Towards new boron carriers for boron neutron capture therapy: Metallacarboranes and their nucleoside conjugates

Article abstract of DOI:10.1016/j.bmc.2005.04.042

Thymidine conjugates containing metallacarborane, {8-[5-(N ³-thymidine)-3-oxa-pentoxy]-3-cobalt bis(1,2-dicarbollide)} ^- (5) and {8-[5-(O⁴-thymidine)-3-oxa-pentoxy]-3-cobalt bis(1,2-dicarbollide)}^- (6) ions and

Full text of DOI:10.1016/j.bmc.2005.04.042

Bioorganic & Medicinal Chemistry 13 (2005) 4168–4175
Towards new boron carriers for boron neutron capture therapy:
metallacarboranes and their nucleoside conjugates
a,
Zbigniew J. Lesnikowski, Edyta Paradowska, Agnieszka B. Olejniczak,
a
a
*
´
Mirosława Studzinska, Petra Seekamp, Uw Schußler, Detlef Gabel,
¨
a
b
b
b
´
e
Raymond F. Schinazic,d and Jaromir Plesek
ˇ
aCenter of Medical Biology, Laboratory of Molecular Virology and Biological Chemistry, Polish Academy of Sciences,
106 Lodowa Street, Lodz 93-232, Poland
bDepartment of Chemistry, University of Bremen, D-28334 Bremen, Germany
cVeterans Affairs Medical Center, Decatur, GA 30033, USA
dLaboratory of Biochemical Pharmacology, Department of Pediatrics, Emory University School of Medicine, Atlanta, GA 30322, USA
eInstitute of Inorganic Chemistry, Academy of Sciences of the Czech Republic, 250-68 Rezˇ, Czech Republic
Received 23 June 2004; accepted 15 April 2005
Available online 5 May 2005
Abstract—Thymidine conjugates containing metallacarborane, {8-[5-(N3-thymidine)-3-oxa-pentoxy]-3-cobalt bis(1,2-dicarbol-
lide)}ꢀ (5) and {8-[5-(O4-thymidine)-3-oxa-pentoxy]-3-cobalt bis(1,2-dicarbollide)}ꢀ (6) ions and several simple [3-cobalt bis(1,2-
dicarbollide)]ꢀ ion (1) derivatives have been studied as potential boron carriers for BNCT. Compound 6 and some nonnucleoside
derivatives of 1 were not toxic above 100 lM. The partition coefficient for both metallacarborane bearing thymidine conjugates 5
and 6 was more than 500 times higher than that of unmodified nucleoside. The cellular uptake studies showed accumulation of com-
pounds 6 in V79 Chinese hamster cells but not of compound 5. The low toxicity of conjugate type of 6 together with its high par-
tition coefficient suggest that judicially designed derivatives of metallacarboranes can be considered as potential boron carriers for
BNCT.
ꢀ 2005 Elsevier Ltd. All rights reserved.
1. Introduction
together with high linear energy transfer (LET) makes
these particles highly lethal to the cells. It is assumed,
that preferential neutron capture of 10B-containing,
tumor selective boron carrier in tumor tissue as com-
pared to healthy cells, due to higher accumulation of
the carrier in the tumor, will cause preferential killing
of cancer cells and induce therapeutic effect (Fig. 1).
Boron neutron capture therapy (BNCT) is a binary sys-
tem for treatment of cancers, based on absorption of
low-energy neutrons by nonradioactive boron-10 (10B)
atoms delivered to neoplastic cells in the form of a boron
carrying drug.1–3 The two independent, separately non-
lethal constituents of this binary modality are radiosen-
sitiser (boron carrier) containing 10B and nonionising
neutron radiation. Capture of neutron by 10B, a stable
isotope, results in the formation of excited boron-11
(11B). 11B is unstable and instantly fissions yielding high
linear energy transfer (LET) lithium-7 (7Li) and ener-
For BNCT to be successful, 109 10B atoms must be local-
ised on or preferably within neoplastic cells (20–40 lg of
B in one gram of tumor), and a sufficient number of ther-
mal neutrons must be delivered to sustain a lethal 10B
(n, alpha) lithium-7 reaction. Therefore, the development
of highly boron loaded, tumor-selective drugs play an
important role if BNCT is to evolve into clinically ac-
cepted treatment for cancer. A number of potential boron
carriers such as boron-containing porphyrins, amino
acids, carbohydrates, nucleic acid bases, nucleosides,
polyamines, DNA groove binders, low-density lipopro-
teins, biopolymers such as peptides, oligophosphates and
oligonucleotides have been synthesised and tested.4–6
The persistent interest in the design and synthesis of
getic a-particles (4He). The kinetic energy of Li and
a-particles is about 2.8 million electron volts (eV) of
energy (100 million times more than was put in). This,
7
Keywords: BNCT; Boron carriers; Metallacarboranes; Nucleoside
conjugates.
*
Corresponding author. Tel.: +48 42 677 12 49; fax: +48 42 677 12
30; e-mail: zlesnik@cbm.pan.pl
0968-0896/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2005.04.042

´
Z. J. Lesnikowski et al. / Bioorg. Med. Chem. 13 (2005) 4168–4175
4169
Figure 1. Theoretical mechanism of BNCT of tumors.
boron-containing nucleosides is that such compounds
may be selectively accumulated in rapidly multiplying
tumor cells, and following their conversion to the corre-
sponding nucleotides, trapped within the cell or ideally,
incorporated into nuclear DNA of tumors.7–11 Regretta-
bly, in spite of the continuous efforts to develop suitable
boron caring drugs the clinical applications of boron
derivatives are currently limited to two molecules, L-4-
(dihydroxyboryl)phenylalanine (BPA) and the sodium
salt of thioborane anion (Na2B12H11SH, BSH).
Recently, we developed the first method for the synthesis
of metallacarborane/nucleoside conjugates and showed
their incorporation into DNA–oligonucleotide.23 Herein,
we present preliminary studies on cytotoxicity of two
metallacarborane containing nucleosides, {8-[5-(N3-thy-
midine)-3-oxa-pentoxy]-3-cobalt bis(1,2-dicarbollide)}ꢀ
(5) and {8-[5-(O4-thymidine)-3-oxa-pentoxy]-3-cobalt
bis(1,2-dicarbollide)}ꢀ (6) ions (Fig. 2); nonnucleoside
part of the above conjugate, [8-(5-hydroxy-3-oxa-pent-
oxy)-3-cobalt bis(1,2-dicarbollide)]ꢀ (7), and compare it
with natural thymidine (3). The partition coefficient of
all the compounds has also been measured. The cytotoxi-
city of several other simple derivatives of [3-cobalt
bis(1,2-dicarbollide)]ꢀ ion (1) (Fig. 3) has also been
determined. In addition, preliminary studies on cellular
uptake of two examples metallacarborane/nucleoside
conjugates 5 and 6 have been performed.
2ꢀ
Boric acid [B(OH)3] and borane clusters (closo-B12H12
or C2B10H12) are the two main types of boron entities
used so far in the synthesis of carrier molecules for
BNCT. To a lesser extent boranes (BHꢀ) and cyano-
boranes (BH2CNꢀ) are used for synthesis of potential
boron caring drugs.12–15 Herein, we propose a new
modifying entity for preparation of potential boron
delivering molecules, namely the [8-substituted 3-cobalt
bis(1,2-dicarbollides)]ꢀ ion and related species.
2. Results
Metallacarboranes, discovered by Hawthorne and co-
workers16–18 represent a vast family of metallocene type
complexes, consisting of deltahedral heteroborane li-
gand(s) and one or more metal atoms such as Co, Fe,
Ni and a variety of others. Most prominent among them
seems to be the [3-cobalt bis(1,2-dicarbollide)]ꢀ ion (1)
[3-Co-(1,2-C2B9H11)2]ꢀ showing remarkable integrity
along with an extensive substitution chemistry reviewed
recently.19 But there are other classes of metallacarbor-
anes as well.20–22 Metallacarboranes, if derived from
large carboranes contain about 1.5 times as much boron
as BSH and 18 times more boron atoms than boric acid.
They are characterized by high lipophilicity, the prop-
erty which may facilitate transport of the carrier mole-
cules across the blood–brain-barrier (BBB) and
improve cellular uptake. The ability of deltahedral li-
gands to form stable complexes with metals allows
incorporation of various metals with different properties
into boron carrier molecule and provides an additional
advantage to high boron load.
The nomenclature of deltahedral sandwich compounds of
the metallacarborane type is very complex and cumber-
some. In this article we will exploit the versatility of the
semi-trivial nomenclature for commo-metalla-bis(icosa-
hedral) sandwich species, coined by the discoverer of this
huge realm of sandwich complexes.16–18 The fundamental
ion [3-Co-(1,2-C2B9H11)2]ꢀ is designated as [3-cobalt
bis(1,2-dicarbollide)]ꢀ (1) (Fig. 2, along with the number-
ing scheme) and its descendants will be considered as its
substitution derivatives (also shown in the figures).
In the respective conjugate acids the solvated proton is
associated with the most basic moiety, that is, usually
the –O–B(8) oxygen. These names will be used through-
out this article and, as much as possible, the respective
code numbers presented in the above list will be pre-
ferred for the sake of brevity and clarity.
{8-[5-(N3-Thymidine)-3-oxa-pentoxy]-3-cobalt bis(1,2-
dicarbollide)}ꢀ (5) and {8-[5-(O4-thymidine)-3-oxa-pento-

´
Z. J. Lesnikowski et al. / Bioorg. Med. Chem. 13 (2005) 4168–4175
4170
Figure 2. Synthesis of [8-dioxane-3-cobalt bis(dicarbollide)]0 (2) and {8-[5-(N3-thymidine)-3-oxa-pentoxy]-3-cobalt bis(1,2-dicarbollide)}ꢀ (5) and {8-
[5-(O4-thymidine)-3-oxa-pentoxy]-3-cobalt bis(1,2-dicarbollide)}ꢀ (6) ions.
xy]-3-cobalt bis(1,2-dicarbollide)}ꢀ (6) ions were
prepared by the alkylation of thymidine at 3-N- and
4-O-position23 with [8-dioxane-3-cobalt bis(dicarbol-
lide)]0 = 8-O(CH2CH2)2O-1,2-C2B9H10-3-Co-10,20-C2B9H11
(2).15 Briefly, alkylation of thymine in fully protected
nucleoside, 50-O-monomethoxytrityl-30-O-acetylthymi-
dyne (4), was performed via activation of 4-O (and com-
petitively 3-N) using sodium hydride followed by 2. The
fully protected products of 3-N and 4-O alkylation were
separated by chromatography on silica gel. The
deacetylation of 30-OH group yielded {8-[5-(N3-50-O-
monomethoxytritylthymidine)-3-oxa-pentoxy]-3-cobalt
bis(1,2-dicarbollide)}ꢀ and {8-[5-(O4-50-O-monometh-
oxytritylthymidine)-3-oxa-pentoxy]-3-cobalt bis(1,2-di-
carbollide)}ꢀ ions, respectively. The 30-deprotected
derivative was detritylated under acidic conditions
leading to deprotected metallacarborane/nucleoside con-
jugate 5 and 6.
ether in vacuo and drying of the residue to a constant
weight yielded 7 quantitatively.25 For the biological
studies 7 obtained as above was additionally purified
by RP-HPLC (Econosil RP C18, 5 lm, 4.7 · 250 mm)
using triethylammonium carbonate buffer (0.1 M,
pH 7.0) containing 60% of acetonitrile. [8-Hydroxy-
3-cobalt bis(dicarbollide)]ꢀ ion (8) was prepared as
described by reductive acetoxylation of [3-cobalt
bis(1,2-dicarbollide)]Cs (1) followed by hydrolysis of
the intermediate26 and [8-phosphate-3-cobalt bis(dicarbo-
llide)]ꢀ ion (9) was prepared in the reaction of 8 with
phosphorus oxychloride followed by hydrolysis of the
obtained intermediate.27 [8-(5-Phosphonate-3-oxa-pent-
oxy)-3-cobalt bis(1,2-dicarbollide)]ꢀ (10) was prepared
in the nucleophilic dioxane ring opening in 2 with di-
methylphosphite anion followed by hydrolysis of the
obtained dimethyl ester of phosphonic acid.28
Cytotoxicity was established by measurement of 50%
inhibition of cell growth (CC50, median cytotoxic
concentration) using plaque technique and MTT stain-
ing.29–32 It was found that {8-[5-(O4-thymidine)-3-oxa-
pentoxy]-3-cobalt bis(1,2-dicarbollide)}ꢀ (6) was low in
toxicity and had a CC50 above 100 lM, {8-[5-(N3-thymi-
dine)-3-oxa-pentoxy]-3-cobalt bis(1,2-dicarbollide)}ꢀ (5)
is characterised by CC50 43–100 lM depending upon the
cell system used. It is worthy to point out that extra-
polation of the dose-dependent cytotoxicity curves for
[8-Dioxane-3-cobalt bis(dicarbollide)]0 (2) was obtained
in the reaction of [3-cobalt bis(1,2-dicarbollide)]Cs (1)16
with dioxane in the presence of dimethyl sulfate (Me2
SO4, DMS) and sulfuric acid.24 [8-(5-Hydroxy-3-oxa-
pentoxy)-3-cobalt bis(1,2-dicarbollide)]ꢀ (7) was pre-
pared by heating of 2 with an excess of NaOH in
THF/H2O mixture with stirring at 60 ꢁC. Then, THF
was removed under vacuum and next 7 was extracted
as the sodium salt into diethyl ether. Removal of diethyl

´
Z. J. Lesnikowski et al. / Bioorg. Med. Chem. 13 (2005) 4168–4175
4171
oxa-pentoxy)-linker (7) has been characterised by
partition coefficient 57.04 11.90 and 46.48 8.19,
respectively.
The cellular uptake experiments showed that com-
pounds 5 and 9 gave boron concentrations within the
range of the blank, 31 and 27 pg/106 cells, respectively
(the blank level of boron was 29 pg/106 cells). For com-
pound 6, accumulation could be confirmed, although at
a low level, 60 pg/106 cells.
3. Discussion
Despite its attractiveness as a modality for treatment of
high grade brain tumors and continuous improvements,
BNCT still remains in the stage of clinical experi-
mentation. Conceptually it is the ultimate optimisation
of an effective agent (high LET radiation), which aims
to deliver selectively to tumor cells. Therefore, to evolve
into clinically accepted treatment for cancer both com-
ponents of binary BNCT system, boron carrier and neu-
tron beam must be considered and optimised. The
developments in epithermal neutron beam technology
from reactor sources have reached a high level of sophis-
tication, and nonreactor epithermal neutron sources can
be made accessible with the availability of additional re-
search and adequate funding in future. However, the
obstacles ahead to assure adequate boron uptake into
tumor cells are still formidable. This in turn, increases
the expectations for a breakthrough in the area of boron
delivering drugs. It seems that both currently used BSH
and BPA, and other boron modified compounds may
not fulfill these expectations. This prompted us to initi-
ate study on metallacarborane derivatives as potential
boron carriers for BNCT. Similar idea was conceived re-
cently by Bregadze et al.35
Figure 3. Nonnucleoside derivatives of [3-cobalt bis(1,2-dicarbollide)]ꢀ
ion (1) substituted at boron-8: [8-(5-hydroxy-3-oxa-pentoxy)-3-cobalt
bis(1,2-dicarbollide)]ꢀ (7), [8-hydroxy-3-cobalt bis(dicarbollide)]ꢀ ion
(8), [8-phosphate-3-cobalt bis(dicarbollide)]ꢀ ion (9) and [8-(5-phos-
phonate-3-oxa-pentoxy)-3-cobalt bis(1,2-dicarbollide)]ꢀ (10).
compound 6 shown in Figure 4 suggests that this com-
pound is not toxic also at concentrations much higher
than 100 lM.
Simple derivatives of [3-cobalt bis(1,2-dicarbollide)]ꢀ
ion (1) tested in the present study can be divided into
two types, one containing neutral ligands such as diox-
ane (2), (5-hydroxy-3-oxa-pentoxy)- (7) or OH- (8) at-
tached to boron-8 of metallacarborane system, and a
second type, containing ionisable phosphate (9) or phos-
phonate (10) group (Fig. 3). Compound 2 was nontoxic
in PBM and Vero cells showing the CC50 above 100 lM
in both cases but toxic in CEM cells (CC50 = 10 lM).
Compound 7 was moderately toxic or toxic in Vero,
MRC-5, PBM and CEM cells showing CC50 of 68, 42,
14 and 3.3 lM, respectively, but was nontoxic in A549
cell line (CC50 above 100 lM). Compounds 8 and 10
did not show toxicity in PBM and Vero cells above
100 lM concentration, and were moderately toxic in
CEM cells (CC50 = 60–88 lM). Compound 9 was non-
toxic in all three cell lines, PBM, CEM and Vero cells
at concentrations higher than 100 lM (Fig. 4, Table 1).
To meet the criteria of effectiveness, there must be a sig-
nificant concentration of boron delivered into tumor tis-
sues by boron caring drug. This requires a high
concentration of boron carrier in blood; therefore, the
low toxicity is an important criterion and first level of
selection of potential boron carriers.
The study on anti-tumor activity of metallacarboranes
was originally initiated more than 20 years ago. Toxic-
ity and pharmacokinetics of potassium salt of [3-cobalt
bis(1,2-dicarbollide)]ꢀ ion (1) in tumor bearing rats36,37
and mutagenicity in Salmonella typhimurium and Dro-
sophila melanogaster were tested.38 Compound 1 was
found cytotoxic, but not mutagenic in the system used.
The in vitro anti-tumor activity of a tin-meta-carborane
derivative {[(1,7-C2B10H11-1-COO)Bu2Sn]2}O2 has
been described.39 Anti-tumor activity of metallacarbor-
anes other than based on dicarba-closo-dodecaboranes
has also been described.22,40 In vitro cytotoxicity of
these complexes against selected human tumor cell
lines ranged from moderate to high. Detected cytotox-
icity of the above compounds makes some of them can-
didates for further studies as potential anti-tumor
agents, but also excludes them as boron carriers for
BNCT.
Partition coefficient for compounds 3, 5–7 and CDU
was established.33,34 We found that the partition coeffi-
cient value for nucleoside conjugates 5 and 6, bearing
metallacarborane complex increased markedly as com-
pared to natural thymidine (3), and was 42.65 5.71,
43.13 8.97 and 0.08 0.02, respectively. For compari-
son, CDU and metallacarborane bearing (5-hydroxy-3-

´
Z. J. Lesnikowski et al. / Bioorg. Med. Chem. 13 (2005) 4168–4175
4172
Figure 4. Dose-dependent cytotoxicity curves in MRC-5, A549 and Vero cells for thymidine (3), [8-(5-hydroxy-3-oxa-pentoxy)-3-cobalt bis(1,2-
dicarbollide)]ꢀ (7) and their conjugates {8-[5-(N3-thymidine)-3-oxa-pentoxy]-3-cobalt bis(1,2-dicarbollide)}ꢀ (5) and {8-[5-(O4-thymidine)-3-oxa-
pentoxy]-3-cobalt bis(1,2-dicarbollide)}ꢀ (6).
In contrast, we have found that conjugates of metalla-
carborane complex type of [3-cobalt bis(1,2-dicarbol-
lide)]ꢀ ion (1) with nucleoside unit, and in some cases
simple nonnucleoside derivatives are characterised by
low cytotoxicity. Thus, compounds 6 and 9 are non-
toxic up to 100 lM in Vero, A549, MRC-5, PBM
and CEM cells, similarly to natural thymidine 3. Metal-
lacarborane/thymidine conjugate containing metallacar-
borane modification attached at 3-N (5) and
nonnucleoside metallacarborane derivative 8 and 10
show moderate toxicity at higher concentrations. For
comparison, for 5-carboranyl-20-deoxyuridine (CDU)

´
Z. J. Lesnikowski et al. / Bioorg. Med. Chem. 13 (2005) 4168–4175
4173
Table 1. Cytotoxicity of [8-dioxane-3-cobalt bis(dicarbollide)]0 (2),
4. Conclusions
thymidine (3), {8-[5-(N3-thymidine)-3-oxa-pentoxy]-3-cobalt bis(1,2-
dicarbollide)}ꢀ (5) and {8-[5-(O4-thymidine)-3-oxa-pentoxy]-3-cobalt
bis(1,2-dicarbollide)}ꢀ (6) and their components: [8-(5-hydroxy-3-oxa-
pentoxy)-3-cobalt bis(1,2-dicarbollide)]ꢀ (7), [8-hydroxy-3-cobalt
bis(dicarbollide)]ꢀ ion (8), [8-phosphate-3-cobalt bis(dicarbollide)]ꢀ
ion (9) and [8-(5-phosphonate-3-oxa-pentoxy)-3-cobalt bis(1,2-dicar-
bollide)]ꢀ (10)
Derivatives of metallacarboranes type of [3-cobalt
bis(1,2-dicarbollide)]ꢀ ion (1) are proposed as modifying
entity potentially useful in construction of novel boron
carriers for BNCT. Nucleoside/metallacarborane conju-
gate and some simple metallacarborane derivatives
substituted at boron-8 are characterised by low toxicity
and high lipophilicity, an advantageous and preferred
property for potential boron delivering drugs.
Compound
Cell system, CC50 (lM)
Vero
A549
MRC-5
PBM
CEM
2
>100
>100
94
ND
>100
82
ND
>100
68
>100
ND
43
10
ND
3
Although only various cobaltacarborane derivatives are
the subject of this paper, we are well aware that a variety
of their ferra- or metalla-analogues could be tailored as
well. We focused on the cobaltacarboranes as on the
models because they are diamagnetic (in contrast to,
e.g., the FeIII analogue) and can be characterised by
multinuclear NMR spectroscopy. However, we recog-
nize that other metal-centres might bring about new
dimension to the subject. The work in this direction is
ongoing in our laboratories.
5
>100
>100
6
>100
68
>100
>100
ND
ND
ND
>100
42
>100
14
7
3.3
8
>100
>100
>100
ND
ND
ND
>100
>100
>100
60
100
88
9
10
CC50 was 17 lM as established in Vero cell line but it
was not toxic up to 100 lM to primary human lympho-
cytes and the glioma cell lines U253 and 9L.41,42 The
toxicity of BPA used clinically in BNCT is much lower
and was assigned as 8.4–8.6 · 10ꢀ3 M in B-16 and TIG-
1–20 cell lines.43
5. Experimental
5.1. Materials
The carboranyl cage is characterised by extremely high
lipophilicity. This feature is recently being exploited in
the application of some borane clusters as hydrophobic
component (pharmacophore) in biologically active mole-
cules providing them with capacity to interact
hydrophobically with other molecules such as proteins
or lipids of cellular membranes. Carborane pharmaco-
phores may also improve the ability of modified mole-
cules to penetrate lipid bilayers and potentially
increase the cellular uptake.44 [3-Cobalt bis(1,2-dicar-
bollide)]ꢀ ion (1) and its derivatives inherit part of car-
borane cluster lipophilicity combined with high affinity
towards inorganic and organic cations. This unique
property is being presently used for selective extraction
of selected cations from diluted water solutions into
organic solvents.19,45
3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium
(MTT), L-glutamine, penicillin G and streptomycin
were bought from Sigma–Aldrich Chemie GmbH
(Germany); dimethylformamide (DMF) was from
Riedel-de Hae¨ (Germany) and sodium dodecylsulfate
(SDS) was from Fluka Chemie GmbH (Switzerland).
Dimethylsulfoxide (DMSO) was bought from Sigma–
Aldrich Chemie GmbH (Germany) and foetal bovine
serum (FBS) was from Gibco BRL (Germany).
EagleÕs minimal essential medium (EMEM) was from
IIET (Institute of Immunology and Experimental
Therapy) PAS (Poland). CellTiter 96 Aqueous One
Solution cell proliferation assay was bought from
Promega (Madison, WI).
5.2. Methods
The lipophilic property of metallacarborane function
and its ability to affect lipophilicity of a carrier molecule
is clearly illustrated by the high partition coefficient of
nucleoside/metallacarborane conjugates. The partition
coefficient of conjugates 5 and 6 is several hundred times
higher than unmodified thymidine and is comparable
with partition coefficient of metallacarborane bearing
(5-hydroxy-3-oxa-pentoxy)-linker (7) and carborane
bearing nucleoside, CDU.
5.2.1. Cells. The human MRC-5 cells (ATCC CCL-171)
and monkey Vero cells (ATCC CCL 81) were grown in
EMEM. Human cell line A549 (ATCC CCL 185) was
propagated in DulbeccoÕs modified EagleÕs medium.
All culture media were supplemented with 10% FBS,
2 mM L-glutamine and 100 units/mL penicillin G—
100 lg/mL streptomycin.
5.2.2. Cytotoxicity of compounds 3, 5–7 in MRC-5, A549
and Vero cells. The compounds (Figs. 2 and 3) were sus-
pended in DMSO or in distilled water and then in
EagleÕs minimal essential medium (EMEM) supple-
mented with 2% of foetal bovine serum. The final con-
centration of DMSO in the culture medium was 0.1%.
The cells were seeded at 1 · 105 cells/well in 96-well tis-
sue culture plates and allowed to proliferate at 39 ꢁC
for 24 h in growth medium. The medium was then re-
placed by new maintenance medium containing different
concentrations of the test compounds (three wells for
The low level of accumulation of compounds 5 and 6 as
it has been shown by cellular uptake studies in V79 Chi-
nese hamster cells is somewhat disappointing. One
should take into consideration, however, that accumula-
tion of boron is cell type dependent and differs even
among subpopulations of malignant glioma cells. There-
fore, studies in other cell types and with different metal-
lacarborane/nucleoside conjugates are needed to select
conjugate with optimised properties.

´
Z. J. Lesnikowski et al. / Bioorg. Med. Chem. 13 (2005) 4168–4175
4174
each concentration). After a 2-day incubation at 37 ꢁC
in 5% CO2, the number of viable cells was determined
by the formazan method (see below). Cytotoxicity of
the compounds is expressed as the 50% cytotoxic con-
centration (CC50), which is the concentration required
to reduce cell growth by 50% of the (untreated) control.
pounds 5 and 6) and 93 mM (compound 9), and 10 lL
was diluted into 10 mL medium to final concentrations
of 77 lM (compounds 5 and 6) and 93 lM (compound
9), respectively. After incubation for 3 h, the cells were
scraped off the dish with a sterile cell wiper and sus-
pended in the incubation medium by vigorous pipetting
through a Pasteur pipette. An aliquot of the cell suspen-
sion was used to determine the number of cell per mL.
The cells were sedimented by centrifugation at 1500
rpm for 8 min, the major part of the supernatant was re-
moved, and the cells were resuspended in the remaining
incubation solution (about 0.5 mL). The cell suspension
was layered, in an Eppendorf 2 mL centrifuge cup, on
top of a bottom layer (0.7 mL) of 20% aqueous
HNO3, onto which a mixture of nine parts silicon oil
(Fluka, AP100, density 1.06 g/mL) and one part mineral
oil (Fluka, density 0.85 g/mL) had been pipetted. The
resulting three-layer ensemble was centrifuged for
3 min at 1500 rpm, whereupon the cells had passed from
the incubation medium through the oil into the bottom
layer and lysed. A sample of the bottom layer was with-
drawn by piercing the centrifuge cup with a hypodermic
needle, and analysed for boron. Removal of cells from
the dish, centrifugation and withdrawal of the bottom
layer was achieved within 20 min. The sample of the
bottom layer was diluted 1:20 with bidistilled water
and filtered through a 0.2 lm disposable filter. Boron
analysis was carried out in a ThermoElectron Element
2 ICP-MS apparatus, measuring both boron-10 and
boron-11. Samples were diluted with water to concen-
tration ranges appropriate for the sensitivity of the
apparatus. Concentrations were determined by the
increment method, comparing the increase of boron sig-
nal after the addition of known concentrations of boric
acid over that of the blank solution. Concentrations are
expressed as pg boron/106 cells (= ppb, assuming a cell
volume of 10ꢀ9 mL per cell). A blank was obtained by
incubating cells in medium not containing any boron
compound, and harvesting and analysing them in the
same way.
5.2.3. Formazan method.30 Colorimetric MTT (3-(4,5-
dimethylthiazol-2-yl)-2,5-diphenyltetrazolium) method
was used. This method is based on the fact that the yel-
low tetrazolium salt MTT is converted to dark blue for-
mazan by living cells but not by dead cells or culture
medium. MTT was diluted in PBS at a concentration
of 5 mg/mL. To measure cell killing or growth, 25 lL
of the MTT dye solution was added to each well of
the 96-well microplate with cell culture. Next, the plates
were incubated for 2 h at 37 ꢁC in 5% CO2. Then,
100 lL of the solvent solution containing: 1/45 mL of
DMF, 2/13.5 g of SDS and 3/55 mL of distilled water,
were added to each well. After overnight incubation at
37 ꢁC, optical density at 550 nm was measured in ELISA
reader Multiscan RC (Labsystems, Finland).
5.2.4. Cytotoxicity of compounds 2, 8–10 in PBM, CEM
and Vero cells.31 PBM, CEM and Vero cells (5 · 104
cells per well) were seeded in 96-well plates in the pres-
ence of increasing concentrations of the test compounds
(Figs. 2 and 3) and incubated at 37 ꢁC with 5% CO2.
After a 5-day incubation, cell viability defined as CC50
was measured using the CellTiter 96 Aqueous One Solu-
tion cell proliferation assay.
5.2.5. Partition coefficient (K) measurement.33 To
100 lM solution of compounds 3, 5–7 and CDU in
deionised H2O containing 5% of CH3OH (1 mL) in
2 mL Eppendorf tube octanol (1 mL) was added. The
resulting mixture was shaken vigorously at room tem-
perature (20–22 ꢁC) for 1 h to ensure that the com-
poundÕs transfer between the two phases was at
equilibrium and then the mixture was left standing for
1 h to separate the phases. Each sample was subse-
quently centrifuged for 5 min at 13,000 rpm then
0.9 mL of H2O or organic solution was transferred to
1 mL UV cell and the UV absorption at k = 260 (com-
pound 3) or 310 nm (compounds 5–7) was measured.
The partition coefficient is defined as the ratio of the
amount of the compound present in the organic phase
to that present in the aqueous phase.
Acknowledgements
This work was supported in part by the Polish Commit-
tee for Scientific Research (KBN), grant number KBN
6P05F 023 20p02. RFS was supported by the Depart-
ment of Energy grant number DEFGO2-96ER62156,
and by the Department of Veterans Affairs. Authors
thank Prof. Z. Błach-Olszewska (Laboratory of Virol-
ogy, L. Hirszfeld Institute of Immunology and Experi-
mental Therapy PAS, Wroclaw) for kindly supplying
A549 and Vero cell lines.
5.2.6. Cellular uptake.46 Cellular uptake was studied
using a filtration technique where the cells were freed
from surrounding medium by centrifugation through
an oil layer. Instead of trichloroacetic acid (TCA) used
in original method, half-concentrated HNO3 was ap-
plied as lysing medium in modified procedure, as precipi-
tation was seen with TCA, which might lead to loss of
boron. V79 Chinese hamster cells were used for the mea-
surement of uptake of boron in cells. The cells were
grown in F10 medium containing 5% newborn calf ser-
um at 37 ꢁC in 5% CO2. For the uptake experiments,
about 1 · 106 cells were seeded into 10 cm diameter Petri
dishes, 24 h before the experiment. The compounds were
dissolved in DMSO at concentrations of 77 mM (com-
References and notes
1. Coderre, J. A.; Turcotte, J. C.; Riley, K. J.; Binns, P. J.;
Harling, O. K.; Kiger, W. S. Technol. Cancer Res. 2003,
T2, 355.
2. Barth, R. F.; Soloway, A. H.; Goodman, J. H.; Gahbauer,
R. A.; Gupta, N.; Blue, T. E.; Yang, W. L.; Tjarks, W.
Neurosurgery 1999, 44, 433.
3. Hawthorne, M. F. Mol. Med. Today 1998, 4, 174.

´
Z. J. Lesnikowski et al. / Bioorg. Med. Chem. 13 (2005) 4168–4175
4175
ˇ
´
ˇ
´
´ˇ
4. Valliant, J. F.; Guenther, K. J.; King, A. S.; Morel, P.;
Schaffer, P.; Sogbein, O. O.; Stephenson, K. A. Coord.
Chem. Rev. 2002, 232, 173.
5. Lesnikowski, Z. J. Eur. J. Org. Chem. 2003, 4489.
6. Sjo¨berg, S.; Carlsson, J.; Ghaneolhosseini, H.; Hedda, L.;
Hartman, T.; Malmquist, J.; Naeslund, C.; Olsson, P.;
Tjarks, W. J. Neuro-Oncol. 1997, 33, 41.
7. Tjarks, W. J. Organomet. Chem. 2000, 614, 37.
8. Lesnikowski, Z. J.; Shi, J.; Schinazi, R. F. J. Organomet.
Chem. 1999, 581, 156.
9. Byun, Y.; Yan, J. H.; Al-Madhoun, A. S.; Johnsamuel, J.;
Yang, W. L.; Barth, R. F.; Eriksson, S.; Tjarks, W. J.
Med. Chem. 2005, 48, 1188.
27. Plesek, J.; Gruner, B.; Cısarova, I.; Baca, J.; Selucky, P.;
´
Rais, J. J. Organomet. Chem. 2002, 657, 59.
28. Plesek, J.; Gruner, B.; Hermanek, S.; Baca, J.; Marecek, V.;
Janchenova, J.; Lhotsky, A.; Holub, K.; Selucky, P.; Rais,
J.; Cisarova, I.; Caslavsky, J. Polyhedron 2002, 21, 975.
29. Mosmann, T. J. Immunol. Methods 1983, 65, 55.
30. Berg, K.; Hansen, M. B.; Nielsen, S. E. APMIS 1990, 98,
156.
31. Stuyver, L. J.; Lostia, S.; Adams, M.; Mathew, J.; Pai, B.
S.; Grier, J.; Tharnish, P.; Choi, Y.; Chong, Y.; Choo, H.;
Chu, C.; Otto, M. J.; Schinazi, R. F. Antimicrob. Agents
Chemother. 2002, 46, 3854.
_
32. Zuk, A. PosteEpy Mikrobiologii 2003, 42, 87.
10. Al-Madhoun, A. S.; Johnsamuel, J.; Yan, J. H.; Ji, W. H.;
Wang, J. H.; Zhuo, J. C.; Lunato, A. J.; Woollard, J. E.;
Hawk, A. E.; Cosquer, G. Y.; Blue, T. E.; Eriksson, S.;
Tjarks, W. J. Med. Chem. 2002, 45, 4018.
33. Dagle, J. M.; Andracki, M. E.; DeVine, R. J.; Walder, J.
A. Nucleic Acids Res. 1991, 19, 1805.
34. Akhtar, S.; Basu, S.; Wickstrom, E.; Juliano, R. L. Nucleic
Acids Res. 1991, 19, 5551.
11. Lunato, A. J.; Wang, J. H.; Woollard, J. E.; Anisuzzaman,
A. K. M.; Ji, W. H.; Rong, F. G.; Ikeda, S.; Soloway, A.
H.; Eriksson, S.; Ives, D. H.; Blue, T. E.; Tjarks, W. J.
Med. Chem. 1999, 42, 3378.
12. Li, P.; Shaw, B. R. J. Org. Chem. 2005, 70, 2171.
13. Hall, A. H. S.; Wan, J.; Shaughnessy, E. E.; Shaw, B. R.;
Alexander, K. A. Nucleic Acids Res. 2004, 32, 5991.
14. Dobrikov, M. I.; Grady, K. M.; Shaw, B. R. Nucleos.
Nucleot. Nucleic Acids 2003, 22, 275.
15. Spielvogel, B.; Powell, V.; Sood, A. Main Group Metal
Chem. 1996, 19, 699.
16. Hawthorne, M. F.; Young, D. C.; Wegner, P. A. J. Am.
Chem. Soc. 1965, 87, 1818.
35. Bregadze, V.; Sivaev, I.; Sjoberg, S. In Research and
Development in Neutron Capture Therapy, Sauerwein, W.;
Moss, R.; Wittig, A., Eds.; Proceedings 10th International
Congress on Neutron Capture Therapy, Essen, Germany,
September 8–13, Bologna, Italy, Monduzzi Editore, 2002,
p 13.
36. Spryshkova, R. A.; Karaseva, L. I.; Brattsev, V. A.;
Serebryakov, N. G. Med. Radiol. 1981, 26, 62.
37. Spryshkova, R. A.; Brattsev, V. A.; Sherman, T. L.;
Stanko, V. I. Med. Radiol. 1981, 26, 51.
38. Miadokova, E.; Trebaticka, M.; Podstavkova, S.; Vlcek,
D. Biologia (Bratislava) 1990, 45, 917.
39. Gielen, M.; Bouhdid, A.; Willem, R.; Bregadze, V. I.;
Ermanson, L. V.; Tiekink, E. R. T. J. Organomet. Chem.
1995, 501, 277.
17. Hawthorne, M. F.; Andrews, T. D. J. Chem. Soc., Chem.
Commun. 1965, 443.
18. Hawthorne, M. F.; Young, D. C.; Andrews, T. D.; Howe,
D. V.; Pilling, R. L.; Pitts, A. D.; Reintjes, M.; Warren, L.
F.; Wegner, P. A. J. Am. Chem. Soc. 1968, 90, 879.
19. Sivaev, I. B.; Bregadze, V. I. Collect. Czech. Chem.
Commun. 1999, 64, 783, and references cited therein.
20. Saxena, A. K.; Hosmane, N. S. Chem. Rev. 1993, 93,
1081.
21. Grimes, R. N. Coord. Chem. Rev. 2000, 200–202, 773.
22. Hall, I. H.; Tolmie, C. E.; Barnes, B. J.; Curtis, M. A.;
Russel, J. M.; Finn, J. M.; Grimes, R. N. Appl. Organo-
met. Chem. 2000, 14, 108.
40. Hall, H. I.; Lackey, C. B.; Kistler, T. D.; Durham, R. W.,
Jr.; Russel, M.; Grimes, R. N. Anticancer Res. 2000, 20,
2345.
41. Yamamoto, Y.; Seko, T.; Nakamura, H.; Niemoto, H.;
Hojo, H.; Mukai, N.; Hashimoto, Y. J. Chem. Soc., Chem.
Commun. 1992, 157.
42. Schinazi, R. F.; Goudgaon, N. M.; Fulcrand, G.; El
Kattan, Y.; Lesnikowski, Z. J.; Moravek, J.; Liotta, D. C.
Int. J. Radiat. Oncol. Biol. Phys. 1994, 28, 1113.
43. Yamamoto, Y.; Nakamura, H. J. Med. Chem. 1993, 36,
2232.
23. Olejniczak, A. B.; Plesek, J.; Kriz, O.; Lesnikowski, Z. J.
Angew. Chem., Int. Ed. 2003, 42, 5740.
24. Plesek, J.; Hermanek, S.; Franken, A.; Cisarova, I.;
Nachtigal, C. Collect. Czech. Chem. Commun. 1997, 62,
47.
44. Endo, Y.; Iijima, T.; Yamakoshi, Y.; Fukasawa, H. y.;
Miyaura, C.; Inada, M.; Kubo, A.; Itai, A. Chem. Biol.
2001, 8, 341.
45. Plesek, J.; Base, K.; Mares, F.; Hanousek, F.; Stibr, B.;
Hermanek, S. Collect. Czech. Chem. Commun. 1984, 49,
2776.
25. Plesek, J.; Gruner, B., unpublished.
ˇ
Organomet. Chem. 2002, 649, 181.
´
ˇ
´
´ˇ
26. Plesek, J.; Gruner, B.; Cısarova, I.; Baca, J.; Fusek, J. J.
46. Wittig, A.; Sauerwein, W. A.; Coderre, J. A. Rad. Res.
2000, 153, 173.