Clarification of the Complexation Behaviour of 2,6-di- O -Methylated ββ -Cyclodextrin and Vitamin E and Radical Scavenging Ability of the Complex in Aqueous Solution

CD–Toc Abstract: The precise understanding of the behaviour of vitamin E ( α -tocopherol; Toc) complexed with cyclodextrin (CD) additives in aqueous solution is a fundamental issue for further development of their aqua-related biological applications. In this study, the solubilisation and complexation behaviours of Toc with methyl-substituted CD derivatives and the radical scavenging ability of the resulting complexes were precisely investigated in water media. Several problems were encountered upon pre-dissolving Toc in an organic solvent prior to the addition to the water media, such as enhancement of the dispersibility and decrease in the complexation capacity. Additionally, dispersions were obtained in some cases when mixing CD and Toc even in the absence of an organic solvent; therefore, to perform the measurements, a transparent solution was prepared via filtration with a nanopore filter. Consequently, unexpectedly, the addition of certain CD methylated derivatives did not always enhance the solubility of Toc significantly. However, 2,6-di- O -methylated β -CD (2,6-DMCD) formed a water-soluble inclusion complex with Toc, effectively enhancing its solubility. A phase solubility study indicated the formation of 1:2 or 1:3 Toc/CD inclusion complexes, and the interaction of 2,6-DMCD with both the chromanol head and the phytol chain of Toc was revealed by 2D ROESY nuclear magnetic resonance analysis. The interaction between 2,6-DMCD and the chromanol head was also confirmed for a 2,6-DMCD–2,2,5,7,8-pentamethyl-6-chromanol inclusion complex. Additionally, a rapid scavenging effect for molecularly dissolved Toc was demonstrated even in a system comprising a chromanol head directly encapsulated by CD. Hence, this work elucidated the precise complexation and radical scavenging ability of 2,6-DMCD–Toc in an aqueous solution, which paves the way for its biological applications.

systems remains limited. As far as we know, no detailed investigation has examined the solubility and complexation behaviour of methylated CD-Toc systems in the absence of organic solvents, which is desirable to gain a deeper insight into the fundamental molecular and anti-oxidative behaviour of the CD-Toc inclusion complex in aqueous systems.
In this study, we describe the effects of CDs, in particular of heptakis 2,6-di-O-methyl -β -CD 2,6-DMCD , on the solubilisation of Toc Fig. 1 , the complexation behaviour with Toc and the radical scavenging ability R.S.A. of the resulting complexes in aqueous solutions. The effect of using an organic solvent 10 vol for pre-dissolving Toc 5 was investigated to understand the different behaviours observed with and without the use of organic solvents. Throughout the study, 2,2,5,7,8-pentamethyl-6-chromanol PMC, Fig. 1 was used for investigating the interaction of the chromanol head with CD and as a reference for the radical scavenging test.

Methods
Electrospray ionisation mass spectroscopy was performed on an Exactive Hybrid Quadrupole Orbitrap mass spectrometer Thermo Fisher Scientific, San Jose, USA in positive mode. The NMR spectra were obtained using a JEOL ECA 500 instrument JEOL Ltd., Tokyo, Japan and deuterated solvents. The numbers of methyl substitutions of 2,6-DMCD, RDMCD and TMβ -CD were calculated on the basis of the integral values in the 1 H NMR spectra. A 2D ROESY NMR experiment was performed in D 2 O at 295 K with a mixing time of 250 ms. Ultraviolet-visible UV-vis absorption spectra were recorded using a V-650 spectrophotometer, JASCO Tokyo, Japan .

Preparation of CD-Toc complex/water systems
To prepare the CD-Toc complex, an excess amount of Toc, D-Toc or the Toc analogue PMC was added to a 4 mL aqueous solution or 10 organic solvent solutions with different CD concentrations. The organic solvent was used for the dissolution of Toc prior to the mixing with CD. After 16 h of stirring at ambient atmosphere, some samples were opaque. Particularly, the CD-added system after mixing exhibited a dispersion state for the organic solvent solutions, whereas the organic solvent-free system afforded a transparent solution with macroscopically segregated Toc domains. This discrimination between the dissolved state and the dispersion state was considered to be greatly important to elucidate the solution behaviour. Therefore, the sample mixtures were filtered through a 0.20 μm nanopore membrane filter DISMIC ® -13HP, Toyo Roshi Kaisha, Ltd., Niigata, Japan to afford sample I, which were transparent for the systems with and without the organic solvent. The Toc concentration in I was determined spectrophotometrically at 291 nm 17,18 .
Additionally, other CD-Toc complex/water samples were prepared using the CD-Toc solids for comparison. The corresponding inclusion solids were prepared via co-precipita- tion and co-evaporation 4, 19, 20 of mixtures of a Toc ethanol solution and 2,6-DMCD aqueous solutions. The preparation and characterisation of the samples are described in Section 2 and Fig. S2 in Supplementary Information. The sample solution prepared by this method was denoted as II.
The equilibrium constant K 1:1 in mM 1 for the 2,6-DMCD-PMC complexes was calculated as Slope/ S 0 1 Slope , where S 0 is the intrinsic solubility of PMC and Slope is the slope of the linear CD-PMC phase solubility diagram 23 .

Radical scavenging test of CD-Toc complex aqueous
solution For the radical scavenging test, the ABTS biradical was used 24 . Briefly, ABTS 117.8 mg was dissolved in water 30.6 mL . After adding potassium persulfate 28.2 mg to activate the ABTS radical, the solution was stored overnight in the dark. The ABTS radical solution was diluted to the appropriate concentration for spectroscopic measurement. An ABTS radical scavenging test was performed by adding each CD solution to the aqueous ABTS solution with and without 10 organic solvent solution. The absorption at 742 nm was measured for 30 min. An aqueous PMC solution in the absence of CD was used as a reference.

Investigation of the solubility enhancement of Toc by
CDs The enhancement of the solubility of Toc by the addition of CD was found to depend highly on the type of CD. No absorption attributable to Toc was observed in the UV-vis spectra in the absence of CDs Fig. 2a , whereas the 2,6-DMCD-added system showed the characteristic absorption of Toc around 291 nm in the water media at a CD concentration of 10 mM, indicating that the Toc solubility was distinctly enhanced by the addition of 2,6-DMCD. The second and third most effective systems were the RDMand Mβ -CD-added systems, respectively. Although other CDs such as β-CD and γ-CD formed highly opaque states in the mixture 7 , trace amounts of Toc were detected in the filtrate Fig. S3 in Supplementary Information , showing their poor solubility enhancement effect.
Consequently, the Toc content in sample I of the α-CD-, β -CD-, γ -CD-, Mβ -CD-, TMβ -CD-, Hβ -CD-and RDMCDadded systems was trace below 0.07 mM even for the RD-MCD-added system , whereas a much higher Toc concentration 0.5 mM was observed for the 2,6-DMCD-added system at a CD concentration of 10 mM. A comparable solubility enhancement was obtained for the 2,6-DMCD-added system in the presence of 100 mM PBS buffer, as confirmed by the Toc absorption observed in the UV-vis spectrum Fig. 2b . Additionally, for the samples prepared by coprecipitation of white powder solids, an apparent solubility enhancement was observed for the 2,6-DMCD-added system Fig. S2 in Supplementary Information . Hence, the 2,6-DMCD-added system enhanced the solubility of Toc in all cases, whereas the addition of other CDs did not always enhance the solubility of Toc significantly. The selection of the suitable CD to provide the appropriate interaction with Toc was crucial, as discussed in the next section.
3.2 Investigation of the CD-Toc complexes formed in water Sueishi et al. proposed the formation of a 1:1 complex of 2,6-DMCD with Toc 5 , in which the interaction of CD and the chromanol ring was unconformable according to a 2D ROESY NMR analysis of a 10 mM 2,6-DMCD and 10 mM Toc mixture in water-mixed media. In their study, an organic solvent acetonitrile was used to dissolve Toc prior to mixing, and no filtration of the mixture was conducted. By contrast, in the present study, the mixture was filtrated to obtain a transparent solution as mentioned in section

2.3.
According to the phase solubility test performed in this study, a non-linear CD/Toc concentration curve with upward tendency was obtained for both 2,6-DMCD and RDMCD-added systems Fig. 3a . In the case of the system without organic solvent, 0.5 mM Toc was dissolved in 10 mM 2,6-DMCD aq, and most of the added Toc approximately 9.5 mM was segregated after 16 h of mixing. Thus, a transparent solution of a mixture of 10 mM 2,6-DMCD and 10 mM Toc could not be obtained Fig. 3a . Moreover, the addition of organic solvents such as methanol and acetonitrile reduced the solubility of Toc Fig. S4 in Supplementary Information , suggesting that the sample reported by Sueishi et al. might contain undissolved species.
Interestingly, twice the concentration of CD was required for the RDMCD-added system to obtain the same Toc concentration level as for the 2,6-DMCD-added system, which revealed the relevance of the methyl substitution at the appropriate positions. Hence, the use of highly purified 2,6-DMCD was important to enhance the Toc solubility in water. Noteworthy, the phase solubility curves were of an A p type according to the definition by Higuchi and Conner 23 , meaning that the solubilisation of the complex was first order with respect to the substrate but second or higher order with respect to the ligand e.g. in 1:2 or 1:3 Toc/CD complexes . This result suggests that the complexation mechanism was different from that proposed by Sueishi et al. Figure 3b shows the 2D ROESY NMR spectra of the 2,6-DMCD-Toc system consisting of a mixture of 20 mM 2,6-DMCD and 2 mM Toc in water in the absence of an organic solvent. In addition to the appearance of intersectional points between the methyl substituents on the chromanol ring and the proton of 2,6-DMCD, cross-peaks between protons on the phytol chain and the sugar proton of 2,6-DMCD 5 were observed for the Toc-added system, which is indicative of the formation of an inclusion complex via interaction of CD with both the chromanol head and the phytol chain of Toc. Thus, it can be concluded that 2,6-DMCD forms an inclusion complex with Toc at 1:2 or 1:3 Toc/CD ratios Fig. 3 , which is most likely responsible for the enhancement of the solubility of Toc. Additionally, the configuration of the phytyl chain did not affect the solubility Fig. S5 in Supplementary Information , suggesting that the interaction is not chirality-dependent. The concentration of the solution was reduced by Ar bubbling treatment Fig. S6 in Supplementary Information , indicating that the complex contributed to the bubble development as a surface active agent.
In the filtrate containing 10 vol methanol or 10 vol acetonitrile, a clear interaction between the chromanol head and CD was also confirmed by 2D ROESY NMR spectroscopy, together with the interaction between the chromanol head and the phytyl chain Fig. S7 in Supplementary  Information ; thus, this interaction must be similar to that of the system without organic solvents. However, this result might be different if the sample at a lower concentration without filtrating the mixture is examined in the presence of an organic solvent because of the formation of different complexes induced by the interface between water and undissolved species or to a reduction of the detection capability of the technique.
The notable solubility behaviour was further confirmed by subjecting the 2,6-DMCD-PMC systems to a phase solubility test. As shown in Fig. 4a, the solubility of PMC was effectively increased for I when using 2,6-DMCD as an additive. A linear increase in the PMC concentration was observed with increasing the CD concentration, affording an A L -type phase solubility diagram 23 that indicates the formation of a 1:1 inclusion complex of PMC and CD. The equilibrium constant K 1:1 was determined to be 4.3 M 1 . A 2D ROESY NMR analysis revealed the presence of a direct interaction between the chromanol ring and the sugar proton of 2,6-DMCD Fig. 4b , supporting the assumption that the inclusion complex was formed via interaction of CD with both the chromanol head of Toc in the 2,6-DMCD-Toc system. In the case of the PMC system, the addition of 10 vol methanol or acetonitrile did not reduce the solubility enhancement effect of 2,6-DMCD, which suggests that these organic solvents preferably interact with the phytyl chain of Toc, preventing to some extent the interaction between 2,6-DMCD and the phytyl chain, thus reducing the solubility of Toc.

ABTS R.S.A. of the Toc and CD mixture
Then, a radical scavenging test was conducted in water 24 . The R.S.A. was expressed as the scavenging ability of PMC in mol, which was evaluated in the water system in the absence of both organic solvent and solubiliser CD and used as a reference. It should be noted that the addition of 2,6-DMCD did not affect the R.S.A. of PMC at the concentrations used in the study. Representative results of the R.S.A. test are shown in Fig. 5.
The addition of a 2,6-DMCD/Toc/water mixture decreased the absorption of the ABTS biradical at 742 nm, revealing the occurrence of radical scavenging effect in water media Fig. 5a . R.S.A. was observed when Toc was present in the sample. Additionally, as shown in Fig. 5b, the 2,6-DMCD-Toc system showed an effective R.S.A. comparable with that of PMC. Noteworthy, the samples I and II for the 2,6-DMCD-Toc systems showed rapid R.S.A. Fig. 5b , indicating that the Toc dissolved by 2,6-DMCD can rapidly access the ABTS biradical, which differs greatly from the R.S.A. of the CD-Toc dispersion systems 7 . Such a system could be used as alternative vehicles to PMC or Trolox, which are often used as artificial water-soluble Toc analogue compounds in biological applications 10, 25 27 .

Conclusion
The formation of CD-Toc inclusion complexes and their radical scavenging effect was demonstrated. Some CD systems formed opaque dispersions, which is more likely when using an organic solvent to dissolve Toc prior to the mixing. To precisely investigate the solution behaviour of the CD-Toc inclusion complexes, filtration was required. Consequently, it was found that the addition of a CD did not always enhance significantly the solubility of Toc, and the selection of the appropriate CD was crucial. 2,6-DMCD was demonstrated to be the most effective solubility enhancer. In aqueous solutions, 2,6-DMCD formed water-soluble 1:2 or 1:3 Toc/CD inclusion complexes via the interaction of both the chromanol ring and the phytol chain of Toc with the CD molecule, which greatly enhanced the solubility of Toc compared with other CDs studied in this work. The solubilised Toc showed rapid access to the ABTS biradical, which renders this system suitable as an alternative vehicle to PMC or Trolox. Studies on the application of the present system as a water-soluble anti-oxidant in cell culture media are currently underway and will be reported in due course.

Supporting Information
This material is available free of charge via the Internet at doi: 10.5650/jos.ess21064