Cyclodextrins: A promising drug delivery vehicle for bisphosphonate

Article abstract of DOI:10.1016/j.carbpol.2016.09.030

Bisphosphonates are well established pharmaceutical drugs with wide applications in medicine. Nevertheless, the side chain and the nature of phosphorous groups could induce a poor aqueous solubility and act on their bioavailability. At the same time, cyclodextrins are cage molecules that facilitate transport of hydrophobic molecules to enhance the intestinal drug absorption of these molecules by forming inclusion complexes. Here we demonstrate that cyclodextrins could be used as a bisphosphonate carrier. The formation of cyclodextrins-bisphosphonate complexes was characterized by 1D and 2D NMR spectroscopy, Isothermal Titration Calorimetry and UV–vis spectroscopy. The results revealed that only the side chain of bisphosphonate was involved in the inclusion phenomenon and its length was a crucial parameter in the control of affinity. Findings from this study suggest that cyclodextrin will be a useful carrier for bisphosphonates.

Full text of DOI:10.1016/j.carbpol.2016.09.030

Accepted Manuscript
Title: Cyclodextrins: A promising drug delivery vehicle for
bisphosphonate
Author: Maelle Monteil Marc Lecouvey David Landy Steven
Ruellan Isabelle Mallard
PII:
S0144-8617(16)31084-0
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.carbpol.2016.09.030
CARP 11557
To appear in:
Received date:
Revised date:
Accepted date:
20-5-2016
7-9-2016
11-9-2016
Please cite this article as: Monteil, Maelle., Lecouvey, Marc., Landy,
David., Ruellan, Steven., Mallard, Isabelle., Cyclodextrins:
&
A
promising drug delivery vehicle for bisphosphonate.Carbohydrate Polymers
http://dx.doi.org/10.1016/j.carbpol.2016.09.030
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Cyclodextrins: A promising drug delivery vehicle for bisphosphonate.
Maelle Monteil1, Marc Lecouvey1, David Landy2, Steven Ruellan2 and Isabelle Mallard2 *
1 Université Paris 13, CSPBAT UMR 7244 CNRS, F-93017 Bobigny,FRANCE
2 ULCO, UCEIV EA 4492, F-59140 Dunkerque, FRANCE
*Corresponding author: Isabelle Mallard
E-mail address: isabelle.mallard@univ-littoral.fr (Isabelle Mallard)
Tel.: +33 0328658257; Fax: 33 0328237605.
1

Highlights
 Successfull cyclodextrin/bisphosphonates complex formation
 Cyclodextrins carriers for bisphosphates
 New systems for drug formulation
Abstract
Bisphosphonates are well established pharmaceutical drugs with wide applications in
medicine. Nevertheless, the side chain and the nature of phosphorous groups could induce a
poor aqueous solubility and act on their bioavailability. At the same time, cyclodextrins are
cage molecules that facilitate transport of hydrophobic molecules to enhance the intestinal
drug absorption of these molecules by forming inclusion complexes.
Here we demonstrate that cyclodextrins could be used as a bisphosphonate carrier. The
formation of cyclodextrins- bisphosphonate complexes was characterized by 1D and 2D
NMR spectroscopy, Isothermal Titration Calorimetry and UV-Vis spectroscopy. The results
revealed that only the side chain of bisphosphonate was involved in the inclusion
phenomenon and its length was a crucial parameter in the control of affinity. Findings from
this study suggest that cyclodextrin will be a useful carrier for bisphosphonates.
Keywords: Bisphosphonates; cyclodextrins; host-guest complex; formation constant.
2

1. Introduction
Bisphosphonates (BPs) are stable analogs of pyrophosphates. Their P-C-P backbone instead
of P-O-P bond in pyrophosphate are designed to resist to enzymatic hydrolysis. Due to their
high affinity to Ca2+ ions, they exhibit a powerful binding affinity to bones (Rogers, Crockett,
Coxon, & Mönkkönen, 2011; Russel, 2011) and especially hydroxyapatite (HAP) (Rogers,
2003). Bisphosphonates containing nitrogen have antiresorptive activity and inhibit the
mevalonate pathway (Russel, 2011; Kavanagh et al., 2006; Merrell, Wakchoure, Lehenkari,
Harris, & Selander, 2007) and the non-nitrogen bisphosphonate mimic pyrophosphate (Russel,
2011). So, bisphosphonates are used for the treatment of metabolic bone disease such as
osteoporosis (Eastell, Walsh, Watts, & Siris, 2011; Lin, 1996) and Paget’s disease (Reid &
Hosking, 2011; Russel, 2011). As a consequence of the effect of solid tumors to develop bone
diseases, they are widely used in anticancer treatment particularly for breast cancer (Muller,
Migianu, Lecouvey, Kraemer, & Oudar, 2005), prostate cancer (Lee, Fong, Singer, &
Guenette, 2001) and multiple myeloma (Shipman, Rogers, Apperley, Russel, & Croucher,
1997). Unfortunately, all the oral forms are poorly absorbed and their biodisponibility
oscillates between 1 to 5%. Their intestinal absorption is low and mainly depends on the
calcium content in food due to their high affinity to metal cations (Lin, 1996). Moreover, the
charged nature of bisphosphonates and their strong hydrophilicity limit their diffusion through
the lipophilic membrane and lead to a fast elimination from the blood.
To circumvent these problems, different carriers were investigated such as liposomes (Chebbi,
Migianu-Griffoni, Sainte Catherine, Lecouvey, & Seksek, 2010), micelles (Seow, Xue, &
Yang, 2007) and nanoparticles (Beneyettou et al., 2011). The chosen carrier should be
suitable for time-controlled delivery, reduce side effects during absorption and decrease the
toxicity of the drugs. We focused our attention on cyclodextrins. Cyclodextrins (CD) are α-
1,4-linked cyclic oligosaccharides of D-glucopyranose with a hydrophilic exterior and a
hydrophobic cavity (Szejtli, 1998) that allow the formation of inclusion complexes with
hydrophobic compounds. Cyclodextrins are non-toxic and biocompatible and are widely used
in soil remediation (Cho et al., 2015) and food formulation (Astray, Gonzalez-Barreiro,
Mejuto, Rial-Otero, & Simal Gandara, 2009; Astray, Mejuto, Morales, Rial-Otero, & Simal
Gandara, 2010). Their ability to form inclusion complex was and are intensively investigated
for drug delivery (Ijaz et al., 2015; Strickley, 2004; Uekama, Hirayama, & Arima, 2006) in
biopharmaceuticals. To this end, the use of cyclodextrin in pharmaceutic formulation is an
3

effective method to enhance the solubility, chemical stability and bioavailability of drugs
(Loftsson, Brewster, & Masson, 2004; Stella & Rajewski, 1997).
In the present work, we have hypothesized that cyclodextrins could be successfully employed
as a carrier for bisphosphonate. To the best of our knowledge, only a few reports concerning
the characterization of cyclodextrin/ bisphosphonate complexes (CD/BP) or the formation of
bisphosphonate functionalized cyclodextrin derivatives were published. A bisphosphonate
called alendronate was linked to cyclodextrin by click chemistry (Hein, Liu, Chen, Cullen, &
Wang, 2010; Liu, Lee, Reinhardt, Marky, & Wang, 2007) for drug delivery. In addition, the
formation of cyclodextrin/bisphosphonate complexes was identified by mass spectrometry
(Biernacka, Betlejewska-Kielak, Witowska-Jarosz, Klosinska-Szmurlo, & Mazurek, 2014) or
Raman microscopy (Daubiné et al., 2009).
This research reports the complete characterization of CD/BP complexes besides zoledronate
(Daubiné et al., 2009) and alendronate (Biernacka et al., 2014). The goal of this study was to
examine if cyclodextrins could be used as molecular carriers for bisphosphonates, and to
determine the influence of the side chain and of the P-C-P backbone on the complexation
phenomenon. Bisphosphonates with long alkyl chain were chosen because of their utilization
against parasitic diseases in particular against Chagas disease (Hudock et al., 2006; Szajnman
et al., 2012), a form of sleeping sickness. This disease is widespread in the world and it is
considered by the World Health Organization to be one of the major parasitic disease.
The inclusion complex formation between bisphosphonate and cyclodextrin was investigated
by 1D and 2D NMR spectroscopy. In addition, isothermal titration calorimetry and UV-
Visible spectroscopy allowed us to determine values of formation constant.
2. Materials and methods
2.1. Materials
-CD, -CD, 2-hydroxypropyl--CD (HP--CD) (degree of substitution=DS=5.6) and
randomly methylated--CD (RAMEB) (DS= 12.6) were purchased from Wacker-Chemie
(Lyon, France). Methyl orange, sodium hydroxide and potassium dihydrogenophosphate were
all of analytical grade and purchased from Acros Organics.
4

Pentadecanoyl chloride, dodecanoyl chloride, myristoyl chloride, undecanoic acid, tridecanoic
acid, dimethylphosphite, trimethylphosphite, methanol, diethylether,
tristrimethylsilylphosphite were purchased from Sigma Aldrich. All products were of
analytical grade.
2.2. Synthesis
All chemical structures of bisphosphonate used in this study are displayed in Fig.1. In the
following, the alkyl chain attached to the P-C-P backbone will be designated as the
bisphosphonate side chain (thus excluding the carbon atom directly linked to phosphorus
atoms). The two digits after M indicate the number of methoxy groups on either of the two P
atoms.
General procedure for synthesis of bisphosphonate tetra acid (BP1, BP2, BP3, BP4, BP5,
BP6).
In a 50 mL round-bottom three-neck flask equipped with a thermometer, acid chloride (2.5
mmol) was added dropwise, under argon, at -5°C, to tris trimethyl silyl phosphite (5 mmol).
When addition was completed, reaction mixture was allowed to stand at room temperature for
31
2 h. The evolution of the reaction was monitored by P {1H} NMR (31P NMR with proton
decoupling). Then, volatile fractions were evaporated under reduced pressure (0.1 Torr)
before being hydrolyzed with methanol. After evaporation, crude products were precipitated
in diethylether.
(1-hydroxyundecane-1, 1-diyl)diphosphonic acid BP10
1H NMR (400ꢀMHz, DMSO): δ 1.89 – 1.70 (m, 2H, CH2), 1.59 – 1.46 (m, 2H, CH2), 1.33 –
3
1.11 (m, 14H, CH2), 0.86 (t, 3H, J=ꢀ8.0ꢀHz, CH3). 13C NMR (100ꢀMHz, DMSO): δ 72.82 (t,
1JꢀP-C=ꢀ154ꢀHz, P-C-P), 33.74, 31.76, 30.53, 29.22, 23.41, 22.59, 14.37. 31P NMR (162ꢀMHz,
DMSO): δ 20.38. Yield: 84%
HRMS (ESI): Calcd for C11H25O7P2 [Mꢀ- Hꢀ] : 331.1079.
(1-hydroxydodecane-1, 1-diyl)diphosphonic acid BP11
1H NMR (400ꢀMHz, DMSO): δ 1.91 – 1.73 (m, 2H, CH2), 1.60 – 1.43 (m, 2H, CH2), 1.11 –
3
1.37 (m, 16H, CH2), 0.85 (t, 3H, J=ꢀ8.0ꢀHz, CH3). 13C NMR (100ꢀMHz, DMSO): δ 72.93 (t,
5

1JP-C=ꢀ144ꢀHz, P-C-P), 31.79, 30.49, 29.71 – 29.36, 29.22, 23.46 (t, 3JP-C=ꢀ6ꢀHz), 22.58, 14.43.
31P NMR (162ꢀMHz, DMSO): δ 20.33. Yield 82%
HRMS (ESI): Calcd for C12H27O7P2 [Mꢀ-ꢀH]: 345.1231.
(1-hydroxytridecane-1, 1-diyl)diphosphonic acid BP12
1H NMR (400ꢀMHz, DMSO): δ 1.90 – 1.70 (m, 2H, CH2), 1.59 – 1.42 (m, 2H, CH2), 1.37 –
3
1.96 (m, 18H, CH2), 0.85 (t, 3H, J=ꢀ8.0ꢀHz, CH3). 13C NMR (100ꢀMHz, DMSO): δ 72.92 (t,
3
1JP-C=ꢀ144ꢀHz, P-C-P), 31.78, 30.49, 29.70 – 29.37, 23.46 (t, JP-C=ꢀ6ꢀHz), 22.57, 14.42. 31P
NMR (162ꢀMHz, DMSO) δ 20.16. Yield 83%
HRMS (ESI): Calcd for C13H29O7P2 [Mꢀ- H]: 359.1389
(1-hydroxytetradecane-1,1-diyl)diphosphonic acid BP13
1H NMR (400ꢀMHz, DMSO): δ 1.92 – 1.72 (m, 2H, CH2), 1.61 – 1.43 (m, 2H, CH2), 1.35 –
3
0.99 (m, 22H, CH2), 0.85 (t, 3H, J=ꢀ8.0ꢀHz, CH3). 13C NMR (100ꢀMHz, DMSO): δ 72.91 (t,
3
1JP-C=ꢀ144ꢀHz, P-C-P), 31.78, 30.51, 29.74 – 29.43, 23.42 (t, JP-C=ꢀ6ꢀHz), 22.57, 14.42. 31P
NMR (162ꢀMHz, DMSO): δ 20.12. Yield 80%
HRMS (ESI): Calcd for C14H31O7P2 [Mꢀ-ꢀH] 373.1545
1-hydroxypentadecane-1,1-diyl)diphosphonic acid BP14
1H NMR (400ꢀMHz, DMSO): δ 1.93 – 1.73 (m, 2H, CH2), 1.61 – 1.44 (m, 2H, CH2), 1.35 –
3
1.11 (m, 22H, CH2), 0.85 (t, 3H, J=ꢀ8.0ꢀHz, CH3). 13C NMR (100ꢀMHz, DMSO): δ 72.92 (t,
3
1JꢀP-C=ꢀ144ꢀHz, P-C-P), 33.97, 31.78, 30.51, 29.72 – 29.38, 29.20, 22.58, 23.47 (t, JꢀP-
C=ꢀ6ꢀHz), 14.42 (CH3). 31P NMR (162ꢀMHz, DMSO): δ 20.15. Yield 77%
HRMS (ESI): Calcd for C15H33O7P2 [Mꢀ-ꢀH] : 387.1705.
(1-hydroxyhexadecane-1,1-diyl)diphosphonic acid (Monteil, Guénin, Migianu, Lutomsky, &
Lecouvey, 2005) BP15
These data are in accordance with those reported in the literature (Monteil et al., 2005)
General procedure for synthesis of bisphosphonate monomethyl esters.
6

(1-hydroxy-1-(hydroxy(methoxy)phosphoryl)hexadecyl)phosphonic acid BP15M01 was
synthesized according the procedure described by Migianu (Migianu, Guénin, & Lecouvey,
2005) and these data are in accordance with those reported in the literature (Migianu et al.,
2005).
General procedure for synthesis of bisphosphonate dialkyl esters.
Dimethyl (1-hydroxyhexadecane-1,1-diyl)bis(hydrogen phosphonate)(Monteil et al., 2005)
BP15M11 was synthesized according the procedure described by (Monteil et al., 2005), and
these data are in accordance with those reported in the literature.
General procedure for synthesis of bisphosphonate trimethyl esters.
Methyl hydrogen (1-(dimethoxyphosphoryl)-1-hydroxyhexadecyl)phosphonate BP15M21
The adequate acid chloride (50 mmol) was added dropwise at -10°C under argon to
trimethylphosphite (5.9 mL, 50 mL). The reaction mixture was then stirred at r.t for 2h (the
31
end of reaction was ascertained by P {1H} NMR or IR spectroscopy). The crude product
was purified as indicated in Table 1 to furnish the corresponding α-keto-phosphonate
dimethyl ester. These data are in accordance with those reported in the literature (Migianu et
al., 2005).
In a 50 mL round-bottom three-neck flask equipped with a thermometer, α-keto-phosphonate
dimethyl ester (5 mmol) was added dropwise, under argon, at -5°C, to dialkyl phosphite (5
mmol). When addition was completed, reaction mixture was allowed to stand at room
temperature. The disappearance of tris(trimethylsilyl) phosphite during the reaction was
31
monitored by P {1H} NMR. The flask containing the solution was placed under an argon
stream. A Pasteur pipette purged with argon was used to take 500 ul of mixture. Sample was
introduced into a NMR tube purged with argon, without deuteriated solvent. If the peak of
P(OSiMe3)3 doesn’t appear on the spectrum, the reaction is over and the volatile fractions
were evaporated under reduced pressure (0.1 Torr) before being hydrolyzed with methanol.
After evaporation, crude products were precipitated in diethyl ether (Migianu, Mallard,
Bouchemal, & Lecouvey, 2004).
1H NMR (400ꢀMHz, CDCl3): δ 3.67, 3.64, 3.60 (d, 9H, 3JP-H= 12Hz, 3 OCH3), 1.94 – 1.82 (m,
3
2H, CH2), 1.65 – 1.51 (m, 2H, CH2), 1.38 – 1.19 (m, 24H, CH2), 0.92 (t, 3H, Jꢀ=ꢀ8.0ꢀHz,
1
CH3). 13C NMR (100ꢀMHz, DMSO): δ 75.05 (t, JꢀP-C=ꢀ150ꢀHz, P-C-P), 53.78, 53.58, 35.10,
7

31.77, 30.31, 29.68 – 29.27, 29.19, 23.22, 22.57, 14.42 (CH3). 31P NMR (162ꢀMHz, CDCl3): δ
23.05 (d, 2JP-P= 42.12 Hz), 21.85 (d, 2JP-P= 42.12 Hz). Yield: 85%
HRMS (ESI): Calcd for C19H41O7P2 [Mꢀ-ꢀH]: 443.2330
2.3. NMR analysis
2D ROESY spectra were recorded at 25°C on a Brucker Advance III 400 spectrometer
equipped with a BBFO probe. The parameters were a mixing time of 500 ms during spin lock
with 64 scans using the states-TPPI method with a 2048 time domain points (2K time
1
domain) in F2 and 256 time increments in F1. H NMR was run with FID resolution of 0.25
Hz/ point, 4s of acquisition time, 1s of recycle delay and 16 scans, for 31P NMR, FID
resolution of 0.97 Hz/ point, 0.5 s of acquisition time, 2s of recycle delay and 16 scans and for
13C NMR, FID resolution of 0.36 Hz/ point, 1.4 s of acquisition time, 2s of recycle delay and
1500 scans.
Solutions of -CD and bisphosphonates were prepared in 99.98% D2O with a concentration of
5 mM and then was mixed to 1:1 ratio.
2.4. ITC analysis
An isothermal calorimeter (ITC200, MicroCal Inc., USA) was used in order to determine
thermodynamic data of inclusion, for each studied complexes, at 298 K. Titration experiments
consisted in 11 additions of the syringe solution (degassed phosphate buffer, 50 mM, pH=6.5,
total volume 40 μL) on the cell content (same buffer, volume equal to 202.8 μL), with a first
small aliquot of 1 μL followed by 10 aliquots of 3.7 μL (injection duration 7.4 s, time interval
75 s, agitation speed 1000 rpm). Heat flow was recorded as a function of time, the first
injection being discarded to eliminate any error induced by material diffusion effects between
the two compartments or backlash in the drive screw mechanism of the syringe. The heat
produced per injection was obtained by measuring each peak area, by means of ORIGIN
software. Titration experiments involved guest solution in the cell (0.25 or 0.5 mM) and host
in the syringe (5 mM). Blank titrations were obtained by injecting host, guest or free buffer
within a buffer solution, respectively. Effective binding signal was then obtained by
subtracting to the titration experiment each host and guest blank experiment and adding the
8

buffer blank experiment (since this latter signal was already taken into account in each of the
two other blank experiments). Complex formation constants (Kf in M-1), binding
stoichiometry (n) and inclusion enthalpies (H in cal.mol-1) were deduced from binding
isotherms (assuming a 1:1 stoichiometry), by nonlinear analyses of complexation heats as a
function of total concentrations. Inclusion entropy and free enthalpy were deduced from the
values of K and H. Binding isotherms treatment were realized by means of a dedicated
homemade program (Bertaut & Landy, 2014).
2.5. UV analysis
Complex formation constant (Kf) between bisphosphonates and cyclodextrins were
characterized by a competition method called also spectral displacement method (Landy,
Fourmentin, Salome, & Surpateanu, 2000). Methyl orange (MO) was used in its basic form
(10-4M in KH2PO4/NaOH buffer pH= 6.2). If a given concentration of bisphosphonate is
added to a solution containing the CD and MO species, it causes a disturbance of the CD-MO
complexation, leading to a difference between the spectra recorded in the absence and
presence of the bisphosphonate. K values of CD/BP can be deduced from this absorbance
difference. Stock solution of methyl orange was set to 0.1 mM. Then solutions of
cyclodextrins were prepared with stock solution of methyl orange with a concentration of 0.5
mM for the -CD and -CD and 0.2 mM for HP--CD and RAMEB respectively. The
concentration of bisphosphonate solution was fixed to 0.5 mM by adding the guest inside CD/
MO solution. Data were treated by the use of a dedicated algorithmic treatment (Landy,
Fourmentin, Salome, & Surpateanu, 2000).
3. Results and discussions
3.1. Synthesis of bisphosphonates
Regarding the synthesis of lipophilic bisphosphonates (BPs), Merck synthesis using PCl3
/H3PO3 is well documented (Kieczykowski, Jobson, Melillo, Reinhold, Grenda, & Shinkai,
1995 ; Lecouvey & Leroux, 2000), but uses harsh acidic conditions combined with heavy
heating and long reaction time. These drastic conditions enable the use of fragile substrates
9

and the preparation of functionalized BPs. As an alternative to this pathway our group
proposed a very mild and one-pot synthesis of BPs from tris(trimethylsilyl)phosphite and acyl
chlorides (Lecouvey, Mallard, Bailly, Burgada, & Leroux, 2001). The interest of this
procedure is that it could be indifferently used for aliphatic and aromatic substrates without
any restriction concerning the nature of the functional group on the side chain.
This method enables the synthesis of various substituted BPs in a one-pot procedure under
mild conditions. Moreover, reaction times were shortened and purifications were easier. This
reaction was successfully extended to anhydride, aliphatic and aromatic substrates in excellent
yields (Guénin, Degache, Liquier, & Lecouvey, 2004). Synthesis of esterified
bisphosphonates is not possible using standard procedure used for non-esterified BPs. Most of
the synthesis previously described usually goes through the making of the bisphosphonate
tetraester which is then dealkylated (Turhanen, Ahlgren, Jarvinen, & Vepsalainen, 2001). The
main drawbacks of these protocols are the thermal and basic instability of bisphosphonate
tetraesters leading to phosphonate-phosphate isomerisation (Fitch & Moedritzer, 1962).
Moreover the regioselective dealkylation to prepare partial esters is hard and does not occur in
good yields (Vepsalainen, 2002). Our group has proposed a very mild and one-pot synthesis
to obtain BP P,P’ and P,P diesters , BP triesters and BP monoesters (scheme 2) (Migianu et al.,
2005; Migianu et al., 2004). We have further exemplified this synthesis to the preparation of
numerous bisphosphonate diesters varying both on the lateral chain and ester functions
(Guénin, Monteil, Bouchemal, Prangé, & Lecouvey, 2007; Monteil et al., 2005)
3.2. NMR spectroscopy
Nuclear magnetic resonance spectroscopy is a major research tool useful for providing
information concerning the complexation between bisphosphonates and cyclodextrins
(Salvatierra, Jaime, Virgili, & Sánchez-Ferrando, 1996).
The structure of -CD represented in Fig. 2., clearly shows that H3 and H5 hydrogens located
in the inner cavity can most readily be linked to effect of guest inclusion (Djedaïni, Zhao Lin,
Perly, & Wouessidjewe, 1990). H3 is located near the wider rim while H5 is on the opposite
side closed to the narrower rim. Usually the guest enters through the wider rim, involving a
most important shift for H3. H1, H2, H4 are set on the outside front of the cavity and should
10

exhibit negligible changes. For H6, the change in chemical shift will be partly dependent on
the guest entry.
As a result, with reference to the proton chemical shift values of the cyclodextrin alone, the
guest and the complex between these two latter, it is possible to get information on the nature
of the inclusion guest orientation in the cyclodextrin and the stability of the complex thus
formed. 1H NMR spectra of the free bisphosphonate guest was not shown on the figure 3 due
to their low solubility in deuterated water. 1H NMR spectra of -CD alone and
bisphosphonate/ CD complex here BP1 were recorded and were depicted in Fig. 3.
1
The values reported in table 1 represent the H chemical shift of -CD in the free and
complex state.
Assessment of the -CD chemical shift in the absence (Schneider, Hacket, Rudiger, & Ikeda,
1
1998) or presence of bisphosphonate by H NMR showed the complex formation between
bisphosphonate and -cyclodextrin. The results demonstrated that for all complexes H3 and
H5 have the largest up field shifts as expected (Schneider et al., 1998), while the complex
formation with bisphosphonate has little effect on H1, H2, H4 and H6 protons of -
cyclodextrin. According to their location outside the cavity, the chemical shift of H1, H2 and
H4 are less affected by complexation. Moreover, the chemical shift of H6 differs slightly
showing that inclusion is not deep enough for the bisphosphonate to strongly interact with H6.
2D ROESY experiment (Forgo & D’Souza, 1998; Veiga, Fernandes, Carvalho, & Geraldes,
2001) was performed to provide informations in order to understand the inclusion
complexation mode and inter and intra molecular interactions between guest atoms and
cyclodextrin. Fig. 4. represents the 2D NMR ROESY spectrum of -cyclodextrin/ BP10
complex. For each studied bisphosphonate/ -cyclodextrin complex, the ROESY spectrum
displays intermolecular cross peaks between H3 and H5 protons of -cyclodextrin and CH3
group and methylene chain of bisphosphonate demonstrating the inclusion of these groups in
the hydrophobic cavity. The same molecular correlation was observed for each complex.
More specifically, the ROESY spectrum indicated two cross peaks between the H3 of the -
cyclodextrin and CH3 group and methylene chain of bisphosphonate and showed significant
correlation between H5 of the -cyclodextrin and methylene chain of bisphosphonate. An
expansion of the ROESY spectrum is displayed in Fig. 5.
11

The presence of cross peak between H3 and methylene and CH3 group of bisphosphonate side
chain demonstrated that the inclusion might take place though the wider rim. The overlap
between H3 and H6 was so important that a cross peak between H6 and CH2 cannot be
excluded and consequently an inclusion though the narrower rim.
The flexibility of the side chain of bisphosphonate allowed the contortion of this last one
inside the cavity allowing the interactions between H3 and H5 and the bisphosphonate. The
end of the alkyl chain may come up along the cavity. For BP11 and BP13, low cross peak
appeared on the bidimensional spectrum involving H3 of cyclodextrin and the two methylene
anchored to P-C-P backbone of bisphosphonate leading to the conclusion that the P-C-P
backbone may be located outside the cavity but very close to the wider rim. The P-C-P
backbone may interact with the surrounding water or with some of the hydroxyl groups of -
cyclodextrin.
3.3. ITC analysis
In order to obtain a quantitative evaluation of the interactions previously highlighted by NMR,
we first investigated the inclusion of BP10 to BP15 with β-CD by isothermal titration
calorimetry. A representative ITC experiment is presented in Fig. 6., in the case of β-
CD/BP10 complex. Injections induced negative signals, with decreasing intensities as the
experiment proceeded. These results assess that a significant interaction take place between
the two partners and that such formation of complexes is an exothermic reaction. As can be
seen, a good fit was obtained between experimental data and theoretical simulation (for
theoretical background see Bertaut & Landy, 2014), on the basis of a 1:1 stoichiometry. The
same trend was observed for each studied host/guest combination. The use of theoretical
models considering the presence of both 1:1 and 2:1 (or 1:2) complexes lead to uncertainties
superior to the obtained values of formation constants for the highest stoichiometry,
suggesting that only 1:1 inclusion compounds are present. Given the important length of such
guest molecules, the fact that no 2:1 complex was observed confirmed that only the
bisphosphonate side chain was available for inclusion, the bisphosphonate part being too
bulky and/or hydrophilic to mobilize a second cyclodextrin.
ITC experiments afforded a complete picture of inclusion thermodynamics. Values of
formation constants, inclusion free enthalpy (G), enthalpy (H) and entropy (-TS) are
listed in Table 3. BP10 to BP15 complexes were characterized by a significant stability, with
free enthalpy ranging from -4954 to -6482 cal/mol. Inclusion was both enthalpy and entropy
12

driven, the entropic contribution being roughly 2 to 3 times higher when compared to the
enthalpic part. The enthalpic stabilization slightly increased with the number of methylene of
the bisphosphonate side chain (i.e. on the P-C-P backbone), from -1263 to -2323 cal/mol
when moving from 10 to 15 carbons. In addition, if we consider free energy of complexation
as a function of the length of the alkyl part, it seems that each added methylene first improved
the β-CD-BP affinity, until a total carbon content equal to 14, after what one can note a slight
decrease of the complex stability. As the chain length increased, the molecular surface buried
within the cyclodextrin cavity became wider, thus enhancing van der Waals interaction and
thus affinity, without significant reduction of the alkyl chain flexibility. For the longest side
chain (i.e. for BP15), a loss of inclusion entropy appeared and was stronger than the rise of
enthalpy. Such behavior may be attributed to a more constrained conformation of the alkyl
chain. It should be also kept in mind that above a certain length, free space is expected to be
no longer available inside the cavity.
3.4. UV-Vis analysis
Formation constants were also evaluated by means of a spectral displacement method.
Introducing a known quantity of BP within a cyclodextrin-methylorange solution indeed lead
to a lower concentration of cyclodextrin-methyl orange complex, and thus to a modification
of the electronic absorption of this solution. As a consequence, affinity between BPs and
cyclodextrins was evaluated from the absorbance difference in the absence and the presence
of BP, assuming a 1:1 stoichiometry in all cases. Corresponding formation constants are
presented in Table 3. These values are in strong agreement with ITC results, which confirms
that only 1:1 complexes were formed and that such spectral displacement method is well
suited to the characterization of cyclodextrins/BP complexes. As this method is easy to handle
in the case of numerous complexes, these competition experiments were applied to all studied
inclusion compounds.
In particular, we evaluated the influence of methyl introduction on the bisphosphonate part,
by considering the inclusion of BP15 to BP15M21 (Table 3), for which 0, 1, 2 and 3 methyl
were introduced respectively. Similar formation constants were observed, demonstrating
again that the bisphosphonate part does not seem to be involved in the inclusion phenomenon.
13

In addition, we investigated the inclusion behavior of modified cyclodextrins, by
characterizing the interaction of RAMEB and HP--CD with BP10 to BP14, in comparison to
the native β-CD. Corresponding stability are presented in Table 3 and expressed in terms of
free energy of inclusion as a function of the number of carbons of the bisphosphonate side
chain, in Fig. 7. Linear trends between free energy and chain length were observed for those
three β-CDs. Slopes of the linear regressions enable the evaluation of added free energy per
methylene (-0.3 to -0.4 kcal/mol). Additionally, the intercept is representative of the inclusion
free energy in the absence of alkyl chain. They are ranging from 0 to 2.2 kcal/mol, which
corresponds to non significant formation constants (inferior to 40M-1), confirming the quasi-
absence of interactions with the bisphosphonate part, stability being essentially provided by
inclusion of the alkyl side chain. If complexation behavior, in terms of order of magnitude,
was similar for the three studied β-CDs, one can note that formation constants were slightly
weaker for RAMEB and HP--CD if compared to the native β-CD. This slight decrease in
affinity might be ascribed to steric hindrance induced by the chains introduced on modified β-
CDs.
Finally, the influence of the cavity size on the inclusion ability was taken into account, by
characterizing the stability of complexes formed between α-CD and BP11 to BP14. As can be
concluded from the results listed in table 3, it seems that the inclusion ability of α-CD and β-
CD are analogous (no significant differences between formation constants, when calculating
the 95% confidence interval by multiplying standard deviations by a Student t value equal to
3.18).
4. Conclusion
The complex formation between bisphosphonates and various cyclodextrins was
demonstrated by complementary methods such as 1D and 2D NMR spectroscopy, UV-visible
and ITC analysis. Results indicated the formation of 1:1 complex of cyclodextrin/
bisphosphonates and the contribution of bisphosphonate side chain to the inclusion
phenomenon. While bisphosphonates bind to various cyclodextrins with a weak influence of
the cavity size and of the studied cyclodextrin derivatives, the bisphosphonate side chain
14

length was the main parameter controlling the complex stability. The high values of formation
constant showed that cyclodextrin could improve the oral bioavailability of bisphosphonates
and could be easily used for novel promising formulation.
References
Astray, G., Gonzalez-Barreiro, C., Mejuro, J.C., Rial-Otero, R., & Simal Gàndara, J. (2009).
A review on the use of cyclodextrins in foods. Food Hydrocolloids, 23 (7), 1631-1640.
Astray, G., Mejuto, J.C., Morales, J., Rial-Otero, R., & Simal Gàndara, J. (2010). Factors
controlling flavors binding constants to cyclodextrins and their applications in foods. Food
Research International, 43 (4), 1212-1218.
Beneyettou, F., Lalatonne, Y., Chebbi, I., Di Benedetto, M., Serfaty, J.M., Lecouvey, M., &
Motte, L. (2011). A multimodal magnetic resonance imaging nanoplatform for cancer
theranostics. Physical Chemistry Chemical Physics, 13 (21), 10020-10027.
Bertaut, E., Landy, D. (2014). Improving ITC studies of cyclodextrin inclusion compounds by
global analysis of conventional and non-conventional experiments. Beilstein Journal of
Organic Chemistry, 10, 2630-2641.
Biernacka, J., Betlejewska-Kielak, K., Witowska-Jarosz, J., Klosińska-Szmurto, E., &
Mazurek, A.P. (2014). Mass spectrometry and molecular modeling studies on the inclusion
complexes between alendronate and-cyclodextrin. Journal of Inclusion Phenomena
Macrocyclic Chemistry, 78, 437-443.
Chebbi, I., Migianu-Griffoni, E., Sainte-Catherine, O., Lecouvey, M., & Seksek, O. (2010). In
vitro assessment of liposomal neridronate on MDA-MB-231 human breast cancer cells.
International Journal of Pharmaceutics, 383 (1-2), 116-122.
Cho, E., Nazir Tahir, M., Choi, J.M., Kim, H., Yu, J.H., & Jung, S.H. (2015). Novel magnetic
nanoparticles coated by benzene and -cyclodextrin bearing dextran, and the sorption of
polycyclic aromatic hydrocarbon. Carbohydrate Polymers, 133, 221-228.
Daubiné, F., Cortial, D., Ladam, G., Armani, H., HaÏkel, Y., Voegel, J.C., Clézardin, P., &
Benkirane-Jessel, N. (2009). Nanostructured polyelectrolyte multilayer delivery systems for
bone metastasis prevention. Biomaterials, 30, 6367-6373.
Djedaïni, F., Zhao Lin, S., Perly, B., & Wouessidjewe, D. (1990). High field nuclear magnetic
resonance techniques for the investigation of a -cyclodextrin : indomethacin inclusion
complex. Journal of Pharmaceutical Sciences, 79 (7), 643-646.
Eastell, R., Walsh, J.S., Watts, N.B., & Siris, E. (2011). Bisphosphonates for postmenopausal
osteoporosis. Bone, 49 (1), 82-8.
15

Fitch, S.J., & Moedritzer, K.J. (1962). NMR study of the P-C(OH)-P to P-C-O-P
rearrangement: tetraethyl-1-hydroxyalkylidenediphosphonates. Journal of the American
Chemical Society, 84, 1876-1879.
Forgo, P., & D’Souza, V.T. (1998). The application of selective ROE experiments to study
solution structures of cyclomaltooligosaccharide derivatives and complexes. Carbohydrate
Research, 306. 473-478.
Graham, R., & Russel, G. (2011). Bisphosphonates : The first 40 years. Bone, 49, 2-19.
Guénin, E., Degache, E., Liquier, J., & Lecouvey, M. (2004). Synthesis of 1-
hydroxymethylene-1,1-bis (phosphonic acids) from Acid Anhydrides: Preparation of a new
cyclic 1-acyloxymethylene-1,1-bis(phosphonic acid). European Journal of Organic
Chemistry, 14, 2983-2987.
Guénin, E., Monteil, M., Bouchemal, N., Prangé, T., & Lecouvey, M. (2007). Synthesis of
phosphonic esters of alendronate, pamidronate and neridronate. European Journal of Organic
Chemistry, 20, 3380-3391.
Hein, C.D., Liu, X .M., Chen, F., Cullen, D.M., & Wand, D. (2010). The synthesis of a
multiblock osteotropic polyrotaxane by copper (I)-catalyzed Huisgen 1,3-dipolar
cycloaddition. Macromolecular Bioscience, 10, 1544-1556.
Hudock, M.P., Sanz Rodriguez, C.E., Song, Y., Chan, J.M., Zhang, Y., Odeh, S., Kosztowski,
T., Leon-Rossel, A., Concepcion, J.L., Yardley, V., Croft, S.L., Urbina, J.A., & Oldfield, E.
(2006). Inhibition of Trypanosoma cruzi hexokinase by bisphosphonates. Journal of
Medicinal Chemistry, 49 (1), 215-223.
Ijaz, M., Matuszczak, B., Rahmat, D., Mahmood, A., Bonengel, S., Hussain, S., Huck, C.W.,
& Bernkop-Schnürch, A. (2015). Synthesis and characterization of thiolated -CD as a novel
mucoadhesive excipient for intra-oral drug delivery. Carbohydrate Polymers, 132, 187-195.
Kavanagh, K.L., Guo, K., Dunford, J.E., Wu, X., Knapp, S., Ebetino, F.H., Rogers, M.J.,
Russel, R.G., & Oppermann, U. (2006). The molecular mechanism of nitrogen containing
bisphosphonates as antiosteoporosis drugs. Proceedings of the National Academy of Sciences
of the United State of America, 103 (20), 7829-34.
Kieczykowski, G. R., Jobson, R. B., Melillo, D. G., Reinhold, D. F., Grenda, V. J., & Shinkai,
I. (1995). Journal of Organic Chemistry, 60, 8310-8312.
Landy, D., Fourmentin, S., Salome, M., & Surpateanu, G. (2000). Analytical improvement in
measuring formation constants of inclusion complexes between -cyclodextrin and phenolic
compounds. Journal of Inclusion Phenomena and Macrocyclic Chemistry, 38 (1-4). 187-198.
Lecouvey, M., & Leroux, Y. (2000). Synthesis of 1-hydroxy-1,1-bisphosphonates.
Heteroatom Chemistry, 11, 556-561.
Lecouvey, M., Mallard, I., Bailly, T., Burgada, R., & Leroux, Y. (2001). A mild and efficient
one pot synthesis of 1-hydroxymethylene-1,1-bisphosphonic acids. Preparation of new tripod
ligands. Tetrahedron Letters, 42, 8475-8478.
16

Lee, M.V., Fong, E.M., Singer, F.R., Guenette, R.S. (2001). Bisphosphonate treatment
inhibits the growth of prostate cancer cells. Cancer Research, 61, 2602-2658.
Lin, J.H. (1996). Bisphosphonates: a review of their pharmacokinetic properties. Bone, 18,
75-85.
Liu, X.M., Lee, H.T., Reinhardt, R.A., Marky, L.A., & Wang, D. (2007). Novel biomineral-
binding cyclodextrins for controlled drug delivery in the oral cavity. Journal of Controlled
Release, 122, 54-62.
Loftsson, T., Brewster, M.E., & Masson, M. (2004). Role of cyclodextrins in improving oral
drug delivery. American Journal of Drug Delivery, 2, 261-275.
Merrell, M.A., Wakchoure, S., Lehenkari, P.P., Harris, K.W., & Selander, K.S. (2007).
Inhibition of the mevalonate pathway and activation of p38 MAP kinase are independly
regulated by nitrogen-containing bisphosphonates in breast cancer cells. European Journal of
Pharmacology, 570, 1-3, 27-37.
Migianu, E., Mallard, I., Bouchemal, N., & Lecouvey, M. (2004). One-pot synthesis of 1-
hydroxymethylene-1,1-bisphosphonate partial esters. Tetrahedron Letters, 45 (23), 4511-4513.
Migianu, E., Guénin, E., & Lecouvey, M. (2005). New efficient synthesis of 1-
hydroxymethylene-1,1-bisphosphonate monomethyl esters. Synlett, 3, 425-428.
Monteil, M., Guénin, E., Migianu, E., & Lutomski, D. (2005). Bisphosphonate prodrugs :
synthesis of new aromatic and aliphatic -1-hydroxy-1,1-bisphosphonate partial esters.
Tetrahedron, 61, 7528-7537.
Muller, S., Migianu, E., Lecouvey, M., Kraemer, M., & Oudar, O. (2005). Alendronate
inhibits proliferation and invasion of human epidermoid carcinoma cells in vitro. Anticancer
research, 25, 2655-2660.
Reid, I.R, & Hosking, D.J. (2011). Bisphosphonates in Paget’s disease. Bone, 49, 89-94.
Rogers, M.J., Crockett, J.C., Coxon, F.P., & Mönkkönen, J. (2011). Biochemical and
molecular mechanisms of action of bisphosphonates. Bone, 49 (1), 34-41.
Rogers, M.J. (2003). New insights into the molecular mechanisms of action of
bisphosphonates. Current Pharmaceutical Design, 9 (32), 2643-2658.
Russel, R. (2011). Bisphosphonates : the first 40 years. Bone, 49, 2-11.
Salvatierra, D., Jaime, C., Virgili, A., & Sánchez-Ferrando, F. (1996). Determination of the
inclusion geometry for the -cyclodextrin/ benzoic acid complex by NMR and molecular
modeling. Journal of Organic Chemistry, 61. 9578-9681.
Schneider, H.J, Hacket, F., Rudiger, V., Ikeda, I. (1998). NMR studies of cyclodextrin
complexes. Chemical Review, 98, 1755-1785.
Seow, W.Y., Xue, J.M., & Yang, Y.Y. (2007). Targeted and intracellular delivery of
paclitaxel using multifonctional polymeric micelles. Biomaterials, 28, 1730-1740.
Shipman, C.M., Rogers, M.J., Apperley, J.F., Russel, R.G., & Croucher, P.I. (1997).
Bisphosphonates induce apoptosis in human myeloma cell lines, a novel antitumor activity.
British Journal of Hamaetology, 98, 665-672.
17

Stella, V.J., & Rajewski, R.A. (1997). Cyclodextrins : their future in drug formulation and
delivery. Pharmaceutical Research, 14, 556-567.
Strickley, R.G. (2004). Solubilizing excipients in oral and injectable formulation.
Pharmaceutical Research, 21, 201-230.
Szajnman, S.H., Rosso, V.S., Malayil, L., Smith, A., Moreno, S.N., Docampo, R., &
Rodriguez, J.B. (2012). 1-(Fluoroalkylidene)-1,1-bisphosphonic acids are potent and selective
inhibitors of the enzymatic activity of Toxoplasma gondii farnesyl pyrophosphate synthase.
Organic and Biomolecular Chemistry, 10 (7), 1424-1433.
Szejtli, J. (1998). Introduction and general overview of cyclodextrins chemistry. Chemical
Reviews, 98 (5), 1743-1753.
Turhanen, P.A., Ahlgren, M.J., Jarvinen, T., & Vepsalainen, J.J. (2001). Bisphosphonate
prodrugs. Selective synthesis of (1-hydroethylidene)-1,1-bisphosphonate partial esters.
Synthesis, 4, 633-637.
Uekama, K., Hirayama, F., & Arima, H. (2006). Recent aspects of cyclodextrin-based drug
delivery systems. Journal of Inclusion Phenomena Macrocyclic Chemistry, 56, 3-8.
Veiga, F.J.B., Fernandes, C.M., Carbalho, R.A., & Geraldes, C.F.G.C. (2001). Molecular
modelling and 1H-NMR : ultimate tools for the investigation of tolbutamide : -cyclodextrin
and tolbutamide : hydroxypropyl--cyclodextrin complexes. Chemical and Pharmaceutical
Bulletin, 49, 1251-1256.
Vepsalainen, J. (2002). Bisphosphonate prodrugs. Current Medicinal Chemistry, 9, 1201-
1208.
18

BP10
BP11
BP12
BP13
BP14
O
OH
C
O
P
O
OH
C
O
P
O
OH
C
O
P
O
OH
C
O
P
O
OH
C
O
P
HO
HO
HO
HO
HO
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
P
P
P
P
P
HO
HO
HO
HO
HO
(CH2)9
CH3
(CH2)10
CH3
(CH2)11
CH3
(CH2)12
CH3
(CH2)13
CH3
BP15
BP15M01
BP15M11
BP15M21
O
P
OH
C
O
P
O
P
OH
C
O
P
O
P
OH
C
O
P
O
P
OH
C
O
P
HO
HO
HO
HO
CH3O
HO
CH3O
CH3O
OH
OH
OCH3
OH
OCH3
OH
OCH3
OH
(CH2)14
CH3
(CH2)14
CH3
(CH2)14
CH3
(CH2)14
CH3
Fig.1. Chemical structure of studied bisphosphonates.
19

Fig. 2. Structures of the glucopyranose repeating unit and the truncated structure of β-
cyclodextrin emphasizing the position of all hydrogens.
20

1
Fig. 3. H NMR spectra of the a) -CD alone, b) -CD/ BP10 complex in D2O, c) spectral
zooming of (b) and d) spectral zooming of (a) between 3.4 and 5.4 ppm.
21

Fig. 4. ROESY spectrum of bisphosphonate -cyclodextrin / P10 complex in D2O.
22

Fig. 5. (a) Expansion of ROESY spectrum of -CD/ BP10 complex (1:1) in D20 and (b)
spectral resolution of ROESY spectrum.
23

Fig. 6. ITC titration of BP10 by β-CD at 25°C. Upper and lower panels display raw
thermogram and binding isotherm (dots: experimental values; solid line: best least square fit),
respectively.
24

Fig.7. Free energy of inclusion (as derived from K in Table 3) as a function of alkyl length,
for the complex formed between BP10 to BP14 with β-CD (blue diamond), RAMEB (red
square) and HP-β-CD (green triangle).
25

Scheme1 Direct methods for the 1-hydroxyalkylidenebisphosphonic acid synthesis.
26

Scheme 2 General methods for the partial ester bisphosphonates synthesis.
27

Table 1: 1H chemical shifts (ppm) for -CD in the free and complex state with BP.
+BP10 +BP11 +BP12 +BP13 +BP14 +BP15 +BP15 +BP15 +BP15
β-CD
M01
5.09
3.92
3.90
3.74
3.66
3.62
M11
5.09
3.91
3.89
3.75
3.66
3.62
M21
5.13
3.94
3.93
3.78
3.69
3.66
H1
H3
H6
H5
H2
H4
5.10
3.99
3.91
3.88
3.68
3.61
5.12
3.95
3.93
3.78
3.69
3.66
5.10
3.92
3.90
3.75
3.66
3.64
5.11
3.94
3.91
3.73
3.68
3.64
5.11
3.95
3.91
3.76
3.67
3.63
5.11
3.94
3.91
3.78
3.69
3.64
5.09
3.93
3.90
3.78
3.67
3.62
28

Table 2. Thermodynamic data of β-CD/BP inclusion complexes obtained by ITC at 25 °C.
K
G
H
-TS
(M-1)
(cal.mol-1)
(cal.mol-1)
(cal.mol-1)
BP10
BP11
BP12
BP13
BP14
BP15
4051 ± 468
-4901 ± 69
-5708 ± 44
-6061 ± 44
-6313 ± 48
-6491 ± 47
-6220 ± 37
-1263 ± 49
-1525 ± 22
-1854 ± 25
-1886 ± 24
-2095 ± 22
-2323 ± 24
-3638 ± 118
-4183 ± 66
-4207 ± 69
-4427 ± 71
-4396 ± 69
-3897 ± 61
15891 ± 1184
28917 ± 2138
44355 ± 3573
59923 ± 4749
37862 ± 2377
29