Page 1
NEW SYNTHETIC APPROACH TO 1,4 - ASYMMETRICALLY FUNCTIONALIZED β - CYCLODEXTRINS AS DRUG CARRIERS AND RECEPTOR MIMICS VIA
AMINO ACIDS DIPEPTIDE ANALOGUES -
Gabriele Lupidi
Dissertação para obtenção do Grau de Mestre em
Química
Orientadores
Pedro Paulo de Lacerda e Oliveira Santos Prof. Enrico Marcantoni
Júri
Presidente: Prof. Maria Matilde Soares Duarte Marques Orientadores: Prof. Pedro Paulo de Lacerda e Oliveira Santos Vogal: Doutora Alexandra Maria Moita Antunes Doutor Corrado Bacciocchi
Janeiro de 2016
Page 2
1
Abstract.
Cyclodextrins make up a family of well-known, synthetic α-1,4 linked cyclic oligosaccharides that have
found widespread use in clinical, industrial and environmental applications. Modified cyclodextrins may
also be used as enzyme models for laboratory research. Our study was focused on the selective
functionalization of two primary hydroxyl groups in positions 1 and 4 of a β-cyclodextrin, with an
Asparagine and a Lysine amino acid residue, respectively.
The project was to control the insertion on the desired positions of the β-cyclodextrin by using a
conveniently sized intermediate, obtained by condensation between two suitable derivatives of the two
amino acids. This idea ensures at the same time both the functionalization of the right position of the β-
cyclodextrin ring and the functionalization of the two positions with the two different amino acids.
Selective protection and functionalization of the two amino acids, in order to have the right substrates
to synthetize the intermediate and to perform the insertion in the cyclodextrin, has been the primary goal
of our work in the laboratory.
Key words:
asymmetric functionalization, β-cyclodextrin, dipeptides, drug carriers, receptor mimics
Page 3
2
Resumo
As ciclodextrinas constituem uma família bem conhecida de oligossacáridos cíclicos sintéticos, com
ligações α-1,4, que têm sido muito utilizados para aplicações clínicas, industriais e ambientais.
Ciclodextrinas modificadas podem também ser usadas como modelos de enzimas para investigação
laboratorial. O presente trabalho focou-se na funcionalização selectiva de dois grupos hidroxilo
primários nas posições 1 e 4 de uma β-cyclodextrina com um resíduo dos amino ácidos asparagina e
lisina, respectivamente.
O projecto consistiu em controlar a inserção nas posições desejadas da β-cyclodextrina utilizando um
intermediário de dimensão conveniente, obtido por condensação entre dois derivados adequados dos
dois amino ácidos. Esta estratégia assegura simultaneamente a funcionalização no local pretendido do
anel da β-cyclodextrina e a funcionalização das duas posições com dois amino ácidos diferentes.
O objectivo principal do trabalho desenvolvido no laboratório foi a protecção e funcionalização selectiva
dos dois amino ácidos, de modo a obter os susbtratos adequados para sintetizar o intermediário e
efectuar a sua inserção na ciclodextrina.
Palavras-chave
β-cyclodextrina, dipéptidos, funcionalização assimétrica, mimetizadores de receptores, transportadores
de fármacos
Page 4
3
Index
1) SUPRAMOLECULAR CHEMISTRY 6
1.1) Introduction on supramolecular chemistry. 6
1.2) “Host” molecules. 7
2) CYCLODEXTRINS 12
2.1) Introduction. 12
3) CYCLODEXTRINS AS DRUG CARRIERS 18
3.1) Introduction. 18
3.2) CDs inclusion complexes. 19
3.3) Advantages of CDs inclusion complexes. 23
3.4) CD applications in the design of some novel delivery system. 26
4) CATALYSIS BY CYCLODEXTRINS 27
4.1) Introduction in cyclodextrin catalysis. 27
4.2) Specificity in cyclodextrin catalysis. 32
5) MODIFIED CDs AS ENZYME MODELS 35
5.1) Introduction on modified CDs as enzyme models. 35
5.2) Chymotrypsin mimics. 36
5.3) Serine protease mimics. 37
5.4) Metalloenzyme mimics. 38
5.5) Enzyme-coenzyme mimics. 40
5.6) Bifunctional or multifunctional enzyme mimics. 46
6) METHODS FOR SELECTIVE MODIFICATIONS OF CYCLODEXTRINS 50
6.1) Overview of methods for modification of cyclodextrins. 50
6.2) Chemistry involved in methods for modification of cyclodextrins. 51
6.3) Primary face modification. 52
6.4) Secondary face modification. 62
7) STRATEGY FOR THE SELECTIVE ASYMMETRICAL DIFUNCTIONALIZATION OF THE
PRIMARY HYDROXYL GROUPS OF A β-CYCLODEXTRIN 66
8) EXPERIMENTAL SECTION 79
8.1) General. 79
8.2) Synthesis of Nε-benzylidene-L-lysine. 80
8.3) Synthesis of Nα-carbobenzyloxy-L-lysine. 81
8.4) Synthesis of Nα-carbobenzyloxy-L-lysine methyl ester. 82
8.5) Synthesis of N-carbobenzyloxy-L-asparagine. 83
8.6) Synthesis of N-carbobenzyloxy-L-asparagine methyl ester. 84
8.7) Synthesis of N-carbobenzyloxy-L-aspartic acid α-methyl ester. 85
8.8) Synthesis of N-carbobenzyloxy-L-homoserine α-methyl ester. 86
9) CONCLUSIONS AND FUTURE WORK 87
REFERENCES 88
Page 5
4
Index of Figures
Figure 1. 7
Figure 2. 8
Figure 3. 8
Figure 4. 9
Figure 5. 9
Figure 6. 10
Figure 7. 10
Figure 8. 11
Figure 9. 13
Figure 10. 13
Figure 11. 14
Figure 12. 15
Figure 13. 16
Figure 14. 17
Figure 15. 21
Figure 16. 22
Figure 17. 22
Figure 18. 28
Figure 19. 31
Figure 20. 34
Figure 21. 35
Figure 22. 36
Figure 23. 40
Figure 24. 41
Figure 25. 42
Figure 26. 47
Figure 27. 49
Figure 28. 50
Figure 29. 51
Figure 30. 59
Index of Schemes
Scheme 1. 28
Scheme 2. 29
Scheme 3. 30
Scheme 4. 30
Scheme 5. 31
Scheme 6. 32
Scheme 7. 34
Scheme 8. 37
Scheme 9. 38
Scheme 10. 41
Scheme 11. 42
Scheme 12. 43
Scheme 13. 46
Scheme 14. 47
Scheme 15. 48
Page 6
5
Scheme 16. 53
Scheme 17. 54
Scheme 18. 54
Scheme 19. 55
Scheme 20. 55
Scheme 21. 56
Scheme 22. 57
Scheme 23. 57
Scheme 24. 58
Scheme 25. 58
Scheme 26. 60
Scheme 27. 61
Scheme 28. 62
Scheme 29. 63
Scheme 30. 63
Scheme 31. 64
Scheme 32. 65
Scheme 33. 68
Scheme 34. 69
Scheme 35. 71
Scheme 36. 71
Scheme 37. 72
Scheme 38. 72
Scheme 39. 73
Scheme 40. 74
Scheme 41. 74
Scheme 42. 74
Scheme 43. 76
Scheme 44. 77
Scheme 45. 77
Index of Tables
Table 1. 14
Table 2. 25
Table 3. 27
Table 4. 33
Table 5. 37
Table 6. 73
Page 7
6
1) SUPRAMOLECULAR CHEMISTRY
1.1) Introduction on supramolecular chemistry
Supramolecular chemistry refers to the area of chemistry beyond the molecules and focuses on
the chemical systems made up of a discrete number of assembled molecular subunits or
components. The forces responsible for the spatial organization may vary from weak
(intermolecular forces, electrostatic or hydrogen bonding) to strong (covalent bonding), provided
that the degree of electronic coupling between the molecular component remains small with respect
to the relevant energy parameters of the component.
As initiated by Prof. J. M. Lehn, who was awarded the 1987 Nobel Prize in Chemistry, the field has
and is the basis for most of the essential biochemical processes of life. It has grown over twenty
years into a major domain of modern teaching, research and technology. It has provided numerous
developments at the interfaces with biology, physics, material science, biomedicine and
nanotechnology, thus giving rise to the emergence and establishment of supramolecular science,
which is, today, a broad multidipliscinary and interdipliscinary domain, providing a highly fertile
ground for creative cooperation of scientists from very different backgrounds.[1,2]
For over 100 years, chemistry has focused primarily on understanding the behaviour of molecules
and their construction from constituent atoms. During last three decades, chemists have extended
their investigation beyond atomic and molecular chemistry in the realm of supramolecular
chemistry.
Broadly speaking, supramolecular chemistry is the study of interaction between, rather than within,
molecules; in other words, chemistry using molecules rather than atoms as building blocks.
F. Vӧgtle has defined supramolecular chemistry as follows:
“In contrast to molecular chemistry, which is predominantly based upon covalent bonding of atoms,
supramolecular chemistry is based upon intermolecular interactions, i.e. on the association of two
or more building blocks, which are held together by intermolecular bonds.”[3]
stating how these “intermolecular bonds”, weaker noncovalent interactions, such as hydrogen
bonding, polar attractions, Van der Waals forces, hydrophilic-hydrophobic interactions, charge-
transfer and host-guest interactions[4] play a key role in this field.
In particular, host-guest chemistry is an important subdivision of supramolecular chemistry in
which usually two or more molecules or ions are held together to form a complex in a unique
structural relationship through intermolecular forces.[5]
Page 8
7
Figure 1. An example of host-guest chemistry.[6]
A molecule (host) can bind another molecule (guest) to produce a “host-guest” complex. The
interactions between host and guest are noncovalent. The term “host-guest” chemistry has been
used to describe a variety of processes occurring in a number of research fields, such as organic,
analytical, biological and organometallic chemistry, and involving molecules and ions with different
structures, dimensions and properties.
It is possible to define host-guest chemistry by considering the common elements that these
processes possesses. In general, host-guest interactions involve the establishment of multiple non-
covalent bonds between a large and geometrically concave organic molecule (the host) and a
simpler organic or inorganic molecule or ion (the guest). Formation of multiple noncovalent bonds
between reactants of similar size or geometry is generally referred to as processes of
supramolecular chemistry, not host-guest chemistry, because inclusion complexes are not formed.
Thus, the geometrical requirements are essential to fit the definition of host-guest chemistry.
1.2) “Host” molecules
Among the different potential “host” molecules suitable to form supramolecular aggregates, we
can list:
Crown ethers. Crown ethers were the first artificial host molecules discovered. They were
accidentally discovered in 1967 by Charles Pedersen, winner of the Nobel Prize in Chemistry in the
same year, who was a chemist working at DuPont Experimental Station in Wilmington (Delaware),
as byproducts of an organic reaction. He was trying to prepare a complexing agent for divalent
cations[7,8] by linking two catecholate groups through one hydroxyl on each molecule (Figure 2).
Page 9
8
Figure 2. Pedersen’s strategy.
Since in the reaction mixture he had a 10% of unreacted catechol, he could isolate a second
molecule. By investigating it’s properties he could find its structure:[9]
Figure 3. Disovery of crown ethers.
Crown ethers are cyclic chemical compounds that consist of a ring containing several ether groups.
The most common crown ethers are oligomers of ethylene oxide, the repeating unit being
ethyleneoxy, i.e., -CH2CH2O-. The term "crown" refers to the resemblance between the structure
of a crown ether bound to a cation, and a crown sitting on a person's head.
Crown ethers strongly bind certain cations, forming complexes. The oxygen atoms are well
situated to coordinate with a cation located inside the ring, whereas the exterior of the ring is
hydrophobic. The resulting cations often form salts that are soluble in nonpolar solvents, and for
this reason crown ethers are useful in phase transfer catalysis.
By incorporating luminescent substituents into their backbone, these compounds have proved to
be sensitive ion probes, as also very low concentrations of the metal provoke measurable changes
in the absorption or fluorescence of the photoactive groups.
Lariat crown ethers. Lariat ethers are macrocyclic polyether compounds having one or more
donor-group-bearing sidearms. In these systems, sidearms can be attached either to carbon[10] or
to nitrogen.[11]
Page 10
9
Figure 4. Example of a Lariat crown ether.
These sidechains are helpful not only to increase the lipophylicity of the whole molecule, but also
to increase the number of possibile binding sites with the metal ion. Moreover they can give a
certain degree of tridimensionality in order to enhance the complexation ability.
Cryptands. Cryptands are a family of synthetic bi- and polycyclic multidentate ligands for a
variety of cations. The Nobel Prize for Chemistry in 1987 was given to Donald J. Cram, Jean-Marie
Lehn, and Charles J. Pedersen for their efforts in discovering and determining uses of cryptands
and crown ethers, thus launching the now flourishing field of supramolecular chemistry.
Figure 5. 2.2.2-Cryptand.
The term cryptand implies that this ligand binds substrates in a crypt, interring the guest as in a
burial. These molecules are three-dimensional analogues of crown ethers but are more selective
and complex the guest ions more strongly. The resulting complexes are lipophilic.
The 3-dimensional interior cavity of a cryptand provides a binding site for
“guest” ions. In contrast to crown ethers, cryptands bind the guest ions
using both nitrogen and oxygen donors. This three- dimensional
encapsulation mode confers some size-selectivity, enabling
discrimination among alkali metal cations.
Cryptands are able to bind otherwise insoluble salts into organic solvents and can also be used as
phase transfer catalysts by transferring ions from one phase to another.[12]
Calixarenes. Calixarenes are a class of macrocycles or cyclic oligomers obtained by
condensation between aldehydes and resorcinols, pyrogallols or para-substituted phenols.
Page 11
10
The word calixarene is derived from calix or chalice, because these molecules are vase-shaped,
and from the word arene that refers to the aromatic building block. Calixarenes have a hydrophobic
cavity, similar to that of cyclodextrins, that can hold smaller molecules or ions containing mainly an
even number n= 4, 6, 8 of aromatics unities. Calixarenes with an odd n value, or higher than 8,
difficult to isolate and obtained in low yields.
Figure 6. An example of a p-tert-butylcalix[4]arene.
This class of compounds is characterized by a wide upper rim, a narrow lower rim and a central
annulus. They are a typical example of host-guest chemistry and efficient sodium ionophores, thus
they are applied as such in chemical sensors. With the right chemistry these molecules exhibit
great selectivity also towards other cations.
Additionally, calixarenes are applied in enzyme mimetics. In particular, tetrathia[4]arene is found to
mimic aquaporin proteins,[13] also known as “water channels”, that are integral membrane proteins
from a larger family of major intrinsic proteins (MIP) forming pores in the membrane of biological
cells. These proteins selectively conduct water molecules in and out of the cell, while preventing
the passage of ions and other solutes.
Calixarenes molecules have also been used for pure aesthetic purposes as in the case of the
creation of a molecular football world cup, as a scientific tribute to the French national team winner
of the 1998 world championship.[14]
Figure 7. Shape relationship of the football world cup (left) with the fullerene-calix[4]arene
conjugate (middle) and the schematic structure representation (right).
Page 12
11
This molecule was obtained by double cyclopropanation of a fullerene with a calix[4]arene
containing two malonamidics substituents in the upper edge.
Dendrimers. Dendrimers are three-dimensional, immensely branched, well-organized
nanoscopic macromolecules (typically 5000-500,000 g/mol). The name has actually derived from
the Greek word “dendron” meaning “tree,” which indicates their unique tree-like branching
architecture.
Dendrimers are defined by three components: a central core, an interior dendritic structure (the
branches), and an exterior surface with functional surface groups. The varied combination of these
components afford products of different shapes and sizes with shielded inner cores, ideal
candidates for applications both in biological and in materials sciences.
Figure 8. Example of the structure of a Dendrimer.
Properties of dendrimers are strictly related to their chemical structures; while the attached surface
groups affect the solubility and chelation ability, the varied cores impart unique properties to the
cavity size, absorption capacity, and capture-release characteristics. Thanks to this, it is possible
to make water-soluble dendrimers, unlike most polymers, by functionalizing their outer shell with
charged species or other hydrophilic groups. Other controllable properties of dendrimers include
toxicity, crystallinity and chirality.[15]
Applications of dendrimers typically involve conjugating other chemical species to the dendrimer
surface that can act as detecting agents (such as a dye molecule), affinity ligands, targeting
Page 13
12
components, radioligands, imaging agents, or pharmaceutically active compounds. Dendrimers
have very strong potential for these applications because their structure can lead to multivalent
systems: one dendrimer molecule has hundreds of possible sites to couple to an active species.
Dendrimers can also be used as solubilizing agents, for example, those with hydrophobic core and
hydrophilic periphery have shown to exhibit micelle-like behavior and have container properties in
solution.[16]
Moreover, they play an important role as drug delivery system. Unlike traditional polymers,
dendrimers have received considerable attention in biological applications due to their high water
solubility,[17] biocompatibility,[18] polyvalency and precise molecular weight. These features make
them an ideal carrier for drug delivery and targeting applications.
There are three methods for using dendrimers in drug delivery:
1. the drug is covalently attached to the periphery of the dendrimer to form dendrimer prodrugs;
2. the drug is coordinated to the outer functional groups via ionic interactions;
3. the dendrimer acts as a unimolecular micelle by encapsulating a pharmaceutical compound
through the formation of a dendrimer-drug supramolecular assembly.
The use of dendrimers as drug carriers by encapsulating hydrophobic drugs is a potential method
for delivering highly active pharmaceutical compounds that may not be in clinical use due to their
limited water solubility and resulting suboptimal pharmacokinetics.[19]
2) CYCLODEXTRINS
2.1) Introduction
Cyclodextrins (CDs) are cyclic oligosaccharides biosynthesized from starch through an
enzymatic process of intramolecular trans-glycosylation by action of the enzyme cyclodextrin
glycosyltransferase (CGTase).[20]
In the next sections, the characteristics and properties of this class of compounds belonging to the
group of “host” molecules are going to be discussed.
Linear molecules, called dextrins, are obtained by enzymatic process. In aqueous solution linear
dextrins arrange to form a helical coil of glycosidic molecules bound together in 1,4 positions. The
CGTase can remove a portion of molecular dextrins and can tie together the two ends of the
fragment, giving rise to the cyclodextrins.
The correct structure of cyclodextrins was discovered by Freudeberg in 1936:[21,22] they are
constituted by several units of glucopyranose linked by α-1,4-glycosidic bonds that impart to the
molecule the typical shape of a truncated hollow cone.
Page 14
13
Figure 9. α-1,4-glycosidic bond
Three most common structures were isolated for the first time by Villiers[23](1891) and
subsequently classified in 1903 by Scardinger:[24] the α-cyclodextrin (α-CD), the β-cyclodextrin (β-
CD) and the γ-cyclodextrin (γ-CD), respectively constituted by 6, 7 or 8 α-glucopyranoside residues.
Figure 10. Structure and conformation of the three most common CDs.
The three-dimensional structure of these cyclodextrins is such that the hydroxyl groups linked to
the carbons C(2) and C(3) are on the edge of the trunk and have the greater diameter (secondary
rim) while the hydroxyl groups linked to the carbon C(6) are on the edge with the smaller diameter
(primary rim).
Obviously, the hydrophilicity of both primary and secondary hydroxyl groups contribute to ensure
the water solubility of the macromolecule. Its internal cavity is rather hydrophobic due to the
presence of hydrogen atoms, bound to the carbon C(3) and C(5), and of the bridged oxygens.
However, it is the seat of an electric dipole moment, with the positive side oriented towards the
primary edge and the negative part oriented towards the secondary
Page 15
14
Figure 11. Three-dimensional representation of CDs.
The size of the macrocycle grows with the increasing number of glucopyranoside residues,[25] which
are structurally characterized by a quite rigid 4C1 chair conformation.
Table 1. Properties of the main cyclodextrins.
Cyclodextrin Mass
Outer
diameter
(nm)
Cavity
Diameter
(nm)
Cavity
volume
(ml/g)
Solubility
H2O
(g/kg)[26]
Hydrate (H2O)[26]
Inner
rim
Outer
rim Cavity External
α, (glucose)6 972 1.52 0.45 0.53 0.10 129.5 2.0 4.4
β, (glucose)7 1134 1.66 0.60 0.65 0.14 18.4 6.0 3.6
γ, (glucose)8 1296 1.77 0.75 0.85 0.20 249.2 8.8 5.4
Previous computational studies,[27] supported by polarimetric data,[28,29] however, show that the
overall frustoconical structure is not rigid, but has a certain degree of flexibility due to the partial
free rotation around the glycosidic bridges, or around the single bonds C(1)-O(4’) and O(4’)-C(4’),
The huge amount of papers published is more than sufficient evidence to show the great interest
accrued to the study of the conformational dynamics of these molecules. Literature is fully
consolidated in the practice of indicating respectively with φ and ψ the dihedral angles relative to
the bonds previously reported (Figure 12).
Furthermore it is common practice to indicate with ω the dihedral angle relative to the -CH2OH
group at position 6, which gives a further degree of conformational freedom of each individual unit
of the glucoside.[30,31]
Page 16
15
Figure 12. Representation of dihedral angles responsible of the dynamic modes of CDs.
The nomenclature and symbols shown above, although they have been introduced by various
works in literature, are now fully accepted by the scientific community and suggested by IUPAC
itself. By convention, the angles φ and ψ are defined respectively as the values of the dihedral
angles O(5)-C(1)-O(4’)-C(4’) and C(1)-O(4’)-C(4’)-C(5’), whose values predetermine the mutual
space orientation of the adiacent glycosidic units that can be seen as resultant of two movements
of these units, seen as rigid elements, which are:
1) A movement of "hover", that consists in the oscillation of the units around the straight line joining
the atoms C(1)---C(4’) and which can be expressed quantitatively by the value assumed by the
pseudo-dihedral angle λ defined by the atoms H(1)-C(1)---C(4’)-H(4’).
This movement brings the secondary hydroxyl groups of the glycosidic units to approach or move
away from the axis of the ideal cavity (obviously the -CH2OH groups moves in the opposite
direction). A simple consideration of the three-dimensional molecular models easily shows that, if
the glycosidic units can be taken as rigid, the value of λ depends on a linear combination of φ and
ψ.
2) A movement of "bending", consisting in a swinging of the unit around a straight line
perpendicular to the axis of the ideal cavity and passing through the glycosidic O(4'). This oscillation
can be expressed quantitatively by the value taken from the pseudo-dihedral angle τ defined by
the atoms C(4)---C(1)-O(4)---C(4’).
This movement brings the secondary hydroxyl groups of adjacent units to approach or move away
from each other, narrowing or widening the edge of secondary cavity (while the primary edge is
changed in the opposite way). A careful consideration of molecular models, however, shows that τ
depends only on φ.
The commonly accepted representation of this class of macrocycles is that of molecules with a
rather regular structure, particularly the β-CD. Its stability seems to be secured by a network of
intramolecular hydrogen bonds that is established on the edge between the secondary hydroxyl
group on the C(2) of each unit of the glycoside (acceptor) and the hydroxyl group on the C(3) of
the adjacent glucoside unit (donor).
Page 17
16
Several studies have shown, however, that such network can be “reversed” with the formation of
new hydrogen bonds between the hydroxyl in position 3 acting as a donor and the hydroxyl function
in position 2 of the adjacent glucoside unit as the acceptor.[32,33]
Figure 13. Structural drawing of β-cyclodextrin showing the intramolecular hydrogen bonds
between the hydroxyl groups of C(2) and C(3).
Anyway, such a network of hydrogen bonds definitely stiffens the macrocycle and the lower
solubility of β-CD with respect to its counterparts α and γ is attributed to this feature. The H-bond
belt is incomplete in the α-CD molecule, because one glucopyranose unit is in a distorted position.
Consequently, instead of the six possible H-bonds, only four can be established simultaneously.
The γ-CD is a non-coplanar, more flexible structure; therefore, it is the more soluble of the three
CDs.
The actual shape of the cyclodextrin is determined predominantly from angles φ, ψ and ω
discussed above. If these were the same on all the macrocycle, the various glycosidic bridges
would assume a coplanar position, so that the CD would have an overall symmetry of type Cn.
Even the NMR spectra of cyclodextrins would make us think about perfectly symmetrical
structures.[33] However, computational studies already mentioned have highlighted the dynamism
and flexibility of these systems.[34] In particular, simulations carried out with molecular dynamics
methods show that the structure has a degree of flexibility as greater as higher is the number of
monomeric units of the macrocycle. The discrepancy between the data collected, and between
their respective interpretations, can be explained by admitting that the apparent symmetry,
highlighted by “slow” research methodologies such as NMR spectroscopy, is actually the result of
an average time between multiple not symmetrical conformations.
The foregoing is intended to emphasize that the simple bonds C(1)-O(4’) and O(1)-C(4’), and the
corresponding dihedral angles φ and ψ, do not assume the same values on all the macrocycle
but, on the contrary, can vary within fairly wide limits.
Page 18
17
Certainly, the most important characteristic of cyclodextrins is their ability to form
supramolecular inclusion complexes with a wide range of organic substrates (guests). This is due
to the fact that the guest migration from the solvent to the hydrophobic cavity of the cyclodextrin
(host) is generally a thermodynamically favoured process. It can be studied using the scheme
proposed by Tabushi,[35] who developed it as a succession of four ideal steps:
1. host desolvation;
2. guest desolvation;
3. host-guest binding;
4. solvent reorganization.
Figure 14. Structure of an α-CD-nitromethane inclusion complex. The nitromethane molecule is
located in the void of the α-CD macrocycle, with the methyl group being disordered over two sites
with equal (50%) occupancy, whereas the corresponding two nitrogen positions coincide in one
fully occupied site. The oxygen atoms of the guest display excessive thermal motions, indicating
that the NO2 group is statistically disordered; refinement of the structure was approximated by six
positions over which the two oxygens are distributed (33% occupancy of each site).[36]
The number of studies conducted on the properties of inclusion has clarified the various factors
that influence this process.[37] Amongst these, we can find desolvation energies of the guest and
the cavity, host-guest nonspecific interactions (van der Waals forces or dipoles) and the eventual
establishment of specific interactions (hydrogen bonding, CH---π interactions).
An immediate consequence of the phenomenon is the more or less significant modification of the
chemical-physical properties of the species included, such as the solubility, the answer to
spectroscopic analyses and, last but not least, the chemical reactivity.
It’s not suprising that this field of research has paved the way for many technical applications and
technologies, particularly in the areas of agribusiness, analytical (stationary phases for
chromatographic applications), pharmaceuticals (drug-carrier) and cosmetics.[38]
Page 19
18
The derivatized cyclodextrins have also proved to be a good model for studying enzyme-mimetic
systems[39] as well as for the synthesis of nanostructured systems.[40]
3) CYCLODEXTRINS AS DRUG CARRIERS
3.1) Introduction
Drug delivery is the method or process of administering a pharmaceutical compound to achieve
a desired therapeutic effect. Drug delivery technologies modify drug release profile, absorption,
distribution and elimination for benefiting in improving product efficacy, safety, patient convenience
and compliance. To be pharmacologically active, all drugs must possess some degree of aqueous
solubility, and most drugs should be lipophilic to permeate the biological membranes via passive
diffusion. The water solubility of any drug is determined by its potency and its type of formulation.[41]
However, if a drug is too hydrophilic, the dissolved drug molecule will not partition from the aqueous
exterior into a lipophilic bio membrane and then permeate the membrane.
It has been proved that more than 40% of drugs failure in the development can be traced because
of poor biopharmaceutical properties, especially poor solubility or poor permeability.[42]
A wide range of different technologies have been evolved recently in order to improve the
solubility of drug candidates. A number of techniques of different nature are being presently utilized
for enhancing the solubility of drugs. The widely available techniques are: micronization,[43] use of
surfactant, salt formation,[44] adsorption,[45] solid dispersion, hot melt extrusion[46] and cyclodextrin
complexation. Among these, cyclodextrins are the most widely used, due to their ability to alter
physical, chemical and biological properties of guest drug molecules through the formation of
“inclusion complexes”.
Complexation is one of the several ways to favorably enhance the physicochemical properties
of pharmaceutical compounds. It is based on the ability of many well known drugs to interact and
form new complex drugs, with altered properties in comparison with a drug alone. So, it may be
loosely defined as the reversible association of a substrate and ligand to form a new species. The
intermolecular forces involved in the formation of complexes are covalent or co-ordinated bonds,
van der Waals forces of dispersion, ion-dipole, dipole-dipole, dipole-induced dipole type
interactions and hydrogen bonding etc.
It offers new possibilities for the improvement of existing drugs with respect to side effects,
therapeutic activity, solubility, and chemical reactivity. Therefore, it is finding increasing application
in the field of pharmacy during the past few decades, and the pharmaceutical technology and
industry have long considered research and development, in the area of complexation, a priority.
Various uses of complexation are enumerated below:[47,48]
Page 20
19
• solubility enhancement of poorly soluble drugs;
• stability enhancement towards various degradation reactions;
• bioavailability enhancement of poorly absorbed drugs;
• dissolution rate enhancement;
• reduction or elimination of toxicity and ulcerogenic effects of drugs;
• conversion of a liquid substance into a solid complex thereby improving its processing
characteristics;
• alteration of chemical reactivity of drug;
• usage as antidote for metal poisoning;
• taste masking of bitter taste drugs;
• masking of unpleasant /obnoxious odor;
• reduction in volatility of drugs;
• co-usage of incompatible drugs;
• improvement in content uniformity;
• ease in manufacturing;
• low cost.
3.2) CDs inclusion complexes
The internal cavity, hydrophobic in nature, is a key feature of the CDs that provides the ability
to form complexes, including a variety of guest molecules. CD inclusion is a stoichiometric
molecular phenomenon in which usually only one molecule interacts with the cavity of the CD
molecule to become entrapped.
Complexation may be thermodynamically seen as the interactions between the different
components of the cyclodextrin and the active drug. The net energetic driving force must be
required for complex formation, which pulls the active drug into the cyclodextrin. The most stable
three dimensional structure of cyclodextrin, as previously described, is a toroid with the larger and
smaller openings presenting hydroxyl groups to the external environment and mostly hydrophobic
functionality lining the interior of the cavity. This unique configuration creates a thermodynamic
driving force, which is required to form complex of active drug with a polar molecules and functional
groups. There are four favorable interactions between cyclodextrin and active drug, which shift the
equilibrium towards complex formation:
1. the displacement of polar water molecules from the apolar cyclodextrin cavity,
2. the increased number of hydrogen bonds formed as the displaced water returns to the larger
pool,
Page 21
20
3. a reduction of repulsive interaction between the hydrophobic active drug and the aqueous
environment, and
4. an increase in hydrophobic interactions as the active drug inserts itself into the apolar
cyclodextrin cavity.
The first complexation equilibrium establishes very rapidly, while the final equilibrium can take
longer to be reached: once inside the cyclodextrin cavity, the active drug makes conformational
adjustments to take maximum advantage of the weak van der Waals forces that exists.[49]
The encapsulation protects the drug molecule against attack by various reactive molecules and in
this way it reduces the rate of hydrolysis, oxidation, steric rearrangement, racemization and even
enzymatic decomposition.[50] In addition, CDs can decrease photo degradation of various light
sensitive drugs.
Many techniques are used to form CD complexes, for example:
Co-precipitation. CD is dissolved in water and the guest is added while stirring the CD solution. By
heating, more CD can be dissolved (20%) if the guest can tolerate the higher temperature. The CD
and guest solution must be cooled under stirring before a precipitate is formed. The precipitate can
be collected by decanting, centrifugation or filtration and washed. The main disadvantage of this
method lies in the scale-up
Slurry complexation. CD can be added to water, as much as 50-60% solids, and stirred. The
aqueous phase will be saturated with CD in solution. Guest molecule will complex with the CD in
solution and, as the CD complex saturates the water phase, the complex will crystallize or
precipitate.
Paste complexation. A small amount of water is added to the guest to form a paste, which is mixed
with the CD using a mortar and pestle. The resulting complex can be dried directly, and milled if
hard mass forms.
Damp mixing. The guest and CD are thoroughly mixed and placed in a sealed container with a
small amount of water. The contents are heated to about 100°C and then removed and dried.
Extrusion. CD, guest and water can be premixed or mixed as added to the extruder. The extruder
complex may dry as it cools or the complex may be placed in an oven to dry. Heat-labile guests
can get decomposed.
Dry mixing. Some guests can be complexed by simply adding the guest to the CD and mixing them
together. This works best with oils or liquid guests.
Page 22
21
In all the above methods, optimization of the amount of water, degree and time of mixing,
temperature and heating time is necessary for each guest.
Different mechanisms play an important role in drug release from the drug-CD complex.
Complexation of the drug (D) to CD occurs through a non-covalent interaction between the
molecule and the CD cavity. This is a dynamic process whereby the drug molecule continuously
associates and dissociates from the host CD. Assuming a 1:1 complexation, the interaction will be
as follows:
Two parameters, the complexation constant (K) and the lifetime of the complex, are very
important for the drug release mechanism.
Dilution. Dissociation due to dilution appears to be a major release mechanism. The example
reported by Piel et al.[51] for miconazole, a more strongly bound drug compared to prednisolone,[52]
supports the probable role of dilution. Dilution is minimal when a drug-CD complex is administered
ophtalmically.
Figure 15. Structure of miconazole.
Competitive displacement. Competitive displacement of drugs from their CD complexes probably
plays a significant role in vivo. Tokumura et al.[53,54] reported that the β-CD complex of a poorly
water-soluble drug, cinnarizine, was more soluble in vitro than cinnarizine alone. Oral
administration of the complex showed less bioavailability than expected, based on the in vitro
dissolution experiments. It was suggested that cinnarizine was too strongly bound to the CD so that
complex dissociation was limiting oral bioavailability. Co-administration of phenylalanine, a
displacing agent, improved the bioavailability of the drug from the complex but not from
conventional cinnarizine tablets.
Page 23
22
Figure 16. Structure of cinnarizine.
Protein binding. Drug binding to plasma proteins may be an important mechanism by which the
drug may be released from a drug-CD complex. It is evident that proteins may effectively compete
with CDs for drug binding and thus facilitate the in vivo release of drugs from drug-CD complexes.
Dilution alone may be effective in releasing free drugs from weak drug-CD complexes but when the
strength of the binding between the drug and CD is increased, a mechanism such as competitive
displacement is at work. Plasma and tissue protein binding may also play a significant role. Frijlink
et al.[55] studied the effect of HP-β-CD (2-hydroxypropyl-β-cyclodextrin, Figure 17) on the
displacement of both naproxen and flurbiprofen from plasma binding sites in vivo. They found that
tissue distribution of flurbiprofen and naproxen was higher when HP-β-CD-drug solution was
administered compare to drug solution in plasma, 10 minutes after parental dose, meaning that
more drug was free from CD solution to distribute to the tissues than from the plasma solution.
Figure 17. Structures of HP-β-CD (top), naproxen (bottom left) and flurbiprofen (bottom right).
Page 24
23
Drug uptake by tissue. A potential contributing mechanism for drug release from CD is preferential
drug uptake by tissues. When the drug is lipophilic and has access to tissue, and is not available
to the CD or the complex, the tissue than acts as a “sink” causing dissociation of the complex based
on simple mass action principles. This mechanism is more relevant for strongly bound drugs or
when the complex is administered at a site where dilution is minimal, e.g., ocular, nasal, sublingual,
pulmonary, dermal o rectal sites.
Change in ionic strength and temperature. In the case of a weak electrolyte, the strength of binding
to CD is dependent on the charged state of the drug, which is dependent on dissociation constant(s)
of the drug and the pH of the environment. For most molecules, the ionized or charged form of the
molecule has poorer binding to CD compared to the non-ionized or neutral form of the drug,
especially when bound to a neutral CD.[56,57] Loftsson et al.[50] and Inoue et al.[58] have shown that
binding of substrate to CD is an exothermic process. Hence, any increase in temperature results in
a weakening of the complex and thus increases the free fraction of substrate. Drug-CD complexes
are usually prepared and stored at/or below room temperature. Since the normal body tissue
temperature can be as high as 37°C, this difference in temperature may be another contributing
factor to drug dissociation in vivo.
3.3) Advantages of CDs inclusion complexes
As we have already described, cyclodextrins inclusion complexes offer a great tool to enhance
the solubility of poorly water-soluble drugs. However, this is not the only advantage CDs bring to
the pharmaceutical industry and they offer many other features:
Enhancement of bioavailability. When poor bioavailability is due to low solubility, CDs are of
extreme value. Preconditions for the absorption of an orally administered drug are its release from
the formulation in dissolved form. When drug is complexed with CD, dissolution rate and,
consequently, absorption are enhanced. Reducing the hydrophobicity of drugs by CD complexation
also improves their percutaneous or rectal absorption. In addition to improving solubility, CDs also
prevent crystallization of active ingredients by complexing individual drug molecules so that they
can no longer self-assemble into a crystal lattice.[59]
Improvement of stability. CD complexation is of immense application in improving the chemical,
physical and thermal stability of drugs. For an active molecule to degrade upon exposure to oxygen,
water, radiation or heat, chemical reactions must take place. When a molecule is entrapped within
the CD cavity, it is difficult for the reactants to diffuse into the cavity and react with the protected
guest.[60]
Page 25
24
Reduction of irritation. Drug substances that irritate the stomach, skin or eye can be encapsulated
within a CD cavity to reduce their irritancy. Inclusion complexation with CDs reduces the local
concentration of the free drug, below the irritancy threshold. As the complex gradually dissociates
and the free drug is released, it gets absorbed into the body and its local free concentration always
remains below the levels that might be irritating to the mucosa.
Prevention of incompatibility: Drugs are often incompatible with each other or with other inactive
ingredients present in a formulation. Encapsulating one of the incompatible ingredients within a CD
molecule stabilizes the formulation by physically separating the components in order to prevent
drug-drug or drug-additive interaction.[61]
Odour and taste masking. Unpleasant odour and bitter taste of drugs can be masked by
complexation with CDs. Molecules or functional groups that cause unpleasant tastes or odours can
be hidden from the sensory receptors by encapsulating them within the CD cavity. The resulting
complexes have no or little taste or odour and are much more acceptable to the patient.
Material handling benefits: Substances that are oils/liquids at room temperature can be difficult to
handle and formulate into stable solid dosage forms. Complexation with CDs may convert such
substances into microcrystalline or amorphous powders, which can be conveniently handled and
formulated into solid dosage forms by conventional production processes and equipment.[62]
This pharmaceutical progess is taking application in many fields, and worldwide there are
already several examples of marketed drugs containing cyclodextrins; some of them are listed
below:
Page 26
25
Table 2. Some examples of marketed products containing cyclodextrins.
Drug Formulation Trade name Company
α-Cyclodextrin
Alprostadil (PGE1) IV solution Prostavasin Ono (Japan)
Cefotiam hexetil HCl Oral tablet Pansporin T Takeda (Japan)
β-Cyclodextrin
Benexate HCl Oral capsule Ulgut Teikoku Kagaku
Sanyou (Japan)
Dexamethasone Dermal ointment Glymesason Fujinaga (Japan)
Nicotine Sublingual tablet Nicorette Pharmacia (Sweden)
Nitroglycerin Sublingual tablet Nitropen Nihon Kayaku
(Japan)
Piroxicam Oral tablet Brexin Chiesi (Italy)
Tiaprofenic acid Oral tablet Surgamyl Roussel-Maestrelli
(Italy)
2-Hydroxypropyl-β-cyclodextrin
Cisapride Suppository Propulsid Janssen (Belgium)
Indomethacin Eye drop solution Indocid Chauvin (France)
Itraconazole Oral and IV solutions Sporanox Janssen (Belgium)
Mitomycin IV solution Mitozytrex SuperGen (USA)
MitoExtra Novartis
(Switzerland)
Randomly methylated β-cyclodextrin
17β-Oestradiol Nasal spray Aerodiol Servier (France)
Chloramphenicol Eye drop solution Clorocil Oftalder (Portugal)
Sulfobutylether β-cyclodextrin
Voriconazole IV solution Vfend Pfizer(USA)
Ziprasidone maleate IM solution Geodon, Zeldox Pfizer(USA)
2-Hydroxypropyl-γ-cyclodextrin
Diclofenac sodium Eye drop solution Voltaren ophtha Novartis
(Switzerland)
Page 27
26
3.4) CD applications in the design of some novel delivery system
The particular properties showed by cyclodextrins either to form complex drugs or to act as
functional carrier materials in pharmaceutical formulations, have opened the field to the possibility
of new applications as potential candidates for efficient and precise delivery of required amounts
of drugs to a targeted site for a necessary period of time. In the last fifteen years, many new tools
have been studied in order to design different release-controlled preparation by employing CD
conjugates in combination with other carriers with different releasing properties.
Liposomes. In drug delivery, the concept of entrapping CD–drug complexes into liposomes
combines the advantages of both CDs (such as increasing the solubility of drugs) and liposomes
(such as targeting of drugs) into a single system, circumventing the problems associated with each
system. Liposomes entrap hydrophilic drugs in the aqueous phase and hydrophobic drugs in the
lipid bilayers and retain drugs en route to their destination.[63] The fact that some lipophilic drugs
may interfere with bilayer formation and stability limits the range and amount of valuable drugs that
can be associated with liposomes. By forming water-soluble complexes, CDs would allow insoluble
drugs to accommodate in the aqueous phase of vesicles and thus potentially increase drug-to-lipid
mass ratio levels, enlarge the range of insoluble drugs amenable for encapsulation (i.e.,
membrane-destabilizing agents), allow drug targeting and reduce drug toxicity. Problems
associated with intravenous administration of CD complexes such as their rapid removal into urine
and toxicity to kidneys, especially after chronic use, can be circumvented by their entrapment in
liposomes.[64]
Nanoparticles. Nanoparticles are stable systems suitable to provide targeted drug delivery and to
enhance the efficacy and bioavailability of poorly soluble drugs. However, the safety and efficacy
of nanoparticles are limited by their very low drug loading and limited entrapment efficiency (with
classical water emulsion polymerization procedures) that may lead to excessive administration of
polymeric material. Two applications of CDs have been found very promising in the design of
nanoparticles: one is increasing the loading capacity of nanoparticles and the other is spontaneous
formation of either nanocapsules or nanospheres by nanoprecipitation of amphiphilic CDs diesters.
Both the new techniques have been reported to be useful due to great interest of nanoparticles in
oral and parenteral drug administration.[65]
Porphyrin–cyclodextrin conjugates as a nanosystem for versatile drug delivery and multimodal
cancer therapy. The porphyrin–CD conjugates were prepared and tested for selective and effective
multifunctional drug delivery and therapy. The porphyrin receptor system combines efficient binding
of the selected drug to the CD cavity and photosensitizing properties of the porphyrin moiety with
high accumulation of the whole complex in cancer tissue.[66] Many therapeutic methods such as
chemotherapy, radiotherapy, photodynamic therapy and immunotherapy are used for treatment of
cancer.[67] These methods can be used as a single therapy or in combination with other therapeutic
approaches; the latter strategy is called combined therapy. The advantage of the combined
Page 28
27
therapies is their higher therapeutic effect compared to the single therapy approach. The most
favourable results are achieved when two or more different modes of treatment are applied
simultaneously.
4) CATALYSIS BY CYCLODEXTRINS
4.1) Introduction in cyclodextrin catalysis
As shown in Table 3, cyclodextrins exhibit catalysis in many organic reactions. A typical rate vs.
concentration plot for the catalysis by cyclodextrin is shown in Figure 18a, which is reminiscent of
enzymatic saturation kinetics. A double reciprocal plot of the same data shows a straight line, just
as an enzymatic reaction does, as shown in Figure 18b. this double reciprocal plot is a direct analog
of a Lineweaver-Burk plot in enzymatic kinetics.
Table 3. Reactions accelerated by cyclodextrins.
Reaction Substrate Acceleration factor
Cleavage of esters
Phenyl esters
300
Mandelic acid esters 1.38
Cleavage of amides Penicillins 89
N-Acylimidazoles 50
Acetanilides 16
Cleavage of organophosphates Pyrophosphates > 200
Methyl phosphonates 66.1
Cleavage of carbonates Aryl carbonates 7.45
Cleavage of sulfates Aryl sulfates 18.7
Decarboxylation Cyanoacetate anions 44.2
α-Ketoacetate anions 3.95
Oxidation α-Hydroxyketones 3.3
Page 29
28
Figure 18. a) The pseudo-first order rate constant for release of phenol from p-nitrophenyl
acetateat pH 10.6 plotted as a function of added β-CD. b) 1/(kobs – kun) for p-nitrophenyl acetate
decomposition is plotted vs. the reciprocal of the β-CD concentration (data from Figure 18a).[68]
Catalysis by cyclodextrins may be divided into two categories:
a. catalysis by the hydroxyl groups, in which the hydroxyl groups of the CD function as intracomplex
catalysts toward the substrates included in the cavity of the cyclodextrin;
b. effect of the reaction field, in which the cavity of the cyclodextrin operates as an apolar and
sterically restricted reaction field.
Both of these catalyses are important in enzymatic reactions.
Catalysis by the hydroxyl groups. Three kinds of catalysis by the hydroxyl groups of cyclodextrins
are known:
1. nucleophilic catalysis;
2. general base catalysis;
3. general acid catalysis.
In nucleophilic catalysis, an anion of a secondary hydroxyl group of the cyclodextrin (CD-OH)
attacks the electrophilic center of the ester substrate included in the cavity of the cyclodextrin,
resulting in the formation of acyl-cyclodextrin together with the realease of the leaving group
(Scheme 1). The catalysis is completed by the regeneration of the cyclodextrin through the
hydrolysis of the acyl-cyclodextrin.
Scheme 1.
Page 30
29
Some studies indicate that the anion of the 3-hydroxyl group rather than the anion of the 2-
hydroxyl group is the nucleophile. NMR studies of phenol-inclusion complexes[69] and the
determination that the tosylation reaction of cyclodextrin takes place under basic conditions[70] lead
to the conclusion that substitution occurs exclusively at the 3-position.
However, Breslow et al.[71] have found an acyl group on the 2- and 3-positions; his reactions were
carried out in basic solution where migration would be expected.
General base catalysis involves enhancement of the nucleophilicity of the water molecule by
the abstraction of a proton. In the cyclodextrin case, general base catalysis was found for the first
time in the hydrolysis of trifluoroethyl ester of p-nitrobenzoate.[72]
There have been no examples of reactions proceeding via general acid catalysis alone by
cyclodextrin. In the hydrolysis of trifluoroacetanilide, however, general acid catalysis enhances the
cleavage of the tetrahedral intermediate formed by nucleophilic attack by a secondary alkoxide ion.
General acid catalysis serves to convert the leaving group from an extremely unstable anion of
aniline to a stable neutral aniline molecule (Scheme 2).[73]
This is the one of the few examples of the cyclodextrin-catalyzed hydrolysis of an amide.
Scheme 2.
Effect of reaction field. Catalysis by cyclodextrins does not always involve the catalytic functions of
the hydroxyl groups. Sometimes, cyclodextrins simply provide the cavities as a reactions field. This
effect is attributable to:
a) a “microdielectric catalysis” due to the apolar character of the cavity or
b) a “conformational catalysis” due to the geometric requirements of inclusion.
Page 31
30
A typical example of microdielectric catalysis by cyclodextrin is the decarboxylation of anions of
activated acids such as α-cyano or β-keto acids.[74] These reactions proceed unimolecularly via
rate-determining heterolytic cleavage of the carbon-carbon bond adjacent to the carboxylate group
(Scheme 3).
Scheme 3.
Cyclodextrin can accelerate the first step through microdielectric catalysis, since the interior of the
cavity has an apolar or either-like atmosphere. In fact, the activation parameters for cyclodextrin-
catalyzed reactions are almost identical to reactions in a 2-propanol-H2O mixture.
An example of conformational catalysis is shown in the intramolecular acyl transfer of 2-
hydroxymethyl-4-nitrophenyl pivalate (Scheme 4).
Scheme 4.
The inclusion of the reactant in the cavity of an α-cyclodextrin produces a 5-6 fold acceleration in
the rate of conversion to the product. α-Cyclodextrin, by virtue of its ability to include the ragent
molecule within a rigid binding site, perturbs the equilibrium between conformers (a) and (b) and
therefore forces the reaction groups to assume a favorable conformation with respect to the
reaction. Thus the binding forces between α-cyclodextrin and the 2-hydroxymethyl-4-nitrophenyl
Page 32
31
pivalate are utilized to effect these conformational restrictions. The inclusion complex of the α-
cyclodextrin with the “transition state analog” of the reaction (Figure 19) is more stable than the α-
cyclodextrin inclusion complex with the reactant by a factor which agrees exactly with the observed
acceleration rate. Hence, the driving force can be attributed entirely to the affinity of the cyclodextrin
for the activated complex.[75]
Figure 19. Structure of the transition state analog.
Another important example of this effect regards Diels-Alder reactions. The Diels-Alder reaction
is an example of an important chemical process for which enzyme catalysis are not available.
Models indicated that β-CD could bind cyclopentadiene into its cavity along with a slim dienophile
such as acrylonitrile (Scheme 5) and for this reason the addition reaction was accelerated the β-
CD.[76-79] By contrast, the smaller α-CD inhibited the reaction by binding only the cyclopentadiene,
and β-CD was also an inhibitor for Diels-Alder reactions with larger components.
Scheme 5. An example of a Diels-Alder reaction promoted by cyclodextrin binding.
Interestingly, this study showed that water itsels also strongly promoted the Diels-Alder reactions,
because of the hydrophobic effect.
An important feature of Diels-Alder reactions in cyclodextrins, and for that matter in water
solution itself, is an increase of the selectivity for products obtained from the most compact
transition states. It has also been seen that the regioselectivity of a Diels-Alder reactions which has
a choice of reaction position, can ben enhanced when the reactions occurs inside a cyclodextrin
cavity (Scheme 6).[80]
Page 33
32
Scheme 6. Cyclodextrin binding promoted regioselectivity of a Diels-Alder reaction.
4.2) Specificity in cyclodextrin catalysis
Catalysis by cyclodextrins often shows specificity, which is characteristic of enzymatic catalysis.
In this case, the specificity are divided into 3 categories:
1. substrate specificity, in which subtle changes in the structure of the substrates have large effect
on catalysis;
2. product specificity, in which the products of the catalysed reactions are highly selective;
3. (D,L)-specificity, in which enantiomeric recognition is made by the cyclodextrin.
Substrate specificity. The most striking specificity by cyclodextrins with respect to the substrate is
found in the hydrolysis of phenyl acetates. As shown in Table 4, the magnitudes of the acceleration
by a α-CD (kcat/Km) for meta- substituted compounds are 29-, 236-, 88- and 13-fold larger than
those for para-substituted compounds, for methyl, tert-butyl, nitro and carboxyl substitution,
respectively. Similar results are obtained also fo β-cyclodextrin.
Page 34
33
Table 4. Catalytic rate constants and acceleration in the α-CD-catalysed hydrolysis oh phenyl
acetates.[68]
Acetate kcat (10-2/sec) kcat/kun Kd (10-2 M)
Phenyl 2.19 27 2.2
m-Tolyl 6.58 95 1.7
p-Tolyl 0.22 3.3 1.1
m-tert-Butylphenyl 12.9 260 0.2
p-tert-Butylphenyl 0.067 1.1 0.65
m-Nitrophenyl 42.5 300 1.9
p-Nitrophenyl 2.43 3.4 1.2
m-Carboxyphenyl 5.55 68 10.5
p-Carboxyphenyl 0.67 5.3 15.0
The larger magnitude of the acceleration of the cleavage of meta-substituted phenyl acetates
by cyclodextrins is due to a smaller activation enthalpy.[81] The magnitude of the acceleration by
the CD is governed by the distance between the nucleophile (3-OH group) and the electrophile
(carbonyl carbon atom). The order of the increase of the acceleration (m-nitrophenyl acetate >>
phenyl acetate > p-nitrophenyl acetate) is identical with that of the decrease of the distance
between the carbonyl carbon atom of the substrate and the O-3 atom of the cyclodextrin.
The strong dependence of the acceleration on the distance is associated with a change in
conformation of the inclusion complex in going from the initial state to the transition state. A
conformational change makes the access of the carbonyl carbon atom to the O-3 atom possible so
that a nucleophilic reaction can result in the formation of the tetrahedral intermediate.
The conformational change should accompany an increase in enthalpy, since the driving force of
the formation of the inclusion complex is equivalent to a decrease of enthalpy. Thus, the
conformational change partially compensates for a decrease in activation entropy coming from loss
of the translational and rotational entropy due to complex formation between α-cyclodextrin and the
substrate prior to chemical transformation. When no conformational change takes place during
reaction, the loss of entropy by complex formation of the substrate with the catalyst shows up as a
decrease of activation enthalpy, because of structural changes of the water molecules around the
substrate and the catalyst. An inclusion complex with a smaller distance, which requires a smaller
conformational change, shows a larger acceleration because of the large magnitude of the
decrease of the activation enthalpy.
Product specificity. Cyclodextrins exhibit remarkable ortho-para selectivity in chlorination of
aromatic compounds by hypochlorous acid (HOCl) (Scheme 7). Chlorination takes place via
formation of a covalent intermediate, a hypochlorite ester of cyclodextrin. In the chlorination of
anisole by HOCl, para-chlorination occurs almost exclusively in the presence of sufficient
cyclodextrin. For example, selectivity in the presence of 9.4 x 10-3 M α-cyclodextrin is 96%, which
is much larger than that in absence of the cyclodextrin (~60%).
Page 35
34
Scheme 7.
In the proposed mechanism, one of the secondary hydroxyl groups reacts with HOCl to form a
hypochlorite ester, which attacks the sterically favorable para position of the anisole molecule
included in the cyclodextrin cavity in an intracomplex reaction. The partecipation of one of the
secondary hydroxyl groups at the C-3 position in the catalysis was shown by the fact that
dodecamethyl-α-CD, in which all the primary hydroxyl groups and all the secondary hydroxyl groups
at the C-2 positions are methylated, exhibit equal or larger ortho-para specificity.[82]
(D,L)-Selectivity. The first observation of asymmetric specificity of cyclodextrins in catalysis was in
the hydrolysis of ethyl mandelates; however, the asymmetric effect was quite small. A much larger
asymmetric specificity was observed in the cyclodextrin-accelerated cleavage of 3-carboxy-2,2,5,5-
tetramehtylpyrrolidin-1-oxy-m-nitrophenyl ester (1), a substrate having an asymmetric carbon atom,
like the mandelates, adjacent to the carbonyl group of the hydrolytically labile ester.[83]
The hydrolysis of 1 in the presence of cyclodextrin proceeds in the same way as that described for
phenyl esters (Scheme 1); i.e. binding of the substrate with cyclodextrin using the phenyl portion of
the substrate, acylation of cyclodextrin, and deacylation of the acyl-CD. The cleavage of racemic 1
in the presence of α-cyclodextrin is biphasic with a fast first step followed by a slower second step.
Because of the different rates at which the two complexed enantiomers are cleaved, the first and
second phases, respectively, correspond to the cleavages of the (+) and (-) enantiomers. This
assignment is based on the fact that both the catalytic rate constats, k2, and the dissociation
constant of the complex, Kd, determined from the dependence of the rates of the first phase on the
cyclodextrin concentration, are equal to those determined by using the optically pure (+)
enantiomer. The k2 for the (+) enantiomer is 6.9 times larger than that for the (-) enantiomer.
Interestingly, in contrast to α-CD, β-CD exhibited no appreciable enantiomeric specificity. The loss
of (D,L)-specificity upon increasing the size of the cyclodextrin cavity indicates that this specificity
is a function of the tightness of binding.
Figure 20. Structire of 3-carboxy-2,2,5,5-tetramehtylpyrrolidin-1-oxy-m-nitrophenyl ester
Page 36
35
A much larger D,L specificity is exhibited by cyclodextrin in the cleavage of chiral
organophosphates such as isopropyl methylphosphonofluoridate (2) and isopropyl p-nitrophenyl
methylphosphonate (3).
Figure 21. Structure of isopropyl methylphosphonofluoridate and
isopropyl p-nitrophenyl methylphosphonate.
The α-cyclodextrin-accelerated cleavages of these organophosphates proceed through
nucleophilic attack of a secondary hydroxyl group of the cyclodextrin on the phosphorous atom,
resulting in a phosphonylated α-cyclodextrin and hydrogen fluoride or p-nitrophenol.
In the cyclodextrin-accelerated cleavage of 2, the catalytic rate constant for the (R)-(-) enantiomer
is 35.6 times larger than that for the (S)-(+) enantiomer. This difference arises from the
stereospecificity of the inclusion complexes, since the (S)-(+) enantiomer, which is less accelerated,
binds to the α-CD more strongly than the (R)-(-) enantiomer, which is more accelerated. It was
proposed that the stereospecific interaction(s) of the included enantiomers with the hydroxyl groups
at the asymmetric C-2 atom and/or the asymmetric C-3 atom of the cyclodextrin in the inclusion
complexes govern this asymmetric catalysis.
5) MODIFIED CDs AS ENZYME MODELS
5.1) Introduction on modified CDs as enzyme models
As previously shown, cyclodextrins in their native forms show many features characteristic of
enzymes: specificity, formation of catalyst-substrate complexes prior to chemical transformation
and large accelerations of reactions in which they are involved.
However, CDs as enzyme models suffer from a shortcoming, namely their only catalytic group
is the hydroxyl group, which restricts the scope of their applicability. Thus, many attempts to
introduce other catalytic functional groups into cyclodextrins have been made. In these modified
cyclodextrins, the introduced groups work as catalytic sites and the cavities of cyclodextrins act as
the binding sites for the substrates.
In a direct study on facial selectivity about the possibility that the cyclodextrin have different
catalytic activity according to the position of the substituent groups, catalysts were prepared with a
Page 37
36
phosphate group attached to either the primary or the secondary face of a β-CD, and it was found
that both were effective.[84] Thus for many purposes either cyclodextrin face is suitable for catalytic
group attachment,[85-87] but there are also examples of substrates that preferentially bind into the
secondary face of β-CD, which is somewhat more open.[88-91] With such substrates, the facial
placement of the catalytic group will matter.
5.2) Chymotrypsin mimics
Attachment of a simple catalytic group to a cyclodextrin can afford interesting enzyme mimics.
For example, Cramer and Mackersen[92] have introduced an imidazole group at C-6 of a β-CD (4)
to mimic the enzymatic activity of chymotrypsin, a proteolytic enzyme acting in the digestive system
of many organisms by facilitating the cleavage of peptide bonds by a hydrolysis reaction. The
reasoning behind this functionalization was that the enzyme has a marked optimum at pH of 7,
indicating the partecipation of a catalytic group with pK ~ 7 in the rate determing step.
Figure 22. Imidazole group at C-6 of a β-CD.
This has shown only a slight rate enhancement compared to a “free” cyclodextrin since the catalytic
group is attached to a primary side (C-6) on the essentially closed face of the toroidal cyclodextrin.
Later efforts to attach an imidazole group to a secondary side at C-2 or C-3 located on the more
open face of cyclodextrin, in order to improve catalysis, have not been successful. The availability
of the selective C-2 tosylate of β-CD (5), via an ingeneous procedure developed by Ueno and
Breslow involving a tosyl transfer reaction,[87] allowed the reaction of 5 with imidazole to sythesise
a catalytically efficient enzyme model of chymotrypsin (Scheme 8).
Page 38
37
Scheme 8.
Preliminary kinetic results for the hydrolysis of p-nitrophenylacetate (Table 5) have shown that the
chemical model 6 of the enzyme chymotrypsin with an imidazole on the secondary side of β-
cyclodextrin has a rate constant 70 times higher than that of β-cyclodextrin with the imidazole
substituent on the primary side.[93]
Table 5. Pseudo-first-order rate constantsa for the hydrolysis of p-nitrophenylacetate.
No. Catalyst 104k (s-1)b
1 β-CD 0.62 ± 0.01
2 β-CD with imidazole on C-6
primary side
12.20 ± 0.01
3 β-CD with imidazole on C-2
secondary side (6)
859.0 ± 2.5
a Rate of release of p-nitrophenol determined spectrometrically at 400 nm in Tris-HCl buffer (0.02
M, pH 7.5), catalyst (0.30 x 10-2 M), p-nitrophenylacetate (0.30 x 10-4 M) with 0.50% (v/v) added
acetonitrile at 25°C.
b Average of three runs.
5.3) Serine protease mimics
Molecule 7 owns an imidazole substituent carrying a benzoate group in a position to imitate the
function of the aspartate ion in the catalytic triad characteristic of serine proteases (Scheme 9).[94-
96]
It was claimed that this compound acted to hydrolyze a bound substrate using the imidazole as a
base with the carboxylate hydrogen bonded to it.
Page 39
38
Scheme 9. Compound 7 and the mechanism designed to mimic catalytic triad in serine
proteases.
However, the data reported did not support this claim but made it clear that the additional functional
groups played no useful role. For instance, the reaction rate was said to be first order in hydroxyde
ion at pH above neutrality, not consistent with catalysis by the attached imidazole group. The
reaction was simply the normal reaction of unsubstituted cyclodextrin with the bound substrate and
actually at a slower rate than for cyclodextrin without the added functionality.
Other researches showed that the imidazole and carboxylate in 7 play no catalytic role but instead
impede the reaction.[97,98]
Thus, there is as yet no true mimic of the serine protease enzymes.
5.4) Metalloenzyme mimics
Chymotrypsin is only moderately effective as an enzyme, and much higher rates are seen with
metalloenzymes. Zinc is especially important in such hydrolytic enzymes. For example, the enzyme
carboxypeptidase A uses zinc in a typical bifunctional role, at the same time activating the carbonyl
towards addition by coordinating with its oxygen and activating water molecule to act as a
nucleophile. Models for this type of process, that act by using metal complexing as the substrate
binding force and a coordinated oxime as the nucleophile, have already been designed; one
example is molecule 8.[99] The geometry of this compound points that the Lewis acidic Zinc and the
basic oxime anion can co-exist without quenching each other; the electrons can flow from one to
the other only through the bridging carbonyl group of the substrate (9). Consequently, the anion of
8 reacts with metal-bound substrate 10 to transfer the acetyl group to the oxime anionic oxygem,
and then the intermediate 11 rapidly hydrolyzes.
Page 40
39
In this process the metal ion is serving multiple functions: it binds the substrate, acidifies the oxime,
coordinates to the carbonyl oxygen of the transferring acetyl group, and then catalyzes hydrolysis
of intermediate 11.
However, catalyst 8 is effective only with substrates that can bind to the metal ion. Researchers
have developed new strategies in order to overcome this disadvantage by attaching this catalyst,
coordinated as its Ni2+ derivative, to the secondary face of α-cyclodextrin obtaining molecule 12.[100]
This was then able to use the metallo-oxime catalyst previously described, but with substrates that
are not metal ligands, simply those that bind hydrophobically into the cyclodextrin cavity. Further
kinetic studies showed a significantly increased rate of hydrolysis of p-nitrophenylacetate, well
beyond that for hydrolysis without the catalyst or for simple acetyl transfer to the cyclodextrin itself.
A very attractive perspective is provided by ligand 13, in which the metal-coordinating group
links two cyclodextrin rings.
Ligand 13, as its Cu2+ complex, gave as much as 105-fold rate acceleration in the ester
hydrolysis.[101,102] With an added nucleophile that also binds to the Cu2+ ion, the reaction is
Page 41
40
accelerated by over 107. The mechanism deduced involves the metal ion acting as a Lewis acid by
coordination to the substrate carbonyl and also delivering a bound hydroxide ion to the ester
carbonyl group. With a cyclodextrin dimer related to 13 Breslow et al. managed to hydrolyze an
ordinary doubly bound ester, not just the more reactive nitrophenyl esters, with catalytic
turnovers.[103]As another example, disubstituted cyclodextrin 14, in which one substituent is a metal-
binding tris(2-aminoethyl)amine group while the other is an imidazole, has been synthesized.[104]
Zn2+ complexed to the tris(2-aminoethyl)amine group gave good rate acceleration in the hydrolysis
of bound catechol cyclic phosphate 15, which was fastest when the two catalytic groups were
attached to opposite sides of the cyclodextrin so they could not bind each other. The geometry of
the complex led to the selective formation of product 16 rather than 17 (Figure 23); both are formed
equally by ordinary hydrolysis without the catalyst.
Figure 23. Disubstituted cyclodextrin as hydrolitic catalyst for catechol cyclic phosphate 15.
5.5) Enzyme-coenzyme mimics
Enzymes frequently use coenzymes to perform catalytic functions not possible with normal
amino acid side chains of the enzyme itself. Thus, it is of interest to attach coenzymes to
cyclodextrins, as mimics of the enzyme-coenzyme combination. The first example was a catalyst
(18) in which pyridoxamine was linked to the primary face of β-CD through a sulphur atom.[105]
Catalyst 18 was able to transform α-keto acids (19) to α-amino acids (20), as pyridoxamine itself
does, but with selectivity (Scheme 10). That is, phenylpyruvic acid was transaminated ca. 100 times
as rapidly as was pyruvic acid by 18, while simple pyridoxamine does not show such selectivity.
This catalyst is selective because of the binding of the phenyl group into the β-CD cavity.
Page 42
41
Scheme 10. The first artificial enzyme (18), that combined a coenzyme with a cyclodextrin binding
group.
To fix pyridoxamine to the cyclodextrin with even better defined geometry, a set of four
compounds 21-(1-4) (Figure 24) was made by reacting a pyridoxaminedithiol with β-CD-6A,6B-
diiodide.
Figure 24. Four catalysts that mimic 18 with extra selectivity.
As molecular models suggested, one pair of isomers 21(1) and 21(2) with the pyridoxamine held
over the CD cavity had a preference for p-substituted phenylpyruvic acids, while the other pair held
the pyridoxamine to the side and preferred m- substituted phenylpyruvic acids substrates.[106]
Enzymes that synthesize amino acids by transamination do so with stereoselectivity. Thus, in
transamination by artificial enzymes there has been much interest in learning how to direct the
proton addition to a particular face of the developing amino acid. The earliest example 18 of such
an enzyme mimic afforded amino acids with some selectivity, because of the chirality of the
cyclodextrin unit. However, more selectivity is expected if the proton is delivered by a chirally
mounted basic group, as in the enzyme.
In a study of such transamination with a chirally mounted base, that does not involve cyclodextrins,
it was found that optically active amino acids could be produced with up to 98% selectivity. [107]
However, less success has attended attempts to extend this to artificial enzymes based on
cyclodextrins. A compound 22 carrying both a pyridoxamine and an ethylenediamine unit attached
to β-CD on neighbouring primary methylene groups was prepared and studied for its ability to form
amino acids from keto acids with chiral selectivity. Although quite good selectivities were
reported,[108] it has proven difficult to duplicate these findings. In some alternate approaches, optical
Page 43
42
induction has indeed been produced with related catalysts (23) (Figure 25) but not in high 90%
selectivities.[109]
Figure 25. Two transaminase mimics that produce amino acids with some optical selectivity.
Pyridoxal phosphate is the coenzyme for many processes involving amino acids, including the
conversion of serine and indole to tryptophan. Compound 24 has been synthesized coupling
pyridoxal to β-CD on its primary side. This artificial enzyme mimicked tryptophan synthase by
coupling a dehydroalanine intermediate formed on the pyridoxal unit to an indole held in the
cyclodextrin ring (Scheme 11).[110]
Scheme 11. Catalyst 24 mimics tryptophan synthase.
Thiamine pyrophosphate is the coenzyme for many important biochemical reactions that
formally require the intermediacy of an acyl anion. This involves the addition of the thiazolium C-2
anion to the carbonyl group of the substrate. Consistent with this, thiazolium salts will catalyze the
benzoin condensation. Since γ-cyclodextrin has a cavity large enough to bind two phenyl rings
simultaneously, an artificial enzyme 25 was synthesized with a thiazolium ring linked to the γ-CD
(Scheme 12).[111,112]
Page 44
43
Scheme 12. Compound 25 catalyzing the benzoin condensation by binding two benzaldehydes
into the γ-CD cavity and coupling them with a thiazolium catalytic group.
It was the most effective catalyst known for benzoin condensation, apparently because it could bind
two benzaldehydes and then link them with catalysis by the thiazolium group.
Flavins are coenzymes for electron-transfer reactions. Several research groups have attached
a flavin to a cyclodextrin, so as to promote electron transfers involving bound substrates.
In one study, Tabushi et al.[113] were able to attach a flavin to a α-CD (26) and it was found to show
preferential electron transfer to nicotinamide derivatives that can bind into the cyclodextrin cavity.
In another work, Ye et al. attached a flavin to either the primary or secondary rim of β-CD and the
oxidation of bound substrates was investigated.[114,115] In particular, the oxidation reaction of several
substituted benzyl alcohols to their corresponding aldehydes, catalyzed by 2-[(7α-O-10-methyl-7-
isoalloxazino)methyl]-β-cyclodextrin (27), by riboflavin (28) and 7,10-dimethylflavin (29) were
studied.
Page 45
44
The reactions catalyzed by 27 were found to be considerably faster than those catalyzed by either
28 or 29. For example, the 27-catalyzed oxidation of p-tert-butylbenzyl alcohol was complete within
2.5 hours, whereas the same reaction catalyzed by 28 or 29 are very slow. It was also observed
that 28 decomposes under photochemical conditions and exhibits up to only 7 turnovers, whereas
the artificial enzyme is more stable[116] and exhibits more than 100 turnovers under the same
conditions. Moreover, the structural similarity of the flavin moiety in 27 and 29 suggests that the
change in the oxidation potentials caused by the substituents on flavin is not responsible for the
rate of acceleration shown, demonstrating in this way the importance of covalent attachment of the
catalytic site to the binding site.
Nicotinamide is the functional component of some coenzymes that perform oxidation-reduction
reactions, often involving hydride transfer. As a first attempt to simulate the action of coenzyme
NADH-dependent enzymes upon tertiary complex formation, Kojima et al. synthesized two
monosubstituted β-CD (30 and 31) and compared the reduction of ninhydrin (2,2,-dihydroxyindane-
1,3-dione, 33) performed by these compounds with respect to the one performed by NADH itself
(32).
The two enzyme mimics 30 and 31 differs in the orientation of the nicotinamide substituent with
respect to the cavity of the β-cyclodextrin. Both molecules were prepared from a β-cyclodextrin
toluene-p-sulfonate as starting material. It has been reported that tosylation of β-CD in aqueous
solution provided monotosylated β-CD regioselectively.[70] Because of the conformation of the
glucose unit of β-CD, the secondary hydroxy-groups are equatorial, and so the tosyl moiety
attached to β-CD is equatorial. In the synthesis of compound 30, the tosyl group was substituted
by the nitrogen atom of the ring of nicotinamide. This reaction proceeds with inversion of
configuration at the tosylated carbon atom, and this the nicotinamide should be axial. On the other
hand, compound 31 was prepared from and intermediate β-CD iodide. The tosyl group was
replaced by I- intermolecularly giving β-CD iodide, and the iodine atom was then substituted by the
ring nitrogen of nicotinamide. Therefore compound 31 should have undergone a double inversion
at the tosylated C-3 atom resulting in retention of its configuration, and so the nicotinamide group
should be equatorial.
As we can see from the structure of the two enzymes, the nicotinamide group of 30 is located out
of the cavity, while that of 31 is presumed to be partilly included in it.
Page 46
45
Studies on the reduction of ninhydrin showed a large rate enhancement, 40- and 60-fold
respectively, for 30 and 31 compared with monomeric NADH, indicating how effectively these
enzymes may react with substrates upon complex formation. The higher reactivity of 31 over 30
may be explained by considering the structural differences. In considering the proximities of the
reaction site and the substrate in these inclusion complexes, the dihydronicotinamide group of 31
may be closer to the substrate than that of 30, which may cause the difference in the reaction
rates.[117]
Coenzyme B12 catalyzes mainly two types of enzyme-catalyzed reactions: isomerases, in which
a hydrogen atom is directly transferred between two adjacent atoms with concomitant exchange of
the second substituent, X, that may be a carbon atom with substituents, an oxygen atom of an
alcohol, or an amine; and methyltransferases, in which we have the transfer of a methyl group
between two molecules.
Two main studies have been developed to try to take an enzyme mimic using cyclodextrin and
vitamin B12. In the first study,[118] B12 was directly linked to a primary methylene group of β-CD by a
carbon-cobalt bond (34). When the B12 dissociates from the cyclodextrin, a cyclodextrin radical 35
is produced.
This mimics the formation of a deoxyadenosyl radical in the enzymatic process, when the B12 unit
dissociates from the ribose linked to its cobalt atom. The cyclodextrin radical was able to abstract
a phenyl selenide group from a substrate bound into the β-CD cavity in 35, just as deoxyadenosine
can abstract a hydrogen atom from a substrate bound to its enzyme (Scheme 13).
Page 47
46
Scheme 13. Coenzyme-like ability of the radical 35, formed from 34, to perform group transfers.
In the second study,[119] a β-CD group was attached to a propionic acid side chain of vitamin B12
(36). It was found that this species could catalyze some rearrangements related to those of the
enzyme and with a preference for substrates that bind into the β-CD cavity.
5.6) Bifunctional or multifunctional enzyme mimics
Enzymes often use acid and base catalysts derived from their amino acid side chains, and it is
common for them to use more than one such group in simultaneous bifunctional or multifunctional
catalysis. Thus, it is of interest to imitate this feature in artificial enzymes. For example, the enzyme
ribonuclease A uses two imidazole groups, histidines 12 and 119, as its principal catalytic groups
in the hydrolysis of RNA. To mimic this enzyme, two imidazole rings were attached to the primary
face of β-CD.[120,121]
Page 48
47
By the use of appropriate bridging groups, it is possible to make disulfonate esters of β-CD on
neighboring glucose units (AB), on units one further apart (AC), or on units separated by two
glucose residues (AD) (Figure 26).[122]
Figure 26. The AB (37), AC (38), AD (39) isomers of a cyclodextrin carrying two imidazoles
on the primary carbons.
These bridged cyclodextrins were converted to the related diiodides, and reaction with imidazole
afforded catalysts 37-39. All three of these enzyme mimics were able to catalyze the hydrolysis of
a cyclic phosphate 40 that could bind well into the β-CD cavity, and all three showed a bell-shaped
pH vs. rate profile with a rate maximum near pH 6.2. This is almost identical to the pH vs. rate
profile for the enzyme ribonuclease itself and indicates that one imidazole functions in its protonated
form while the other is unprotonated (Scheme 14). Isotope effect studies showed that the two
catalytic groups were operating simultaneously.[123]
Scheme 14. Cyclodextrin bis(imidazoles) catalyzing the hydrolysis of substrates 40 and 41.
In the classical mechanism for the enzyme, the hydrolysis of the cyclic phosphate intermediate in
RNA cleavage involves water delivery to the phosphorous atom by the unprotonated imidazole,
while the leaving group is protonated by the imidazolium ion. If the enzyme mimics used a similar
mechanism, the AD isomer 39 would be the most active, since it has the best geometry for this
mechanism. However, it was found that the best catalyst for the hydrolysis of the cyclic phosphate
Page 49
48
40 was the AB isomer 37. This indicated that the function of the imidazolium ion was to protonate
the phosphate anionic oxygen, which it can be reached better by 37 than by other catalyst isomers.
A study was made on the importance of a tight fit of substrate into the binding cavity for such
enzyme model systems.[124] The substrates were either the tert-butyl derivative 40 or an analogue
41 with a methyl group instead. The catalysts were all AB diimidazoles, but α, β, or γ-cyclodextrins.
The strongest binding was seen with the tert-butylated substrate 40 into the β-CD derivative 37,
and this combination also gave the fastest rate of hydrolysis. It was also the most selective, since
only product 42 could be detected; the other catalyst-substrate combination afforded mixtures of
products 42 and 43.
Being available a set of cyclodextrin catalysts carrying two imidazoles in different geometries, it
is possible to investigate other reactions catalyzed by simultaneous acid-base proton transfers.
One process examined was the enolization of a bound ketone, p-tert-butylacetophenone (44),
which binds well into β-CD (Scheme 15).[125,126]
The reaction showed a bell-shaped pH vs. rate curve, indicating that both the imidazole and
imidazolium ion played a catalytic role. It was found that the best isomer for the enolization,
monitored by deuterium exchange, was the AD isomer. This indicated what the preferred geometry
is for proton abstraction from carbon, an important matter not easily determined without the
geometric information furnished by these bifunctional catalysts. The same catalyst set has also
been examined, and found effective, in two intramolecular aldol condensations involving keto
aldehyde 45 and dialdehyde 46.[127,128]
Scheme 15. Cyclodextrin bis(imidazoles) catalyzing enolization of 44 and aldol condensation of
45 and 46.
Another very interesting example of bifunctional enzyme models is the one provided from the
recent synthesis of a glycoside hydrolases enzyme mimicked by difunctionalized cyclodextrins.
Glycoside hydrolases (also called glycosidases) assist in the hydrolysis of glycosidic bonds in
sugars. The inspiration for the design of the CD glycosidases came from the world of natural
glycosidases which in their catalysis typically encompass two catalytically active carboxylate
Page 50
49
groups in their active site.[129] This acid/base-catalysis principle spawned the invention of
carboxylate cyclodextrin glycosidases, that were able to catalyze aryl glycoside bond hydrolysis
reactions with rate enhancements of up to 1000 times, following the enzyme-characteristic
Michaelis-Menten rate law pattern.[130,131] Their mechanism of catalysis is proposed to involve
electrostatic stabilization of the transition state by the carboxylate groups, with subsequent
nucleophilic substitution with phosphate (Figure 27).
Figure 27. Proposed mechanism of carboxylate CD hydrolysis of aryl glycosides
in presence of phosphate buffer.
Other cyclodextrins enzymes that act as artificial glycosidases are the trifluoromethyl alcohol CDs,
but they do so only with modest activity.[132] The greater breakthrough in CD artificial glycosides
came with the discovery of the CD cyanohydrins, affording up to 8000 times increase in reaction
rate.[133-135] Synthesis of 7A,7D-dicyanohydrin-β-CD, in which the cyanohydrin functionality is
positioned one carbon atom further away from the cavity than in the known 6A,6D-dicyanohydrin
CDs, seems to offer a particularly efficient enzyme model.[136] In the proposed mechanism for the
catalyzed reaction, the electron-withdrawing effect of the nitrile group acidifies the cyanohydrin
hydroxy, easing the donation of this alcohol proton to the substrate glycosidic oxygen, thereby
facilitating bond cleavage (Figure 28).
Page 51
50
Figure 28. Proposed mechanism of cyanohydrin CD catalysis.
6) METHODS FOR SELECTIVE MODIFICATIONS OF CYCLODEXTRINS
6.1) Overview of methods for modification of cyclodextrins
Methods for selective modification of cyclodextrin can be divided into three categories:
1. the “clever” method, where the chemistry of cyclodextrin is exploited to get the desired product
by the shortest route;
2. the “long” method, where a series of protection and deprotection steps have taken place in order
to selectively reach the position which would otherwise not be selectively accessible;
3. the “sledgehammer” method, where cyclodextrin is indiscriminately reacted to give a mixture of
products and the desired one is laboriously separated out from other isomers and homologues by
chromatographic methods.
Given the choice between three categories, one would always choose the first strategy because it
is most productive and least painful; however, a method in the first category is not always available
when a modified cyclodextrin of a specific structure is needed.
Page 52
51
6.2) Chemistry involved in methods for modification of cyclodextrins
It is important to understand the various chemical factors that are involved in these methods for
modification of cyclodextrins to fully appreciate and be able to apply them in the elaboration of new
synthetic pathways towards unprecedented products.
In the present discussion some physical and chemical properties, i.e. the difference in the CD
dimension going from the lower to the upper rim, the hydrophobicity of the cavity or the belt of
intramolecular hydrogen bonds between hydroxyl groups at 2- and 3-position of adjacent glucose
units have been analyzed. However, it is important to point out that two primary factors need to be
considered in the chemistry of cyclodextrins for their modification: the nucleophilicity of the hydroxyl
groups and the ability of CDs to form complexes with the reagents used. All modifications of
cyclodextrins take place at the hydroxyl groups. Since these groups are nucleophilic in nature, the
initial reaction, which directs the regioselectivity and the extent of modifications (mono, di, tri, etc.)
of all subsequent reactions, is an electrophilic attack on these positions.
Of the three types of hydroxyl groups present in cyclodextrins, those at the 6-position are the
most basic (and often the most nucleophilic), those at the 2-position are the most acidic, and those
at the 3-position are the most inaccessible. Thus, under normal circumstances, an electrophilic
reagent attacks the 6-position (I in Figure 29). It is important to recognize that more reactive
reagents will attack the hydroxyl groups less selectively. Thus, more reactive reagents will not only
react with the hydroxyl group at the C-6, but also with those on the seconday side. An example of
this is that the less reactive reagent tert-butyldimethylsilyl chloride (TBDMSCl) will react selectively
with hydroxyl groups at the 6-position,[137] while the more reactive trimethylsilyl chloride (TMSCl)
will react with all the hydroxyl group indiscriminately.[138]
Figure 29. Scheme of the methods for modification of cyclodextrins.
Page 53
52
Since the hydroxyl groups at the 2-position are the most acidic, they will be the first to get
deprotonated.[139] The oxyanion thus formed is more nucleophilic than the non-deprotonated
hydroxyl groups at the 6-position (II in Figure 29). However, this situation is complicated by proton
transfers between these two positions which can lead to a product mixture.
An interesting factor affecting the chemistry of the hydroxyl groups is provided by the ability of
cyclodextrins to form complexes (III). If the electrophilic reagent forms a complex with cyclodextrin,
then the orientation of the reagent within the complex introduces an additional factor in determining
the nature of the product. If the complex formed is very strong, then the predominant product formed
will be dictated by the orientation of the reagent withing the complex. On the other hand, if the
complex is weak, then the product formation will be directed by the relative nucleophilicities of the
hydroxyl groups. It is also important to note that the solvents play an important role in determining
the strength and the orientation of the complex between the reagent and cyclodextrin. Thus, tosyl
chloride reacts with α-cyclodextrin in pyridine to give the 6-tosylated product, whereas in aqueos
base, it gives the 2-tosylated product.[140] The size of the cyclodextrin cavity also has pronounced
effect on the strength and the orientation of the complex and affects the product of the reaction.
For example, in aqueous solutions, tosyl chloride reacts with α-CD to give the 2-substituted product,
whereas with β-cyclodextrin, it give the 6-substituted product.[141]
A strategy used to avoid complications due to binding of the reagent into the cavity of
cyclodextrin, and thus give products which are not expected by their normal nucleophilicity, is to
protect the hydroxyl groups and direct the incoming reagent exclusively to the open –OH groups.
For example, if one protects the 2-position of cyclodextrin, the incoming electrophile can be directed
to the 6-position (IV in Figure 29). An example of this is the protection of the secondary side of
cyclodextrin as the first step, and then the reaction of the primary side with alkyl halides.[138]
Similarly, protection of the primary side enables to direct the incoming electrophile exclusively to
hydroxyl group at the 2-position (V).
6.3) Primary face modification
Since primary hydroxyl groups are more nucleophilic than their secondary counterparts, they
are easily modified into other functional groups. Selective permodification of all the primary hydroxyl
groups is relatively easier than mono-, di-, or tri-substitution because symmetrical substitution is
achieved when the reaction is allowed to run for a longer time with appropriate amount of reagents.
Regioisomerism further complicates this situation when selective di- or tri-substituted cyclodextrins
are to be prepared. Usually these products require solvent and time consuming chromatographic
purification.
Strong electrophiles such as alkyl, phosphoryl, silyl, sulfonyl, or carboxylic acid chlorides react
with hydroxyl groups of CD to produce an alkylated, silylated, sulfonated, or acetylated product
Page 54
53
along with an acid, which is neutralized by using a basic solvent or a weak base. The reason for
this is that cyclodextrins are stable under basic conditions whereas they decompose in the
presence of a strong acid. All these reagents are very reactive and attack the hydroxyl groups of
cyclodextrins indiscriminately producing mono-, di-, or tri-substituted products. Increasing the size
of these reagents is not helpful in controlling the selectivity; even a bulky group like trityl (47) is not
selective and gives a mixture of products which require chromatographic treatment.[142,143]
Among all the sbustituents, the sulfonyl group acts as a good leaving group and can be
displaced by nucleophiles to synthesize useful derivatives. 6-Sulfonates serve as precursor for the
preparation of the 6-deoxycyclodextrin compounds. A large number of nucleophiles attack the
carbon atom at the 6-position in these sulfonates to give the corresponding modified CD.
Monosubstitution at the 6-position of cyclodextrins. The most popular method for
monomodifications at the 6-position of cyclodextrins is to convert the cyclodextrins into its mono-6-
sulfonylcyclodextrin conjugate and then perform a nucleophilic attack of a reagent containing the
appropriate group. These monosulfonates are prepared by reacting 1 equiv. of benzene or p-
toluenesulfonyl chloride with cyclodextrin in pyridine or DMF containing a base (Scheme 16).
Scheme 16.
Monotosylation of cyclodextrin is often a nonselective process and produces a mixture of primary
as well as seconday side tosylated products along with di- or tri-tosylated derivatives.Thus,
depending on the desired purity of the final product, it requires extensive purification. The yield of
the final product is often reduced because the tosylate can undergo and exchange by chloride ions
or an elimination process to give either a 3,6-anhydro compound or an alkene. Pyridine, a non-
user-friendly solvent of choice for this reaction, forms a pyridinium complex with the cavity and
Page 55
54
complicates the workup process. However, the major advantage of this solvent is its ability to direct
the reaction to the 6-position as compared to DMF where sulfonation occurs on both faces of
cyclodextrin. Despite all these problems and drawbacks, monotosylates have been extensively
investigated.[144] One of the best developed methods for the synthesis of monotosylcyclodextrins is
to treat cyclodextrin with tosyl chloride in 1:1 equivalent ratio in aqueous alkaline medium for a short
time, to give the mono-6-tosylate in fairly good yield.
The direct synthesis of monothiocyclodextrins with aromatic thiol and unprotected cyclodextrin in
DMF or pyridine can be performed by a thio-Mitsunobu reaction. This reaction gives a mixture of
mono-, di-, and tri-substituted products which are purified by chromatography (Scheme 17).[145]
Scheme 17.
Mono-tosyl derivatives of cyclodextrins are particularly useful for the possibility to convert them into
mono-deoxy-iodo-CDs (Scheme 18),[146] which can be readily reacted with nucleophile to obtain
mono functionalized cyclodextrins.
Scheme 18.
Page 56
55
Thanks to the reactivity of these compounds, a group of Italian researchers have selectively
functionalized a primary hydroxyl of a β-cyclodextrin with different substituents like Anserine (β-
alanyl-3-methyl-L-histidine), Carnosine (β-alanyl-L-histidine) and Homocaronisine (γ-aminobutyryl-
L-histidine) without using any protecting group (Scheme 19).[147]
Scheme 19.
In addition, monoamino derivatives of cyclodextrin are of particular interest. These are
conveniently obtained from monoazides of cyclodextrin by reduction with triphenylphosphine in the
presence of aqueous ammonia.[86] Monoazides of cyclodextrin are indirectly obtained by heating
the monotosylate with sodium or lithium azide salt in DMF (Scheme 20).
Scheme 20.
A direct approach to make monoazides is through Vilsmeier-Haack type reactions in which
cyclodextrins are heated with NaN3 containing triphenylphosphine in DMF.[148]
Page 57
56
Monoamines show greater solubility and react with isocyanate without a need to protect the primary
hydroxyl groups to produce isothiocyanatocyclodextrins (Scheme 21).[149]
Scheme 21.
Monoamines are invaluable in attaching desired groups to the primary side side of cyclodextrins
via carbodiimide (DCC) coupling technology. This strategy has been used to connect various sugar
units such as β-D-glucose, β-D-galactose, α-D-mannose and β-D- and L-fructose to cyclodextrins
through alkyl chains. Monoamines condense also with D- or L-N-dansylleucine in DMF containing
DCC and 1-hydroxybenzotriazole at room temperature to form D- or L-mono-6-(N-
dansylleucylamino)-6-deoxy- β-cyclodextrin in 50% yield (55 and 56).[150]
Monoaldehydic cyclodextrins are an important class of derivatives because they provide a route
for further modifications. The monoaldehyde has been synthesized by oxidizing 6-tosyl-β-
cyclodextrin in DMSO with collidine added as a non-nucleophilic base. Further oxidation of the
monoaldehyde leads to the corresponding carboxylic acid. If we react the monoaldehyde with
hydroxylamine or hydrazine we can also produce a monooxime or a monohydrazone derivatives
(Scheme 22).[151,152]
Page 58
57
Scheme 22.
However, the monoaldehyde can be synthesized directly by reacting cyclodextrins with Dess-Martin
periodinane (DMP) in 85-100% yield (Scheme 23).[153] This process avoids complications in the
synthesis of monotosylcyclodextrin previously mentioned.
Scheme 23.
Page 59
58
Alkyl ethers of cyclodextrins cannot be synthesized from tosylates because nucleophiles in this
case (alkoxide ions) act as strong bases which pick up protons from hydroxyl groups at the 3-
position and produce the 3,6-anhydro compound by ring inversion (Scheme 24).
Scheme 24. Conversion of a 6-substituted cyclodextrin to a 3,6-anhydrocyclodextrin.
Alkyl ethers of β-CD are obtained by the “long” method in which the primary side is first protected
by TBDMS. This is followed by permethylation of the secondary face (both 2- and 3-positions),
desilylation of the primary side, and then monotosylation of the primary side. The reaction of an
alkoxide ion with this protected tosylate gives the desired alkyl ether on the primary side without
the formation of the 3,6-anhydro derivative (Scheme 25).[154]
Scheme 25.
The main problem with this approach is that methyl groups on the secondary side cannot be easily
removed. This limitation can be overcome by using acetyl groups to protect the secondary side,
which can be subsequently hydrolyzed under aqueous alkaline conditions to afford the final
product. This “long” method has been made shorter by directly protecting the secondary side using
TBDMSCl without first protecting the primary side, a protection strategy that takes advantage of the
Page 60
59
acidity of the hydroxyl groups at the 2-position to selectively deprotonate them. In this way, their
nucleophilicity is increased, as the reactivity towards TBDMSCl compared to that of the hydroxyl
groups on the primary side. The resultant protected cyclodextrin can now be exploited for selective
alkylation of the primary face.[155] The advantage of this approach is that the protecting groups
TBDMS is easily removed under mild conditions once the desired modification on the primary side
are completed.
Disubstitution at the 6-position of cyclodextrins. Disubstituted cyclodextrins are obtained by using
more than 1 equivalent of reagent with cyclodextrin under suitable conditions to give a mixture of
products. It is a cumbersome process that affords the formation of positional isomers and regio-
isomers that require extensive purification by HPLC. Statistical calculations suggest that
disubstitution can produce 33 regioisomers in the case of β-cyclodextrins,[156] which indicates the
enormous complexity of the process. As in the sulfonation reactions in general,[157] these can also
be plagued by substitutions on the secondary side.
A particularly efficient method to obtain disulfonated cyclodextrins is by reaction of arenesulfonyl
chlorides with cyclodextrins to give AB, AC, and AD isomers.[158] Although these disulfonyl chlorides
give a mixture of regioisomers, they show a distinct regiospecificity based on their structures. An
elegant method to control the regiospecificity to produce AB, AC, or AD isomers by the use of the
geometry of the reagents has been described.[159] For example, as shown in Figure 37, trans-
stilbene and biphenyl-based capping reagents preferentially give the AD isomers,[160]
benzophenone-based reagents give AC isomers, and 1,3-benzenedisulfonyl chlorides (especially
the electron-rich 4,6-dimethoxybenzene-1,3-disulfonyl chloride[161]) give the AB isomers.
Figure 30. Use of the geometry of reagents to direct the regiospecificity in disubstitution of
cyclodextrins.
Page 61
60
Disulfonated cyclodextrins are important intermediates in the synthesis of disubstituted
cyclodextrins. The strategy is to displace the ditosylates with nucleophiles in a manner similar to
the reaction of monotosylates. It is important to note that the positional isomerism (AB, AC, AD) of
the ditosylate is retained in these disubstituted derivatives.
Another strategy for synthesis of substituted cycldodextrins is to convert the ditosylates to
dideoxy-diiodo derivatives by a reaction with KI and then treat these diiodides with appropriate
nucleophiles. An example is provided by the synthesis of the difunctionalized β-CD (6A,6D-dideoxy-
6A,6D-di(2-aminoethanethio)-β-cyclodextrin (Scheme 26).[162]
Scheme 26.
This approach provides also the advantage that disulfonated or dihalogenated cyclodextrins
react with alkanethiolates in aqueous or DMF medium to give thioethers of cyclodextrins. Thiolate
Page 62
61
ions act as good nucleophiles arther than strong bases and do no produce elimination products or
3,6-anhydro compounds.
Another very particular and interesting approach is the one started from the discovery of an
original strategy to regioselectively deprotect sugars[163] that allowed the possibility of differentiate
cyclodextrins.[164] A fully benzyl-protected CD is selectively bis-deprotected by using
diisobutylaluminium hydride (DIBAL-H) to afford 6A,6D-diols 57 and 58 in 82 and 83% yield from α-
and β-CDs, respectively (Scheme 27).
Scheme 27.
A remarkable feature of this reaction is that it is a rare example of selective CD bis-functionalisation
being as efficient on both α- and β-CDs.[165]
In order to better explore the potentiality of this approach, further studies on DIBAL-H promoted
selective deprotection have been carried out. The first step was the functionalization of the α-CD
57 on positions 6A,6D by an appropriate OR group resistant to DIBAL-H. The aim of this work was
to address if it was possible to duplicate the deprotection in a regioselective manner, and as result
Page 63
62
only two diametrically opposed hydroxyl groups were found to be deprotected, by analogy with the
first deprotection process (Scheme 28).
Scheme 28.
6.4) Secondary face modification
The secondary side is more crowded than the primary side due to the presence of twice the
number of hydroxyl groups. Hydrogen bonding between hydroxyl groups at the 2- and 3-position
makes them rigid and less flexible with respect to C-6 hydroxyl groups. All these factors make the
Page 64
63
secondary side less reactive and harder to functionalize selectively than the primary face. During
the course of a reaction, as the degree of substitution increases, the secondary side becomes even
more crowded and this results in steric hindrance towards the incoming nucleophile forcing the
attacking group toward the other face and decreasing selectivity. Positional isomerism on the
secondary side further complicates the situation.
Monosubstitution at the 2-position of cyclodextrins. An authentic sample of mono 2-tosyl-β-
cyclodextrin was first prepared using a group transfer strategy.[87] m-Nitrophenyl tosylate was
reacted with cyclodextrin in DMF/aqueous buffer at pH 10. This reaction proceeds via complex
formation to transfer the tosyl group to the 2-position. The tosyl group gets transferred preferentially
to the 2-position due to the orientation of the reagent within the host molecule (Scheme 29).
Scheme 29.
The hydroxyl groups at the 2-positions are more acidic than those at 6-position, and this feature
has been exploited by using NaH as strong base under anhydrous conditions for selective
tosylation at the 2-position.[140] Yields in all these cases are affected by the elimination of the
sulfonate group due to its good leaving behaviour. The elimination of the tosyl group at the 2-
position by the hydroxyl groups affords the manno-2,3-epoxycyclodextrin (Scheme 30).
Scheme 30.
Formation of an inclusion complex by the cavity of cyclodextrins plays a very prominent role in
determining the reaction site for the incoming group. An included reagent may react with the
hydroxyl groups at the 2-, 3-, or 6- positions depending on the nature of the complex. Some of
these problems can be overcome by protection of the primary side before tosylation of the
secondary side. For example, per-6-silyl-mono-2-tosylcyclodextrin is synthesized by the reaction
of 6-silylated cyclodextrin with tosyl chloride in THF with NaH as a base in 32% yield.[166] An
Page 65
64
advantage of this strategy is that the reaction as well as the purification steps can be carried out in
organic solvents and the desilylation can be easily carried out to yield the desired tosylate (Scheme
31).
Scheme 31.
A mono 2-amino cyclodextrin derivative was prepared starting from a perbenzoylated β-CD by
a series of reactions including selective de-O-benzoylation at one of the 2-position, oxidation, oxime
formation and reduction of the oxime to obtain the mono-2-amino-β-cyclodextrin (Scheme 32).[167]
Page 66
65
Scheme 32.
Monosubstitution at the 3-position of cyclodextrins. Direct and selective reaction at the 3-position
is a very complicated and difficult process because of higher reactivity of 2- and 6-hydroxyl groups.
Most modifications at the 3-position are obtained by a reaction of a nucleophile with manno-mono-
2,3-epoxycyclodextrin. Tosyl chloride has been reacted indiscriminately with cyclodextrin to give
the mono 3-tosylate of α- and β-cyclodextrins along with 2- or 6-tosylates as previously mentioned.
The products are then separated and purified by chromatographic techniques. However, 2-
naphthalenesulfonyl and 3-nitrobenzenesulfonyl chlorides have been shown to give a higher yield
of 3-substituted cyclodextrin. CDs are reported to form complexes with these reagents to direct the
electrohpile toward the hydroxyl group at either 3- or 2-position. In aqueous acetonitrile solution,
the former reagent gives 29% yield of the 3-sulfonate and only 3.9% and 0.4% of 2- and 6-
sulfonates, respectively.
In general, nucleophiles react with manno-mono-2,3-epoxycyclodextrin to give substitution at
the 3-position.
Alkylamino derivatives at 3-position are prepared by ring opening of the manno-2,3-epoxide. Mono-
3-aminocyclodextrins are synthesized by treating the epoxyde with aqueous ammonia solution at
room temperature. Similarly, hydroxylamine reacts with manno-2,3-epoxide to give 3-
hydroxylaminocyclodexrin. The hydroxylamino derivative is converted to the corresponding oxime
by air oxidation and can be hydrolyzed by NaHSO3 to obtain a cyclodextrin ketone.[168] Mono-3-
thiocyclodextrin is synthesized by treating the epoxide with sodium thioacetate and subsequently
hydrolyzing this thioester with aqueous alkaline solution.
Page 67
66
7) STRATEGY FOR THE SELECTIVE ASYMMETRICAL DIFUNCTIONALIZATION OF THE
PRIMARY HYDROXYL GROUPS OF β-CYCLODEXTRIN
Some computational studies have shown that the distance between the primary hydroxyl groups
in positions 1,4 of a β-cyclodextrin, if suitably functionalized with amino acid residues, allow to build
enzyme models particularly useful for the development of new drugs or for the study of interactions
between these receptors with some biologically active molecules.
The aim of this project was, first, to develop a synthetic strategy to obtain the β-cyclodextrin 59
difunctionalized in positions 6A and 6D with one unit of L-Asparagine and one of L-lysine,
respectively.
As already described in chapter 6, some strategies to synthesize 1,4 (6A – 6D) difunctionalized
β-cyclodextrins are known in literature. One of them is based on the preparation of a perbenzylated
cyclodextrin and subsequently, through the use of DIBAL-H, selectively deprotect the two desired
position obtaining the diol 58.
Page 68
67
One might think to get to the target molecule 59 starting from 58 through chemical functionalization
of cyclodextrins. However, it is clear that it remais difficult to functionalize selectively a hydroxyl
with Asparagine and the other with Lysine. At this point a possible strategy may be to functionalize
the two free hydroxyl groups in two different steps. Some methods that allow to obtain monoesters
from aliphatic diols,[169] although they have never been applied in the chemistry of carbohydrates
are present in literature.
This fact offers the possibility to proceed in two ways. The first consists in a monoesterification with
an amino acid residue to obtain a monosubstituted cyclodextrin (60) and the proceed with a second
monoesterification, making the second amino acid to react to obtain compound 61. Another
hypothesis is to synthesize a difunctionalized CD with the same amino acid (62) and, with a
subsequent monotransesterifiction, achieve the derivative 61. A final debenzylation would provide
target cyclodextrin 59 (Scheme 33).
At this point all the possible isomers should be separated with a chiral chromatographic column.
This strategy, as well as having the problem of a poor selectivity, has the disadvantage of the
Page 69
68
necessity to carry out the two steps of protection and deprotection, which would lengthen a
synthesis already constituted by several steps and this could further affect the final yield.
Scheme 33.
Another route is the one that follows the capping mechanism of the position 1 and 4 of the
cyclodextrin using arenesulfonyl chlorides. In particular, the use of a biphenyl-4,4’-disulfonyl
Page 70
69
chloride would lead to the synthesis of derivative 63 without the protection of other free hydroxyl
groups (Scheme 34).[162]
Certainly, the possibility to avoid the protection of the whole cyclodextrin allows to get to the
difunctionalized β-CD through a more sustainable strategy. However, we must not underestimate
that the unprotected β-CDs are not very soluble both in water and in most of the organic solvents.
Scheme 34.
Through this strategy, however, we cannot overcome the problem of obtaining selectively a β-CD
with two different functionalities as desired for the molecule 59. In any case, the idea of a capped
cyclodextrin has suggested a possible strategy to introduce a derivative, analogous to the biphenyl-
4,4’-disulfonyl chloride, constituted by the two amino acids required for the functionalization of the
β-cyclodextrin target, which, thanks to subsequent transformations, allow us to get molecule 59.
By computational approximate calculations we found that the distance between two sulfur atoms
in molecule 64 is equal to 10,49 Å.
Page 71
70
Therefore, through the formation of an imine deriving from the two amino acids, it is possible to
think about getting a β-CD capped in both positions 1 and 4 as shown in structure 65. Next, with
the hydrolysis of the imine, it is possible to get the difunctional cyclodextrin 66. With subsequent
oxidation of the aldehyde functionality and amidation, we can obtain molecule 59 (Scheme 35).
This strategy allows us not only to avoid protection / deprotection steps of the entire cyclodextrin
but also to have in the same step at the same time both the two different amino acids bound to the
CD.
10,49Å
Page 72
71
Scheme 35.
The dimer could be synthesized by reaction of the aldehyde 67, resulting from the Aspartate,[170]
with suitably protected Lysine 68. By performing calculations similar to those conducted for the
disulfonyl chloride 64, we can derive the dimensions of this imine being 10,71Å, a value very close
to that of the disulfonyl derivative. For this reason this seemed a potential synthetic strategy
(Scheme 36).
Scheme 36.
First of all we had to funtionalize both amino acids in order to have the right substrates to
synthesize the dimer. We started with the protection of the α-NH2 of the Lysine. Since this amino
acid possesses two NH2 groups, in order to selectively protect the desired one we have to “lock”
the terminal amine by reacting the aminoacid with benzaldehyde forming imine 70. This reaction is
almost quantitative since the terminal amino group reacts readily with benzaldehyde both for its
position (less steric indrance) and for its enhanced nucleophilicity with respect to the other NH2
which is next to an electron withdrawing group. Once we had imine 70, in a single step, we carried
out the benzyl chloroformate (CbzCl) protection of the free amino group and, by adding
concentrated HCl and heating the reaction mixture, we could hydrolize the imine unlocking the
terminal NH2 (Scheme 37). Since different reactions were carried out under acid o basic pH, we
10,71Å
Page 73
72
chose Cbz as protecting group because it is resistant to these conditions and is simply removed by
Pd/C catalyzed hydrogenolysis.
Scheme 37.
We also decided to convert the carboxylic acid to its methyl ester derivative in order to lower the
possibility of side reactions during all the steps of the synthesis (Scheme 38). We experienced
many difficulties to obtain good yields and purity of the desired products even following procedures
already present in literature, so we focused a lot in improving all the steps of the functionalization
of both the amino acids.
Scheme 38.
The first attempt to carry out this reaction was performed by adding 2 eq. of SOCl2 to methanol
at -5°C and, after some minutes, we added the Nα-(Carbobenzoxy)-L-lysine at the same
temperature and the reaction mixture was stirred for 5h during which it was allowed to warm to
room temperature. Since the yield obtained with this process were not satisfying at all, we decided
to perform different synthetic experimental conditions to obtain the methyl ester derivative (Table
6). These conditions were not used in literature before.
Page 74
73
Table 6.
N. Solvent Reactants Temperature Time Yield
1 MeOH SOCl2 (2 eq.) r.t 5h n/a
2 MeOH SOCl2 (2 eq.) r.t. Overnight 15%
3 MeOH SOCl2 (2 eq.) +
Amberlist 15
r.t. 5h n/a
4 MeOH SOCl2 (2 eq.) +
Amberlist 15
r.t. Overnight 15%
5 MeOH SOCl2 (2 eq.) +
MgSO4
r.t. Overnight 20%
6 MeOH SOCl2 (4eq) r.t. Overnight
35%
7 MeOH SOCl2 Reflux Overnight 50%
At the same time, we proceeded also with the functionalization of the L-Asparagine. In this case
the synthesis requires more steps compared to the one for the L-lysine because this time we have
not only to protect some functionalities but also to convert the amide group to the aldehyde
conjugate. Even in this strategy we decided to protect the amino group with Cbz both for the
advantage reported above and for the possibility to deprotect the two amino acids in one single
step (Scheme 39).
Scheme 39.
After obtaining methyl ester 73, we converted this substrate to the carboxylic acid derivative N-
(Benzyloxycarbonyl)-aspartic acid α-methyl ester (74) by treating the amide with t-BuONO (Scheme
40).
Page 75
74
Scheme 40.
The next two steps involve reduction of the carboxylic acid to alcohol and oxidation of the
hydroxyl group to the corresponding aldehyde (67) by using Dess-Martin Periodinane (DMP).
Scheme 41.
Afterwards, the strategy continues with the condensation reaction between the free amino group
of Lysine and the aldehyde derived from Asparagine to obtain the desired imine (69). By
hydrolyzing methyl ester we can re-establish the carboxylic acid moieties to obtain the dicarboxylic
acid (76, Scheme 42).
Scheme 42.
Page 76
75
This provides us the substrate to perform a coupling between the hydroxyl groups of the β-CD and
the –COOH of the dimer by performing a Steglich esterification, promoted by DCC (N,N’-
Dicyclohexylcarbodiimide) and DMAP (4-Dimethylaminopyridine).
Steglich esterification is a mild reaction which allows the conversion of sterically demanding and
acid labile substrates through constitution of an activated and very good leaving group on the
carboxylate hydroxyl group.
With amines, the reaction proceeds without problems to the corresponding amides because amines
are more nucleophilic. If the esterification is slow, a 1,3-rearrangement of the O-acyl intermediate
to an N-acylurea, which is unable to further react with the alcohol occurs, diminishing the final yield
or complicating purification of the product. DMAP suppresses this reaction, acting as an acyl
transfer reagent. The reaction mechanism is described as follows:
What we expect to happen, since the dimension of the molecule (76) are similar to those of the
disulfonil dichloride derivative (64), is that the synthesized dimer would react selectively, or almost
selectively, with the two primary hydroxyl groups in positions 1 and 4 of the cyclodextrin to have
the molecule (77). At this point, preferentially under mildly acid hydrolysis conditions, we should
obtain the difunctionalized cyclodextrin (78) in which the two amino acid precursors are already in
the right positions: one is just a Cbz-protected L-lysine, the other one is an aldehyde which may be
seen as a L-Asparagine derivative (Scheme 43).
Page 77
76
Scheme 43.
What is required to do next is to re-establish the amide functionality from the aldehyde. This can
be achieved following two different routes. The first involves the direct amidation of the aldehyde.
Several methods are known in literature to do so,[171] but none has been used with substrates similar
to ours. The second is a more classic way that follows a first oxidation of aldehyde to carboxylic
acid, followed by conversion of the –COOH to -CONH2 (Scheme 44).
Page 78
77
Scheme 44.
A final step of deprotection by Pd/C catalyzed hydrogenolysis should lead to the desired 1,4
asymmetrically functionalyzed β-cyclodextrin 59 (Scheme 45).
Scheme 45.
Page 79
78
During my thesis work in laboratory, my personal task has been to selectively functionalize the two
amino acids following the presented strategy, focusing on finding the best reaction conditions and
improving all the steps in order to obtain better yields and more pure substances.
Page 80
79
8) EXPERIMENTAL SECTION
8.1) General
Reactions were monitored through thin layer chromatography on Merck silica gel plates
Kieselgel 60 F254. The separation and purification of compounds were performed through flash
chromatography on silica gel Merck (0,040-0,063 mm). Characterization of products was performed
through 1H and 13C nuclear magneic resonance. NMR spectra were acquired with a spectrometer
Varian Mercury Plus 400, operating at 400 MHz, using CDCl3, D2O and DMSO-d6 as solvents.
Chemical shifts are expressed in δ (ppm) referred to the undeuterated solvent. The following
abbreviation are used: s = singlet, d = doublet, t = triplet, q = quartet, dd = double doublet, tt = triple
triplet, m = multiplet, br = broad. Substrates, reagents and solvents were acquired from common
commercial sources and used as received. If dry solvents were used, these were dried according
to standard procedures.
Page 81
80
8.2) Synthesis of Nε-benzylidene-L-lysine (70)
To a solution of L-lysine (1.46 g, 10 mmol) in 2N lithium hydroxide (4 ml) at 0°C was added
benzaldehyde (0.84 ml, 12 mmol, 1.2 eq.). The reaction flask was stirred at the same temperature
until the benzaldehyde had dissolved and a white solid precipitated. After standing in the
refrigerator for several hours, the solid was filtered, washed with cold ethanol and dried over CaCl2
under vacuo for one day to give Nε-benzylidene-L-lysine as a white solid in 80% yield. It was used
in the next step without further purification. The spectroscopic data of the product obtained are in
accordance to the literature.[172]
Nε-benzylidene-L-lysine
White solid (1,868 g, 80%).
m.p. = 187-189 °C
1H NMR: (400 MHz, D2O) δH = 1.12-1.29 (m, 2H), 1.43-1.60 (m, 4H), 2.83 (t, 2H), 3.25 (m, 1H),
3.47 (m, 1H), 7.30-7.62 (m, 5H), 7.82 (s, 1H).
Page 82
81
8.3) Synthesis of Nα-carbobenzyloxy-L-lysine (71)
Nε-benzylidene-L-lysine (1.868 g, 8 mmol) was dissolved in a mixture of 1N sodium hydroxide
solution (7.8 ml) and ethanol (7.8 ml) below -5 °C. A cooled mixture of 1N sodium hydroxyde
solution and ethanol (1:1, 31 ml) and benzyl chloroformate (1.52 ml, 10,4 mmol, 1.3 eq.) were
added in two portions over 5 minutes at -10°C with vigorous stirring. The mixture was stirred at the
same temperature for 10 min and then at room temperature for further 30 min.
Concentrated hydrochloric acid (2.34 ml) was added and the resulting mixture was heated at 50 °C
for 30 min. The mixture was then extracted with diethyl ether (3x20 ml) and the aqueous layer was
adjusted to pH 6.2 with a NaOH/KH2PO4 buffer. The resulting mixture was concentrated in vacuo
to a volume approximately half of the starting volume. After standing in the refrigerator for one day,
some white crystals precipitated. The reimaining solvent was removed in vacuo and the solid was
dried over CaCl2 in vacuo for one day to give Nα-carbobenzyloxy-L-lysine in 78% as a white solid,
which was used in next step without purification. The spectroscopic data of the product obtained
are in accordance to the literature.[172]
Nα-carbobenzyloxy-L-lysine
White solid (1.743 g, 78%)
1H NMR: (400 MHz, D2O) δH = 1.22-1.41 (m, 2H), 1.49-1.79 (m, 4H), 2.9 (t, 2H), 3.61 (m, 1H), 5.0
(m, 2H), 7.25-7.34 (m, 5H).
13C NMR : (100MHz, D2O) δC = 21.8 (γ-CH2), 26.0 (δ-CH2), 30.9 (β-CH2), 39.0 (ε-CH2), 54.1 (α-
CH2), 66.4 (CH2-Ar), 127.0 (Ar), 127.8 (Ar), 128.4 (Ar), 135.8 (Ar), 157.6 (NCO-Cbz), 178 (CO).
Page 83
82
8.4) Synthesis of Nα-carbobenzyloxy-L-lysine methyl ester (68)
SOCl2 (0.583 ml, 8 mmol, 4eq.) was added dropwise to MeOH (6 ml) at 0 °C and stirred 5 min. Nα-
Cbz-L-lysine (0.562 g, 2 mmol) was added and the reaction stirred 15 min at 0 °C and at reflux
overnight. MeOH was removed under reduced pressure, and the residue dissolved in sat. NaHCO3
(50 ml) and extracted with CH2Cl2 (3x40 ml). The combined organic layers were dried (Na2SO4)
and the solvent was removed in vacuo to afford a yellowish oil, which was used in next step without
purification. The spectroscopic data of the product obtained are in accordance to the literature.[173]
Nα-carbobenzyloxy-L-lysine methyl ester
Yellow solid (0.3 g, 51%)
1H NMR: (400 MHz, CDCl3) δH = 1.45-1.86 (m, 6H), 2,87 (t, 2H), 3.68 (s, 3H), 4.31 (m, 1H), 5.1 (s,
2H), 7.38-7.47 (m, 5H).
13C NMR : (100MHz, CDCl3) δC = 22.3 (γ-CH2), 26.7 (δ-CH2), 30.8 (β-CH2), 40.4 (ε-CH2), 51.9
(OCH3) 54.7 (α-CH2), (CH2-Ar), 127.1 (Ar), 128.1 (Ar) 128.5 (Ar), 157.4 (NCO-Cbz), 174.5 (CO).
Page 84
83
8.5) Synthesis of N-carbobenzyloxy-L-asparagine (72)
To a mixture of L-asparagine (1.32 g, 10 mmol) and MgO (0.86 g, 21 mmol, 2.1 eq.) in 10 ml of
H2O was added in 4 portions at 5 °C benzyl chloroformate (1.713 ml, 12 mmol, 1.2 eq.). After stirring
15 min at the same temperature, the thick reaction mixture was stirred at r.t. for 3 hours. It was then
acidified with 2N HCl to pH 1-2, filtered and the solid was washed with H2O and dried over CaCl2
giving 2.1 g of product. The entire material was recrystallized from 40 ml of MeOH to yield 1.3 g of
N-carbobenzyloxy-L-asparagine. Recrystallization in the same manner of the residue obtained from
the concentration of mother liquors to dryness gave other 0.65 g of the product for a total yield of
73%. The spectroscopic data of the product obtained are in accordance to the literature.[174]
N-carbobenzyloxy-L-asparagine
White solid (1.95 g, 73%)
m.p. = 164-165 °C
1H NMR: (400 MHz, DMSO) δH = 2.41-2.58 (m, 2H), 4.41-4.49 (m, 1H), 5.14 (s, 2H), 7.11-7.41 (m,
6H), 12.23 (br, 1H).
Page 85
84
8.6) Synthesis of N-carbobenzyloxy-L-asparagine methyl ester (73)
To a solution of N-carbobenzoxy-L-asparagine (1.95 g, 7.3 mmol) in dry DMF (18ml) was added
NaHCO3 (1.23 g, 14.6 mmol, 2 eq.) and iodomethane (1.135 ml, 18.25 mmol, 2.5 eq.). The white
suspension was stirred 24h at r.t. After the reaction time had elapsed, the obtained mixture was
diluted with water (1x40 ml) and was extracted with EtOAc (3x30 ml). The combined organic layers
were washed with H2O (1x30 ml) and brine (1x30 ml). After drying (Na2SO4) the solvent was
evaporated under reduced pressure giving a white solid in 60% yield that was used in next step
without further purification. The spectroscopic data of the product obtained are in accordance to
the literature.[175]
N-carbobenzyloxy-L-asparagine methyl ester
White solid (1.23 g, 60%)
m.p. = 152-153 °C
1H NMR: (400 MHz, DMSO) δH = 2.41-2.59 (m, 2H), 3.61 (s, 3H), 4.38-4.45 (m, 1H), 5.02 (s, 2H)
6.91 (br s, 1H), 7.25-7.38 (m, 6H), 7.61 (d, 1H).
13C NMR : (100MHz, CDCl3) δC = 36.94 (CHCH2CO), 50.68 (NHCHCO), 52.10 (OCH3), 65.6
(PhCH2O), 127.83 (Ar), 127.95 (Ar), 128.48, (Ar), 137.01 (Ar), 155.9 (OCONH), 170.89, 172.31
(CONH2, COOCH3).
Page 86
85
8.7) Synthesis of N-carbobenzyloxy-L-aspartic acid α-methyl ester (74)
To a hot solution of N-carbobenzoxy-L-asparagine methyl ester (1.23 g, 4.3 mmol) in MeCN (13
ml) was added tert-butyl nitrite (1.051 ml, 8.6 mmol, 2 eq.) in one portion. The yellow solution was
heated at reflux for several hours. After conversion of the starting material to its corresponding acid,
checked by TLC; the solvent was removed under reduced pressure. The dark yellow oil was taken
up with NaHCO3 (5%, 30 ml) and washed with EtOAc (2x30 ml). After acidification with conc. HCl
to pH 1-2, the product was extracted into CH2Cl2 (3x40 ml) and dried (Na2SO4). The solvent was
removed in vacuo to obtain N-carbobenzyloxy-L-aspartic acid α-methyl ester in 83% yield. The
spectroscopic data of the product obtained are in accordance to the literature.[175]
N-carbobenzyloxy-L-aspartic acid α-methyl ester
Yellow oil (1.02 g, 83%)
1H NMR: (400 MHz, DMSO) δH = 2.58 (dd, J = 16.45, 7.09, 1H), 2.71 (dd, J = 16.67, 5.55, 1H), 3.61
(s, 3H), 4.38-4.45 (m, 1H), 5.01 (s, 2H), 7.28-7.4 (m, 5H), 7.73 (d, 1H), 12.49 (br s, 1H).
Page 87
86
8.8) Synthesis of N-carbobenzyloxy-L-homoserine α-methyl ester (75)
N-carbobenzyloxy-L-aspartic acid α-methyl ester (0.2 g, 0.7 mmol) was dissolved in dry THF (1 ml)
and cooled to -5°C with an ice-salt bath. Borane in THF (1 M, 1.422 ml, 2 eq.) was added dropwise
over a period of 30 minutes. The solution was allowed to warm up to room temperature and stirring
was continued overnight. After checking the reaction TLC, excess borane was quenched with citric
acid (10% w/v, 40 ml) and the mixture was extracted with Et2O (4x50 ml). The combined organic
layers were washed with brine, dried (Na2SO4) and the solvent was removed in vacuo to afford a
yellow oily crude product, which was subject to column chromatography eluting with Exane:EtOAc
60:40 to obtain pure 75 as a yellow oil in 48% yield. The spectroscopic data of the product obtained
are in accordance to the literature.[175]
N-carbobenzyloxy-L-homoserine α-methyl ester
Yellow oil (91 mg, 48%)
1H NMR: (400 MHz, DMSO) δH = 1.65-1.7 (m, 1H), 1.79-1.85 (m, 1H), 3.36-3.46 (m, 2H), 3.61 (s,
3H), 4.18-4.21 (m, 1H), 4.54 (t, 1H), 5.02 (s, 2H), 7.28-7.39 (m, 5H), 7.64 (d, 1H).
13C NMR : (100MHz, DMSO) δC = 33.9 (CHCH2), 51.1 (NHCHCO), 51.9 (OCH3), 57.1 (CH2OH),
65.6 (PhCH2O), 127.85 (Ar), 127.93 (Ar), 128.48, (Ar), 137.01 (Ar), 156.2 (OCONH), 173.35
(COOCH3).
Page 88
87
9) CONCLUSIONS AND FUTURE WORK
A new strategy for the total synthesis of asymmetrically selectively difunctionalized β-
cyclodextrins was designed entirely by our research group. The first part, concerning the
functionalization of amino acids in order to have compatible substrates for the formation of the
dimer and the bonding with the CD, has been carried on during my period in laboratory.
Next short-term goals will cover surely the completion of the functionalization of amino acids with
the last oxidation step and subsequently the formation of the dimer. Afterwards, the attack on the
cyclodextrin will be tried with conditions used for the Steglich esterification and, if it’s going to work,
hydrolysis of the imine present in the dimer will be performed and the desired amino acids
functionalization will be restored.
Also, one can concretely think to improve some steps of the synthesis to optimize the reaction
conditions and obtain more satisfactory yields.
Clearly, the idea of a novel synthesis is very fascinating from a scientific perspective and the idea
of being able to achieve a product useful to the pharmaceutical industry is extremely challenging.
Unfortunately, we can not be sure of the actual effectiveness of the strategy before testing it
experimentally; however, alternative ways to reach the desired product have already been
evaluated and ready to be carried out if the main strategy fails.
Page 89
88
REFERENCES
[1] J. M. Lehn, Science. 1993, 260, 1762.
[2] J. M. Lehn, Supramolecular chemistry. 1995, Wiley-VCH.
[3] F. Vӧgtle, Supramolecular chemistry. An introduction. 1993, Wiley.
[4] D. Philips, J. F. Stoddart. Angew. Chem. Int. Ed. 1996, 35, 1154.
[5] M. Vincentil, E. Pelizzettil, A. Secchi, E. Dalcanale, Pure Appl. Chem. 1995, 67, 1075.
[6] S. Anderson, H. L. Anderson, A. Bashall, M. McPartlin, J. K. M. Sanders, Angew. Chem. Int. Ed.
1995, 34, 1096.
[7] C. J. Pedersen, J. Am. Chem. Soc. 1967, 89, 7017.
[8] C. J. Pedersen, J. Am. Chem. Soc. 1967, 89, 2495.
[9] C. J. Pedersen, Nobel lecture. December 8, 1987.
[10] D. M. Dishong, C. J. Diamond, M. I. Cinoman, G. W. Gokel, J. Am. Chem. Soc. 1983, 105,
586.
[11] R. A. Schultz, B. D. White, D. M. Dishong, K. A. Arnold, G. W. Gokel, J. Am. Chem. Soc. 1985,
107, 6659.
[12] D. Landini, A. Maia, F. Montanari, P. Tundo, J. Am. Chem. Soc. 1979, 101, 2526.
[13] P. K. Thallapally, G. Lloyd, J. L. Atwood, L. J. Barbour, Angew. Chem. Int. Ed. 2005, 44, 3848.
[14] A. Soi, A. Hirsch, New J. Chem. 1998, 22, 1337.
[15] H. M. Bechara, G. K. Derkara, F. V. Manvia, V. K. Nanjwade, Eur. J. Pharm. Sci.
2009, 38, 185.
[16] J. M. J. Frechet, Science 1994, 263, 1710.
[17] (a) D. Soto-Castro, J. A. Cruz-Morales, P. Guadarrama, Bioorg. Chem. 2012, 41, 13. (b) F.
Diederich, B. Felber, Proc. Natl. Acad. Sci. U S A. 2002, 99, 4778.
[18] R. Duncan, L. Izzo, Adv. Drug. Deliv. Rev. 2005, 57, 2215.
[19] M. Najlah, S. Freeman, D. Attwood, A. D’Emanuele, Int. J. Pharm. 2007, 336, 183.
[20] A. Biwer, G. Antranikian, E. Heinzle, Appl. Microbiol. Biotechnol. 2002, 59, 609.
[21] K. Freudeberg, G. Blomquist, L. Ewald, K. Soff, Ber. Dtsch. Chem. Ges. 1936, 9, 1258.
[22] K. Freudeberg, M. Mayers-Delins, Chem. Ber. 1938, 71, 1596.
[23] A. Villers, Compt. Rend. 1891,112, 536.
[24] F. Z. Scardinger, Unters. Nahr, U. Genessm. 1903, 6, 865.
[25] J. Szejtly, Chem. Rev. 1998, 98, 1743.
[26] E. Sabadini, T. Cosgrovea, F. do Carmo Egídio, Carbohydr. Res. 2006, 341, 270.
[27] K. B. Lipkowitz, J. Org. Chem. 1991, 56, 6357.
[28] P. Lo Meo, F. D’Anna, S. Riela, M. Gruttadauria, R. Noto, Tetrahedron Lett. 2006, 47, 9099.
[29] P. Lo Meo, F. D’Anna, S. Riela, M. Gruttadauria, R. Noto, Tetrahedron. 2007, 63, 9163.
[30] D. Cremer, J. A. Pople, J. Am. Chem. Soc. 1975, 97, 1354.
[31] K. B. Lipkowitz, Chem. Rev. 1998, 98, 1829.
[32] W. Saenger, J. Jacob, K. Gessler, T. Steiner, D. Hoffman, Chem. Rev. 1998, 98, 1787.
[33] H. J. Schneider, F. Hacket, V. Rüdiger, H. Ikeda, Chem. Rev. 1998, 98, 1755.
Page 90
89
[34] (a) H. Dodziuk, J. Mol. Struct. 2002, 614, 33. (b) B. Mayer, G. Marconi, C. Klein, G. Kohler, P.
Wolschann, J. Inclusion Phenom. Mol. Recognit. Chem. 1997, 29, 79.
[35] I. Tabushi, Y. Kiyosuke, T. Sugimoto, K. Yamamura, J. Am. Chem. Soc. 1978, 100, 916.
[36] T. Nakagawa, S. Immel, F. W. Lichtenthaler, H. J. Lindner, Carbohydr. Res. 2000, 324, 141.
[37] (a) F. D’Anna, P. Lo Meo, S. Riela, M. Gruttadauria, R. Noto, Tetrahedron. 2001, 57, 6823. (b)
F. D’Anna, P. Lo Meo, S. Riela, M. Gruttadauria, R. Noto, Tetrahedron. 2002, 58, 6039. (c) I.
Tabushi, T. Mitzutani, Tetrahedron. 1987, 43, 1439. (d) L. Liu, Q. X. Guo, J. Inclusion Phenom.
Macrocyclic Chem. 2002, 42, 1. (e) M. Wedig, S. Laug, T. Christians, M. Thunhorst, U. Holzgrabe,
J. Pharm. Biomed. Anal. 2002, 27, 531. (f) M. V. Rekharsky, Y. Inoue, J. Am. Chem. Soc. 2000,
122, 4418. (g) M. V. Rekharsky, Y. Inoue, J. Am. Chem. Soc. 2002, 124, 813. (h) K. Kano, H.
Hasegawa, J. Am. Chem. Soc. 2001, 123, 10616. (i) M. V. Rekharsky, M. P. Mayhew, R. N.
Goldberg, P. D. Ross, Y. Yamashoji, Y. Inoue, J. Phys. Chem. B. 1997, 101, 87. (j) P. Lo Meo, F.
D’Anna, S. Riela, M. Gruttadauria, R. Noto, Org. Biomol. Chem. 2003, 1, 1584. (k) P. Lo Meo, F.
D’Anna, S. Riela, M. Gruttadauria, R. Noto, Tetrahedron. 2004, 60, 9099.
[38] (a) K. Uekama, F. Hirayama, T. Irie, Chem. Rev. 1998, 98, 2045. (b) M. Singh, R. Sharma,
Biotechnol. Adv. 2002, 20, 341. (c) A. R. Hedges, Chem. Rev. 1998, 98, 2035. (d) K. B. Lipkowitz,
R. Corner, M. A. Peterson, A. Morreale; J. Shakelford, J. Org. Chem. 1998, 63, 732. (e) S. Li, W.
C. Purdy, Chem. Rev. 1992, 92, 1457. (f) A. M. Stalcup, E. Schneriderman, J. Chromatography B.
2000, 745, 83. (g) K. Takahashi, Chem. Rev. 1998, 98, 2013.
[39] R. Breslow, S. D. Dong, Chem. Rev. 1997, 98, 199.
[40] (a) J. Liu, W. Ong, E. Romàn, M. J. Lynn, A. E. Kaifer, Langmuir. 2000, 16, 3000. (b) J. Liu, J.
Alvarez, W. Ong, E. Romàn, A. E. Kaifer, Langmuir. 2001, 17, 6762. (c) J. Liu, W. Ong, A. E. Kaifer,
Langmuir. 2002, 18, 5981.
[41] T. Loftsson, J. Incl. Phenom. Macrocycl. Chem. 2002, 44, 63.
[42] M. E. Davis, Nat. Rev. 2004, 3, 1023.
[43] S. M. Manali, Int. J. Pharm. Res. Dev. 2011, 3, 35.
[44] http://www.drugdeliverytech.com/cgi-bin/articles.cgi?idArticle=262.
[45] K. A. Khan, Drug Dev. Ind. Pharm. 1981, 7, 421.
[46] M. K. Gupta, Y. C. Tseng, D. Goldman, R. H. Bogner, Pharm. Res. 2002, 19, 1663.
[47] T. Loftsson, D. Duchene, Int. J. Pharm. 2007, 329, 1.
[48] S. Li, K. Connors, J. Pharm. Sci. 1983, 72, 1333.
[49] S. T. Birhade, V. H. Bankar, P. D. Gaikwad, S. P. Pawar, Int. J. Pharm. Sci. and Drug. Res.
2010, 2, 199.
[50] T. Loftsson, M. E. Brewster, J. Pharm. Sci. 1996, 85, 1017.
[51] G. Piel, T. Van Hees, B. Evard, L. A. Delattre, Proceedings of the Ninth International
Symposium on Cyclodextrins. May 31-June 3, 1998, Santiago de Compostela, Spain, Kluwer
Academic Publisher, Dordrecht, p. 223.
[52] K. Okimoto, R. A. Rajewski, K. Uekama, J. A. Jona, V. J. Stella, Pharm. Res. 1996, 13, 256.
[53] T. Tokomura, Y. Tsushima, M. Kayano, Y. Machida, T. Nagai, J. Pharm. Sci. 1985, 74, 496.
Page 91
90
[54] T. Tokomura, M. Nanba, Y. Tsushima, K. Tatsuishi, M. Kayano, Y. Machida, T. Nagai, J.
Pharm. Sci. 1986, 75, 391.
[55] H. W. Frijlink, E. J. F. Franssen, A. C. Eissens, R. Oosting, C. F. Lerk, D. K. F. Meijer, Pharm.
Res. 1991, 8, 380.
[56] T. Backensfeld, B. W. Muller, K. Kolter, Int. J. Pharm. 1991, 74, 85.
[57] F. Y. Liu, D. O. Kildsig, A. K. Mitra, Pharm. Res. 1992, 9, 1671.
[58] Y. Inoue, T. Hakushi, Y. Liu, L. H. Tong, B. J. Shen, D. S. Jin, J. Am. Chem. Soc. 1993, 115,
475.
[59] H. Arima, K. Yunomae, K. Mirake, T. Irie, F. Hirayama, K. Uekama, J. Pharm. Sci. 2001, 90,
690.
[60] H. Arima, T. Miyaji, T. Irie, F. Hirayama, K. Uekama, Eur. J. Pharm. Sci. 2001, 6, 53.
[61] M. J. Arias, J. R. Mayano, P. Munoz, J. M. Gines, A. Justo, F. Giordano, Drug. Dev. Ind. Pharm.
2000, 26, 253.
[62] K. Asai, M. Morishita, H. Katsuta, S. Hosoda, K. Shinomiya, M. Noro, Int. J. Pharm. 2002, 246,
25.
[63] L. H. Emara, R. M. Badr, A. A. Elbary, Drug. Dev. Ind. Pharm. 2002, 28, 795.
[64] B. Evrard, P. Chiap, P. DeTullio, F. Ghalmi, G. Piel, T. Van Hees, J. Control Release. 2002,
85, 45.
[65] A. R. Hedges, Chem. Rev. 2005, 98, 2035.
[66] F. Hirayama, K. Uekama, Adv. Drug. Deliv. Rev. 2007, 36, 125.
[67] A. C: Jain, M. C: Adeyeye. Int. J. Pharm. 2005, 212, 177.
[68] R. L. VanEtten, J. F. Sebastian, G. A. Clowes, M. L. Bender, J. Am. Chem. Soc. 1967, 89,
3242.
[69] R. J. bergeron, P. S. Burton, J. Am. Chem. Soc. 1982, 104, 3664.
[70] K. S. Inozuka, M. Kojima, K. Hattori, F. Toda, Bull. Chem. Soc. Jpn. 1980, 53, 3221.
[71] R. Breslow, M. F. Czarniecki, J. Emert, H. Hamaguchi, J. Am. Chem. Soc. 1980, 102, 762.
[72] M. Komiyama, S. Inoue, Chem. Lett. 1979, 1101.
[73] M. Komiyama, M. L. Bender, J. Am. Chem. Soc. 1977, 99, 8021.
[74] T. S. Straub, M. L. Bender, J. Am. Chem. Soc. 1972, 94, 8875.
[75] D. W: Griffiths, M. L. Bender, J. Am. Chem. Soc. 1973, 95, 1679.
[76] D. Rideout, R. Breslow, J. Am. Chem. Soc. 1980, 102, 7812.
[77] H. J. Schneider, N. K. Sangwan, Angew. Chem. Int. Ed. Engl. 1987, 26, 896.
[78] H. J. Schneider, N. K. Sangwan, J. Chem. Soc., Chem. Commun. 1986, 1787.
[79] T. Hudlicky, G. Butora, S. P. Fearnley, A. G. Gum, J. Chem. Soc., Perkin Trans. 1 1995, 1,
2393.
[80] S. W. Chung, J. Y. Wang, J. Chem. Soc., Chem. Commun. 1995, 971.
[81] M. Komiyama, M. L. Bender, J. Am. Chem. Soc. 1978, 100, 4576.
[82] R. Breslow, H. Kohn, B. Siegel, Tetrahedron Lett. 1976, 1645.
[83] K. Flohr, R. M. Paton, E. T. Kaiser, J. Am. Chem. Soc. 1975, 97, 1209.
[84] B. Siegel, A. Pinter, R. Breslow, J. Am. Chem. Soc. 1977, 99, 2309.
Page 92
91
[85] A. W. Coleman, P. Zhang, H. Parrot-Lopez, C. C. Ling, M. Miocque, L. Mascrier, Tetrahedron
Lett. 1991, 32, 3997.
[86] S. Hanessian, A. Benalil, C. J. Laferriere, J. Org. Chem. 1995, 60, 4786.
[87] A. Ueno, R. Breslow, Tetrahedron Lett. 1982, 23, 3451.
[88] M. Maletic, H. Wennemerds, D. Q. McDonald, R. Breslow, W. C. Still, Angew. Chem. Int. Ed.
Engl. 1996, 35, 1490.
[89] Z. Yang, R. Breslow, Tetrahedron Lett. 1997, 38, 6171.
[90] M. A. Mortellaro, W. K. Hartmann, D. G. Nocera, Angew. Chem. Int. Ed. Engl. 1996, 35, 1945.
[91] B. K. Hubbard, L. A. Beilstein, C. E. Heath, C. J. Abelt, J. Chem. Soc. Perkin Trans. 1 1996, 2,
1005.
[92] F. Cramer, G. Mackensen, Angew. Chem. Int. Ed. Engl. 1966, 5, 601.
[93] K. R. Rao, T. N. Srinivasan, N. Bhanumathi, P. B. Sattur, J. Chem. Soc., Chem. Commun.
1990, 10.
[94] M. L. Bender, J. Inclusion Phenom. 1984, 2, 433.
[95] V. T. D’Souza, K. Hanabusa, T. O’Leary, R. C. Gadwood, M. L. Bender, Biochem. Biophys.
Res. Commun. 1985, 129, 727.
[96] M. Komiyama, E. J. Breaux, M. L. Bender, Bioorg. Chem. 1977, 6, 127.
[97] R. Breslow, S. Chung, Tetrahedron Lett. 1989, 30, 4353.
[98] S. Zimmerman, Tetrahedron Lett. 1989, 30, 4357.
[99] R. Breslow, D. Chipman, J. Am. Chem. Soc. 1965, 87, 4195.
[100] R. Breslow, L. E. Overman, J. Am. Chem. Soc. 1970, 92, 1075.
[101] R. Breslow, B. Zhang, J. Am. Chem. Soc. 1992, 114, 5882.
[102] B. Zhang, R. Breslow, J. Am. Chem. Soc. 1997, 119, 1676.
[103] J. Yan, R. Breslow, Tetrahedron Lett. 2000, 41, 2059.
[104] S. D. Dong, R. Breslow, Tetrahedron Lett. 1998, 39, 9343.
[105] R. Breslow, M. Hammond, M. Lauer, J. Am. Chem. Soc. 1980, 102, 421.
[106] R. Breslow, J. W. Canary, M. Varney, S. T. Waddell, D. Yang, J. Am. Chem. Soc. 1990, 112,
5212.
[107] C. S. Zimmerman, A. W. Czarnik, R. Breslow, J. Am. Chem. Soc. 1983, 105, 1694.
[108] I. Tabushi, Y. Kuroda, M. Yamada, H. Higashimura, R. Breslow, J. Am. Chem. Soc. 1985,
107, 5545.
[109] R. Breslow, J. Chmielewsky, D. Foley, B. Johnson, N. Kumabe, M. Varney, R. Mehra,
Tetrahedron. 1988, 44, 5515.
[110] W. Weiner, J. Winkler, S. C. Zimmerman, A. W. Czarnik, R. Breslow, J. Am. Chem. Soc.
1985, 107, 4093.
[111] D. Hilvert, R. Breslow, Bioorg. Chem. 1984, 12, 206.
[112] R. Breslow, E. Kool, Tetrahedron Lett. 1988, 29, 1635.
[113] I. Tabushi, M. Kodera, J. Am. Chem. Soc. 1987, 109, 4734.
[114] H. Ye, W. Tong, V. T. D’Souza, J. Am. Chem. Soc. 1992, 114, 5470.
[115] H. Ye, W. Tong, V. T. D’Souza, J. Chem. Soc., Perkin Trans. 1 1994, 2431.
Page 93
92
[116] W. Tong, H. Ye, D. Rong, V. T. D’Souza, J. Comput. Chem. 1992, 13, 614.
[117] M. Kojima, F. Toda, K. Hattori, J. Chem. Soc., Perkin Trans. 1 1981, 1647.
[118] R. Breslow, P. J. Duggan, J. P. Light, J. Am. Chem. Soc. 1992, 114, 3982.
[119] M. Rezac, R. Breslow, Tetrahedron Lett. 1997, 38, 5763.
[120] E. Anslyn, R. Breslow, J. Am. Chem. Soc. 1989, 111, 5972.
[121] R. Breslow, Acc. Chem. Res. 1991, 24, 317.
[122] I. Tabushi, Y. Kuroda, A. Mochizuki, J. Am. Chem. Soc. 1980, 102, 1152.
[123] E. Anslyn, R. Breslow, J. Am. Chem. Soc. 1989, 111, 8931.
[124] R. Breslow, C. Schmuck, J. Am. Chem. Soc. 1996, 118, 6601.
[125] R. Breslow, A. Graff, J. Am. Chem. Soc. 1993, 115, 10998.
[126] R. Breslow, J. Mol.Catal. 1994, 91, 161.
[127] J. M. Desper, R. Breslow, J. Am. Chem. Soc. 1994, 116, 12081.
[128] R. Breslow, J. Desper, Y. Huang, Tetrahedron Lett. 1996, 37, 2541.
[129] D. L. Zechel, S. G. Withers, Acc. Chem. Res. 2000, 33, 11.
[130] C. Rousseau, F. Ortega-Caballero, L. U. Nordstrøm, B. Christensen, T. E. Petersen, M. Bols,
Chem. Eur. J. 2005, 11, 5094.
[131] C, Rousseau, N. Nielsen, M. Bols, Tetrahedron Lett. 2004, 45, 8709.
[132] J. Bjerre, T. H. Fenger, L. G. Marinescu, M. Bols, Eur. J. Org. Chem. 2007, 704.
[133] F. Ortega-Caballero, J. Bjerre, L. S. Laustsen, M. Bols, J. Org. Chem. 2005, 70, 7217.
[134] F. Ortega-Caballero, C. Rousseau, B. Christensen, T. E. Petersen, M. Bols, J. Am. Chem.
Soc. 2005, 127, 3238.
[135] F. Ortega-Caballero, M. Bols, Can. J. Chem. 2006, 84, 650.
[136] J. Bjerre, M. Bols, Eur. J. Org. Chem. 2010, 3487.
[137] K. Takeo, K. Uemura, H. J. Mitoh, J. Carbohydr. Chem. 1988, 7, 293.
[138] F. Cramer, G. Mackensen, K. Kensse, Chem. Ber. 1969, 102, 494.
[139] D. Rong, V. T. D’Souza, Tetrahedron Lett. 1990, 31, 4275.
[140] K. Fujita, S. Nagamura, T. Imoto, Tetrahedron Lett. 1984, 25, 5673.
[141] K. Takahashi, K. Hattori, F. Toda, Tetrahedron Lett. 1984, 25, 3331.
[142] M. Tanaka, Y. Kawaguchi, T. Niinae, T. Shozo, J. Chromatogr. 1984, 314, 193.
[143] S. Cottaz, H. Driguez, Synthesis. 1989, 755.
[144] R. C. Petter, J. S. Salek, C. T. Sikorsky, G. Kumaravel, F. T. Lin, J. Am. Chem. Soc. 1990,
112, 3860.
[145] F. Sallas, P. Leroy, A. Marsura, A. Nicolas, Tetrahedron Lett. 1994, 35, 6079.
[146] R.P. Bonomo, V. Cucinotta, F. D’Alessandro, G. Impellizzeri, G. Maccarrone, G. Vecchio, E.
Rizzarelli, Inorg. Chem. 1991, 30, 2708.
[147] F. Bellia, A. M. Amorini, D. La Mendola, G. Vecchio, B. Tavazzi, B. Giardina, V. Di Pietro, G.
Lazzarino, E. Rizzarelli, Eur. J. Med. Chem. 2008, 43, 373.
[148] A. R. Khan, P. Forgo, K. J. Stine, V. T. D’Souza, Chem. Rev. 1998, 98, 1977.
[149] J. M. Garciá Fernández, C. O. Mellet, S. Maciejewski, J. Defaye, J. Chem. Soc., Chem.
Commun. 1996, 2741.
Page 94
93
[150] H. Ikeda, M. Nakamura, N. Ise, F: Toda, A. Ueno, J. Org. Chem. 1997, 62, 1411.
[151] J. B. Huff, C. Bieniarz, J. Org. Chem. 1994, 59, 7511.
[152] J. Yoonm S. Hong, K. A. Martin, A. W. Czarnik, J. Org. Chem. 1995, 60, 2792.
[153] M. J. Cornwell, J. B. Huff, C. Bieniarz, Tetrahedron Lett. 1995, 36, 8371.
[154] G. Yi, J. S. Bradshaw, B. E. Rossiter, A. Malik, W. Li, M. L. Lee, J. Org. Chem. 1993, 58,
4844.
[155] S. Tian, V. T. D’Souza, Tetrahedron Lett. 1994, 35, 9339.
[156] S. Tian, Ph.D. Dissertation, University of Missouri – St. Louis, 1997, 26.
[157] P. R. Ashton, P. Ellwood, I. Staton, J. F. Stoddart, J. Org. Chem. 1991, 56, 7274.
[158] I. Tabushi, Y. Kuroda, K. Yokota, L. C. Yuan, J. Am. Chem. Soc. 1981, 103, 711.
[159] I. Tabushi, K. Yamamura, T. Nabeshima, J. Am. Chem. Soc. 1984, 106, 5267.
[160] G. Yi, J. S. Bradshaw, B. E. Rossiter, S. L. Rees, P. Petersson, K. E. Markides, M. L Lee, J.
Org. Chem. 1993, 58, 2561.
[161] R. Breslow, J. W. Canary, M. Varney, S. T. Waddell, D. Yang, J. Am. Chem. Soc. 1990, 112,
5212.
[162] S. J. Hwang, N. C. Bellocq, M. E. Davis, Bioconjugate Chem. 2001, 12, 280.
[163] M. Sollogoub, S. K. Das, J. M. Mallet, P. Sinaӱ, C. R. Acad. Sci. Ser. IIc. 1999, 441.
[164] T. Lecourt, A. Hérault, A. J. Pearce, M. Sollogoub, P. Sinaӱ, Chem. Eur. J. 2004, 10, 2960.
[165] D. Armspach, L. Poorters, D. Matt, B. Bernmerad, F. Balegroune, L. Toupet, Org. Biomol.
Chem. 2005, 3, 2588.
[166] E. V. van Dienst, B. H. M. Snellink, I. von Piekartz, M. H. B. G. Gansey, F. Venema, M. C.
Feiters, R. J. M. Nolte, J. F. J. Engbersen, D. N. Reinhoudt, J. Org. Chem. 1995, 60, 6537.
[167] N. Sakairi, H. Kuzuhara, T. Okamoto, M. Yakima, Bioorg. Med. Chem. 1996, 4, 2187.
[168] M. A. Mortellaro, S. Hong, D. T. Winn, A. W. Czarnik, Bioorg. Med. Chem. Lett. 1994, 4, 2035.
[169] H. Shargi, M. H. Sarvari, Tetrahedron. 2003, 59, 3627.
[170] V. Mitsakos, S. R. A. Devenish, P. A. O’Donnell, J. A. Gerrard, C. A. Hutton, Bioorg. Med.
Chem. 2011, 19, 1535.
[171] N. K. Dunlap, J. Drake, A. Ward, T. L. J. Salyard, L. Martin, J. Org. Chem. 2008, 73, 2928.
[172] B. Hudger, U. Dyer, G. Hutton, M. Woods. Tetrahedron Lett. 1996, 37, 6399.
[173] J. Hu, M. J. Miller, J. Org. Chem. 1994, 59, 4858.
[174] M. Manning. J. Med. Chem. 1971, 14, 1143.
[175] P. A. Ottersbach. Org. Lett. 2013, 15, 448-451.