Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2025 Apr 9;147(16):13851–13858. doi: 10.1021/jacs.5c02055

Enzyme-Mediated Dynamic Combinatorial Chemistry Enables Large-Scale Synthesis of δ-Cyclodextrin

Kasper H Hansen 1, Andreas Erichsen 1, Dennis Larsen 1, Sophie R Beeren 1,*
PMCID: PMC12022984  PMID: 40202199

Abstract

graphic file with name ja5c02055_0006.jpg

α-, β-, and γ-cyclodextrins (CDs) are macrocycles formed from six, seven, and eight α-1,4-linked d-glucopyranose units and are industrially produced on ton scales for use as hosts for bioactive guests in foods, cosmetics, and pharmaceuticals. Large-ring cyclodextrins, with more than eight glucose units, have been known for decades but never isolated in more than milligram quantities. We report a scalable method to synthesize δ-CD, formed from nine glucose units, in high yield (>40%), high purity (>95% purity without chromatography), and unprecedented quantities (multigram scale). We exploit a superchaotropic dodecaborate template, B12Cl122–, to direct the selective synthesis of δ-CD from within an enzyme-mediated dynamic combinatorial library of interconverting cyclodextrins. Our single-step reaction uses a recyclable template, cheap starting materials, and a commercial ‘food-grade’ enzyme and can thus give access to large quantities of δ-CD. This work will enable the first large-scale investigations of the properties and applications of this little-known larger CD.

Introduction

Cyclodextrins (CDs), or more specifically, α-, β-, and γ-CD, are the most extensively studied and widely utilized macrocyclic host molecules available.1−3 These cyclic oligosaccharides, formed from six, seven, and eight α-1,4-linked d-glucopyranose units, were discovered more than a century ago, and their structures, properties, and derivatization have been rigorously studied throughout the years, especially since the 1970s when their industrial ton-scale production became possible.1−3 These three CDs, and their artificially modified analogues, bind hydrophobic and chaotropic guests,4,5 which has made them ubiquitous hosts in the field of supramolecular chemistry. Moreover, they serve to encapsulate, solubilize, stabilize, and deliver bioactive guests for a wide array of applications in the pharmaceutical, food, and cosmetics industries.6−8 Amid vast scientific and industrial advances, the basic CD toolbox, consisting of α-, β-, and γ-CD, has remained unchanged. Large-ring CDs, formed from more than eight glucose units, were discovered many years ago,9−12 but until now, no scalable methods to produce larger CDs have been developed. The organic synthesis of δ-CD, formed from nine glucose units, has been achieved in a 26-step synthesis starting from maltotriose,13 and smaller CDs with 3–5 glucose units have also been accessed using elaborate organic synthesis.14,15 In this work, we present the first multigram synthesis of a large-ring CD, namely, δ-CD or cyclomaltononaose (Figure 1).

Figure 1.

Figure 1

Open in a new tab

δ-CD synthesis from an enzyme-mediated dynamic combinatorial library (DCL). (a) Schematic showing the DCL achieved by the action of CGTase on α-CD (or any other α-1,4-glucan source) and the effect of using B12Cl122– as a template. (b) Reaction scheme for the synthesis of δ-CD from α-CD. (c) Structures of α-, β-, γ-, and δ-CD in space-filling models based on reported X-ray crystallography,37,52−54 along with truncated cone illustrations that represent the shape and size of the interior surface of the CDs,48 as well as idealized icosahedral structures of the dodecaborate anions. (d) Chromatograms (HPLC-ELSD) showing the product distribution obtained at equilibrium in the reaction of α-CD (10 mg/mL) with CGTase in the absence and presence of Na2B12Cl12, Na2B12Br12, or Na2B12I12 (5 mM). G1 = d-glucose, G2 = maltose, G3 = maltotriose, etc.

The industrial production of α, β, and γ-CD is carried out via the enzymatic breakdown of starch by cyclodextrin glucanotransferase (CGTase, EC 2.4.1.19).16,17 While α, β, and γ-CD are the major products of this reaction, small amounts of large-ring CDs are also formed, mainly in the early stages of the reaction.18 Large-ring CDs with 9–31 glucose units have been isolated from complex CD mixtures by chromatography,19−24 and recently, the Zimmermann group achieved increased selectivity for large-ring CD production using engineered CGTases.25,26 CGTase catalyzes both fast, reversible transglycosylation and slow hydrolysis of α-1,4-glucosidic linkages.16,27

We have previously shown how the action of CGTase on an α-1,4-glucan generates a dynamic mixture of interconverting CDs and linear α-1,4-glucans—a dynamic combinatorial library (DCL).28 The CDs form a kinetically trapped subsystem that operates under pseudo-thermodynamic control wherein the product distribution reflects the relative stabilities of the CDs, and this distribution can be altered by the addition of templates that bind to and stabilize specific CDs.28−36 We have recently utilized this system with a bolaamphiphile template to achieve the enzymatic synthesis of δ-CD in a 7% yield, where δ-CD was isolated from a biased mixture of CDs using preparative HPLC.33

While accumulated knowledge of the binding properties of δ-CD is still quite limited, a handful of studies have shown that δ-CD can form weak inclusion complexes with cycloundecanone,37 I3–, sodium dodecasulfate,38 some steroidal compounds,20 and some hydrophobic drugs.39 It can help to solubilize fullerenes,40−43 discriminate between enantiomers of small racemic drugs,44−46 and form [2]-, [3]-, and [4]-pseudorotaxanes with bolaamphiphiles with alkyl chain axles.33 Of particular relevance to this work, Nau and co-workers have discovered the extremely high-affinity binding of superchaotropic halogenated dodecaborate cluster dianions, B12Cl122–, B12Br122–, and B12I122–, to γ-, δ-, and ε-CD, formed from 8, 9, and 10 glucose units, respectively.47,48 The strong binding of polyoxometallates to γ-CD has also been attributed to the chaotropic effect.49−51

Herein, we present a scalable, dynamic enzymatic synthesis of δ-CD that employs Na2B12Cl12 as a recyclable template to amplify δ-CD from within a dynamic mixture of interconverting CDs (Figure 1a,b). α-CD can be converted to δ-CD in 40% yield in a single reaction step to obtain a high-purity product (>95% purity) without the need for chromatography. Detailed binding studies using NMR spectroscopy and isothermal titration calorimetry (ITC) show how γ-, δ-, and ε-CD form various 1:1 and 2:1 complexes with the halogenated dodecaborate cluster dianions. Simulations of the DCL product distributions reveal how the delicate interplay between relative binding strengths, binding stoichiometry, and the relative stabilities of the free cyclodextrins can explain the extremely selective templating effect achieved using B12Cl122–.

Results and Discussion

Enzyme-Mediated DCLs of CDs with Halogenated Dodecaborate Templates

A series of enzyme-mediated DCLs were set up to test the ability of dodecaborate anions to act as templates for the enzymatic synthesis of large-ring CDs. α-CD (10 mg/mL) was treated with CGTase in the presence of Na2B12Cl12, Na2B12Br12, or Na2B12I12 (5 mM) in phosphate-buffered water (50 mM, pH 7.5) at room temperature. The distribution of products formed in these DCLs was monitored over time using HPLC with evaporative light scattering detection (ELSD) (Figures S11, S13, and S14). Figure 1d displays chromatograms showing the distributions of α-1,4-glucan products formed once equilibrium had been reached (48 h for the templated libraries and 6 h for the untemplated library). In the untemplated library, α-CD, β-CD, and γ-CD were formed as the main products, and only trace amounts of large-ring CDs were detected. In the presence of Na2B12Cl12, remarkably selective production of δ-CD occurred such that δ-CD accounted for 84% of the CD composition. Some short linear glucans and glucose were also formed, which is due to an unavoidable background hydrolysis reaction.28 In both the DCLs templated with Na2B12Br12 and Na2B12I12, the dominant product was γ-CD, but a significant proportion of δ-CD (35%) was formed in the presence of Na2B12Br12 and small amounts of both δ-CD (7%) and ε-CD (9%) were formed in the library templated with Na2B12I12.

The highly selective conversion of α-CD to δ-CD seen in the Na2B12Cl12-templated DCL is comparable with some of the best template-directed macrocycle syntheses reported to date.55,56 It is especially remarkable, given the structural similarities between the different-sized CDs and the reported high affinity of each dodecaborate anion for several different CDs.47,48 In order to fully understand these templating effects, therefore, we decided to carry out a thorough investigation of the binding interactions of Na2B12Cl12, Na2B12Br12, and Na2B12I12 with γ-CD, δ-CD, and ε-CD, using a combination of 1H NMR spectroscopy and ITC. For this purpose, δ-CD and ε-CD were isolated from preparative-scale DCLs templated with Na2B12Cl12 and Na2B12I12, respectively, using preparative HPLC with a hydrophilic interaction liquid chromatography (HILIC) column.

Investigation of the Binding Interactions between CDs and Dodecaborate Anions

1H NMR titrations were performed by titrating the dodecaborate salts Na2B12X12 (X = Cl, Br, or I) into D2O solutions of γ-CD, δ-CD, and ε-CD. Previous work has shown that α- and β-CD only exhibit weak nonspecific interactions to these dodecaborate salts,47,48 which was confirmed for B12Cl122– (Figures S18–S21). In the titrations of δ-CD with each of the dodecaborate salts, binding was observed in slow exchange on the NMR chemical shift timescale (Figures 2a and S28–S33). Strong binding interactions with 1:1 stoichiometry were clearly indicated with affinities higher than could be determined using 1H NMR titrations (i.e., above ca. 106m–1), and so, the association constants (Ka) were determined using ITC (Figures 2f and S45–S47). Consistent with previous reports,48 the highest affinity for δ-CD was found for B12Br122– (Ka = 1.2 × 107 M–1), followed by B12Cl122– (Ka = 8.7 × 106 M–1) and B12I122– (Ka = 9.2 × 105 M–1) (Table 1). In the NMR titrations of ε-CD with Na2B12Cl12 and Na2B12Br12, binding was observed in a fast exchange regime, and moderate to strong 1:1 binding was determined (Ka = 2.7 × 104 M–1 and Ka = 2.5 × 105 M–1, respectively) (Figures S34–37). Very strong binding of ε-CD to Na2B12I12 was seen in slow exchange, so again we turned to ITC to determine the 1:1 association constant (Ka = 3.4 × 106 M–1) (Figures S38, S39, and S48).

Figure 2.

Figure 2

Open in a new tab

Binding studies with CDs and dodecaborates. (a–c) Partial 1H NMR spectra showing the anomeric region from the titrations of δ-CD (0.01 mM) with Na2B12Cl12, γ-CD (0.3 mM) with Na2B12Cl12, and γ-CD (0.01 mM) with Na2B12Br12 in D2O, respectively. (d,e) Partial DOSY NMR spectra showing the anomeric region of γ-CD alone and in the presence of Na2B12Cl12 and Na2B12Br12, respectively. (f) Baseline-corrected heat rate and fit obtained for the ITC titration of δ-CD with Na2B12Cl12 in H2O. (g) Fit to the changes in chemical shift (Δδ) during the titration of γ-CD with Na2B12Cl12. (h) Fit obtained in the titration of γ-CD with Na2B12Br12.

Table 1. Association Constants (Ka) for Binding of Dodecaborate Anions to γ-CD, δ-CD, or ε-CDa.

  Ka (M–1) for dodecaborate guestsb
Host   B12Cl122– B12Br122– B12I122–
γ-CD Ka (1:1) 2.6 × 104 5.5 × 104 2.4 × 105
  Ka (2:1) 1.9 × 104 2.1 × 106 2.1 × 104
δ-CDc Ka (1:1) 8.7 × 106 1.2 × 107 9.2 × 105
ε-CD Ka (1:1) 2.7 × 104 2.3 × 105 3.4 × 106c
Open in a new tab
a

Determined by 1H NMR spectroscopy titration in D2O, unless stated otherwise. Uncertainties on the fits are given in Section S9.

b

All as their disodium salts.

c

Determined by ITC in H2O.

The 1H NMR titrations with γ-CD showed greater complexity and revealed a hitherto undetected47 2:1 binding stoichiometry in the solution phase ((γ-CD)2·B12X122–). In the titrations with Na2B12Cl12 and Na2B12I12, binding was observed in fast exchange and the presence of complexes with two different stoichiometries was clearly indicated by the changing direction of the chemical shift perturbations upon addition of dodecaborate anions (Figures 2b, S22, S23, S26, and S27). Job plots likewise indicated 2:1 binding interactions, and in both cases, the chemical shift changes fit well to a 2:1 binding model (Figure 2g). For B12Cl122, moderate 1:1 and 2:1 binding constants were measured (Ka(1:1) = 2.6 × 104 M–1, Ka(2:1) = 1.9 × 104 M–1), while for B12I122–, the first binding was an order of magnitude higher (Ka(1:1) = 2.4 × 105 M–1, Ka(2:1) = 2.1 × 104 M–1). DOSY NMR experiments were performed on solutions of γ-CD and Na2B12Cl12 at varying ratios, showing that with substoichiometric equivalents of Na2B12Cl12, the observed diffusion coefficient was smaller than with higher equivalents of Na2B12Cl12, which is consistent with the formation of the (γ-CD)2·B12Cl122– and γ-CD·B12Cl122– complexes in fast exchange (Figures 2d and S41).

In the titration with Na2B12Br12, we observed an unusual situation, where the formation of the 1:1 complex was seen in fast exchange, while the 2:1 complex was formed in slow exchange (Figures 2c and S24 and S25). DOSY NMR experiments showed clearly the formation of two discrete complexes, γ-CD·B12Br122– and (γ-CD)2·B12Br122– with different diffusion coefficients (Figures 2e and S42). By monitoring both the chemical shift changes and the integrals of each signal (and building upon methodology we have developed previously to handle a similar situation albeit with 1:1 and 1:2 binding in fast and slow exchange),33 we could determine association constants, Ka(1:1) = 5.5 × 104 M–1 and Ka(2:1) = 2.1 × 106 M–1 (Figure 2h and Section S11). The strong and highly cooperative 2:1 binding of γ-CD to B12Br122– suggests that two γ-CD molecules can optimally encapsulate and thereby desolvate one molecule of B12Br122–, alleviating the superchaotropic effect the dodecaborate ion exerts on the bulk water.57 This solution-state binding stoichiometry is consistent with the reported crystal structure for this complex.47 It is noteworthy that for the complexation of γ-CD with all three dodecaborates, large downfield shifts of the H3 but not the H5 proton signals were observed (Figure S40). This is indicative of shallow binding where only the protons inside the CD cavity close to the wide rim are perturbed and consistent with the formation of a 2:1 sandwich complex ((γ-CD)2·B12X122–).58 In contrast, for the complexation of δ-CD with B12Cl122– and B12Br122–, large downfield shifts for the H5 protons, which are located near the narrow rim of the CD, were observed, indicating that the guest penetrates deeply into the cavity, as would be expected for a strong 1:1 complex.

DCL Simulations

We were surprised to find that although B12Cl122– appeared to be the most efficient template for the enzymatic synthesis of δ-CD, B12Br122– had the highest affinity for δ-CD. However, DCLs are complex networks of interconnected equilibria and can exhibit counterintuitive behavior.59 The product distribution reached at equilibrium is representative of the lowest energy of the entire system and, thus, is influenced not only by the strength of the individual binding interactions but also by the stoichiometry of the binding interactions, the number of building blocks (glucose units) required to assemble each library member, and the intrinsic stability of the different library members. To rationalize the observed product distributions, therefore, we used the DCLSim software developed in the Otto and Sanders groups to simulate DCLs of CDs according to the model shown in Figure 3a.59 We utilized our measured binding constants and determined relative formation constants for α-, β-, γ-, δ-, and ε-CD from the product distribution in an untemplated library ( Section S13). The DCLs were simulated at a total CD concentration of 4 mg/mL as it takes up to 48 h for the DCLs templated with dodecaborate anions to reach equilibrium, and by this time, approximately 60% of the CDs have been hydrolyzed to form G1–G4. Previous work has shown that pseudo-thermodynamic control operates for the CD subsystem, and so the linear glucans can be disregarded in the simulation.28Figure 3b shows a comparison between the simulated and experimental product distributions obtained for DCLs templated with 5 mM Na2B12Cl12, Na2B12Br12, and Na2B12I12.

Figure 3.

Figure 3

Open in a new tab

(a) DCLSim model used to simulate the product distribution in DCLs templated with dodecaborates. (b) Comparison between observed CD distributions at 48 h (Obsd.) and the simulated values (Sim.) at the same total CD concentration for the untemplated DCL and DCLs templated with 5 mM Na2B12Cl12, Na2B12Br12, or Na2B12I12. (c) DCLSim predictions of CD distribution (lines) and observed CD distributions (data points) as a function of template concentration, assuming a total CD concentration of 4 mg/mL.

A very good correlation between our experimentally determined yields of γ-, δ-, and ε-CD and the simulated results was found, which confirms the validity of the binding constants and stoichiometries that we determined. Compared to the simulations, marginally higher concentrations of α-CD and β-CD were obtained in the experiments, but this can be explained by the fact that the system is not truly at equilibrium. Short linear glucans are constantly exiting and re-entering the kinetically trapped CD subsystem, and there is a kinetic preference for them to form smaller CDs before eventual conversion to the template-stabilized larger CDs.28,36

Simulated DCLs were generated with varying concentrations of each template, and the predicted distributions of α-CD−ε-CD are plotted in Figure 3c (lines). In the DCLs templated with B12Br122– and B12Cl122–, we found that there exists a competition between the formation of 1:1 complexes with δ-CD or 2:1 complexes with γ-CD. For B12Br122–, the 2:1 complex with γ-CD is highly stable, and this results in the production of similar quantities of γ-CD and δ-CD when a 5 mM template concentration is employed. Simulations predict that with fewer equivalents of the template, γ-CD production will be favored, while increasing the template concentration favors δ-CD, albeit never reaching the yields of δ-CD obtained with B12Cl122– (Figure S54). For DCLs templated with B12Cl122–, the highly selective production of δ-CD can be explained by the fact that B12Cl122– forms a much weaker 2:1 complex with γ-CD. The simulation indicated that 5 mM Na2B12Cl12 would be sufficient to give a high δ-CD yield. This result was confirmed experimentally by setting up a series of DCLs with different concentrations of the Na2B12Cl12 template. A sharp decrease in δ-CD yield was seen below 4 mM Na2B12Cl12, and it was observed that higher template concentrations (8–10 mM) caused slower evolution of the libraries and no benefit to the δ-CD yield (Figure 3c (data points) and Figure S12). For DCLs templated with B12I122–, simulations predicted the formation of mainly γ-CD and only very small quantities of δ-CD and ε-CD even at high template concentrations. In the untemplated DCL, δ-CD and ε-CD make up less than 1% and 0.5% of the total CD concentration, respectively, while γ-CD accounts for 12.5% (Section S13). The product distribution in the untemplated DCL reflects the intrinsic stability of each library member. The concentration of library members in templated DCLs is then a function of their stability relative to one another and their relative affinity for the template. It seems that the affinity of B12I122– for these large-ring CDs is insufficiently high to cause their amplification in the DCL due to competitive binding by γ-CD and the much higher intrinsic stability of γ-CD.

Large-Scale Synthesis of δ-CD

Having determined the optimal template concentration for δ-CD production, the next step toward a scalable synthesis was to optimize the reaction time. The evolution of the product distribution in the DCL templated with 5 mM Na2B12Cl12 was monitored closely. Figure 4a displays chromatograms showing the gradual conversion of α-CD to β-CD, γ-CD, and finally δ-CD over 48 h. The changing CD distribution and concentrations of all glucans in the mixture (including linear and cyclic species) are plotted in Figure 4b. The highest selectivity toward δ-CD was observed after 48 h once the system had reached equilibrium and almost all γ-CD had been converted to δ-CD. However, the highest yield of δ-CD was obtained after 24 h as significant hydrolysis occurred during the second day, leading to a buildup of G1–G4 and a lower total CD concentration. Since ease of purification is a critical consideration when developing a scalable synthesis, we judged that a longer reaction time is optimal, despite a small sacrifice in δ-CD yield, as the separation of short linear glucans from δ-CD by precipitation is relatively trivial, while the separation of γ-CD from δ-CD is challenging and requires chromatography.

Figure 4.

Figure 4

Open in a new tab

Optimization of δ-CD synthesis. (a) Chromatograms (HPLC-ELSD) showing the evolution of a DCL wherein α-CD was treated with CGTase in the presence of Na2B12Cl12 (5 mM) in sodium phosphate buffer at pH 7.5. G1 = d-glucose, G2 = maltose, G3 = maltotriose, etc. (b) Distribution of glucans formed in the DCL as a function of time: (top) relative concentrations of CDs formed; (bottom) concentrations of all glucans formed. Smoothed lines are added to guide the eye.

We have thus identified a scalable process for δ-CD production that yields high-purity δ-CD without the use of chromatography and with efficient template recovery (Figure 5).60 The reaction is carried out by combining α-CD (10.0 g), Na2B12Cl12 (3.0 g), and a commercial CGTase stock solution in water (1 L) at pH 7.5 and at 30 °C. After 42 h, the reaction mixture is boiled for 15 min to denature the enzyme, concentrated to 20% of the original reaction volume, and then δ-CD is precipitated by addition of acetone. Purification by three reprecipitations gives δ-CD in at least 40% yield with a purity of at least 95%, according to 1H NMR spectroscopy (Figure S1). 11B NMR spectroscopy is used to confirm the absence of any template in the precipitate. δ-CD with greater than 99% purity can be obtained by further purification using preparative HPLC with a HILIC column (Figure S3). Importantly, the template can be recovered with 90% efficiency and used in subsequent reaction cycles. The filtrate obtained after precipitation of δ-CD is concentrated and acidified with aqueous HCl. Et3N is added to precipitate [Et3NH]2[B12Cl12], and then NaOH is used for conversion to the water-soluble Na2B12Cl12. The template has been used in at least 3 sequential reaction cycles with no decrease in the efficiency of δ-CD production. Additionally, the purity of the recovered template was confirmed using 1H and 11B NMR spectroscopy (Figure S7).

Figure 5.

Figure 5

Open in a new tab

Schematic of the scalable reaction process for δ-CD synthesis from an α-glucan source, including the template recovery process (in blue arrows).

To further showcase the utility of Na2B12Cl12 for the enzymatic synthesis of δ-CD, we tested starch as the α-1,4-glucan starting material as starch is the raw material typically used in the industrial synthesis of small CDs and would be a logical starting material for industrial-scale synthesis of δ-CD. Performed at an analytical scale, the reaction proceeded in a similar manner as when starting from α-CD (Figure S15). After 24 h, δ-CD was the main cyclic product. Small quantities of 6-O-α-d-glucopyranosyl-δ-CD were also formed, which can be explained by the fact that amylopectin, the main component of starch, has a dendrimer-like structure with long α-1,4-glucan chains connected by α-1,6-linked branching points. This branched CD side product could be avoided by addition of two debranching enzymes, isoamylase and pullulanase, alongside CGTase during the synthesis (Figure S16).

Conclusions

In summary, we exploited the chaotropic effect in an enzyme-mediated DCL to access unprecedented quantities of a large-ring cyclodextrin. Where δ-CD has previously been obtained in milligram quantities,13,20,33 we have synthesized grams of this rare macrocycle using a cheap starting material (α-CD), a ‘food-grade’ enzyme, and a recyclable template. Due to the unusually high selectivity of this template-directed reaction, δ-CD can be isolated in >95% purity and >40% yield by simple precipitation. The template can be recovered by precipitation with high efficiency, and the recovered template has been used successfully for several subsequent δ-CD syntheses on milligram to 10 g scales. Careful analysis of the binding interactions between B12Cl122–, B12Br122–, and B12I122– and γ-CD, δ-CD, and ε-CD, and simulation of product distributions in DCLs, revealed that the high selectivity of this B12Cl122–-templated synthesis results as much from the high affinity of B12Cl122– for δ-CD as from the lack of a competing strong 2:1 binding to γ-CD. We believe that our methodology could eventually enable the industrial-scale production of δ-CD so that in the future, δ-CD may find its place alongside α-CD, β-CD, and γ-CD, as a versatile host for a wide range of applications.

Acknowledgments

High-field NMR spectra were acquired at the NMR Center DTU funded by the Villum Foundation. We thank Dr. Kasper Enemark-Rasmussen for help with DOSY experiments, and we thank Prof. Peter Westh for the use of his ITC and Marta Iglesia Escarpizo-Lorenzana for assistance with the instrument.

Glossary

Abbreviations

CD

cyclodextrin

CGTase

cyclodextrin glucanotransferase

DCL

dynamic combinatorial library.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacs.5c02055.

  • Experimental procedures, HPLC reaction monitoring of enzyme-mediated DCLs, NMR titrations, ITC experiments, details of simulations, and characterization data to support the conclusions (PDF)

Author Contributions

† K.H.H. and A.E. contributed equally. The manuscript was written through contributions of all authors.

Financial support is gratefully acknowledged from the European Research Council (ERC-Starting Grant ENZYME-DCC), Villum Foundation (grant no. 42091), Carlsberg Foundation (CF19-0510), and the Independent Research Fund Denmark (grant number 1054-00082B).

The authors declare no competing financial interest.

Supplementary Material

ja5c02055_si_001.pdf (6.6MB, pdf)

References

  1. Morin-Crini N.; Fourmentin S.; Fenyvesi É.; Lichtfouse E.; Torri G.; Fourmentin M.; Crini G. 130 years of cyclodextrin discovery for health, food, agriculture, and the industry: a review. Environ. Chem. Lett. 2021, 19, 2581–2617. 10.1007/s10311-020-01156-w. [DOI] [Google Scholar]
  2. Crini G. Review: A history of cyclodextrins. Chem. Rev. 2014, 114, 10940–10975. 10.1021/cr500081p. [DOI] [PubMed] [Google Scholar]
  3. Szejtli J. Introduction and general overview of cyclodextrin chemistry. Chem. Rev. 1998, 98, 1743–1753. 10.1021/cr970022c. [DOI] [PubMed] [Google Scholar]
  4. Rekharsky M. V.; Inoue Y. Complexation thermodynamics of cyclodextrins. Chem. Rev. 1998, 98, 1875–1917. 10.1021/cr970015o. [DOI] [PubMed] [Google Scholar]
  5. Assaf K. I.; Nau W. M. Large anion binding in water. Org. Biomol. Chem. 2023, 21, 6636–6651. 10.1039/D3OB00975K. [DOI] [PubMed] [Google Scholar]
  6. Del Valle E. M. M.; Cyclodextrins E. M. M. Cyclodextrins and their uses: a review. Process Biochem. 2004, 39, 1033–1046. 10.1016/s0032-9592(03)00258-9. [DOI] [Google Scholar]
  7. Sharma N.; Baldi A. Exploring versatile applications of cyclodextrins: An overview. Drug Delivery 2016, 23, 729–747. 10.3109/10717544.2014.938839. [DOI] [PubMed] [Google Scholar]
  8. Hedges A. R. Industrial applications of cyclodextrins. Chem. Rev. 1998, 98, 2035–2044. 10.1021/cr970014w. [DOI] [PubMed] [Google Scholar]
  9. Larsen K. L. Large cyclodextrins. J. Inclusion Phenom. Macrocyclic Chem. 2002, 43, 1–13. 10.1023/A:1020494503684. [DOI] [Google Scholar]
  10. Saenger W.; Jacob J.; Gessler K.; Steiner T.; Hoffmann D.; Sanbe H.; Koizumi K.; Smith S. M.; Takaha T. Structures of the common cyclodextrins and their larger analogues – beyond the doughnut. Chem. Rev. 1998, 98, 1787–1802. 10.1021/cr9700181. [DOI] [PubMed] [Google Scholar]
  11. Endo T. Large-ring cyclodextrins. Trends Glycosci. Glycotechnol. 2011, 23, 79–92. 10.4052/tigg.23.79. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Ueda H. Physicochemical properties and complex formation abilities of large-ring cyclodextrins. J. Inclusion Phenom. Macrocyclic Chem. 2002, 44, 53–56. 10.1023/A:1023055516398. [DOI] [Google Scholar]
  13. Wakao M.; Fukase K.; Kusumoto S. Chemical synthesis of cyclodextrins by using intramolecular glycosylation. J. Org. Chem. 2002, 67, 8182–8190. 10.1021/jo025887r. [DOI] [PubMed] [Google Scholar]
  14. Nakagawa T.; Ueno K.; Kashiwa M.; Watanabe J. The stereoselective synthesis of cyclomaltopentaose. A novel cyclodextrin homologue with D.P. five. Tetrahedron Lett. 1994, 35, 1921–1924. 10.1016/S0040-4039(00)73196-0. [DOI] [Google Scholar]
  15. Ikuta D.; Hirata Y.; Wakamori S.; Shimada H.; Tomabechi Y.; Kawasaki Y.; Ikeuchi K.; Hagimori T.; Matsumoto S.; Yamada H. Conformationally supple glucose monomers enable synthesis of the smallest cyclodextrins. Science 2019, 364, 674–677. 10.1126/science.aaw3053. [DOI] [PubMed] [Google Scholar]
  16. Uitdehaag J. C. M.; Van Der Veen B. A.; Dijkhuizen L.; Dijkstra B. W. Catalytic mechanism and product specificity of cyclodextrin glycosyltransferase, a prototypical transglycosylase from the α-amylase family. Enzyme Microb. Technol. 2002, 30, 295–304. 10.1016/S0141-0229(01)00498-7. [DOI] [Google Scholar]
  17. Biwer A. A. G.; Antranikian G.; Heinzle E. Enzymatic production of cyclodextrins. Appl. Microbiol. Biotechnol. 2002, 59, 609–617. 10.1007/s00253-002-1057-x. [DOI] [PubMed] [Google Scholar]
  18. Terada Y.; Yanase M.; Takata H.; Takaha T.; Okada S. Cyclodextrins are not the major cyclic α-1,4-glucans produced by the initial action of cyclodextrin glucanotransferase on amylose. J. Biol. Chem. 1997, 272, 15729–15733. 10.1074/jbc.272.25.15729. [DOI] [PubMed] [Google Scholar]
  19. Fujiwara T.; Tanaka N.; Kobayashi S. Structure of δ-Cyclodextrin 13.75H2O. Chem. Lett. 1990, 19, 739–742. 10.1246/cl.1990.739. [DOI] [Google Scholar]
  20. Miyazawa I.; Ueda H.; Nagase H.; Endo T.; Kobayashi S.; Nagai T. Physicochemical properties and inclusion complex formation of δ-cyclodextrin. Eur. J. Pharm. Sci. 1995, 3, 153–162. 10.1016/0928-0987(95)00006-Y. [DOI] [Google Scholar]
  21. Endo T.; Ueda H.; Kobayashi S.; Nagai T. Isolation, purification, and characterization of cyclomaltododecaose (ν-cyclodextrin). Carbohydr. Res. 1995, 269, 369–373. 10.1016/0008-6215(94)00359-N. [DOI] [Google Scholar]
  22. Wakamiya A.; Endo T.; Nagase H.; Ueda H.; Kobayashi S.; Nagai T. Isolation and Purification of Cyclomaltononaose (δ-CD) from CELDEX SG-30. Yakuzaigaku 1997, 57, 220–223. [Google Scholar]
  23. Endo T.; Nagase H.; Ueda H.; Kobayashi S.; Nagai T. Isolation, Purification, and Characterization of Cyclomaltodecaose (ϵ-Cyclodextrin), Cyclomaltoundecaose (ζ-Cyclodextrin) and Cyclomaltotridecaose (θ-Cyclodextrin). Chem. Pharm. Bull. 1997, 45, 532–536. 10.1248/cpb.45.532. [DOI] [Google Scholar]
  24. Koizumi K.; Sanbe H.; Kubota Y.; Terada Y.; Takaha T. Isolation and characterization of cyclic α-(1→4)-glucans having degrees of polymerization 9–31 and their quantitative analysis by high-performance anion-exchange chromatography with pulsed amperometric detection. J. Chromatogr. A 1999, 852, 407–416. 10.1016/S0021-9673(99)00643-3. [DOI] [PubMed] [Google Scholar]
  25. Sonnendecker C.; Zimmermann W. Change of the product specificity of a cyclodextrin glucanotransferase by semi-rational mutagenesis to synthesize large-ring cyclodextrins. Catalysts 2019, 9, 242. 10.3390/catal9030242. [DOI] [Google Scholar]
  26. Sonnendecker C.; Melzer S.; Zimmermann W. Engineered cyclodextrin glucanotransferases from Bacillus sp. G-825–6 produce large-ring cyclodextrins with high specificity. Microbiologyopen 2019, 8, e00757 10.1002/mbo3.757. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Tewari Y. B.; Goldberg R. N.; Sato M. Thermodynamics of the hydrolysis and cyclization reactions of alpha-, beta-, and gamma-cyclodextrin. Carbohydr. Res. 1997, 301, 11–22. 10.1016/S0008-6215(97)00073-6. [DOI] [PubMed] [Google Scholar]
  28. Larsen D.; Beeren S. R. Enzyme-mediated dynamic combinatorial chemistry allows out-of-equilibrium template-directed synthesis of macrocyclic oligosaccharides. Chem. Sci. 2019, 10, 9981–9987. 10.1039/C9SC03983J. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Larsen D.; Bjerre P. M.; Beeren S. R. Light-controlled out-of-equilibrium assembly of cyclodextrins in an enzyme-mediated dynamic system. Chem. Commun. 2019, 55, 15037–15040. 10.1039/C9CC08452E. [DOI] [PubMed] [Google Scholar]
  30. Yang S.; Larsen D.; Pellegrini M.; Meier S.; Mierke D. F.; Beeren S. R.; Aprahamian I. Dynamic enzymatic synthesis of γ-cyclodextrin using a photoremovable hydrazone template. Chem 2021, 7, 2190–2200. 10.1016/j.chempr.2021.05.013. [DOI] [Google Scholar]
  31. Erichsen A.; Larsen D.; Beeren S. R. Chaotropic and Kosmotropic Anions Regulate the Outcome of Enzyme-Mediated Dynamic Combinatorial Libraries of Cyclodextrins in Two Different Ways. Front. Chem. 2021, 9, 1–8. 10.3389/fchem.2021.721942. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Samuelsen L.; Larsen D.; Schönbeck C.; Beeren S. R. pH-Responsive templates modulate the dynamic enzymatic synthesis of cyclodextrins. Chem. Commun. 2022, 58, 5152–5155. 10.1039/D1CC06554H. [DOI] [PubMed] [Google Scholar]
  33. Erichsen A.; Peters G. H. J.; Beeren S. R. Templated Enzymatic Synthesis of δ-Cyclodextrin. J. Am. Chem. Soc. 2023, 145, 4882–4891. 10.1021/jacs.3c00341. [DOI] [PubMed] [Google Scholar]
  34. Sørensen J.; Hansen E. L.; Larsen D.; Elmquist M. A.; Buchleithner A.; Florean L.; Beeren S. R. Light-controlled enzymatic synthesis of γ-CD using a recyclable azobenzene template. Chem. Sci. 2023, 14, 7725–7732. 10.1039/D3SC01997G. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Larsen D.; Beeren S. Tuning the outcome of enzyme-mediated dynamic cyclodextrin libraries to enhance template effects. Chem.—Eur. J. 2020, 26, 11032–11038. 10.1002/chem.202001076. [DOI] [PubMed] [Google Scholar]
  36. Larsen D.; Beeren S. R. Building up cyclodextrins from scratch - templated enzymatic synthesis of cyclodextrins directly from maltose. Chem. Commun. 2021, 57, 2503–2506. 10.1039/D1CC00137J. [DOI] [PubMed] [Google Scholar]
  37. Harata K.; Akasaka H.; Endo T.; Nagase H.; Ueda H. X-ray structure of the δ-cyclodextrin complex with cycloundecanone. Chem. Commun. 2002, 1968–1969. 10.1039/B205249K. [DOI] [PubMed] [Google Scholar]
  38. Larsen D.; Erichsen A.; Masciotta G.; Meier S.; Beeren S. R. Quantitative determination of the binding capabilities of individual large-ring cyclodextrins in complex mixtures. Chem. Commun. 2024, 60, 2090–2093. 10.1039/D3CC05897B. [DOI] [PubMed] [Google Scholar]
  39. Ueda H.; Wakamiya A.; Endo T.; Nagase H.; Tomono K.; Nagai T. Interaction of cyclomaltononaose (δ-CD) with several drugs. Drug Dev. Ind. Pharm. 1999, 25, 951–954. 10.1081/DDC-100102256. [DOI] [PubMed] [Google Scholar]
  40. Furuishi T.; Ohmachi Y.; Fukami T.; Nagase H.; Suzuki T.; Endo T.; Ueda H.; Tomono K. Enhanced solubility of fullerene (C60) in water by inclusion complexation with cyclomaltononaose (δ-CD) using a cogrinding method. J. Incl. Phenom. Macrocycl. Chem. 2010, 67, 233–239. 10.1007/s10847-009-9689-4. [DOI] [Google Scholar]
  41. Furuishi T.; Fukami T.; Nagase H.; Suzuki T.; Endo T.; Ueda H.; Tomono K. Improvement of solubility of C70 by complexation with cyclomaltononaose (δ-cyclodextrin). Pharmazie 2008, 63, 54–57. [PubMed] [Google Scholar]
  42. Suvegh K.; Fujiwara K.; Komatsu K.; Marek T.; Ueda T.; Vertes A.; Braun T. Positron lifetime in supramolecular gamma- and delta-cyclodextrin–C60 and – C70 compounds. Chem. Phys. Lett. 2001, 344, 263–269. 10.1016/S0009-2614(01)00798-9. [DOI] [Google Scholar]
  43. Furuishi T.; Endo T.; Nagase H.; Ueda H.; Nagai T. Solubilization of C70 into Water by Complexation with δ-Cyclodextrin. Chem. Pharm. Bull. 1998, 46, 1658–1659. 10.1248/cpb.46.1658. [DOI] [Google Scholar]
  44. Sonnendecker C.; Thurmann S.; Przybylski C.; Zitzmann F. D.; Heinke N.; Krauke Y.; Monks K.; Robitzki A. A.; Belder D.; Zimmermann W. Large-Ring Cyclodextrins as Chiral Selectors for Enantiomeric Pharmaceuticals. Angew. Chem., Int. Ed. 2019, 58, 6411–6414. 10.1002/anie.201900911. [DOI] [PubMed] [Google Scholar]
  45. Larsen K. L.; Endo T.; Ueda H.; Zimmermann W. Inclusion complex formation constants of α-, β-, γ-, δ-, ε-, ζ-, η- and θ-cyclodextrins determined with capillary zone electrophoresis. Carbohydr. Res. 1998, 309, 153–159. 10.1016/S0008-6215(98)00130-X. [DOI] [Google Scholar]
  46. Larsen K. L.; Zimmermann W. Analysis and characterisation of cyclodextrins and their inclusion complexes by affinity capillary electrophoresis. J. Chromatogr. A 1999, 836, 3–14. 10.1016/S0021-9673(99)00091-6. [DOI] [Google Scholar]
  47. Assaf K. I.; Ural M. S.; Pan F.; Georgiev T.; Simova S.; Rissanen K.; Gabel D.; Nau W. M. Water structure recovery in chaotropic anion recognition: High-affinity binding of dodecaborate clusters to γ-cyclodextrin. Angew. Chem., Int. Ed. 2015, 54, 6852–6856. 10.1002/anie.201412485. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Assaf K. I.; Gabel D.; Zimmermann W.; Nau W. M. High-affinity host-guest chemistry of large-ring cyclodextrins. Org. Biomol. Chem. 2016, 14, 7702–7706. 10.1039/C6OB01161F. [DOI] [PubMed] [Google Scholar]
  49. Khlifi S.; Marrot J.; Haouas M.; Shepard W. E.; Falaise C.; Cadot E. Chaotropic Effect as an Assembly Motif to Construct Supramolecular Cyclodextrin–Polyoxometalate-Based Frameworks. J. Am. Chem. Soc. 2022, 144, 4469–4477. 10.1021/jacs.1c12049. [DOI] [PubMed] [Google Scholar]
  50. Falaise C.; Moussawi M. A.; Floquet S.; Abramov P. A.; Sokolov M. N.; Haouas M.; Cadot E. Probing Dynamic Library of Metal-Oxo Building Blocks with γ-Cyclodextrin. J. Am. Chem. Soc. 2018, 140, 11198–11201. 10.1021/jacs.8b07525. [DOI] [PubMed] [Google Scholar]
  51. Falaise C.; et al. “Host in Host” Supramolecular Core–Shell Type Systems Based on Giant Ring-Shaped Polyoxometalates. Angew. Chem., Int. Ed. 2021, 60, 14146–14153. 10.1002/anie.202102507. [DOI] [PubMed] [Google Scholar]
  52. Harata K. The Structure of the Cyclodextrin Complex. V. Crystal Structures of alpha-Cyclodextrin Complexes with p-Nitrophenol and p-Hydroxybenzoic acid. Bull. Chem. Soc. Jpn. 1977, 50, 1416–1424. 10.1246/bcsj.50.1416. [DOI] [Google Scholar]
  53. Harata K. The Structure of the Cyclodextrin Complex. XVII. Crystal Structure of beta-Cyclodextrin -Hexamethylene-tetramine (1:1) Complex Hexahydrate. Bull. Chem. Soc. Jpn. 1984, 57, 2596–2599. 10.1246/bcsj.57.2596. [DOI] [Google Scholar]
  54. Harata K. The structure of the cyclodextrin complex. XX. Crystal structure of uncomplexed hydrated γ-cyclodextrin. Bull. Chem. Soc. Jpn. 1987, 60, 2763–2767. 10.1246/bcsj.60.2763. [DOI] [Google Scholar]
  55. Marcos V.; Stephens A. J.; Jaramillo-Garcia J.; Nussbaumer A. L.; Woltering S. L.; Valero A.; Lemonnier J.-F.; Vitorica-Yrezabal I. J.; Leigh D. A. Allosteric initiation and regulation of catalysis with a molecular knot. Science 2016, 352, 1555–1559. 10.1126/science.aaf3673. [DOI] [PubMed] [Google Scholar]
  56. Vial L.; Ludlow R. F.; Leclaire J.; Pérez-Fernández R.; Otto S. Controlling the biological effects of spermine using a synthetic receptor. J. Am. Chem. Soc. 2006, 128, 10253–10257. 10.1021/ja062536b. [DOI] [PubMed] [Google Scholar]
  57. Assaf K. I.; Nau W. M. The chaotropic effect as an assembly motif in chemistry. Angew. Chem., Int. Ed. 2018, 57, 13968–13981. 10.1002/anie.201804597. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Haouas M.; Falaise C.; Leclerc N.; Floquet S.; Cadot E. NMR spectroscopy to study cyclodextrin-based host-guest assemblies with polynuclear clusters. Dalton Trans. 2023, 52, 13467–13481. 10.1039/D3DT02367B. [DOI] [PubMed] [Google Scholar]
  59. Corbett P. T.; Otto S.; Sanders J. K. M. Correlation between Host–Guest Binding and Host Amplification in Simulated Dynamic Combinatorial Libraries. Chem.—Eur. J. 2004, 10, 3139–3143. 10.1002/chem.200400300. [DOI] [PubMed] [Google Scholar]
  60. Beeren S. R.; Hansen K. H.; Erichsen A.; Larsen D.. Method for the production of large-ring cyclodextrins. WO 2024042228, 2024.

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ja5c02055_si_001.pdf (6.6MB, pdf)

Articles from Journal of the American Chemical Society are provided here courtesy of American Chemical Society

RESOURCES