together with a tea beverage, the more caffeine in the beverage, the less precipitate of the 1 : 1 complex of risperidone and EGCg forms, and therefore the weaker the effect it has on the efficacy of risperidone. Furthermore, it was found that not only caffeine but also compounds with a xanthine derivative, such as diprophylline, theobromine, and theophylline, have an effect to solubilize the precipitate formed by risperidone oral solution and the tea beverage, thereby reducing the amount of risperidone that precipitates.

Conflict of Interest

The authors declare no conflict of interest.

REFERENCES

• 1)  Rikunabi-yakuzaishi.jp. “Yakuzaishi Hiyari▪Hatto▪Hot”: <https://rikunabi-yakuzaishi.jp/contents/hiyari/008/>, accessed 27 October, 2024.
• 2)  Goromaru T, Fujita K, Mizumoto M, Ishizu T. Quantification of risperidone contained in precipitates produced by tea catechins using nuclear magnetic resonance. Chem. Pharm. Bull., 71, 134–139 (2023).
• 3)  The Ministry of Health, Labour and Welfare, Japanese Pharmacopoeia 18th Edition, p. 151 (2021).
• 4)  Seshadri R, Dhanraj N. Proc. International Flavor Conference, Porto Karras Greece. Frontiers of flavor : proceedings of the 5th International Flavor Conference, Porto Karras, 1–3 July, 1987, Chalkidiki, Greece.
• 5)  Tsutsumi H, Sato T, Ishizu T. Stereochemical structure of the complex of (-)-epigallocatechin 3-gallate and caffeine. Chem. Lett., 41, 1669–1671 (2012).
• 6)  Ishizu T, Tsutsumi H, Kinoshita Y, Mukaida H, Sato T, Kajitani S. Properties of precipitate of creaming down by (-)-epigallocatechin-3-O-gallate and caffeine. Chem. Pharm. Bull., 62, 552–558 (2014).
• 7)  Ishizu T, Tsutsumi H, Sato T. Mechanism of creaming down based on chemical characterization of a complex of caffeine and tea catechins. Chem. Pharm. Bull., 64, 676–686 (2016).
• 8)  Ishizu T. Development of high-order functions using (-)-epigallocatechin-3-O-gallate in water. Chem. Pharm. Bull., 68, 1143–1154 (2020).
• 9)  Zoglio MA, Maulding HV Jr, Windheuser JJ. Complexes of ergot alkaloids and derivatives. I. The interaction of caffeine with ergotamine tartrate in aqueous solution. J. Pharm. Sci., 58, 222–225 (1969).
• 10)  Ikeda H, Moriwaki H, Yukawa M, Iwase Y, Aki H. Mechanism of interaction between risperidone and tea catechin(1)complex formation of risperidone with epigallocatechin gallate. Yakugaku Zasshi, 130, 1589–1595 (2010).
• 11)  Tsutsumi H, Tanabe H, Ishizu T. Chiral recognition of pharmaceuticals having a xanthine skeleton by (-)-epigallocatechin-3-O-gallate in water. Chem. Pharm. Bull., 66, 620–623 (2018).
• 12)  Ishizu T, Tokunaga M, Fukuda M, Matsumoto M, Goromaru T, Takemoto S. Molecular capture and conformational change of diketopiperazines containing proline residues by epigallocatechin-3-O-gallate in water. Chem. Pharm. Bull., 69, 585–589 (2021).