Synthesis of a carborane-containing cholesterol derivative and evaluation as a potential dual agent for MRI/BNCT applications

Article abstract of DOI:10.1039/c3ob42414f

In this study the synthesis and characterization of a new dual, imaging and therapeutic, agent is proposed with the aim of improving the efficacy of Boron Neutron Capture Therapy (BNCT) in cancer treatment. The agent (Gd-B-AC01) consists of a carborane un

Full text of DOI:10.1039/c3ob42414f

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
Synthesis of a carborane-containing cholesterol
derivative and evaluation as a potential dual agent
for MRI/BNCT applications†
Cite this: Org. Biomol. Chem., 2014,
12, 2457
Diego Alberti,‡a Antonio Toppino,‡b Simonetta Geninatti Crich,*a Chiara Meraldi,b
Cristina Prandi,b Nicoletta Protti,c,d Silva Bortolussi,c,d Saverio Altieri,c,d Silvio Aimea
and Annamaria Deagostinob
In this study the synthesis and characterization of a new dual, imaging and therapeutic, agent is proposed
with the aim of improving the efficacy of Boron Neutron Capture Therapy (BNCT) in cancer treatment.
The agent (Gd-B-AC01) consists of a carborane unit (ten boron atoms) bearing a cholesterol unit on one
side (to pursue the incorporation into the liposome bi-layer) and a Gd(III)/1,4,7,10-tetraazacyclododecane
monoamide complex on the other side (as a MRI reporter to attain the quantification of the B/Gd concen-
tration). In order to endow the BNCT agent with specific delivery properties, the liposome embedded
with the MRI/BNCT dual probes has been functionalized with a pegylated phospholipid containing a folic
acid residue at the end of the PEG chain. The vector allows the binding of the liposome to folate receptors
that are overexpressed in many tumor types, and in particular, in human ovarian cancer cells (IGROV-1).
An in vitro test on IGROV-1 cells demonstrated that Gd-B-AC01 loaded liposomes are efficient carriers for
the delivery of the MRI/BNCT probes to the tumor cells. Finally, the BNCT treatment of IGROV-1 cells
showed that the number of surviving cells was markedly smaller when the cells were irradiated after
internalization of the folate-targeted GdB10-AC01/liposomes.
Received 3rd December 2013,
Accepted 11th February 2014
DOI: 10.1039/c3ob42414f
www.rsc.org/obc
calculate the radiation dose. One of the problems confronting
the successful clinical implementation of BNCT is the
Introduction
Boron Neutron Capture Therapy (BNCT) is a binary treatment difficulty of quantitatively mapping the distribution of the
that combines low energy neutron irradiation with the pres- boron containing molecules in patients before and during the
ence of boron-containing compounds at the target cells. This treatment. To date there has been no non-invasive way to
methodology has been proposed for several pathologies, evaluate the boron concentration in diseased and healthy
although most of the efforts have been focused on cancer.1–3 tissues of the subject undergoing the irradiation. Thus, dose
BNCT has been under investigation for many years, but it has calculations are based on boron content values in blood,
not entered the routine clinical practice yet. Since this treat- tumour, and normal tissue obtained from biodistribution
ment is effective only on cells internalizing large amounts of studies performed beforehand.4,5 Blood samples can be taken
boron (at least 20–30 ppm), the principal cause of its limited just before and even during irradiation to measure the tissue
success mainly depends on the insufficient delivery of NCT boron concentration, assuming the tumour/blood ratios
compounds to the target cells. A key issue deals with the assessed in previous biodistribution studies. However the
measurement of the local boron concentration that is crucial boron distribution varies among patients and therefore large
to determine the optimal neutron irradiation time and to uncertainties exist in the tumour-to-blood boron concentration
ratio. The recent development of more sensitive imaging tech-
niques and molecular medicine protocols, based on the use of
site specific agents able to carry large amounts of both thera-
aDepartment of Molecular Biotechnology and Health Sciences, University of Torino,
via Nizza 52, 10126 Torino, Italy. E-mail: simonetta.geninatti@unito.it;
peutic and diagnostic agents, can contribute to enhance the
BNCT efficacy and to become more competitive with the estab-
lished tumour treatment protocols.6 Among the compounds
developed for BNCT, much attention has been devoted to car-
boranes, icosahedral cages containing several boron atoms,
which could increase the payload of boron in the tumour
Fax: +39 011 6705687
bDepartment of Chemistry, University of Torino, via P. Giuria 7, 10125 Torino, Italy
cDepartment of Nuclear and Theoretical Physics, University of Pavia, Pavia, Italy
dNuclear Physics National Institute (INFN), Section of Pavia, Pavia, Italy
†Electronic supplementary information (ESI) available: 1H NMR, 13C NMR and
DEPT spectra of products 1–5 and 8. See DOI: 10.1039/c3ob42414f
‡These authors contributed equally.
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 2457–2467 | 2457

View Article Online
Paper
Organic & Biomolecular Chemistry
cells.7–11 Boron containing compounds can be further functio- disadvantage of the use of therapeutic and imaging agents as
nalized with lipophilic moieties in order to permit their independent molecules is that they maintain the same biodis-
specific and efficient delivery through the incorporation into tribution only if the liposome remains intact. After liposome
targeted nanoparticles. The synthesis of new dual agents for degradation their biodistribution will be different and depen-
applications in MRI/BNCT where a carborane unit is linked to dent on their size, hydrophilic character, residual charges, etc.
a lipophilic chain and to a Gd–DOTA complex through amidic In order to overcome this problem, in this study a dual agent,
bonds has been developed in our laboratory (AT101).12 Accord- in which the imaging (Gd–DOTA) and the BNCT (carborane)
ing to the synthetic design, the lipophilic probe AT101 binds components are part of the same molecule, is proposed. The
to LDLs (low density lipoproteins) and accumulates in tumor incorporation of the molecule into the liposome bilayer has
cells characterized by an up-regulation of LDL transporters. been pursued by a functionalization with a cholesterol residue.
The exploitation of LDLs as biological vectors was indeed the Cholesterol is a standard component of the liposome bi-layer
key to obtain a high concentration of B atoms per g in the usually introduced into the phospholipid formulation in order
tumor cells together with a good selectivity between tumor and to increase the particle stability. In contrast, complexes
healthy tissues. In vivo MR image acquisition showed that the bearing only a palmitic chain, as the parent AT101 previously
amount of boron taken up in the tumor region was above the investigated by our group, cannot be steadily incorporated into
threshold required for a successful NCT treatment. More the liposome bi-layer and the complex may be released ran-
recently, with the purpose of simplifying the synthetic pro- domly. Synthetic mimics of cholesterol represent an important
cedure but maintaining, and possibly improving, the favorable molecular tool in the fields of bioorganic/medicinal chemistry
properties of AT101, we designed a new structure containing a and chemical biology. Furthermore, carborane cholesterol
triazole linker instead of the amidic function (MEA/01).13 Its derivatives were incorporated into unilamellar liposomes
adducts were tested “in vitro” on B16 melanoma cells and the showing selective localization in tumours22,23 and they might
obtained intracellular Gd/B concentration appears sufficient be useful vehicles for transporting boron to the tumor site. In
for further “in vivo” BNCT studies.
this study, the liposomes embedded with MRI/BNCT dual
An alternative promising approach to deliver efficiently probes have been functionalized with a pegylated phospho-
boron containing agents to tumors consists of the use of lipo- lipid, containing a folic acid residue at the end of the PEG
somes. In fact, there are many examples in the literature of for- chain, as a specific carrier to transport large amounts of B/Gd
mulations containing boron compounds encapsulated into the dual agents to tumor cells, in particular human ovarian cancer
aqueous cavity or, after conjugation with lipids, incorporated cells overexpressing folate receptors. Finally, in order to
into the liposomal bilayer.14–16 A specific class of liposomes, improve the efficacy of the dual agent, the DOTA-monoamide
named “long time circulating” polyethylene glycol (PEG)- ligand has been complexed with gadolinium enriched in the
coated liposomes, is able to accumulate in solid tumors as a 157Gd isotope (95%) that has the largest (255 000 b) thermal
consequence of microvascular permeability and defective lym- neutron capture cross-section among all the stable isotopes,
phatic drainage.17–19 The extent of passive extravasation is and it is about 66 times as large as 10boron (3838 b).24
directly dependent on the prolonged residence time of lipo-
somes in the bloodstream. As far as the internalization step is
concerned, one should note that in the absence of a receptor
mediated mechanism, the internalization of the intact lipo-
some into tumor cells does not take place. Therefore, in the
Results and discussion
Synthesis of the B/Gd dual agent
absence of liposome active targeting the boron containing Since the hydroxyl group of cholesterol is sterically hindered
molecule has first to be released in the tumor extracellular and then unreactive, in the first synthetic step it has been
matrix in order to allow its diffusion into cells. Alternatively, functionalized, by a nucleophilic substitution, with a hydroxyl
the active targeting of liposomes to tumor cells is obviously a propyl group in order to make easier the linkage to the carbor-
promising strategy to improve drug delivery to tumors. To this ane cage. As shown in Scheme 1, cholesterol was transformed
purpose, liposomes have to be functionalized with specific into the corresponding mesylate (1) in quantitative yields (as
1
ligands able to bind and to internalize through receptors assessed by the detection of the signal at 2.97 ppm in the H
expressed on tumor vasculature or on tumor cells. Recently, spectrum pertinent to the mesylate CH3). Intermediate 1 was
Nakamura and co-workers developed liposomes containing reacted with propan-1,3-diol in anhydrous dioxane at 120 °C
mercaptoundecahydrododecaborate (BSH) in the inner overnight. Despite the harsh applied conditions, the product
aqueous cavity and 10% distearoyl boron lipid (DSBL) incor- was obtained in 27% yield after chromatographic purification
porated into the phospholipid bilayer.20,21 The proposed because of the steric hindrance of the alcoholic group. Alcohol
system appears particularly efficient because the liposome 2 was then oxidized to the corresponding carboxylic acid, fol-
shell itself contains boron in addition to the boron encapsu- lowing the classical Jones reaction conditions with an overall
lated in the aqueous cavity. Furthermore, they performed the yield of 78%. The formation of 3-(cholesten-5-en-3β-yloxy)-pro-
in vivo real-time tracking of 10% DSBL liposomes by MRI pionic acid (3, Scheme 1) was confirmed by the detection of a
using liposomes encapsulating Gd–DOTA, observing a high resonance at 170 ppm in the 13C spectrum pertinent to the car-
signal intensity enhancement in the tumor region. The bonyl group. The COOH moiety was then coupled to C-(2-
2458 | Org. Biomol. Chem., 2014, 12, 2457–2467
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Scheme 1 C-(2-Benzyloxyethyl)-C’-[cholesten-5-en-3β-yloxy-(3-propionacetoamidomethyl)]-o-carborane (4). Reaction conditions and yields: (a)
MsCl, Et3N, Et2O, 0 °C → r.t (>99%); (b) HO(CH2)3OH, dioxane, 120 °C (27%); (c) 3 M CrO3 in H2SO4, acetone, r.t. (80%); (d) dicyclohexylcarbodiimide,
CH2Cl2, rt (79%) or N-hydroxysuccinimide, dicyclohexylcarbodiimide, CH2Cl2, rt (64%).
benzyloxy)-ethyl-C′-aminomethyl-o-carborane 7 (prepared as
Alcohol 5 was readily oxidized to the corresponding carboxylic
previously reported).12 Firstly, the coupling was attempted acid 6 by CrO3 in acetone–sulfuric acid solution (Scheme 2).
according to the procedure proposed by Kaminski,25 but the
The structure of derivative 6 was supported by the H and
1
desired product was not obtained. A two-step protocol, where 13C NMR data. The H NMR spectrum showed the disappear-
the acid was activated with N-hydroxysuccinimide (NHS) in the ance of the signal at 3.75 ppm assigned to the HOCH2-group,
presence of DCC, was then applied with the formation of with the concomitant growth in the 13C NMR spectrum of the
product 4 (64% yield after chromatographic purification). 1H signal at 172 ppm corresponding to the carbonyl group.
1
NMR analysis confirmed the proposed structure. In particular,
In order to obtain the molecule containing both the BNCT
the resonance of the α-methylenic group shifted from and MRI moieties, derivative 6 was reacted with N-tert-butyl-
3.39 ppm in the amino derivative 7 to 4.06 ppm in the coup- DOTAMA-C6-NH2 using the two-step procedure previously
ling product.
described for the preparation of C-(2-benzyloxyethyl)-C′-[cho-
At this point the functionalization of the other arm of the o- lesten-5-en-3β-yloxy-(3-propionacetoamidomethyl)]-o-carborane
carborane has been tackled. As reported in Scheme 2, the (4) (Scheme 2). Carborane 6 was then activated via NHS in the
benzylic protecting group was selectively removed by Pd/C cata- presence of DCC and after filtration of DCU, reacted with the
lyzed hydrogenation.26 Different catalyst concentrations and suitable amine affording C-[N-tert-butyl(DOTAMA-C6-)-carba-
solvents were tested in order to avoid the concomitant moylmethyl]-C′-[cholesten-5-en-3β-yloxy-(3-propionacetoamido-
hydrogenation of the cholesterol double bond. It was found methyl)]-o-carborane (8) in 67% overall yield. The structure
that 5% of Pd/C in EtOH–CH2Cl2 2/1 provide the best con- was completely characterized and confirmed by ESI mass spec-
ditions, affording after 24 h the desired product 5 in 92% trometry with the detection of the characteristic ion at 1327.87
overall yield.
(M + H+). After removal of tert-butyl ester groups (CF3CO2H/
Scheme 2 Synthesis of C-[(DOTAMA-C6)carbamoylmethyl]-C’-[cholesten-5-en-3β-yloxy-(3-propionacetoamidomethyl)]-o-carborane (9). Reac-
tion conditions and yields: (a) H2, Pd/C, EtOH, CH2Cl2, rt (92%); (b) 3 M CrO3 in H2SO4, acetone, r.t. (80%); (c) N-hydroxysuccinimide, dicyclohexyl-
carbodiimide, i-PrEt2N, CH2Cl2, rt (42%); (d) CF3COOH, CH2Cl2, r.t., (95%).
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 2457–2467 | 2459

View Article Online
Paper
Organic & Biomolecular Chemistry
Scheme 3 Synthesis of Gd(III)-C-[N-(DOTAMA-C6)carbamoylmethyl]-
C’-2-[1,3-bis(hexadecyloxy)propan-2-yloxy]acetoamidomethyl-o-
caborane complex (10, Gd-B-AC01). (a) GdCl3, H2O/MeOH.
CH2Cl2), the Gd(III) complex 10 was finally obtained by adding
stoichiometric amounts of GdCl3 in an aqueous solution of
the ligand at pH 6.5 (Scheme 3).
Fig. 1 1/T1 1H NMRD profile (0.01–70 MHz, pH 7, and 25 °C) of Gd-
○
■
Liposome preparation and characterization
B-AC01 ( ) and of Gd-B-AC01/liposomes ( ). The profile was acquired
using a 0.32 mm Gd complex solution and normalized to the relaxivity
(mm−1 s−1).
Liposomes were prepared by means of the standard thin lipid
film hydration method27 using the following components:
POPC, CHOL, Gd-B-AC01, DSPE-PEG2000, and DSPE-PEG2000-
Folate. Three types of liposomes containing the same amount
of POPC (71%) and different amounts of cholesterol and Gd-
B-AC01, namely 15.5 : 7.5%, 13 : 10%, and 8 : 15%, respectively,
were prepared. In order to increase the stability and limit their
non-specific phagocytic uptake, all the liposomes contain 6%
of pegylated phospholipid (DSPE-PEG2000). The millimolar
relaxivity (21.5 MHz, 25 °C) of Gd-B-AC01 incorporated into the
recorded for an aqueous solution of Gd-B-AC01 (0.32 mM) and
compared with the profile obtained for Gd-B-AC01 embedded
in the liposome membrane (Fig. 1). Both profiles show a relax-
ivity hump at about 30 MHz, typical of slowly moving systems.
The relaxivity enhancement observed with respect to the free
complex is limited by the relatively long exchange time τM of
water directly coordinated to the metal ion (inner sphere
water) as previously shown for several Gd complexes bearing a
DOTA-monoamide ligand structure.33,34 The presence of a
relaxivity peak between 10 and 60 MHz of the Gd-B-AC01
aqueous solution confirms that the complex forms stable
micelles. A titration (T1 vs. [Gd-B-AC01]) indicated a critical
micelle concentration (cmc) lower than 15 μM. The linear
dependence of relaxation rates between 300 and 15 μM indi-
cates that over this concentration range the complex remains
in the micellar form. Lower concentrations are under the
detection limit of the relaxometric method.
liposome formulation was 15.3 0.7 (mmol L−1 −1 s−1. This
)
value is relatively high as a consequence of both the decrease
of the molecular mobility upon the insertion into the liposome
membrane and the relatively fast water exchange rate through
the phospholipid bilayer due to the presence in the formu-
lation of a high % of unsaturated phospholipids (POPC). The
average liposome hydrodynamic diameter was 125 10 nm for
all investigated formulations as detected by dynamic light scat-
tering measurements. In order to accumulate the Gd-B-AC01
loaded liposome selectively in the tumor cells, 1% of DSPE
phospholipid conjugated to a folate residue has been added to
the phospholipid formulation. Overexpression of folate recep-
tors (FRs) is well established in many cancer cells (ovary,
brain, kidney, breast and lung)28,29 thus identifying this recep-
tor as a potential target for many types of ligand directed
cancer therapeutics.30 FR may be further qualified as a tumor-
specific target, since it generally becomes accessible to intrave-
nous administered drugs only after malignant transformation.
Folate receptors (FRs) are normally expressed on cancer cells
up to 107 FRs per cell but the rate of receptor recycling is rela-
tively low.31,32 Folate conjugates bind to folate receptors with
high affinity (Kd ≈ 10−9 M) and they are internalized through
receptor-mediated endocytosis. The physicochemical pro-
perties of the liposome with 1% folate-phospholipid are very
similar to the untargeted control liposome (without folic acid).
Uptake experiments on IGROV-1 cells in vitro and MRI
analysis
Liposomes were tested in vitro on Human Ovarian Carcinoma
IGROV-1 cells to evaluate the amount of Gd taken up by cells
after 24 hours incubation by measuring the metal concen-
tration by ICP-MS analysis. These cells express a significant
level of FRα (0.5–1 × 106 receptors per cell).35 Liposome stabi-
lity was demonstrated by the absence of non-specific release of
Gd-B-AC01 to tumor cells. For this purpose, cells were incu-
bated, for 24 h at 37 °C, in the presence of non-targeted Gd-
B-AC01/liposomes containing increased % of Gd-B-AC01 from
7.5 to 15%. The results reported in Fig. 2A show that the
amount of Gd associated non-specifically with IGROV-1 cells is
directly proportional to the % loading of Gd-B-AC01 in the
liposome. This is due to the high affinity of cholesterol for cell
membranes that causes the translocation of a small but signifi-
NMRD profiles
1/T1 nuclear magnetic resonance dispersion (NMRD) profiles, cant % of the complex from the liposome to the cell mem-
over a range of Larmor frequencies 0.01 < MHz < 70, have been brane. By reducing the % of Gd-B-AC01 with respect to the
2460 | Org. Biomol. Chem., 2014, 12, 2457–2467
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 2 (A) Gd moles released to IGROV-1 cell membranes by non-targeted Gd-B-AC01/liposomes prepared with an increased amount (%) of
GdAC01 after 24 h incubation Gd-B-AC01 in the presence of cells. The amount of cell associated Gd has been determined by ICP-MS. (B) Internaliz-
○
■
ation of Gd chelates in IGROV-1 cells incubated in the presence of an increasing amount of non-targeted ( ) and folate-targeted ( ) liposomes for
24 hours at 37 °C.
other phospholipids, it is possible to reduce the number of
released complexes bound specifically to the cell surface that
can cause the decrease of the ratio of the internalized boron
by tumors with respect to the surrounding healthy cells. For
these reasons the formulation containing 7.5% of Gd-
B-AC01 has been selected to perform the FR targeting study on
FR expressing IGROV-1 cells. The Gd moles internalized by
IGROV-1 cells, measured by ICP-MS after incubation in the
presence of the increased concentration of folate-targeted lipo-
somes, were compared with those obtained using non-targeted
ones. Fig. 2B shows that by increasing the concentration of tar-
geted and non-targeted liposomes incubated with IGROV-1
Fig. 3 Weighted spin echo images, recorded at 1 (A) and 7 (B) T, of an
cells, the targeted liposomes quickly reach complete saturation
agar phantom containing IGROV-1 cells labeled with folate-targeted
and non-targeted Gd-B-AC01/liposomes: (a) non-incubated control
of their folate uptake. This behavior demonstrates a high
affinity of folate targeted liposomes for the FR receptor. More-
over, it indicates that the non-specific Gd-B-AC01 release from
non-targeted liposomes does not invalidate the comparison if
one works with relatively low liposome concentrations.
cells; (b) and (c) cells incubated for 24 h with non-targeted liposomes of
5 (b) and 10 (c) μm; cells incubated with folate targeted liposomes of
5 (d) and 10 (e) μm. Cells are pelleted at the bottom of glass capillaries.
In order to assess whether the amount of Gd internalized
in target cells is enough to permit MRI visualization, T1
weighted images of glass capillaries containing cellular pellets
obtained by incubating ca. 1 × 106 IGROV cells with increasing
amounts of targeted and non-targeted Gd-B-AC01/liposomes
were acquired at 1 and 7 T (Fig. 3). At 1 T, there is a dominant
effect of the long reorientational motion of the paramagnetic
macromolecule that causes a dramatic relaxation rate increase
with a consequent signal intensity (SI) enhancement increase
as shown in the NMRD profiles of Fig. 1. Clearly, the SI
enhancements of labeled cells measured at 1 T (SI = 80% after
5 μM Gd incubation) were markedly higher than those
measured at 7 T (SI = 45% after 5 μM Gd incubation). The
mean SI enhancement (%) of the target tissues was calculated
according to eqn (1):
The recorded SI of the targeted liposome in both cell lines
is higher with respect to non-targeted ones. This demonstrates
the specific accumulation of folate targeted liposomes in
tumor cells with respect to control liposomes. The amount of
internalized boron measured by ICP-MS was 38 and 9 ppm for
folate-targeted and non-targeted liposomes, respectively. The
former amount is significantly higher than the minimum
boron concentration necessary to perform BNCT. We can con-
clude that Gd-B-AC01 loaded liposomes are efficient carriers
for the delivery of MRI/BNCT probes to tumor cells.
BNCT treatment of IGROV-1 cells
Fig. 4 shows the percent of cells that survived the neutron
irradiation. Two cell controls have been used in this exper-
iment. The first did not receive any neutron irradiation or
boron and the second control was irradiated without boron.
The number of surviving cells was significantly lower when
SI % enhancement ¼ ððmean SIPOST ꢀ mean SIPREÞ=
ð1Þ
mean SIPREÞ ꢁ 100
where SIPRE and SIPOST correspond to the SI measured in cells had been irradiated after folate-targeted GdB10-AC01/
untreated or control cells (SIPRE)and in liposome treated cells liposome internalization whereas in the absence of boron the
(SIPOST), respectively.
effect of neutron irradiation is almost negligible. Furthermore,
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 2457–2467 | 2461

View Article Online
Paper
Organic & Biomolecular Chemistry
anti-tumor drugs (such as doxorubicin) in order to improve
the efficacy of the treatment.
Experimental procedures
General methods
Flasks and all equipment used for the generation and reaction
of moisture-sensitive compounds were dried using an electric
heater under Ar. THF was distilled from benzophenone ketyl,
anhydrous Et2O was distilled from LiAlH4 and anhydrous
CH2Cl2 from CaH2 prior to use. BuLi (1.6 M in hexanes) was
obtained from Aldrich. (Bmim)+Cl− was purchased from
Solvent Innovation GmbH. Decaborane was bought from
KATCHEM spol. s r.o. All commercially obtained reagents and
solvents were used as received. Products were purified by pre-
parative column chromatography on Macherey Nagel silica gel
for flash chromatography, 0.04–0.063 mm/230–400 mesh.
Reactions were monitored by TLC using silica gel on
TLC-PET foils Fluka, 2–25 mm, layer thickness 0.2 mm,
medium pore diameter 60 Å. Carboranes and their derivatives
were visualized on TLC plates using a 5% PdCl2 aqueous solu-
1
tion in HCl. H NMR spectra were recorded at 200 MHz, and
13C NMR spectra at 50.2 MHz. Data were reported as follows:
chemical shifts in ppm with tetramethylsilane as an internal
standard, integration, multiplicity (s = singlet, d = doublet, t =
triplet, q = quartet, dd = double–doublet, m = multiplet, br =
broad), coupling constants (Hz), and assignment. 13C NMR
spectra were measured with complete proton decoupling.
Chemical shifts were reported in ppm with the residual
solvent as an internal standard. GC-MS spectra were obtained
on a mass selective detector HP 5970 B instrument operating
at an ionizing voltage of 70 eV connected to a HP 5890 GC
with a cross-linked methyl silicone capillary column (25 m ×
0.2 mm × 0.33 mm film thickness). ESI MS spectra were
obtained on a Waters micromass ZQ spectrometer equipped
with an ESI ion source. IR spectra were recorded on a Perkin
Elmer BX FT-IR. C-(2-Benzyloxyethyl)-C′-aminomethyl-o-carbor-
ane (7) and N-tert-butylDOTAMA-C6-NH2 were synthesized as
described in the literature and their spectroscopic data corre-
spond to those reported.12
Fig. 4 (A) Number of IGROV-1 cells that survived the BNCT treatment;
(B) proliferation curve of IGROV-1 cells re-plated one day after the BNCT
treatment.
the proliferation rate of the boron treated cells that survived
the irradiation is significantly lower than both control cells.
Although the cells have not been completely killed, the
purpose of the experiments has been satisfied because these
measurements were meant to verify that the new formulations
had potentiality as vectors for BNCT. Further studies with dose
escalation are planned.
Conclusions
In summary, the dual MRI/BNCT probe described in this
(3β)-Cholest-5-en-3-methanesulfonate (1). Cholesterol (12.9
paper can be exploited to accumulate B and Gd in target mmol, 5 g) was dissolved in 100 mL of anhydrous Et2O and
tumor cells. The new formulation has shown to have relevant cooled to 0 °C in a 250 mL three-necked round bottom flask,
characteristics: (1) the ability to selectively concentrate high and then Et3N (1.5 equiv., 19.4 mmol, 2.7 mL) followed by
amounts of B in the tumor cells; (2) the possibility to quantify MsCl (19.4 mmol, 1.5 mL) were added. The reaction mixture
the boron concentration by MRI; (3) the cholesterol versatility was stirred for 1 h, quenched with H2O and then extracted
to be incorporated into different particles can be exploited to with Et2O (3 × 30 mL). The combined extracts were washed
perform a multi-step treatment based on the administration of with H2O (2 × 30 mL), dried and evaporated under reduced
different drugs targeted to different overexpressed receptors; pressure leaving 5.9 g (98%) of a pale yellow oil which was at
(4) the amount of internalized B is enough to perform an once used for the following reaction. Found C, 72.37; H, 10.40;
efficient BNCT treatment and the selective targeting of B using S, 6.92. Calc. for C28H48O3S: C, 72.36; H, 10.41; S, 6.90. νmax
folate-targeted liposomes is reported to reduce significantly (neat)/cm−1 2940, 1467, 1353, 1172; δH (200 MHz; CDCl3,
the uptake by healthy cells in the surrounding regions; (5) the Me4Si): 0.65 (3H, s, CH3CH(CH2)3), 0.70–2.05 (40H, m, CH3CH
use of liposomes as nanoplatforms for the delivery of Gd and (CH2)3, (CH3)2CH, CH3 ring, CH2 ring, CH ring), 2.50 (2H, m,
B agents will permit the simultaneous delivery of standard CH2CHvC), 3.03 (1H, m, CH3SO2), 4.50 (1H, m, CHO), 5.41
2462 | Org. Biomol. Chem., 2014, 12, 2457–2467
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
(1H, bd, CH2CHvC, J = 5.5 Hz); δc (50.2 MHz; CDCl3, Me4Si):
C-(2-Benzyloxyethyl)-C′-[cholesten-5-en-3β-yloxy-(3-propiona-
11.6 (CH3), 18.5 (CH), 18.9 (CH), 20.8 (CH2), 22.4 (CH), 22.6 cetoamidomethyl)]-o-carborane (4). In a 100 mL three-necked
(CH), 23.6 (CH2), 24.0 (CH2), 27.7 (CH3), 28.0 (CH2), 28.7 round bottom flask, 3-(cholesten-5-en-3β-yloxy)-propionic acid
(CH2), 31.5 (CH), 31.6(CH2), 35.5 (CH), 35.9 (CH2), 36.1 (CH2), (3) (1.27 mmol, 0.58 g) was dissolved in 40 mL of dry
36.6 (CH), 38.4 (CH2), 38.9 (CH2), 39.3 (CH2), 39.4 (CH2), 42.1 CH2Cl2, and then C-(benzyloxy)-ethyl-C′-aminomethyl-o-car-
(Cq), 49.7 (CH3), 55.9 (CH), 56.4 (CH), 81.7 (CH), 123.4 (Cq), borane (1.26 mmol, 0.38 g) and N,N′-dicyclohexylcarbodii-
138.5 (CH). m/z (ESI+) 465.33 [M + H]+.
mide (1.51 mmol, 0.31 g) were added at room temperature.
3-((3β)-Cholest-5-en-3-yloxy)-propanol (2). According to the The resulting mixture was stirred at room temperature over-
literature procedure,27 (3β)-cholest-5-en-3-methanesulfonate (1) night. The reaction mixture was filtered and was washed
(12.8 mmol, 5.93 g) was dissolved in 100 mL of anhydrous with 3% NaHCO3 (1 × 30 mL) and brine (1 × 30 mL). The
dioxane in a 250 mL three-necked round bottom flask. 1,3-Pro- organic layer was dried, and the solvent evaporated under
pandiol (319 mmol, 24.3 g) was added and the resulting reduced pressure. The crude was purified on silica gel (pet-
mixture was stirred overnight at 120 °C. Then the solution was roleum ether–diethyl ether 80/20) giving 0.7 g (73%) of a
cooled down at room temperature and the solvent was white solid.
removed under reduced pressure. The residue was dissolved in
M.p. = 76–79 °C. Found C, 67.45; H, 9.84; S. Calc. for
50 mL of CH2Cl2 and washed with H2O. The organic layer was C42H73B10NO3: C, 67.43; H, 9.83. νmax (neat)/cm−1 3332, 2581,
separated, washed with NaHCO3 (1 × 50 mL), water (1 × 1660; δH (200 MHz; CDCl3, Me4Si): 0.68 (3H, s, CH3CH(CH2)3),
50 mL), and brine (1 × 50 mL), dried over anhydrous K2CO3, 0.70–2.49 (40H + 10H, m, CH3CH(CH2)3, (CH3)2CH, CH3 ring,
and evaporated under reduce pressure. The crude was purified CH2 ring, CH ring, BH), 2.40 (2H, t, J = 5.2 Hz, CH2CH2CONH),
on silica gel (petroleum ether–diethyl ether 50/50) giving 2.69 (2H, t, J = 6.0 Hz, PhCH2OCH2CH2), 3.23 (1H, m, CHO),
1.54 g (27%) of a white solid. Found C, 81.05; H, 11.80. Calc. 3.68 (4H, m, PhCH2OCH2CH2, OCH2(CH2)2CONH), 4.07 (2H,
for C30H52O2: C, 81.02; H, 11.79. νmax (neat)/cm−1 3434, 2934, d, J = 6.4 Hz, CONHCH2), 4.51 (2H, s, PhCH2O), 5.38 (1H, bd,
2098, 1642; δH (200 MHz; CDCl3, Me4Si): 0.67 (3H, s, CH3CH J = 4.8 Hz, CH2CHvC), 7.31 (1H, t, J = 6.0 Hz, CONH), 7.39
(CH2)3), 0.70–2.49 (42H, m, CH3CH(CH2)3, (CH3)2CH, CH3 ring, (5H, bs, Ph); δc (50.2 MHz; CDCl3, Me4Si): 11.6 (CH3), 18.6
CH2 ring, CH ring), 2.60 (1H, bs, OH), 3.19 (1H, tt, J = 8.5, 5.2, (CH3), 19.2 (CH3), 20.9 (CH2), 22.4 (CH3), 22.7 (CH3), 23.7
CHO), 3.67 (2H, t, J = 5.8 Hz, HOCH2CH2O), 3.77 (2H, t, J = 5.4 (CH2), 24.1 (CH2), 27.8 (CH), 28.1 (CH2), 31.7 (CH), 31.8 (CH2),
Hz, HOCH2CH2O), 5.33 (1H, bd, J = 5.2 Hz, CH2CHvC); δc 35.1 (CH2), 35.6 (CH2), 36.0 (CH), 36.7 (CH2), 36.8 (Cq), 36.9
(50.2 MHz, CDCl3, Me4Si): 11.6 (CH3), 18.5 (CH3), 19.0 (CH3), (CH2), 38.8 (CH2), 39.3 (CH2), 39.6 (CH2), 41.2 (CH2), 42.1 (Cq),
20.8 (CH2), 22.3 (CH), 22.5 (CH), 23.8 (CH2), 24.0 (CH2), 27.6 50.0 (CH), 56.0 (CH), 56.6 (CH), 63.2 (CH2), 68.2 (CH2), 72.9
(CH), 28.0 (CH), 31.6 (CH2), 32.3 (CH2), 35.6 (CH), 36.0 (CH2), (CH2), 78.4 (Cq), 79.1 (Cq), 79.4 (CH), 121.9 (CH), 127.6 (CH),
36.4 (Cq), 37.0 (CH2), 38.8 (CH2), 39.2 (CH2), 39.5 (CH2), 42.0 127.7 (CH), 128.3 (CH), 137.3 (Cq), 140.1 (Cq), 171.2 (Cq); m/z
(Cq), 49.9 (CH), 56.0(CH), 56.4 (CH), 60.2 (CH2), 65.8 (CH2), (ESI+) 749.15 [M + H]+.
76.4 (CH), 121.3 (CH), 140.3 (Cq); m/z (ESI+) 467.70 [M + Na]+.
3-(Cholesten-5-en-3β-yloxy)-propionic acid (3). 3-((3β)- toamidomethyl)]-o-carborane (5). In a 50 mL two-necked
C-(Hydroxyethyl)-C′-[cholesten-5-en-3β-yloxy-(3-propionace-
Cholest-5-en-3-oxy)-propanol (2) (0.63 g, 1.41 mmol) was dis- round bottom flask, C-(2-benzyloxyethyl)-C′-[cholesten-5-en-
solved in 21 mL of acetone, and then a 3 M solution of CrO3 3β-yloxy-(3-propionacetoamidomethyl)]-o-carborane (4) (0.18 g,
(4 equiv., 5.6 mmol, 0.56 g) in H2SO4 (7 mL) was added drop- 0.24 mmol) was dissolved in 20 mL of a mixture of EtOH–
wise at 0 °C. The reaction mixture was stirred overnight at rt, CH2Cl2 (50–50), and then Pd/C (10%) was wetted with a few
and then quenched with H2O. The solvent was evaporated drops of water and added. The reaction mixture was stirred
under reduced pressure and the mixture was extracted with overnight at rt in a H2 saturated atmosphere, filtered and then
CH2Cl2 (5 × 10 mL); the organic layers were washed once with the solvent was evaporated under reduced pressure giving
brine (1 × 10 mL), dried and evaporated giving 0.49 g (76%) of 0.14 g (88%) of a white solid. M.p. = 103–106 °C. Found C,
a white solid. Found C, 78.57; H, 11.00. Calc. for C30H50O3: C, 63.90; H, 10.25; S. Calc. for C35H67B10NO3: C, 63.88; H, 10.26;
78.55; H, 10.99. νmax (neat)/cm−1 3400, 2938, 1715; δH νmax (neat)/cm−1 3433, 2583, 1655, 1587; δH (200 MHz; CDCl3,
(200 MHz; CDCl3, Me4Si): 0.68 (3H, s, CH3CH(CH2)3), Me4Si): 0.70 (3H, s, CH3CH(CH2)3), 0.80–2.40 (41H + 10H, m,
0.70–2.05 (38H, m, CH3CH(CH2)3, (CH3)2CH, CH3 ring, CH2 CH3CH(CH2)3, (CH3)2CH, CH3 ring, CH2 ring, CH ring, OH,
ring, CH ring), 2.05–2.50 (2H, m, CH2CH2COOH), 2.62 (2H, t, BH), 2.50 (2H, t, J = 5.0 Hz, HOCH2CH2), 2.63 (2H, t, J = 6.2 Hz,
J = 6.4 Hz, CH2COOH), 3.21 (1H, m, CHO), 3.76 (2H, t, J = CH2CH2CONH), 3.24 (1H, m, CHO), 3.73 (2H, t, J = 6.4 Hz,
6.4 Hz, CH2OH), 5.35 (1H, bd, J = 5.2 Hz, CH2CHvC); δc OCH2(CH2)2CONH), 3.80 (2H, t, J = 5.2 Hz, HOCH2CH2), 4.11
(50.2 MHz, CDCl3, Me4Si): 11.6 (CH3), 18.6 (CH3), 19.1 (CH3), (2H, d, J = 6.6 Hz, CONHCH2), 5.39 (1H, bs, CH2CHvC), 7.36
20.8 (CH2), 22.3 (CH3), 22.6 (CH3), 23.6 (CH2), 24.0 (CH2), 27.8 (1H, bt, CONH); δc (50.2 MHz; CDCl3, Me4Si): 11.6 (CH3), 18.5
(CH2), 28.0 (CH), 31.6 (CH2), 31.8 (CH), 35.0 (CH2), 35.6 (CH), (CH3), 19.2 (CH3), 20.9 (CH2), 22.4 (CH3), 22.6 (CH3), 23.6
36.0 (CH2), 36.6 (CH2), 36.9 (CH2), 38.7 (CH2), 39.4 (CH2), 39.5 (CH2), 24.1 (CH2), 27.8 (CH), 28.0 (CH2), 31.7 (CH), 31.8 (CH2),
(CH2), 42.1 (Cq), 49.9 (CH), 55.9 (CH), 56.6 (CH), 62.8 (CH2), 33.7 (CH), 35.6 (CH2), 36.0 (CH), 36.6 (CH2), 36.8 (Cq), 37.4
79.3 (CH), 121.5 (CH), 140.4 (Cq), 176.9 (Cq); m/z (ESI+) 459.72 (CH2), 38.7 (CH2), 39.3 (CH2), 39.5 (CH2), 41.4 (CH2), 42.1 (Cq),
[M + H]+.
49.9 (CH), 55.9 (CH), 60.8 (CH2), 63.1 (CH2), 78.1 (Cq), 79.0
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 2457–2467 | 2463

View Article Online
Paper
Organic & Biomolecular Chemistry
(Cq), 79.6 (CH), 122.0 (CH), 140.0 (Cq), 171.7 (Cq); m/z (ESI+)
C-[(DOTAMA-C6-)-carbamoylmethyl]-C′-[cholesten-5-en-3β-yloxy-
(3-propionacetoamidomethyl)]-o-carborane (9). In a 25 mL
659 [M + H]+.
[Cholesten-5-en-3β-yloxy-(3-propionacetoamidomethyl)-o-car- round bottom flask, 247 mg (0.186 mmol) of C-[N-tert-butyl
boranyl]-acetic acid (6). C-(Hydroxyethyl)-C′-[cholesten-5-en- (DOTAMA-C6)-carbamoylmethyl]-C′-[cholesten-5-en-3β-yloxy-(3-
3β-yloxy-(3-propionacetoamidomethyl)]-o-carborane (5) (0.48 g, propionacetoamidomethyl)]-o-carborane (8) were cooled to
0.73 mmol) was dissolved in 21 mL of acetone, and then a 3 M 0 °C, dissolved in 5 mL of a mixture of CF3COOH–CH2Cl2
solution of CrO3 (4 equiv., 2.9 mmol, 0.29 g) in H2SO4 (7 mL) (50–50) and stirred for 4 h at rt. After evaporation of
was added carefully at 0 °C. The reaction mixture was left over- CF3COOH–CH2Cl2, 204 mg (1.99 mmol) of viscous colorless oil
night at rt, and then quenched with H2O. The solvent was were obtained (95%). Found C, 59.21; H, 9.17; N, 8.49. Calc.
evaporated under reduced pressure and the mixture was for C57H105B10N7O2: C, 59.19; H, 9.15; N, 8.48; νmax (neat)/cm−1
extracted with CH2Cl2 (5 × 10 mL), the organic layers were 3390, 2931, 2581, 1741, 1676, 1586; δH (400 MHz; CD3SOCD3,
washed once with brine (1 × 10 mL), dried and evaporated Me4Si): 0.65 (3H, s, CH3CH(CH2)3), 0.80–2.40 (51H + 10H, m,
giving 0.42 g (85%) of pale yellow oil. Found C, 62.53; H, 9.76; CH3CH(CH2)3, (CH3)2CH, CH3 ring, CH2 ring, CH ring,
N, 2.09. Calc. for C35H65B10NO4: C, 62.56; H, 9.75; N, 2.08; νmax NHCH2CH2NH, NHCOCH2(CH2)4CH2NHCO, BH), 3.10–3.80
(neat)/cm−1 3316, 2932, 2613, 2577, 1739, 1586, 1552; δH (13H, m, CH2COOtBu, NHCOCH2(CH2)4CH2NHCO, CHO), 3.79
(200 MHz; CDCl3, Me4Si): 0.70 (3H, s, CH3CH(CH2)3), (2H, bt, OCH2(CH2)2CONH), 4.00 (2H, d,
J = 6.2 Hz,
0.80–2.45 (41H + 10H, m, CH3CH(CH2)3, (CH3)2CH, CH3 ring, CONHCH2), 5.30 (1H, bs, CH2CHvC), 7.30 (1H, m, NHCO),
CH2 ring, CH ring, BH), 2.63 (2H, t, J = 5.6 Hz, CH2CH2CONH), 8.60 (2H, m, 2 × NHCO); δc (100.4 MHz; CD3SOCD3, Me4Si):
3.24 (1H, m, CHO), 3.79 (2H, t, J = 5.6 Hz, OCH2(CH2)2CONH), 11.6 (CH3), 18.5 (CH3), 19.3 (CH3), 20.8 (CH2), 22.4 (CH3), 22.7
4.11 (2H, d, J = 4.4 Hz, CONHCH2), 5.39 (1H, bs, CH2CHvC δc (CH3), 23.1 (CH2), 23.8 (CH2), 24.7 (CH2), 28.4 (CH2), 28.4
(50.2 MHz; d6DMSO, Me4Si): 12.3 (CH3), 19.0 (CH3), 19.6 (CH3), (CH2), 28.5 (CH2), 29.5 (CH2), 31.6 (CH), 31.7 (CH2), 32.9
22.9 (CH2), 23.1 (CH3), 23.2 (CH3), 24.5 (CH2), 25.9 (CH2), 28.0 (CH2), 35.6 (CH), 36.0 (CH2), 36.9 (Cq), 37.1 (CH2), 38.1 (CH2),
(CH), 29.7 (CH2), 32.0 (CH), 33.9 (CH2), 35.8 (CH), 36.2 (CH2), 36.3 38.5 (CH2), 38.9 (CH2), 39.3 (CH2), 39.6 (CH2), 41.4 (CH2), 41.6
(CH), 36.9 (CH2), 37.0 (Cq), 39.0 (CH2), 40.6 (CH2), 41.6 (CH2), (CH2), 42.0 (2 × CH2), 46.0–54.0 (8 × CH2), 50.0 (CH), 55.4 (2 ×
42.4 (CH2), 42.7 (CH2), 43.1 (Cq), 48.1 (CH), 50.2 (CH), 56.2 (CH2), CH2), 56.0 (CH), 56.6 (CH), 57.0 (Cq), 63.8 (2 × CH2), 65.8 (Cq),
56.8 (CH2), 76.0 (Cq), 78.7 (Cq), 81.6 (CH), 121.5 (CH), 141.0 (Cq), 75.5 (Cq), 79.0 (2 × CH), 80.0 (Cq), 121.7 (CH), 140.7 (Cq),
168.9 (Cq); 171.3 (Cq); m/z (ESI+) 696 [M + Na + H]+.
C-[N-tert-Butyl(DOTAMA-C6-)-carbamoylmethyl]-C′-[choles- (Cq); m/z (ESI+) 1350 [M + Na + H]+.
166.3 (Cq), 171.1 (Cq), 171.3 (Cq), 171.4 (Cq), 171.9 (Cq), 172.2,
ten-5-en-3β-yloxy-(3-propionacetoamidomethyl)]-o-carborane
Gd(III)-C-[N-(DOTAMA-C6)carbamoylmethyl]-C′-2-[1,3-bis(hexa-
(8). According to the previous procedure described for com- decyloxy)propan-2-yloxy]acetoamidomethyl-o-caborane complex
pound
3,
[cholesten-5-en-3β-yloxy-(3-propionacetoamido- (10). In a 100 mL round bottom flask, 70 mg of C-
methyl)-o-carboranyl]-acetic acid (6) (0.2 g, 0.29 mmol) was [(DOTAMA-C6)-carbamoylmethyl]-C′-[cholesten-5-en-3β-yloxy-
reacted with N-tert-butylDOTAMA-C6-NH2 (0.19 g, 0.29 mmol). (3-propionacetoamidomethyl)]-o-carborane (9) (0.055 mmol)
The crude was purified on silica gel (eluant: CH2Cl2–MeOH and 1 equivalent of GdCl3 were dissolved in water at rt. The pH
96–4, then CH2Cl2–MeOH 80–20) affording 0.15 g (42%) of solution was checked and maintained at 6.5 with 1 M NaOH
pale yellow oil. Found C, 62.57; H, 9.82; N, 7.41. Calc. for aqueous solution. The solution was stirred for 24 h, and then
C69H129B10N7O10: C, 62.55; H, 9.81; N, 7.40; νmax (neat)/cm−1 the pH was adjusted to 8.5 with 1 M NaOH aqueous solution
3294, 2933, 2575, 1730, 1665, 1552; δH (200 MHz; CDCl3, and the mixture was stirred for 2 h, and then filtered with a
Me4Si): 0.70 (3H, s, CH3CH(CH2)3), 0.80–2.50 (78H + 10H, m, 0.2 mm syringe filter, the pH was adjusted to 7 with a 1 M HCl
CH3CH(CH2)3, (CH3)2CH, CH3 ring, CH2 ring, CH ring, aqueous solution and the solvent evaporated. Inorganic salts
NHCH2CH2NH, NHCOCH2(CH2)4CH2NHCO, BH), 1.50 (27H, s, were removed by gel filtration on a Sephadex® G10 column.
COOtBu),
3.10–3.80
(13H,
m,
CH2COOtBu, The solvent was removed affording 63 mg of the complex
NHCOCH2(CH2)4CH2NHCO, CHO), 3.79 (2H, bt, OCH2 (0.043 mmol, 40%). Found C, 46.68; H, 7.88; S. Calc. for
(CH2)2CONH), 4.11 (2H, d, J = 6.2 Hz, CONHCH2), 5.39 (1H, C12H24B10O2: C, 46.73; H, 7.84; νmax (neat)/cm−1 3329, 2923,
bd, CH2CHvC), 8.20 (1H, bs, NHCO), 8.70 (2H, m, 2 × NHCO); 2540, 1681, 1620, 1409, 1180; m/z (ESI+) 1451.90 [M + H]+.
δc (50.2 MHz; CDCl3, Me4Si): 11.6 (CH3), 18.5 (CH3), 19.3
Liposome preparation and characterization
(CH3), 20.9 (CH2), 22.4 (CH3), 22.7 (CH3), 23.7 (CH2), 24.1
(CH2), 25.2 (CH2), 27.8 (9 × CH3), 28.4 (CH2), 28.4 (CH2), 28.5 All the phospholipids used in the liposome preparation
(CH2), 29.5 (CH2), 31.6 (CH), 31.7 (CH2), 32.9 (CH2), 35.6 (CH), (POPC:
36.0 (CH2), 36.9 (Cq), 37.1 (CH2), 38.1 (CH2), 38.5 (CH2), 38.9 DSPE-PEG 2000 methoxy: 1,2-distearoyl-sn-glycero-3-phos-
(CH2), 39.3 (CH2), 39.6 (CH2), 41.4 (CH2), 41.6 (CH2), 42.0 (2 × phoethanolamine-N[methoxy(polyethylene glycol)-2000]
1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine,
CH2), 46.0–54.0 (8 × CH2), 50.0 (CH), 55.4 (2 × CH2), 56.0 (CH), ammonium salt, DSPE-PEG 2000-Folate {1,2-distearoyl-sn-
56.6 (CH), 57.0 (Cq), 63.8 (2 × CH2), 65.8 (Cq), 75.5 (Cq), 79.0 glycero-3-phosphoethanolamine-N-[folate(polyethylene glycol)-
(2 × CH), 80.0 (Cq), 81.7 (Cq), 81.8 (Cq), 121.7 (CH), 140.7 (Cq), 2000] ammonium salt} were purchased from Avanti Polar
166.3 (Cq), 171.1 (Cq), 171.3 (Cq), 171.4 (Cq), 171.9 (Cq), 172.2, Lipids (Alabaster, AL). Cholesterol (Chol) was purchased from
(Cq); m/z (ESI+) 1350 [M + Na + H]+.
Sigma-Aldrich (St. Louis, MO). Gd-B-AC01 was synthesized
2464 | Org. Biomol. Chem., 2014, 12, 2457–2467
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
according to the above procedure. Liposomes were prepared by detached with 0.02% EDTA (Lonza). For the in vitro uptake
following the thin lipid film hydration method.36 The follow- experiments, about 4.5 × 105 of IGROV were seeded in 6 cm
ing liposome formulations were prepared using POPC : Chol : diameter culture dishes. After 6 h, the medium was removed
Gd-B-AC01 : DSPE-PEG2000
Lipo-Ctrl 1: 71 : 8 : 15 : 6 : 0 molar ratio. Lipo-Ctrl 2: incubated for 24 h with Lipo-Folate or Lipo-Ctrl. At the end of
71 : 13 : 10 : 6 : 0 molar ratio. Lipo-Ctrl 3: incubation, cells were washed three times with 5 mL ice-cold
methoxy : DSPE-PEG2000-folate; and it was replaced with RPMI w/o folate. After 24 h, cells were
71 : 15.5 : 7.5 : 6 : 0 molar ratio. Liposomes with folate (Lipo- PBS, detached with 0.05% trypsin and 0.02% EDTA in PBS and
Folate) were prepared by adding 1% molar ratio of DSPE-PEG transferred into glass capillaries for MRI analysis (see below).
2000 Folate to all previous described formulations; moreover The Gd content of IGROV-1 cells was determined by induc-
DSPE-PEG 2000 was added in 5% molar ratio instead of 6%. tively coupled plasma mass spectrometry (ICP-MS) (Element-2;
Briefly, all lipids were dissolved in chloroform and the organic Thermo-Finnigan, Rodano, Milan, Italy). Sample digestion was
solution was slowly evaporated for removing the solvent until a performed with 2 mL of concentrated HNO3 (70%) under
thin film was formed. The film was then hydrated at 55 °C microwave heating (Milestone MicroSYNTH Microwave labsta-
with 1.5 mL of a saline buffer containing 5 mM Hepes and tion equipped with an optical fiber temperature control and
150 mM NaCl, pH = 7.4 (HBS). The suspension of multilamel- HPR-1000/6 M six position high pressure reactor, Bergamo,
lar vesicles (MLV) was extruded (Lipex extruder, Northern Italy). After digestion, sample volumes were brought to 3 mL
Lipids Inc., Canada) four times on a polycarbonate filter with with ultrapure water and samples were analyzed by ICP-MS. The
a pore diameter of 400 nm, four times on a 200 nm filter and protein concentration was determined from cell lysates by the
three times on a 100 nm filter. The final liposome suspension Bradford assay, using bovine serum albumin as a standard.
was dialyzed two times at 4 °C against a 100-fold volume of
MRI. MR images were acquired at 1 and 7 T field. MR
HBS. The amount of Gd-B-AC01 incorporated into the lipo- images at 1 T were acquired with an Aspect M2 High Perform-
some was determined by 1H NMR T1 measurements at ance MRI System (Aspect Magnet Technologies Ltd, Netanya,
21.5 MHz, 25 °C (Stelar Spinmaster, Mede, Italy) of the minera- Israel) consisting of a NdFeB magnet, equipped with a 35 mm
lized complex solution (in HCl 6 M at 120 °C for 16 h). The solenoid Tx/Tr coil of inner diameter 35 mm. This system is
hydrated mean diameter of liposomes was determined using a equipped with fast gradient coils (gradient strength, 450 mT m−1
Malvern dynamic light-scattering spectrophotometer (Malvern at 60 A; ramp time, 250 μs at 160 V) with a field homogeneity of
Instruments, Malvern, UK). All samples were analyzed at 25 °C 0.2–0.5 gauss. MR images at 7 T were acquired on a Bruker
in filtered (cutoff = 200 nm) HBS buffer (pH = 7). The polydis- Avance300 spectrometer equipped with a Micro 2.5 microimaging
persity index (PDI) for all the liposomes prepared in this work probe (Bruker BioSpin, Ettlingen, Germany) using two birdcage
was smaller than 0.2. The liposomes were used within two resonators with 30 and 10 mm inner diameters.
weeks from the preparation and stored at 4 °C under argon.
For recording MR images in vitro, cells were pelleted at the
1H relaxation rate measurements. The longitudinal water bottom of glass capillaries placed in a phantom embedded of
proton relaxation rates were measured on the Stelar Spinmas- high gelling agarose gel (1% in PBS). MR images were acquired
ter spectrometer (Stelar, Mede, Italy) operating at 21.5 MHz by using a standard T1 weighted multislice spin echo sequence,
means of the standard inversion recovery technique (16 exper- using the following parameters: TR/TE/NEX 250/4/6, resolution
iments, 2 scans). A typical 90° pulse width was 3.5 ms and the 78 μm, slice thickness 1 mm at 7 T and TR/TE/NEX 250/7/16,
reproducibility of the T1 data was 0.5%. The 1/T1 nuclear resolution 78 μm, slice thickness 1 mm at 1 T, respectively. At
magnetic relaxation dispersion (NMRD) profiles of water both field strengths a standard saturation recovery protocol
protons were measured over a continuum of magnetic field was used to measure T1 relaxation times.
strength from 0.00024 to 0.5 T (corresponding to 0.01–20 MHz
proton Larmor frequency) on the fast field-cycling relaxometer
Cell irradiation
(Stelar Spinmaster FFC 2000; Stelar) equipped with a silver Seven flasks, four with IGROV-1 cells previously incubated for
magnet. The relaxometer operates under complete computer 24 h in the presence of folate targeted GdAC01/liposomes, at
control with an absolute uncertainty in the 1/T1 values of 1%. 10 μM Gd concentration, and three with non-treated control
The typical field sequences were used: the non-polarized cells were irradiated in the thermal column of the TRIGA Mark
sequence between 40 and 8 MHz and the pre-polarized II reactor at the University of Pavia, Italy. The cells cultured in
sequence between 8 and 0.01 MHz. The observation field was the presence of liposomes were washed with cold PBS before
set at 13 MHz. Sixteen experiments of two scans each were irradiation. At the end of the irradiation, the medium was
used for the T1 determination for each field.
removed, it was replaced with RPMI and flasks were placed at
Cell culture and uptake experiments. Human ovarian carci- 37 °C in a humidified atmosphere of 5% CO2.
noma cell line (IGROV-1) was kindly provided by Dr Claudia
The irradiation position had been previously characterized
Cabella (Bracco Imaging, Colleretto Giacosa TO). IGROV-1 cells from the point of view of neutron flux distribution by means
were cultured in RPMI (Lonza) supplemented with 10% (v/v) of thin activation foils.37 At a reactor power of 250 kW the
FBS, 2 mM glutamine, 100 U mL−1 penicillin, and 100 U mL−1 thermal neutron flux in air at that position is (1.17 0.03) ×
streptomycin. Cells were incubated at 37 °C in a humidified 1010 cm−2 s−1, while the epithermal and fast components are
atmosphere of 5% CO2. At 80% confluence, cells were at least two orders of magnitude lower. The flux is roughly
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 2457–2467 | 2465

View Article Online
Paper
Organic & Biomolecular Chemistry
constant (less than 1%) along the vertical direction, thus the
flasks were superposed and irradiated at the same time. The
irradiation time has been fixed at 15 minutes at a reactor
power of 30 kW, corresponding to a thermal neutron fluence
of 1.26 × 1012 cm−2. The radiation dose absorbed by cells
treated with 10 μM Gd was 6.2 Gy and the dose absorbed by
control cells was 0.6 Gy.
8 J. F. Valliant, K. J. Guenther, A. S. King, P. Morel,
P. Schaffer, O. O. Sogbein and K. A. Stephenson, Coord.
Chem. Rev., 2002, 232, 173–230.
9 G. Wu, R. F. Barth, W. Yang, R. J. Lee, W. Tjarks,
M. V. Backer and J. M. Backer, Anti-Cancer Agents Med.
Chem., 2006, 6, 167–184.
10 E. L. Crossley, E. J. Ziolkowski, J. A. Coderre and
L. M. Rendina, Mini Rev. Med. Chem., 2007, 7, 303–313.
Proliferation assay. The day after irradiation cells were
detached with 0.02% EDTA and the trypan blue exclusion test 11 I. Viñas and C. Teixidor, Future Med. Chem., 2013, 5, 617–
of cell viability was performed. Then, around 6 × 105 IGROV-1
619.
cells from each differently treated flask were seeded in 10 cm 12 S. Aime, A. Barge, A. Crivello, A. Deagostino, R. Gobetto,
diameter culture dishes. After 1, 2, 3 and 4 days, cells were
washed with PBS, detached with 0.05% trypsin and 0.02%
C. Nervi, C. Prandi, A. Toppino and P. Venturello, Org.
Biomol. Chem., 2008, 6, 4460–4466.
EDTA in PBS and transferred into falcon tubes. Then, cells 13 A. Toppino, M. E. Bova, S. G. Crich, D. Alberti, E. Diana,
were sonicated for 30″ at 30% power in ice and the protein
concentration from cell lysates was determined by the Brad-
ford method, using bovine serum albumin as a standard.
A. Barge, S. Aime, P. Venturello and A. Deagostino, Chem-
istry, 2013, 19, 720–727.
14 P. J. Kueffer, C. A. Maitz, A. A. Khan, S. A. Schuster,
N. I. Shlyakhtina, S. S. Jalisatgi, J. D. Brockman, D. W. Nigg
and M. F. Hawthorne, Proc. Natl. Acad. Sci. U. S. A., 2013,
110, 6512–6517.
Conflict of interest
15 H. Nakamura, in Methods Enzym., ed. N. Duzgunes, 2009,
pp. 179–208.
The authors declare no competing financial interest.
16 S. Altieri, M. Balzi, S. Bortolussi, P. Bruschi, L. Ciani,
A. M. Clerici, P. Faraoni, C. Ferrari, M. A. Gadan, L. Panza,
D. Pietrangeli, G. Ricciardi and S. Ristori, J. Med. Chem.,
2009, 52, 7829–7835.
Acknowledgements
This research was performed in the framework of the EU COST 17 V. P. Torchilin, Adv. Drug Delivery Rev., 2008, 60, 548–558.
Action TD1004, and supported by MIUR (PRIN 2009235JB7), 18 A. P. Pathak, D. Artemov, B. D. Ward, D. G. Jackson,
by the University of Torino (code D15E11001710003 project:
Innovative Nanosized Theranostic Agents), by the University of
M. Meeman and Z. M. Bhujwalla, Cancer Res., 2005, 65,
1425–1432.
Genova (Progetto San Paolo; Title: Validazione di molecole per 19 M. R. Dreher, W. G. Liu, C. R. Michelich, M. W. Dewhirst,
il rilascio tumore specifico di farmaci e la valutazione contest-
uale della risposta mediante imaging funzionale), and by Con-
F. Yuan and A. Chilkoti, J. Natl. Cancer Inst., 2006, 98, 335–
344.
sorzio Interuniversitario di Ricerca in Chimica dei Metalli dei 20 H. Koganei, M. Ueno, S. Tachikawa, L. Tasaki, H. S. Ban,
Sistemi Biologici (CIRCMSB). We thank Franco Fedeli for the
synthesis of N-tert-butylDOTAMA-C6-NH2.
M. Suzuki, K. Shiraishi, K. Kawano, M. Yokoyama,
Y. Maitani, K. Ono and H. Nakamura, Bioconjugate Chem.,
2013, 24, 124–132.
21 H. Nakamura, Future Med. Chem., 2013, 5, 715–730.
22 M. Bialek-Pietras, A. B. Olejniczak, S. Tachikawa,
H. Nakamura and Z. J. Lesnikowski, Bioorg. Med. Chem.,
2013, 21, 1136–1142.
References
1 R. F. Barth, J. Neuroonc., 2003, 62, 1–5.
2 R. F. Barth, J. A. Coderre, M. G. H. Vicente and T. E. Blue, 23 G. L. Pan, S. Oie and D. R. Lu, Pharm. Res., 2004, 21, 1257–
Clin. Cancer Res., 2005, 11, 3987–4002. 1262.
3 M. F. Hawthorne and M. W. Lee, J. Neuroonc., 2003, 62, 33– 24 N. Dewi, et al., Biomed. Pharmacother., 2013, 67, 451–457.
45. 25 Z. J. Kaminski, Synthesis, 1987, 917–920.
4 J. Laakso, M. Kulvik, I. Ruokonen, J. Vahatalo, R. Zilliacus, 26 J. Schimmel, M. I. P. Eleuterio, G. Ritter and R. R. Schmidt,
M. Farkkila and M. Kallio, Clin. Chem., 2001, 47, 1796–
1803.
5 M. Kortesniemi, T. Seppala, I. Auterinen and S. Savolainen,
Appl. Radiat. Isot., 2004, 61, 823–827.
Eur. J. Org. Chem., 2006, 1701–1721.
27 A. Bajaj, P. Kondaiah and S. Bhattacharya, Bioconjugate
Chem., 2008, 19, 1640–1651.
28 E. I. Sega and P. S. Low, Cancer Metast. Rev., 2008, 27, 655–
664.
6 S. Geninatti-Crich, A. Deagostino, A. Toppino, D. Alberti,
P. Venturello and S. Aime, Anti-Cancer Agents Med. Chem., 29 W. Xia and P. S. Low, J. Med. Chem., 2010, 53, 6811–6824.
2012, 12, 543–553.
7 A. H. Soloway, W. Tjarks, B. A. Barnum, F. G. Rong,
30 Y. J. Lu and P. S. Low, Adv. Drug Delivery Rev., 2002, 64,
342–352.
R. F. Barth, I. M. Codogni and J. G. Wilson, Chem. Rev., 31 B. A. Kamen and A. K. Smith, Adv. Drug Delivery Rev., 2004,
1998, 98, 1515–1562.
56, 1085–1097.
2466 | Org. Biomol. Chem., 2014, 12, 2457–2467
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
32 C. M. Paulos, J. A. Reddy, C. P. Leamon, M. J. Turk and 35 N. Kamaly, T. Kalber, M. Thanou, J. D. Bell and
P. S. Low, Mol. Pharm., 2004, 66, 1406–1414. A. D. Miller, Bioconjugate Chem., 2009, 20, 648–655.
33 L. Tei, A. Barge, S. G. Crich, R. Pagliarin, V. Negri, 36 F. Olson, C. A. Hunt, F. C. Szoka, W. J. Vail and
D. Ramella, G. Cravotto and S. Aime, Chemistry, 2010, 16,
8080–8087.
D. Papahadjopoulos, Biochim. Biophys. Acta, 1979, 557, 9–
23.
34 C. Grange, S. Geninatti-Crich, G. Esposito, D. Alberti, 37 N. Protti, S. Bortolussi, M. I. M. Prata, P. Bruschi, S. Altieri
L. Tei, B. Bussolati, S. Aime and G. Camussi, Cancer Res.,
2010, 70, 2180–2190.
and D. W. Nigg, Trans. Am. Nucl. Soc., 2012, 107, 1269–
1272.
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 2457–2467 | 2467