Synthesis of modified cyclic and acyclic dextrins and comparison of their complexation ability

Article abstract of DOI:10.3762/bjoc.10.301

We compared the complex forming ability of α-, β- and -cyclodextrins (α-CD, β-CD and -CD) with their open ring analogs. In addition to the native cyclodextrins also modified cyclodextrins and the corresponding maltooligomers, functionalized with neutral 2

Full text of DOI:10.3762/bjoc.10.301

Synthesis of modified cyclic and acyclic dextrins and
comparison of their complexation ability
Kata Tuza*1, László Jicsinszky1,2, Tamás Sohajda1, István Puskás1 and Éva Fenyvesi1
Letter
Open Access
Address:
Beilstein J. Org. Chem. 2014, 10, 2836–2843.
1CycloLab Cyclodextrin R&D Laboratory Ltd, Illatos út 7, Budapest,
1097, Hungary and 2Dipartimento di Scienza e Tecnologia del
Farmaco, Universitá di Torino, via P. Giuria 9, Turin, 10125, Italy
doi:10.3762/bjoc.10.301
Received: 15 July 2014
Accepted: 14 November 2014
Published: 02 December 2014
Email:
Kata Tuza* - tuza@cyclolab.hu
This article is part of the Thematic Series "Superstructures with
cyclodextrins: Chemistry and applications II".
* Corresponding author
Keywords:
Guest Editor: G. Wenz
acyclodextrin; cyclodextrin; ibuprofen; maltodextrin; maltoheptaose;
maltohexaose; maltooctaose
© 2014 Tuza et al; licensee Beilstein-Institut.
License and terms: see end of document.
Abstract
We compared the complex forming ability of α-, β- and γ-cyclodextrins (α-CD, β-CD and γ-CD) with their open ring analogs. In
addition to the native cyclodextrins also modified cyclodextrins and the corresponding maltooligomers, functionalized with neutral
2-hydroxypropyl moieties, were synthesized. A new synthetic route was worked out via bromination, benzylation, deacetylation
and debenzylation to obtain the 2-hydroxypropyl maltooligomer counterparts. The complexation properties of non-modified and
modified cyclic and acyclic dextrins were studied and compared by photon correlation spectroscopy (PCS) and capillary elec-
trophoresis (CE) using model guest compounds. In some cases cyclodextrins and their open-ring analogs (acyclodextrins) show
similar complexation abilities, while with other guests considerably different behavior was observed depending on the molecular
dimensions and chemical characteristics of the guests. This was explained by the enhanced flexibility of the non-closed rings. Even
the signs of enantiorecognition were observed for the chloropheniramine/hydroxypropyl maltohexaose system. Further studies are
planned to help the deeper understanding of the interactions.
Introduction
The complexing ability of amylose to iodine and fatty acids is al. found that there is complex formation but the complex
well known in the literature [1-3]. The question is raised forming ability of closed ring CDs is 2–3 orders of magnitudes
whether the maltooligomers containing 6, 7, or 8 glucose units higher than that of the non-cyclic, open-chain analogs [5]. Both
(G6, G7 and G8), derived from the three native α-, β-, and maltohexaose and maltoheptaose (G6 and G7) caused spectral
γ-cyclodextrins, respectively, are also suitable complexing changes for guests like p-cresol and methyl orange in the
agents. These maltooligomers are often considered as linear UV–vis region. They also found that using a CPK molecular
dextrins unable to form inclusion complexes [4]. Komiyama et model both G6 and G7 form a favorably cyclic conformation
2836

Beilstein J. Org. Chem. 2014, 10, 2836–2843.
and this explains the weak complex formation contrary to the tion spectroscopy (PCS) is a potent tool to study the aggrega-
smaller-sized non-cyclic analogs (G3, G4 and G5) not showing tion of guest molecules in the presence of the hosts [12].
any interaction. In another study of Bettinetti et al. G7 was
found to wrap up naproxen, taking on a cyclic conformation and In this work first the interaction of some model drugs was
forming a ‘pseudo’ inclusion complex of lower binding constant studied in the presence of the native cyclodextrins and the
than with β-CD [6]. The helical conformations may act as corresponding acyclodextrins. Also, modified cyclodextrins and
‘semicavities’ as proposed by Schurig’s group [7] who found the corresponding maltooligomers functionalized with
even enantioseparation with the acetylated/silylated G6 and G7 2-hydroxypropyl (HP) moieties, that are HP-α-CD, HP-β-CD,
as chiral selectors in gas chromatography. In some cases these HP-γ-CD, HP-G6, HP-G7 and HP-G8 were synthesized and
acyclodextrins showed similar or even better enantioresolution investigated as chiral additives for the separation of model com-
compared to their cyclic counterparts commonly utilized in pounds by CE.
chiral separations [8].
Results and Discussion
One of the most effective techniques to explore if there are We compared the complexing ability of the three native
interactions between guests and hosts – particularly in the case cyclodextrins and their acyclic analogs with CE by calculating
of ionizable guest molecules – is capillary electrophoresis. The their apparent 1:1 binding constants (Kapp) [13]. The results are
change in the mobility of a guest in the electric field in the pres- listed in Table 1.
ence of a host indicates the interaction [9-11].
While some of the model compounds showed no interaction
Another evidence for molecular interaction can be the dimin- with most hosts studied (like vinpocetine or others not
ished aggregation of the guest drug; therefore photon correla- displayed), and some of them complexed by CDs only
Table 1: Kapp values of different dextrin-drug systems (25 °C).a
Guest drug
α-CD
G6
<5
β-CD
G7
<5
γ-CD
G8
<5
2-Phenylethylamine
<5
520
300
Ibuprofen
135
<5
290
<5
2300
200
<5
320
240
280
<5
Vinpocetine
<5
Mefloquine
Nateglinide
<5
<5
85
<5
15
<5
<5
10
300
1400
620
300
aKapp values were calculated with x-reciprocal method [13], results are displayed in M−1 units.
2837

Beilstein J. Org. Chem. 2014, 10, 2836–2843.
(2-phenylethylamine, mefloquine), two of the tested drugs, – yields aggregates sized in the range of 100 nm (Figure 1b).
ibuprofen and nateglinide showed significant interaction with Adding a sub-stoichiometric amount of G8 to the saturated
both the cyclic and acyclic analogs. While the apparent binding ibuprofen solution (Figure 1c), the aggregates were still present.
constant of the ibuprofen/β-CD system is the highest, all the However, by adding G8 in equimolar amount, the aggregates
acyclic derivatives showed significant affinity towards this were no longer detectable; the assumed associate of the drug
guest irrespective of the number of glucopyranose units. Unlike and the maltooligomer exists in molecularly dispersed state.
CDs, the number of the glucose units might not be of primary Based on the results of the PCS analysis, it was concluded that
importance in the case of ibuprofen/acyclodextrin interactions. the interaction of the drug with G8 has an explicit stoichiom-
An even more surprising finding was observed for nateglinide, etry (1:1 mol/mol), which shows similarity to a typical CD-drug
as the highest interaction affinity was determined with the G6 interaction.
acyclic dextrin, followed by β-CD, G7 and α-CD. In general, it
can be concluded that maltooligomers have to be considered as Based on the promising results of the preliminary complexation
potential complexation and interacting agents in the future that studies with some model drugs, modified cyclodextrins and the
may substitute CDs in certain systems.
corresponding acyclodextrins were synthesized (Scheme 1) in
order to compare their complexation behavior.
Systematic aggregation studies were performed to corroborate
the existence of molecular interactions between G8 and The structure elucidation of the O-benzyl intermediary prod-
ibuprofen as an example (Figure 1) using PCS technique. The ucts and the final HP-maltooligomers was performed by NMR
size distribution function of G8 in 1% solution (Figure 1a) (see Supporting Information File 1). Both proton and carbon
shows that G8 does not form aggregates in itself in distilled assignments are based on the DEPT-ed-HSQC spectra.
water, since the peak found at around 1.5 nm well corresponds
to the dimension of a single maltooligomer molecule. A satu- The NMR spectra confirm the complete 1-O-benzylation at the
rated solution of ibuprofen in distilled water – on the other hand glycosidic oxygen (Figures S1–S6, Supporting Information
Figure 1: Aggregate size analysis of aqueous solutions of G8 and ibuprofen by PCS. a) G8 in 1% solution; b) saturated ibuprofen solution;
c) combining ibuprofen and G8 in substoichiometric ratio (1:0.75); d) combining ibuprofen and G8 in stoichiometric ratio (1:1).
2838

Beilstein J. Org. Chem. 2014, 10, 2836–2843.
Scheme 1: Schematic representation of the preparation of HP-substituted maltooligomers.
File 1). Figures S7–S12 confirm the random 2-hydroxypropyla- traoses in the corresponding HP-oligomers was not successful.
tion and the lack of aromatic signals show the successful The chromatogram in Figure S14 (Supporting Information
removal of the benzyl group from the target compounds. From File 1) shows HP-maltohexaose compared to maltohexaose and
proton spectra (Figures S7, S9, and S11, Supporting Informa- O-benzylmaltohexaose. We assumed that the 2-hydroxypropy-
tion File 1) the degree of substitution (DS) values were calcu- lated derivatives have the same purity as the intermediary
lated: 4.1, 4.3 and 4.3 for HP-G6, HP-G7, and HP-G8, respect- benzylated derivatives.
ively.
The affinity of HP-maltooligomers and the corresponding CD
HPLC studies showed that the purity of the intermediary prod- derivatives have been tested towards a group of guest mole-
ucts was 86.5, 79.0 and 75.6% for 1-O-benzylmaltohexaose, cules, selected based on their highly different structure and
1-O-benzylmaltoheptaose and 1-O-benzylmaltooctaose, respect- different ionic properties under these circumstances. The elec-
ively. The possible impurities (maltohexaose, maltopentaose, trophoretic method was the same as the one applied for the
maltotetraose, 1-O-benzylmaltopentaose, 1-O-benzylmaltote- native compounds. The guest candidates were chlorpheni-
traose, etc.) were separated. The chromatogram of 1-O-benzyl- ramine, histidine, ibuprofen, camphoric acid, and chrysan-
maltohexaose is shown as an example (Figure S13, Supporting themic acid. Most of the guest compounds were racemates in
Information File 1). The quantitative evaluation of the chro- order to detect enantiorecognition ability if there is any. In most
matograms are summarized in Table S3 (Supporting Informa- cases the maltooligomers and the cyclodextrin derivatives
tion File 1).
behaved similarly, although the enantiorecognition properties
frequently differed. For example, the enantiomers of chrysan-
The determination of purity of the hydroxypropyl derivatives is themic acid were only recognized by HP-β-CD whereas the
difficult because of the overlapping peaks yielded by products only host resulting in a partial chiral separation of chlorpheni-
having various DS and their α- and β-anomers. Separation of ramine enantiomers was HP-G6 (Figure 2, fifth and first
the possible smaller maltooligomer impurities, like HP-malto- electropherogram, respectively). These preliminary results
heptaoses, HP-maltohexaoses, HP-maltopenaoses, HP-maltote- demonstrate that these derivatives have to be considered
2839

Beilstein J. Org. Chem. 2014, 10, 2836–2843.
Figure 2: Electropherograms of tested model drugs in the presence of 2-hydroxypropylated acyclic and cyclic dextrins (assignments of the peaks are
given in Table 2).
not only as possible complexation agents but also as potential the mobility of the detected peak is the average of that of the
chiral selectors in the future. Electropherograms of 2-hydroxy- free and complexed species weighted by the molar fraction, the
propyl derivatives and five guests are presented below more pronounced this mobility change is, the stronger inter-
(Figure 2).
action is suggested. This indicates that in the case of chlor-
pheniramine (positively charged at this pH) prolongation of the
Conclusions regarding the interaction affinities can be drawn migration time suggests the stronger interaction, while for the
based on Figure 2 and the apparent binding constants. For other guests (negatively charged at this pH) the contrary.
exploring phenomena in Figure 2 it should be noted that the
closer a certain host can mobilize the analyte to the range of the As a demonstration of the results, the association constants
electroosmotic flow (EOF, the migration range of neutral calculated from the electrophoretic mobilities of the five guests
species in the system) the stronger the apparent interaction is. are presented in Table 2. As it was expected the closed ring
The mobility of the detected analytes is based on the charge-to- structure is advantageous for some guest molecules resulting in
mass ratio. When an interaction takes place between a neutral 5–10 times higher association constants as compared to the
host and a charged analyte, the mass of the associate increases proper open structures (e.g., ibuprofen with HP-α-CD/HP-G6
while the average charge decreases. Consequently, the asso- and HP-β-CD/HP-G7). In some cases only a slight difference
ciate migrates closer to the EOF. As in each electropherogram was obtained (e.g., chrysantemic and camphoric acid with
2840

Beilstein J. Org. Chem. 2014, 10, 2836–2843.
Table 2: Kapp values of different dextrin-drug systems (25 °C).a
Guest drug
HP-
α-CD
G6
β-CD
G7
γ-CD
G8
70
Chlorpheniramine (a)
250
220
200
130
75
Histidine (b)
Ibuprofen (c)
<5
<5
<5
< 5
<5
<5
610
110
3500
450
280
210
Chrysanthemic acid (d)
Camphoric acid (e)
150
45
130
55
580
880
610
720
25
165
60
540
aKapp values were calculated with double reciprocal method [13], results are displayed in M−1 units.
HP-β-CD/HP-G7) and an example was found when the open Conclusion
structure was superior to the closed one (chrysantemic acid with Non-substituted or derivatized linear dextrins can interact with
HP-γ-CD/HP-G8).
small biologically active agents. In some cases this interaction
was found to be stronger than or comparable to that with
With chlorpheniramine, CD derivatives yielded in all cases cyclodextrins. The enhanced affinities of the non-cyclic dextrins
inevitably stronger interactions than the corresponding can be explained by their higher flexibility compared to the
maltooligomers. It can be observed, that the hydroxypropylated rigid structure of cyclodextrins. Based on the preliminary find-
CDs and the corresponding acyclodextrins possessed similar ings of this study, the acyclodextrins should be considered as
affinity towards the analyte. In the case of histidine, none of the potential complexation agents and chiral selectors. A deeper
investigated hosts showed detectable affinity. The highest insight into the complexation should reveal the differences in
stability was determined with HP-β-CD for ibuprofen and the the structures of the associates that will support the observed
CDs showed superiority to acyclodextrins. Similar patterns as in tendencies in interaction behavior. The synthesis and investi-
the case of chlorpheniramine could not be observed: the order gations of further derivatives are in progress.
of affinity among CDs and maltooligomers differed signifi-
cantly. As for chrysanthemic acid, the strongest interaction was Experimental
found with the HP-G7 maltooligomer. For this analyte the Materials: Cyclodextrin derivatives, peracetylated maltohexa-,
acyclodextrins showed comparable affinity for the complexa- -hepta-, and -octaoses (>95% purity) were products of Cyclolab
tion to the CD derivatives. Investigating camphoric acid the ten- Ltd., Budapest, Hungary. All other reagents were purchased
dency of chlorpheniramine could be observed, namely that CD from Fluka, Merck and Molar Chemicals and used without
and the maltooligomers with the same number of glucose units further purification. All reactions were monitored by TLC.
had similar affinity towards the analyte except HP-γ-CD and TLC: Merck 5554 Silicagel 60/F254, 5:1 v/v alcohol free chlo-
HP-G8, the latter having much lower affinity.
roform/acetone for bromination, 85:15 v/v alcohol free chloro-
2841

Beilstein J. Org. Chem. 2014, 10, 2836–2843.
form:acetone for benzylation and 10:7 v/v dioxane:satd. aq HP-(DS~4)-cyclodextrins: Identical synthetic methods were
ammonia for hydroxypropylation and debenzylation; visualiza- used, as for the preparation of the corresponding acyclodextrin
tion: 15% cc. H2SO4 in 96% EtOH.
derivatives. DS = 4 was calculated based on the reactant ratios.
Synthesis of HP-maltooligomers
CE equipment: Agilent 7100, capillary: uncoated silica, 25 cm
1-Bromo-peracetyl-maltooligomers, according to [14]: effective length, 20 kV voltage, 50 × 4 mbar·s hydrodynamic
Peracetylated G6, G7 and G8 maltooligomers (3.5 mmol) were injection. Background electrolyte: 30 mM phosphate buffer pH
dissolved in methylene chloride (DCM, 30 mL) at 0 °C, HBr in set to 7.2.
acetic acid (105 mmol, 48% w/v) was added, and the reaction
mixtures were kept at 0 °C for 4 h. The products were isolated PCS Equipment: Malvern Zetasizer Nano ZS, UK
by liquid–liquid extraction (DCM/water). Evaporation yielded
6.1/7.0/7.7 g, respectively (90–95%).
Supporting Information
1-O-Benzyl-peracetyl-maltooligomers with slight modifica-
Supporting Information File 1
tion of [15]: Acetobromo-G6, G7 and G8 maltooligomers
(3 mmol) were dissolved in benzyl alcohol (BnOH, 120 mmol),
freshly prepared dry silver carbonate (6 mmol) was added and
the suspensions were stirred in the dark at room temperature for
8 h. The solids were filtered off and the products were isolated
by precipitation in diisopropyl ether (DIPE). The crude prod-
ucts were purified by chromatography (DCM/acetone 9:1).
Yield: 3.4/3.8/4.2 g (55–60%).
Additional NMR and HPLC data.
[http://www.beilstein-journals.org/bjoc/content/
supplementary/1860-5397-10-301-S1.pdf]
Acknowledgements
The authors would like to thank Katalin Csabai for HPLC
results. Technical assistance of Ibolya Helembai and Klára Sai-
Halász is highly appreciated and acknowledged.
1-O-Benzyl maltooligomers according to a modified method
of [16]: 1-O-Benzyl-G6, G7 and G8 maltooligomer peracetates
(1.5 mmol) were dissolved in methanol (MeOH, 70 mL),
sodium methoxide (4.5 mmol) was added and the solution was
stirred at room temperature for 12 h. The reaction mixture was
evaporated to dryness. The crude products were dissolved in
water and treated with strong cation ion exchanger. The prod-
ucts were obtained by evaporation. Yield: 1.4/1.5/1.7 g
(80–85%).
References
1. Noltemeyer, M.; Saenger, W. Nature 1976, 259, 629–632.
doi:10.1038/259629a0
2. Fanta, G. F.; Shogren, R. L.; Salch, J. H. Carbohydr. Polym. 1999, 38,
1–6. doi:10.1016/S0144-8617(98)00104-0
3. Lay Ma, U. V.; Floros, J. D.; Ziegler, G. R. Carbohydr. Polym. 2011, 83,
1869–1878. doi:10.1016/j.carbpol.2010.10.055
4. Gabelica, V.; Galic, N.; De Pauw, E. J. Am. Soc. Mass Spectrom. 2002,
13, 946–953. doi:10.1016/S1044-0305(02)00416-6
5. Komiyama, M.; Hirai, H.; Kobayashi, K.
Makromol. Chem., Rapid Commun. 1986, 7, 739–742.
doi:10.1002/marc.1986.030071110
1-O-Benzyl-HP (DS~4)-maltooligomers: 1-O-Benzyl-G6, G7
and G8 maltooligomers (0.25 mmol) were dissolved in aqueous
solutions of NaOH (2.5/1.2/1.2 mmol; 1/0.5/0.5 mL water), 1,2-
propylene oxide (1.6 mmol) was added and the solutions were
stirred at 5 °C for 24 h. As the TLC showed no further changes,
the reaction mixtures were treated with strong ion exchanger,
and then evaporated to dryness. The products were isolated by
precipitation in acetone from aqueous solutions. Yield: 0.22/
0.24/0.27 g (~65%).
6. Bettinetti, G. P.; Mura, P.; Melani, F.; Rillosi, M.; Giordano, F.
J. Inclusion Phenom. Mol. Recognit. Chem. 1996, 25, 327–338.
doi:10.1007/BF01045002
7. Sicoli, G.; Jiang, Z.; Jicsinszky, L.; Schurig, V. Angew. Chem. 2005,
117, 4161–4164. doi:10.1002/ange.200500509
8. Sicoli, G.; Pertici, F.; Jiang, Z.; Jicsinszky, L.; Schurig, V. Chirality
2007, 19, 391–400. doi:10.1002/chir.20383
9. Smith, N. W.; Evans, M. B. J. Pharm. Biomed. Anal. 1994, 12,
579–611. doi:10.1016/0731-7085(93)E0012-C
10.Chankvetadze, B.; Endresz, G.; Blaschke, G. Electrophoresis 1994, 15,
804–807. doi:10.1002/elps.11501501113
HP-maltooligomers according to [17]: 1-O-Benzyl-HP G6,
G7 and G8 -maltooligomers (0.15 mmol) were suspended in
EtOH (3 mL), ammonium formiate (0.75 mmol) and 10% Pd/C
(0.1 g) were added and the solutions were boiled for 30 min.
Catalysts were filtered off and washed with water and the
filtrates were treated with strong ion exchanger and charcoal.
Evaporation of solvents yielded 0.1/0.1/0.1 g of solids, respect-
ively (~55%).
11.Chankvetadze, B.; Lindner, W.; Scriba, G. K. E. Anal. Chem. 2004, 76,
4256–4260. doi:10.1021/ac0355202
12.Garnero, C.; Zoppi, A.; Genovese, D.; Longhi, M. Carbohydr. Res.
2010, 345, 2550–2556. doi:10.1016/j.carres.2010.08.018
13.Rundlett, K. L.; Armstrong, D. W. J. Chromatogr., A 1996, 721,
173–186. doi:10.1016/0021-9673(95)00774-1
2842

Beilstein J. Org. Chem. 2014, 10, 2836–2843.
14.Lemieux, R. U. Tetra-O-acetyl-a-D-glucopyranosyl Bromide. In
Methods in Carbohydrate Chemistry; Whistler, R. L.; Wolfrom, M. L.,
Eds.; Academic Press Inc.: New York and London, 1963; Vol. 2,
pp 221–222.
15.Lindberg, B. Acta Chem. Scand. 1949, 3, 151–156.
doi:10.3891/acta.chem.scand.03-0151
16.Thompson, A.; Wolfrom, M. L. Deacetylation. In Methods in
Carbohydrate Chemistry; Whistler, R. L.; Wolfrom, M. L., Eds.;
Academic Press Inc.: New York and London, 1963; Vol. 2, pp 215–220.
17.Jicsinszky, L.; Iványi, R. Carbohydr. Polym. 2001, 45, 139–145.
doi:10.1016/S0144-8617(00)00319-2
License and Terms
This is an Open Access article under the terms of the
Creative Commons Attribution License
(http://creativecommons.org/licenses/by/2.0), which
permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
The license is subject to the Beilstein Journal of Organic
Chemistry terms and conditions:
(http://www.beilstein-journals.org/bjoc)
The definitive version of this article is the electronic one
which can be found at:
doi:10.3762/bjoc.10.301
2843