Improvement of Pharmaceutical Properties of Isoprenoid Compounds through the Formation of Cyclodextrin Pseudorotaxane-Like Supramolecules

Abstract

The purpose of this study was to design cyclodextrin (CyD)-based pseudorotaxane-like supramolecular complexes with various isoprenoid compounds, such as reduced coenzyme Q10 (R-CoQ10), squalene, tocotrienol, and teprenone, and to evaluate their pharmaceutical properties. Squalene, tocotrienol, and teprenone formed precipitates with β-CyD and γ-CyD in aqueous solution, whereas R-CoQ10 formed precipitates with γ-CyD aqueous solution. The results of powder X-ray diffraction and ¹H-NMR analyses indicated that these precipitates are derived from pseudorotaxane-like supramolecular complexes. The photostability of teprenone was markedly improved by complexation with CyDs, especially in the γ-CyD system. In addition, the dispersion rates of teprenone in the γ-CyD system were higher than those in the β-CyD system, compared with the corresponding physical mixtures. In conclusion, pharmaceutical properties such as photostability and dispersion rates of isoprenoid compounds were improved by the formation of pseudorotaxane-like supramolecular complexes with β-CyD and/or γ-CyD.

Isoprenoid compounds play an important role as a physiologically active substance. Therefore, a large number of isoprenoid compounds such as lycopene, β-carotene, vitamin A, D, E, K, teprenone and coenzyme Q10 (CoQ10), etc. have been utilized as foods and active pharmaceutical ingredients.¹⁾ However, solubility of isoprenoid compounds in water is extremely low due to their highly hydrophobic isoprenoid side chain, resulting in a low oral bioavailability.²⁾ Therefore, adequate pharmaceutical formulations are necessary to improve aqueous solubility and oral bioavailability of isoprenoid compounds. In addition, a number of isoprenoid compounds are light-sensitive, thus improvement of their photostability is required.³⁾

Cyclodextrins (CyDs), cyclic oligosaccharides, are able to form inclusion complexes with hydrophobic drugs, improving their water solubility, stability and bioavailability.^4,5) CyDs can also form inclusion complexes with a number of isoprenoid compounds, such as CoQ10,^6–9) vitamin A, D, E, K,^10–12) β-carotene,¹³⁾ lycopene,¹⁴⁾ and teprenone,¹⁵⁾ and then improve their water solubility, stability and bioavailability. Likewise, Uekama and colleagues reported that 2,6-di-O-methyl-β-CyD (DM-β-CyD) or γ-CyD markedly improves solubility and oral bioavailability of CoQ10 in rats, dogs or humans.^16–18) Thus, CyDs have been utilized to improve the pharmaceutical properties of isoprenoid compounds.

CyDs are known to form neckless-like supramolecular assemblies with linear polymers such as polyethylene glycol (PEG) and polypropylene glycol (PPG), namely polypseudorotaxanes.^19–22) It is interesting to note that, CyDs form polypseudorotaxanes in a size-dependent manner, i.e., α-CyD forms polypseudorotaxane with PEG, but not with PPG. Meanwhile, β- and γ-CyDs form polypseudorotaxanes with PPG. Recently, we reported that the pseudorotaxane-like supramolecular complex (PLSC) of CoQ10 with γ-CyD is prepared by co-precipitation or kneading of both compounds in water, because an isoprenoid unit has a similar structure to PPG.^23,24) PLSC formation could be useful to improve the pharmaceutical properties of CoQ10, since γ-CyD forms PLSC with CoQ10 through inclusion complexation with a hydrophobic and unstable isoprenoid unit. In fact, the dispersion rate of CoQ10 from a powder state and the release rate of CoQ10 from hydrophilic ointment were markedly increased by the formation of PLSC with γ-CyD. Thus, a formation of the PLSC is useful to improve pharmaceutical properties of CoQ10. However, few reports are available on the formation of the PLSCs with various isoprenoid compounds except for CoQ10. In the present study, we investigated the PLSCs formation with four kinds of isoprenoid compounds, i.e., reduced CoQ10 (R-CoQ10), squalene, tocotrienol and teprenone (Fig. 1). Photostability and dispersion rates of teprenone in the PLSCs were also investigated.

Fig. 1. Chemical Structures of Isoprenoid Compounds Used in This Study

Experimental

Materials

α-, β- and γ-CyDs were obtained from Nihon Shokuhin Kako (Tokyo, Japan). R-CoQ10, squalene, tocotrienol and teprenone were purchased from Kaneka (Osaka, Japan), Kishimoto Special Liver Oil Co., Ltd. (Tokyo, Japan), Eisai Food & Chemical Co. (Tokyo, Japan) and Wako Pure Chemical Industries, Ltd. (Osaka, Japan), respectively. PPG (average molecular weight=1000) was obtained from Wako Pure Chemical Industries, Ltd. All other chemicals and solvents were of analytical reagent grade and double distilled water was used throughout the study.

Preparation of PLSCs of Isoprenoid Compounds with CyDs

The PLSCs of isoprenoid compounds with CyDs were prepared by the co-precipitation method.^23,24) R-CoQ10, squalene, tocotrienol and teprenone were added to aqueous α-CyD (145 mg/mL), β-CyD (18.5 mg/mL) and γ-CyD (232 mg/mL) solutions at molar ratios of 1 : 7, 1 : 5, 1 : 4 and 1 : 4 (isoprenoid compounds/CyDs), respectively. Here, we previously revealed that γ-CyD forms PLSC with CoQ10 at a molar ratio 5/1, implying that ca. two isoprenoid units tread into one CyD molecule.^23,24) Therefore, a higher ratio of CyD was reacted with isoprenoid compounds to prepare PLSCs. After sonication by a US-4 sonicator (AS ONE, Osaka, Japan) for 30 min, the suspensions were agitated at 25°C for 5 d (R-CoQ10) or 1 d (squalene, tocotrienol and teprenone) under nitrogen gas atmosphere in the dark. Here, reaction rate of CyDs with R-CoQ10 (powder state) was slower than those with the other isoprenoid compounds (liquid state), hence, R-CoQ10 was reacted with CyDs for 5 d. After centrifugation (2193 G, 10 min), the supernatants were removed. The resulting precipitates were dried under reduced pressure overnight. In the case of R-CoQ10, the resulting precipitates were washed with diethyl ether twice to remove the free R-CoQ10. In the case of the other isoprenoid compounds (liquid state), free drugs were easily removed by removing the supernatants.

To determine the changes in a yield of CyD PLSCs with isoprenoid compounds as a function of the ratio of CyDs to isoprenoid compounds, various volumes of 18.5 mg/mL β-CyD aqueous solution or 232 mg/mL β-CyD aqueous solution was added to a certain amount of isoprenoid compounds. Final molar ratios of CyD and isoprenoid compounds were 0–8 (R-CoQ10/CyD), 0–6 (squalene/CyD), 0–5 (tocotrienol/CyD) and 0–6 (teprenone/CyDs), respectively. The subsequent procedure is the same as that described above.

Preparation of Physical Mixtures of Isoprenoid Compounds with CyDs

Physical mixtures of isoprenoid compounds with CyDs were simply prepared by mixing the solid CyD and isoprenoid compounds without solvent. The molar ratios were 1 : 5 (R-CoQ10/CyD), 1 : 3 (squalene/CyD), 1 : 3 (tocotrienol/CyD) and 1 : 2 (teprenone/CyDs), respectively.

Measurements

Powder X-ray diffraction patterns of CyD PLSCs of isoprenoid compounds were measured using a Rigaku RINT 2500 VL X-ray diffractometer (Tokyo, Japan) under the following conditions: Ni-filtered CuKα radiation (1.542 Å), 40 kV, 40 mA, divergent slit of 1.74 mm (1°), scanning slit of 0.94 mm (1°), receiving slit of 0.15 mm, and goniometer angular increment of 1°/min. ¹H-NMR spectra were taken at 25°C on a JEOL JNM-R500 spectrometer operating at 500 MHz, using a 5 mm sample tube. We used deuterated dimethyl sulfoxide (DMSO-d₆) as a solvent, and the DMSO signal was used as an internal reference for ¹H-NMR.

Photostability Study

The powder samples of CyD PLSCs or physical mixtures with teprenone were put into test tubes, and the samples were placed under the light (6000 lx) and nitrogen gas atmosphere. At appropriate intervals, the samples were dissolved in 0.25 mL of DMSO and were diluted with 25 mL of 96% methanol. The content of intact teprenone was measured by HPLC under the following conditions: a Hitachi L-7100 pump, a Hitachi L-7400 UV detector at 214 nm, a Hitachi D-2500 ChromatoIntegrator (Tokyo, Japan), a Waters Symmetry^® C18 column (4.6 mm i.d.×250 mm, Milford, MA, U.S.A.), a mobile phase of methanol–water (96 : 4, v/v), and a flow rate of 1.0 mL/min.

Dispersion Studies

The in vitro dispersion property of teprenone from CyD PLSCs was examined by the dispersed-amount method.^23,24) The powder sample (equivalent to 10 mg teprenone, <100 mesh) was added to 100 mL of degassed water at 37°C. The suspension was stirred at 100 rpm/min. At appropriate intervals, 1.0 mL of the dispersion medium was collected and filtered (MWCO 0.8 µm, DISMIC-25 CS, Toyo Roshi, Tokyo, Japan). The filtrate (0.5 mL) was diluted with 1.0 mL of 96% methanol, and then a concentration of teprenone was measured by HPLC.

Size Distribution Study

The aqueous suspensions containing CyD PLSCs or physical mixtures with teprenone (0.03 mg/mL) were vigorously agitated. The particle sizes of the samples were determined by a Zetasizer nano (Malvern Instruments, Worcestershire, U.K.) at 25°C.

Statistical Analysis

Data are given as the mean±standard error (S.E.). Statistical significance of mean coefficient for the studies was performed by ANOVA followed by Scheffe’s test to perform multiple comparisons. p-Values for significance were set at 0.05.

Results and Discussion

Preparation of PLSCs with Various Isoprenoid Compounds

In general, CyDs form precipitates with linear molecules when both components form polypseudorotaxanes.^19,20,22) Figure 2 shows the photographs of suspensions after mixing the isoprenoid compounds and α-, β- or γ-CyD in water. α-CyD tended to disperse the isoprenoid compounds in water, and formed no precipitates derived from PLSCs. This may be due to the small cavity of α-CyD to include isoprenoid units. On the other hand, β-CyD formed precipitates with squalene, tocotrienol and teprenone, although the yield of the precipitates was negligible in the R-CoQ10 system. In the case of γ-CyD system, all of the isoprenoid compounds formed precipitates with high yield. In the present experimental conditions, R-CoQ10 was added to the CyD solutions in a powder state. In addition, the concentration of the β-CyD solution (18.5 mg/mL) was lower than that of γ-CyD solution (232 mg/mL). Thus, reaction efficiency of the β-CyD/R-CoQ10 system may be lower than that of the γ-CyD/R-CoQ10 system or the β-CyD/other isoprenoid compound systems. Therefore, R-CoQ10 could not efficiently form precipitates with β-CyD. These results suggest that β- and γ-CyDs form PLSCs with the isoprenoid compounds, except for the β-CyD and R-CoQ10 system.

Fig. 2. Photographs of Suspensions Containing Isoprenoid Compounds and CyDs

Powder X-Ray Diffraction

To confirm the formation of PLSCs, powder X-ray diffraction patterns of the precipitates obtained in Fig. 2 were examined (Fig. 3). The diffraction patterns of the precipitates consisting of β-CyD/squalene, β-CyD/tocotrienol and β-CyD/teprenone were different from those of physical mixtures, but roughly the same as that of β-CyD/PPG polypseudorotaxane. In the case of the β-CyD/R-CoQ10 system, it was difficult to measure a powder X-ray diffraction because of its low yield. Likewise, the diffraction patterns of the precipitates consisting of γ-CyD and the isoprenoid compounds were also different from those of physical mixtures, and roughly the same as that of γ-CyD/PPG polypseudorotaxane. Moreover, the diffraction patterns of β- and γ-CyD PLSCs with the isoprenoid compounds were indexed on the basis of the two-dimensional hexagonal and tetragonal unit cells.^24,25) As a result, the observed d-spacings of β- and γ-CyD PLSCs corresponded with the values calculated with the equations deriving from hexagonal and tetragonal structures, respectively (data not shown). These results indicate that β- and γ-CyD PLSCs with the isoprenoid compounds form hexagonal and tetragonal columnar channels, respectively.

Fig. 3. Powder X-Ray Diffraction Patterns of CyD PLSCs with Isoprenoid Compounds

(a) β-CyD, (b) β-CyD/squalene physical mixture, (c) β-CyD/squalene PLSC, (d) β-CyD/tocotrienol physical mixture, (e) β-CyD/tocotrienol PLSC, (f) β-CyD/teprenone physical mixture, (g) β-CyD/teprenone PLSC, (h) β-CyD/PPG polypseudorotaxane, (i) γ-CyD, (j) γ-CyD/R-CoQ10 physical mixture, (k) γ-CyD/R-CoQ10 PLSC, (l) γ-CyD/squalene physical mixture, (m) γ-CyD/squalene PLSC, (n) γ-CyD/tocotrienol physical mixture, (o) γ-CyD/tocotrienol PLSC, (p) γ-CyD/teprenone physical mixture, (q) γ-CyD/teprenone PLSC, (r) γ-CyD/PPG polypseudorotaxane.

Stoichiometry

To estimate the stoichiometry of β- and γ-CyD PLSCs with the isoprenoid compounds, we examined a yield of the PLSCs at the various molar ratios of CyD/isoprenoid compounds²³⁾ (Fig. 4). The yields of γ-CyD PLSC with R-CoQ10 increased as the ratio increased, and reached a plateau at a molar ratio of ca. 5 (γ-CyD/CoQ10). Meanwhile, the yields of β- and γ-CyD PLSCs with squalene, tocotrienol and teprenone reached a plateau at a molar ratio of ca. 3.

Fig. 4. Changes in a Yield of CyD PLSCs with Isoprenoid Compounds as a Function of the Ratio of CyDs to Isoprenoid Compounds

Each point represents the mean±S.E. of 3 experiments.

Then, to clarify the detail of stoichiometry of β- and γ-CyD PLSCs with the isoprenoid compounds, ¹H-NMR spectra of the PLSCs were measured in DMSO-d₆. The peaks of CyDs and isoprenoid compounds were observed in the ¹H-NMR spectra (data not shown), and molar ratios of CyDs/isoprenoid compounds were calculated from the integral values of both components. The number of β- or γ-CyD molecules threaded onto a squalene, tocotrienol or teprenone molecule was estimated to be 2–3 (Table 1). In the case of R-CoQ10, the number of γ-CyD molecule was ca. 5. These results roughly corresponded with those of Fig. 4. Importantly, the length of fully extended CoQ10 molecule is ca. 50 Ų⁶⁾ and the depth of cavity of CyDs is 7.9 Å. These facts indicate that at most 6.3 CyD molecules could be threaded onto a CoQ10 molecule, corresponding to the results of Table 1. Thus, these results suggest that the number of CyDs threaded onto the isoprenoid compounds depends on the length of the isoprenoid side chain. Meanwhile, γ-CyD seemed to include one isoprenoid chain, although γ-CyD and some guest molecules form inclusion complexes with 1 : 2 stoichiometry. In fact, γ-CyD forms polypseudorotaxane with double-stranded PEGs.²⁷⁾ In the case of PPG, γ-CyD, however, it forms solid polypseudorotaxane with single-stranded PPG,²¹⁾ and the isoprenoid chain has similar structure to PPG. Thus, we speculate that γ-CyD mainly forms PLSCs with single-stranded isoprenoid compounds. However, we only determined the stoichiometry of solid PLSCs in the present study. Hereafter, we should clarify detailed stoichiometry of PLSCs in both solid and liquid states.

Table 1. Stoichiometry of CyD PLSCs with Isoprenoid Compounds

Isoprenoid compound	CyD	CyDnumbera)
R-CoQ10	γ-CyD	5.46
Squalene	β-CyD	2.96
	γ-CyD	2.68
Tocotrienol	β-CyD	2.52
	γ-CyD	3.41
Teprenone	β-CyD	2.13
	γ-CyD	2.33

a) The number of CyDs units involved in the PLSCs with isoprenoid compounds.

Photostability

Previously, we examined the pharmaceutical properties of γ-CyD PLSC of CoQ10, having a long isoprenoid chain.^23,24) In the present study, we chose relatively short isoprenoid compounds, i.e., teprenone, and evaluated some pharmaceutical properties of β- and γ-CyD PLSCs with teprenone.

Firstly, to confirm the photostability of teprenone in PLSCs, we examined the time courses for disappearance of teprenone under light (Fig. 5). The stability of teprenone in both β- and γ-CyD PLSCs was markedly improved, compared to the corresponding physical mixtures. In particular, the stabilizing effect of γ-CyD PLSC was superior to that of β-CyD PLSC. According to the results of Table 1, a coverage of γ-CyD to teprenone was higher that of β-CyD. Therefore, γ-CyD PLSC would appear to show a superior photostability to β-CyD PLSC. In addition, a surface area of the powder could affect photostability of teprenone in PLSCs. Therefore, powder morphology of the PLSCs should be clarified.

Fig. 5. Time Courses for Disappearance of Teprenone in CyD Physical Mixture or PLSC

Each point represents the mean±S.E. of 3–8 experiments. * p<0.05, compared with physical mixture.

Dispersion Property

Figure 6 shows the dispersion profiles of teprenone from β- and γ-CyD PLSCs in water at 37°C. Dispersion of teprenone increased in both β- and γ-CyD systems, compared to the corresponding physical mixtures. In the case of β-CyD PLSC, the dispersion of teprenone increased at an early stage, but it decreased thereafter. β-CyD could form PLSC with teprenone at lower concentration than γ-CyD as described above. Therefore, β-CyD may form precipitates derived from PLSC with teprenone released in the medium. On the other hand, γ-CyD PLSC showed a superior dispersion profile to β-CyD PLSC for at least 60 min. Anyhow, β- and γ-CyD PLSCs increased the dispersion of teprenone in water.

Fig. 6. Dispersion Profiles of Teprenone from CyD Physical Mixture or PLSC in Water at 37°C

Each point represents the mean±S.E. of 3–7 experiments. * p<0.05, compared with physical mixture.

Here, we measured the particle sizes of β- and γ-CyD PLSCs with teprenone in water (Table 2). The mean particle sizes were 405 nm and 456 nm in the β- and γ-CyD systems, respectively. In contrast, the teprenone/β-CyD or teprenone/γ-CyD physical mixture formed micro-order particles in water. Thus, we can state that β- and γ-CyD PLSCs form submicron particles, resulting in their high dispersibility.

Table 2. Particle Sizes of CyD PLSCs with Teprenone in Water

Sample	Mean diameter (nm)
Teprenone/β-CyD physical mixture	>5000
Teprenone/β-CyD PLSC	405±10
Teprenone/γ-CyD physical mixture	1868±707
Teprenone/γ-CyD PLSC	456±12

Each value represents the mean±S.E. of 3–4 experiments.

Conclusion

The present study showed that various isoprenoid compounds form the PLSCs with β- and γ-CyD. In addition, β- and/or γ-CyD PLSCs, especially γ-CyD PLSC, improved pharmaceutical properties such as photostability and dispersibility of teprenone. These findings suggest that β- and γ-CyDs may be useful as excipients for isoprenoid compounds.

Acknowledgment

The authors thank Nihon Shokuhin Kako Co., Ltd. (Tokyo, Japan) for providing parent CyDs.

Conflict of Interest

This study was funded by FANCL Corporation (Kanagawa, Japan). Dr. Masatoshi Honjo and Chihiro Iwamoto are researchers of FANCL Corporation (Kanagawa, Japan).

References

• 1) Nowicka B., Kruk J., Biochim. Biophys. Acta, 1797, 1587–1605 (2010).
• 2) Beg S., Javed S., Kohli K., Recent Pat. Drug Deliv. Formul., 4, 245–255 (2010).
• 3) Matsuda Y., Masahara R., J. Pharm. Sci., 72, 1198–1203 (1983).
• 4) Uekama K., Hirayama F., Irie T., Chem. Rev., 98, 2045–2076 (1998).
• 5) Davis M. E., Brewster M. E., Nat. Rev. Drug Discov., 3, 1023–1035 (2004).
• 6) Hatanaka J., Kimura Y., Lai-Fu Z., Onoue S., Yamada S., Int. J. Pharm., 363, 112–117 (2008).
• 7) Prosek M., Butinar J., Lukanc B., Fir M. M., Milivojevic L., Krizman M., Smidovnik A., J. Pharm. Biomed. Anal., 47, 918–922 (2008).
• 8) Uekaji Y., Onishi M., Nakata D., Terao K., Paananen A., Partanen R., Yoshii H., J. Phys. Chem. B, 118, 11480–11486 (2014).
• 9) Miyamoto S., Kawai A., Higuchi S., Nishi Y., Tanimoto T., Uekaji Y., Nakata D., Fukumi H., Terao K., Chem-Bio. Informatics J., 9, 1–11 (2009).
• 10) Pitha J., Life Sci., 29, 307–311 (1981).
• 11) Szejtli J., Bolla-Pusztai E., Tardy-Lengyel M., Szabo P., Ferenczy T., Pharmazie, 38, 189–193 (1983).
• 12) Lin H. S., Chean C. S., Ng Y. Y., Chan S. Y., Ho P. C., J. Clin. Pharm. Ther., 25, 265–269 (2000).
• 13) Anantachoke N., Makha M., Raston C. L., Reutrakul V., Smith N. C., Saunders M., J. Am. Chem. Soc., 128, 13847–13853 (2006).
• 14) Vertzoni M., Kartezini T., Reppas C., Archontaki H., Valsami G., Int. J. Pharm., 309, 115–122 (2006).
• 15) Saruta N., Fukami T., Furuishi T., Suzuki T., Tomono K., Adv. Powder Technol., 21, 326–330 (2010).
• 16) Ueno M., Ijitsu T., Horiuchi Y., Hirayama F., Uekama K., Acta Pharm. Nord., 1, 99–104 (1989).
• 17) Gao X., Nishimura K., Hirayama F., Arima H., Uekama K., Schmid G., Terao K., Nakata D., Fukumi H., Asian J. Pharm. Sci., 1, 95–102 (2006).
• 18) Terao K., Nakata D., Fukumi H., Schmid G., Arima H., Hirayama F., Uekama K., Nutr. Res., 26, 503–508 (2006).
• 19) Harada A., Kamachi M., J. Chem. Soc., Chem. Commun., 1990, 1322–1323 (1990).
• 20) Harada A., Kamachi M., Macromolecules, 23, 2821–2823 (1990).
• 21) Harada A., Okada M., Li J., Kamachi M., Macromolecules, 28, 8406–8411 (1995).
• 22) Harada A., Hashidzume A., Yamaguchi H., Takashima Y., Chem. Rev., 109, 5974–6023 (2009).
• 23) Nishimura K., Higashi T., Yoshimatsu A., Hirayama F., Uekama K., Arima H., Chem. Pharm. Bull., 56, 701–706 (2008).
• 24) Higashi T., Nishimura K., Yoshimatsu A., Ikeda H., Arima K., Motoyama K., Hirayama F., Uekama K., Arima H., Chem. Pharm. Bull., 57, 965–970 (2009).
• 25) Ohmura M., Kawahara Y., Okude K., Hasegawa Y., Hayashida M., Kurimoto R., Kawaguchi A., Polymer, 45, 6967–6975 (2004).
• 26) Di Bernardo S., Fato R., Casadio R., Fariselli P., Lenaz G., FEBS Lett., 426, 77–80 (1998).
• 27) Harada A., Li J., Kamachi M., Nature (London), 370, 126–128 (1994).