Design, synthesis, and preliminary biological evaluation of 6-o-glucose-azomycin adducts for diagnosis and therapy of hypoxic tumors

Article abstract of DOI:10.1021/jm2017336

Several 2-nitroimidazole-based molecules (NIs) are used as clinical hypoxic tumor radiodiagnostics, but they are not effective as radiosensitizers/ radiochemotherapeutics. These NIs permeate tumor cells nonselectively via diffusion, and in therapy, where high doses are required, their dose limiting toxicities preclude success. The synthesis and preliminary in vitro evaluations of three glucoazomycins, members of a novel class of C6-O-glucose-linked- azomycin conjugates that are putative substrates of glucose transport proteins (GLUTs) and possess hypoxia-selective radiosensitization features, are now reported. The hypoxia-dependent upregulation of several GLUTs provides a rational basis to develop these glucoazomycins because more selective uptake in hypoxic cells would decrease systemic toxicities at effective doses. Calculated partition coefficients (ClogP, ?1.70 to ?2.99) predict rapid in vivo clearance for low systemic toxicity. In vitro experimental data show that glucoazomycins are radiosensitizers and that they competitively inhibit glucose uptake.

Full text of DOI:10.1021/jm2017336

Article
pubs.acs.org/jmc
Design, Synthesis, and Preliminary Biological Evaluation of 6-O-
Glucose−Azomycin Adducts for Diagnosis and Therapy of Hypoxic
Tumors
Piyush Kumar,*,† Gennady Shustov,§ Hong Liang,§ Vladimir Khlebnikov,§ Weizhong Zheng,†
Xiao-Hong Yang,† Chris Cheeseman,‡ and Leonard I. Wiebe†
‡
†Department of Oncology and Department of Physiology, Faculty of Medicine and Dentistry, University of Alberta, Edmonton,
Alberta T6G 1Z2, Canada
§NAEJA Pharmaceutical Inc., Edmonton, Canada
S
* Supporting Information
ABSTRACT: Several 2-nitroimidazole-based molecules (NIs) are used as
clinical hypoxic tumor radiodiagnostics, but they are not effective as
radiosensitizers/radiochemotherapeutics. These NIs permeate tumor cells
nonselectively via diffusion, and in therapy, where high doses are required,
their dose limiting toxicities preclude success. The synthesis and preliminary in
vitro evaluations of three glucoazomycins, members of a novel class of C6-O-
glucose-linked-azomycin conjugates that are putative substrates of glucose
transport proteins (GLUTs) and possess hypoxia-selective radiosensitization
features, are now reported. The hypoxia-dependent upregulation of several
GLUTs provides a rational basis to develop these glucoazomycins because more selective uptake in hypoxic cells would decrease
systemic toxicities at effective doses. Calculated partition coefficients (ClogP, −1.70 to −2.99) predict rapid in vivo clearance for
low systemic toxicity. In vitro experimental data show that glucoazomycins are radiosensitizers and that they competitively inhibit
glucose uptake.
effect quantified as the sensitizer enhancement ratio (SER). A
number of oxygen mimetic sensitizers with acceptable SERs
have been synthesized and tested,10,11 and among these, 2-
nitroimidazoles (azomycins) with single electron reduction
INTRODUCTION
■
Hypoxia was identified over 50 years ago as an important factor
in biological responsiveness to ionizing radiation1 and as a
complicating phenomenon in the radiotherapy of solid
tumors.2−4 Hypoxia has been recognized as a principle
determinant of the microenvironment of a number of
pathological conditions including stroke, myocardial infarction,
diabetes, arthritis, and transplant rejection; it is the direct
consequence of regional ischemia arising from poor vasculature
and compression of blood and lymphatic vessels in tumors.
Ischemia reduces the delivery of nutrients and oxygen to
affected areas.5 Hypoxia gives rise to an aggressive cancer cell
phenotype that promotes resistance to drugs and low linear
energy transfer (LET) ionizing radiation, elevates metastatic
potential, and upregulates glycolysis, transport mechanisms,
and an almost unlimited array of intermediary and effector
proteins.6,7
Many classes of compounds have been tested as hypoxic cell
theranostics, that is, as radiosensitizers (therapeutics) and as
radiopharmaceuticals (diagnostics) of hypoxic tumor.8 The
main approach has been to mimic the effects of oxygen, i.e., to
trap radiation damage to vital cellular macromolecules, thereby
mimicking the oxygen effect. Most hypoxic mammalian cells are
2.5−3 times more radioresistant than normally oxygenated cells
(the oxygen enhancement ratio; OER). Adams postulated that
the oxygen effect could be mimicked by electron-affinic
compounds (oxygen mimetics) in the absence of oxygen,9 an
potentials (E1 ) of about −390 mV have offered the best
7
balance between cytotoxicity and radiosensitizing efficacy.12
Bioreductive activation of the nitro group in these sensitizers
leads to their covalent binding with the intracellular macro-
molecules in hypoxic cells, resulting in their selective and
enhanced accumulation therein. The lipophilic/hydrophilic
balance provided by the azomycin N-1 “tail” plays a pivotal
role in trans-membrane diffusion, tissue biodistribution, and
clearance from the body, with optimal partition coefficients (log
P) below 1 to avoid deposition into lipoidal tissue (i.e., brain)
and peripheral neurotoxicity.13,14 Among the many 2-nitro-
imidazole-based radiosensitizers, misonidazole (Ro-07−0582,
1-(2-nitro-1-imidazolyl-1)-3-methoxypropanol, MISO) ap-
peared to be one of the most efficacious azomycin radio-
sensitizers in preclinical tests15 but eventually failed because of
dose-limiting toxicities at clinically effective concentrations.16
Several radiohalogenated azomycins (Figure 1), including
123IAZA,17,18 [18F]FMISO,19 and [18F]FAZA,20−22 are used
clinically for the diagnosis of hypoxic tumors,23 hypoxia-
selective mapping,24 in staging and restaging of the disease, and
Received: December 22, 2011
Published: June 18, 2012
© 2012 American Chemical Society
6033
dx.doi.org/10.1021/jm2017336 | J. Med. Chem. 2012, 55, 6033−6046

Journal of Medicinal Chemistry
Article
Chemistry. C6-O-linked glucose−azomycin adducts were
designed so that the C1 −OH group remains unsubstituted and
the GLUT interaction is not compromised, since it is reported
that C1 glycosides are not substrates for GLUTs.32 1-α-D-O-
Methyl-2,3,4-tri-O-benzyl-6-O-((oxiran-2-yl)methyl)-
glucopyranose 2 was obtained in almost quantitative yield by
base-catalyzed reaction of 1-α-D-O-methyl-2,3,4-tri-O-benzyl-
glucopyranose 1 with epibromohydrin in DMF. The use of
racemic epibromohydrin led to the formation of a diastereo-
meric mixture of 2. Proton spectroscopy of 2 demonstrated two
sets of magnetic resonances, confirming the formation of the
H8R and H8S diastereomers. No attempt was made to separate
the diastereomers of the intermediate products. Coupling of 2
with 2-nitroimidazole in the presence of Na2CO3 in refluxing
ethanol yielded methyl-1-α-D-(2,3,4-tri-O-benzyl)-6-O-(3-[2-
nitro-1H-imidazolyl]-2-hydroxypropyl)glucopyranose 3a along
with the bicyclic derivative 3b (Scheme 1) that formed as a
Figure 1. Structure of clinically used azomycin-based hypoxia-selective
radiopharmaceuticals and their ClogP values. FMISO and FAZA are
used in positron emission tomographic imaging (PET), whereas IAZA
is used for single photon emission computed tomography (SPECT).
the therapy planning for patients.25 The challenge has been,
and remains, to design an electron-affinic radiosensitizer that
will attain efficacious concentrations in target (i.e., hypoxic
tumor), yet not reach toxic levels in other tissues.26−30 Until
recently, no mechanism was known that could be exploited to
achieve this balance, as existing radiosensitizers nonselectively
permeate cells via diffusion. The discovery that a number of
transport elements, particularly the facilitative glucose transport
proteins (GLUTs), are upregulated by hypoxia31 has created an
opportunity to conceptualize and design new azomycin adducts
that may exploit this hypoxia-upregulated transport while
Scheme 1. Preparation of Glucose-6-O-Linked Azomycin
a
Conjugates
retaining optimal E1 and hypoxia-dependent adduct forming
7
properties of the azomycins. Glucose-radiosensitizer adducts,
specifically glucose−azomycin (glucoazomycin) conjugates, are
logical choices for this design, but limitations to the nature and
site of substitution32 on glucose invoke substantial constraints
and challenges to harnessing such molecules effectively. The
synthesis and preliminary in vitro radiosensitization studies of
novel 2-nitroimidazolyl−glucose conjugates coupled at C6-O-
via a 3-carbon propyl linker (glucoazomycins, Figure 2) that are
putative substrates of GLUTs are now reported. Preliminary
biological data and design considerations for this new class of
molecules are discussed.
a
Reagents and conditions: (a) NaH/DMF, 0−22 °C, 1 h,
epibromohydrin addition, 5 → 22 °C, 20 h; (b) Na2CO3, EtOH, 2-
nitroimidazole, reflux, 15 h; (c) 2-nitroimidazole, Cs2CO3, EtOH,
reflux, 27 h.
Figure 2. Structures and the ClogP values of 8, 11, and 13.
result of the intramolecular substitution of the nitro group in 3a
(combined yield 78%, ratio 3a:3b is 2.7:1). Compounds 3a and
3b were separated by column chromatography and both were
identified as the diastereomeric mixtures, with each product
RESULTS AND DISCUSSION
■
The principle objectives of this work were to develop a new
class of bioreductively activated, hypoxia-selective molecules
that are substrates for transport by hypoxia-upregulated GLUTs
in solid tumors. GLUTs-facilitated permeation would offer
increased concentrations of these molecules in the target
tissues, leading to enhanced hypoxia-selective binding by the
tumor cell and resulting in improved hypoxia-specific
theranostic properties. The chemistry and preliminary bio-
logical evaluations of 1-α/β-D-(6-O-(9-[2-nitro-1H-imidazolyl]-
8-hydroxypropyl)glucopyranose (glucoazomycin, GAZ, 8), 1-
α/β-D-(6-O-(9-[2-nitro-1H-imidazolyl]-8-fluoropropyl)-
glucopyranose (fluoroglucoazomycin, F-GAZ, 11), and 1-α/β-
D-(6-(((8,9E)-9-(2-nitro-1H-imidazol-1-yl)allyloxy)methyl)-
glucopyranose (allylglucoazomycin, A-GAZ, 13) are now
reported.
1
displaying the signals for two diastereoisomers in its H NMR
spectrum. HPLC analysis confirmed 98% purity for the
diastereomers of 3a, the desired product that separated as
yellow solid.
It was also observed that when 2-nitroimidazole was coupled
with 2 in the presence of cesium carbonate, a milder base, the
reaction yielded 3a exclusively (45% yield) and no traces of 3b
were found. The 1H NMR spectrum of 3a verified the presence
of two diastereomers (8R and 8S generated from the
diastereomeric mixture of 2) because it displayed dual
resonances for several protons. The 13C NMR spectrum also
demonstrated dual resonances for several carbons, specifically
for C9 (δ 52.16 and 52.24), C8 (δ 69.24 and 69.34), C5 (δ
70.00 and 70.04), C4 (δ 77.19 and 77.27), and for C1 (δ 98.05
6034
dx.doi.org/10.1021/jm2017336 | J. Med. Chem. 2012, 55, 6033−6046

Journal of Medicinal Chemistry
Article
a
Scheme 2. Preparation of Compounds 8 and 11
a
Reagents and conditions: (a) Ac2O, pyridine, 22 °C, 20 h; (b) trimethylsilyl triflate, Ac2O, −40 °C → 22 °C, 18 h; (c) 0.1N NaOH, MeOH, 22 °C,
20 min; (d) DAST, THF, −20 °C → 22 °C, 20 h.
and 98.11 ppm). The mass spectrum (MS) showed the
molecular ion for 3a, providing additional evidence for its
formation. This compound was the key intermediate to
synthesize the target molecules 8, 11, and 13; the removal of
protecting groups in 3a led to the formation of 8, while its C8
fluorination transformed 3a into the protected fluoroglucoazo-
mycin, 9, that was used for the synthesis of 11 (Scheme 2).
Moreover, the introduction of a nosyl group at C8 −OH in 3a
gave methyl-1-α-D-(2,3,4-tri-O-benzyl)-6-O-(9-[2-nitro-1H-imi-
dazolyl]-8-(4-nitrobenzenesulfonyloxy)propyl)glucopyranose 4
(Scheme 3), a precursor of 13. Acetylation of 4 to the
to synthesize 8 therefore involved the replacement of protective
ether (benzyl) groups in the sugar moiety of 3a by ester
(acetyl) groups so that mild alkaline conditions could be
applied to hydrolyze the latter. This route involved the
acetylation of 3a using acetic anhydride/pyridine that converted
3a to the corresponding C8-O-acetate derivative, the 8R and 8S
isomers of methyl-1-α-D-(2,3,4-tri-O-benzyl)-6-O-(9-[2-nitro-
1H-imidazolyl]-8-O-(acetyl)propyl)glucopyranose 6 in 99%
yield. Appearance of two distinct s at δ 1.91 and 1.96 in the
1H NMR spectrum of 6 indicated the incorporation of acetyl
function at C8−OH in 3a. Alternate procedures to synthesize 6
involved the reaction of 4 with tetrabutylammonium acetate or
cesium acetate in anhydrous toluene, but these reactions, in
addition to 6, always led to a secondary olefinic product
methyl-1-α-D-(2,3,4-tri-O-benzyl)-6-(((8,9E)-9-(2-nitro-1H-
imidazol-1-yl)allyloxy)methyl)glucopyranose 12a (Scheme 4).
When the diastereomeric mixture of 6 was treated with
excess acetic anhydride and trimethylsilyl triflate at 0 °C, all
benzyl groups in the glucose moiety, as well as the methyl
group at the anomeric C1 position, were replaced with acetate
groups. Acetylation at C1 was not stereoselective and therefore
led to the formation of C1α- and C1β-acetates of 7, each with
8R and 8S diastereomers (6 being a mixture of 8R and 8S
stereomers). The total isolated yield (both α- and β- anomers)
of 1-D-(1,2,3,4-tetra-O-acetyl)-6-O-(9-[2-nitro-1H-imidazolyl]-
Scheme 3. Preparation of Compound 5, a Useful
a
Intermediate for Synthesizing F-18 Labeled 11
1
a
8-O-(acetyl)propyl)glucopyranose 7 was 58%. The H NMR
Reagents and conditions: (a) nosyl chloride, Et3N, DMAP, CH2Cl2,
spectrum of 7 revealed the α/β diastereomeric ratio to be
84:12.5. The appearance of a d (J2,1 = 3.6 Hz) at δ 6.32 and a
dd at δ 5.09 (J1,2 = 3.6 Hz, J3,2 = 10.0 Hz) in the H NMR
−10 °C, 14 h; (b) trimethylsilyl triflate, Ac2O, −40 °C → 14 °C, 4 h.
1
corresponding pentaacetate 1-α/β-D-(1,2,3,4-tetra-O-acetyl)-6-
O - ( 9 - [ 2 - n i t r o - 1 H - i m i d a z o l y l ] - 8 - O - ( 4 -
nitrobenzenesulfonyloxy)propyl)glucopyranose 5 provided the
optimal precursor for nucleophilic radiofluorination to prepare
the positron-emitting radiofluorinated 18F-11.
The synthesis of 8 was initially attempted by boron
trichloride-assisted debenzylation of 3a, followed by the
removal of the 1-O-methyl group using 1 N HCl at 80 °C,
but this led to a very complicated mixture of the products. The
MS analysis of this mixture, however, indicated the formation of
the desired product, albeit in low yields. An alternate approach
spectrum reflected the formation of the α-7 isomer (R and S
mixture), while these protons for the corresponding β-7 isomer
appeared at δ 5.68 (J2,1 = 8.0 Hz) and 5.14 ppm, respectively.
The presence of dual resonances for C2 at δ 71.13 and 71.16
for the α-7 isomer, and at δ 74.09 and 74.16 for β-7,
respectively, in the C-13 NMR spectrum of 7 further confirmed
the presence of α-7[8S], α-7[8R], β-7[8S], and β-7[8R]
diastereomers as the products. A few other protons and carbons
also displayed dual resonances, of which the signals for acetyl
and imidazole protons and carbons were clearly distinct.
6035
dx.doi.org/10.1021/jm2017336 | J. Med. Chem. 2012, 55, 6033−6046

Journal of Medicinal Chemistry
Article
a
NMR spectrum of 5 confirmed the replacement of benzyl
protective groups with acetyl functions at the glucose −OHs. In
addition, the 1-O-methyl group in 4 was also replaced by acetyl,
although not stereoselectively. As a result, the formation of α-5
Scheme 4. Preparation of Compound 13
1
and β-5 anomers (anomeric ratio, 13:1 by H NMR) was seen
during this debenzylative-acetylation process, with each anomer
containing 8R and 8S diastereomers (94% combined yield,
purity >99% as determined by HPLC).
Attempts to synthesize 11 from nosylate 4 using n-Bu4NF as
the fluorinating agent were unsuccessful. The presence of
neighboring protons at C7 and C9 led to intramolecular
elimination under basic conditions, resulting in the formation of
a mixture of olefins 12a and methyl-1-α-D-(2,3,4-tri-O-benzyl)-
6-(((7,8E)-9-(2-nitro-1H-imidazol-1-yl)allyloxy)methyl)-
glucopyranose 12b (5:1 ratio by NMR). Such competitive
elimination is known to occur in the presence of the strongly
basic fluoride ion if the rate of fluorination is slow.33,34 Because
these efforts to synthesize 11 from the nosylates 4 and 5 were
not successful, the DAST-assisted fluorination of 3a was
pursued (Scheme 2). This afforded the desired fluorinated
product, methyl-1-α-D-(2,3,4-tri-O-benzyl)-6-O-(9-[2-nitro-1H-
imidazolyl]-8-fluoropropyl)glucopyranose, 9, as a mixture of 8R
and 8S fluorinated diastereomers in 78% yield with ∼98%
combined chemical purity. The incorporation of fluorine
creates strong spin−spin interactions with the protons and
the carbons at the site of attachment as well as with neighboring
nuclei.35 The H8 resonance for the compound 9 appeared at δ
4.81 ppm, shifted downfield by 0.73 ppm in comparison to the
corresponding nonfluorinated precursor 3a. It displayed a
complex splitting pattern due to its coupling with the
neighboring protons at C7 and C9 and also with fluorine
that was introduced at C8 (JF,H9 = 47 Hz). The F−H coupling
constants with H7 (JF,H7 = 31 Hz) and H9 (JF‑H9 = 20 Hz) are
in accordance with the coupling constants reported for similar
environments.36 The 19F NMR spectrum of 9 further confirmed
the incorporation of fluorine at C8, and the fluorine signal (δ
−191.88 ppm) demonstrated the spin−spin couplings with the
protons substituted at C7, C8, and C9 as one would expect
a
Reagents and conditions: (a) TBAF (1 M in THF), 22 °C, 17 h; (b)
DBU·AcOH, DMF, 22 °C, 24 h; (c) CsF, DMF, 120 °C, 17.5 h; (d)
DBU, MeCN, 22 °C, 14 h; (e) trimethylsilyl triflate, Ac2O, −40 to 22
°C, 14 h; (f) 0.1 M NaOH, MeOH, 22 °C, 20 min.
The two diastereomers of α-7 and β-7 were separated by
HPLC, where the 8R and 8S stereoisomers for α-7 (isomeric
composition 83.9%) appeared at retention time of 11.02 min
and for β-7 (isomeric composition 14%) at retention time of
12.44 min, respectively. Methanolic NaOH-based (0.1 N)
hydrolysis easily removed the protective acetyl groups from 7
(four diastereomers) and afforded a mixture of α-8 and β-8
anomers in 62% yield, again each containing a mixture of 8R
and 8S stereoisomers. Proton and carbon resonances for α-8
displayed a d (J2,1 = 3.6 Hz) at δ 5.10 for H1 proton and a
signal at δ 92.87 for C1, respectively, while β-7 demonstrated
1
based on the H NMR spectrum.
Compound 9, on treatment with the Lewis acid TMS-triflate
as described earlier, afforded 1-α/β-D-(1,2,3,4-tetra-O-acetyl)-6-
O-(9-[2-nitro-1H-imidazolyl]-8-fluoropropyl)glucopyranose 10
the resonances for H1 proton at δ 4.48 (dd, J2,1 = 8.0 Hz, J3,2
=
9.2 Hz) and for C1 carbon at δ 97.09 ppm, respectively. HPLC-
based retention times were 7.03 and 6.97 min for α-8 and β-8,
respectively (Figure 3).
3a was also the key intermediate for synthesizing 5, the
nosylate precursor of 18F-11. Thus, nosylation of 3a in the
presence of dimethylaminopyridine (DMAP) and triethylamine
(Et3N) afforded an 8R and 8S diastereomeric mixture of 4
(because 3a was mixture of diastereomers) in 98% yield
1
in satisfactory (54%) yield. The H NMR spectrum of 10
revealed the presence of both α-10 and β-10 anomers (isomeric
ratio; 91:9), and each anomer displayed resonances for the 8R
and 8S stereoisomers. HPLC analysis of this mixture resolved
α-10 (retention time 11.08 min, 74.5% and 12.57 min, 12%)
and β-10 diastereomers (retention time 9.19 min, 2% and 9.34
1
1
min, 9.5%), respectively. H, 13C, and 19F NMR spectra, H
spin−spin decoupling, NOE experiments, and the chemical
shift values for each nucleus all provided structural verification
for 10 and confirmed the presence of four diastereomers. The
strongest proof for the formation of α-[8R]-10, α-[8S]-10, β-
[8R]-10, and β-[8S]-10 diastereomers came from dual
1
(Scheme 3). The H NMR spectrum confirmed the formation
of this product as H8 displayed a multiplet (m) at δ 5.02 that
was shifted downfield by 0.94 ppm in comparison to the
unsubstituted 3a (Table 1). The dual resonances for several
protons and carbons due to the presence of 8R and 8S
resonances for several protons and carbons in the H and 13C
1
stereomers of 3a were observed in both H and 13C NMR
1
NMR spectra, respectively. In addition, the ratios for the α/β
anomers determined from the 1H NMR spectra (91:9)
corresponded closely with the composition determined by
HPLC (87:11.5). Fluorine at C8 was not affected during the
transformation of 9 to 10, and the corresponding F−H spin−
spin split patterns were visible in H8, H7, H7′, H9, and H9′
protons. The 13C NMR spectrum of 10 also displayed two d at
δ 90.31 (JF‑C = 179.2 Hz) and δ 90.60 (JF‑C = 177.3 Hz) for the
spectra. Because time is of the essence in developing short-lived
positron-emitting radiopharmaceuticals (PERs), and because
acetylated precursors can be rapidly deacetylated under mild
alkaline conditions, 4 was converted to the corresponding
tetraacetyl derivative 5 using trimethylsilyl triflate (9 mol equiv)
and an excess of acetic anhydride at −40 °C (Scheme 3). The
disappearance of benzylic methylene and aromatic protons and
1
the appearance of proton signals for acetyl groups in the H
6036
dx.doi.org/10.1021/jm2017336 | J. Med. Chem. 2012, 55, 6033−6046

Journal of Medicinal Chemistry
Article
Figure 3. HPLC chromatograms of glucoazomycins 8, 11, and 13 indicating their purity and the retention times (Ace C18, 5 μm, 150 mm × 4.6 mm
column [Agilent]; mobile phase, gradients of aqueous 0.1% H3PO4 solution and acetonitrile, flow rate 1.0 mL/min).
1
Table 1. H and 13C NMR Chemical Shifts and Respective Coupling Constants for α- and β-Isomers of 8, 11, and 13
chemical shifts in δ ppm and related coupling constants J (Hz)
proton
H-8
carbon
C9
compd
H-7, H-7′
H-9, H-9′
C7
C8
C2-Im
145.33
C4-Im
126.90
C5-Im
128.09
α-8 (R
and S)
3.44, 3.54 (two dd)
4.07 (m)
4.48, 4.74 (two ddd)
52.14, 52.19 68.83
68.78
β-8 (R
and S)
3.54, 3.71 (two dd)
3.86 (m)
4.44, 4.70 (two ddd)
52.24, 52.26 68.86
68.78
145.33
149.06
126.90
131.25
128.20
131.93
α-11 (R
and S)
3.78, 3.81 (two ddd, JF,7 = JF,7′
29.7)
=
5.05 (dddd, JF,8
48.0)
=
=
4.74 (ddd, JF,9 = 16.0)
4.86 (ddd, JF,9 = 16.0)
54.09 (d,
JF,C
23.5)
94.61 (d, JF,C 74.15 (d,
= 176.1) JF,C
21.4)
94.57 (d, JF,C 74.29 (d,
= 176.9) JF,C
=
=
β-11 (R
and S)
3.78, 3.87 (two ddd, JF,7 = 16.0,
JF,7′ = 27.5)
4.90 (dddd, JF,8
48.0)
4.75 (ddd, JF,9 = 22.8)
4.83 (ddd, JF,9′ = 16.0)
54.18 (d,
149.06
131.25
132.04
JF,C
22.7)
69.96
69.96
=
=
23.5)
124.06
124.13
α-13
β-13
4.26 (m)
6.28 (dd)
7.62 (d, J8,9 = 14)
7.58 (d, J8,9 = 14)
123.69
123.78
143.97
buried
125.68
125.70
127.98
buried
4.26 (m, H7 and H7′)
6.24 (dd, J7,8
=
5.6, J9,8 = 14.0)
8R and 8S diastereomers of the α-anomer; the signals for the
corresponding β-10 carbons were of very low intensity, and
thus not marked. The 19F NMR spectrum confirmed these
interpretations with two m signals at δ −192.33 and −193.91
ppm, which showed the presence of four diastereomers (8R and
8S, for each α- and β-isomer). In 1H-NOE equilibrium
experiments on 10, irradiation of H1α enhanced H2α
resonance by 12%, H8α irradiation enhanced the signals for
H9, H9′, H7, and H7′ (7.5% combined), while the irradiation of
H9 proton affected H9′ (27%), H8, and surprisingly also
imidazolyl H5 (4% and 2.9%, two stereoisomeric signals)
indicating that H9′ and H5 imidazole protons are spatially
6037
dx.doi.org/10.1021/jm2017336 | J. Med. Chem. 2012, 55, 6033−6046

Journal of Medicinal Chemistry
Article
oriented close enough to interact with each other. This
interaction, which could influence the electronegativity
(reducibility) of the imidazole-NO2 function, was not seen in
the corresponding deacetylated compound 11.
restrictive stereochemistry due to an olefinic bond in 13 has a
“retaining” impact on the α/β ratio during the deacetylation
step because this ratio changed much less (71.4% α-isomers in
13 and 91% α- in 12) in comparison to the compounds 8
(50.4% α-isomers in comparison to 84% in 7) and 11 (53.3%
α-isomers in comparison to 74.5% in 10).
Biological Testing In Vitro. Bioreductive activation is the
critical step leading to adduct formation,37 which in turn is the
basis for the diagnosis (imaging) and radiosensitization (as
adjunct therapy to XRT) of hypoxic tumor.38 Compounds 8
and 11 demonstrated moderate hypoxia-selective biological
potential, which suggested that the stereochemistry at C8 was
expected to have a minimal influence on cytotoxicity,
radiosensitization, and GLUT-1-facilitated transmembrane
transport. Consequently, only the stereomeric mixtures (α/β/
8R/8S) of each compound were tested.
Cytotoxicity (IC50). This study included evaluation of 8 and
11, against EMT-6, HeLa, MOO6X, KBALB, and KBALB-STK
cells, and of 13 against EMT-6 cells in cell culture using an
MTT assay, which was compared to IAZA, a known
nitroimidazole-based radiosensitizer. With the exception of 8
against MOO6X cells, in which it was about 10 times more
toxic than IAZA, all compounds had cytotoxic effects in the
0.5−1 mM range, similar to IAZA (Table 2).
Compound 10 was deacetylated using 0.1 N NaOH to afford
four diastereomers (α-8R, α-8S, β-8R, and β-8S) of 11 in >98%
yield (combined for all diastereomers). An HPLC chromato-
gram of this mixture demonstrated the presence of four
diastereomers for 11, which were not completely separated
(Figure 3). The MS of this mixture displayed the molecular ion
1
(M + 1+ 100%) peak, validating the formation of 11. The H
NMR spectrum also revealed dual resonances (chiral C8) for
most of the protons for both α- and β-11. Distinct F−H and
1
F−C coupling patterns were also visible in both H and 13C
NMR spectra, indicating that the fluorine was not lost during
the alkali-based deacetylation. Proton−proton decoupling
spectra and NOE experiments confirmed the assignment of
the protons in 11.
Separation of the diastereomeric mixture of 10 and 11 was
attempted by normal silica gel chromatography as well as by
HPLC chromatography using C-18 reverse phase and columns
with chiral stationary phase, varying gradients of the eluents
(hexanes−isoPrOH), and changes in the flow rate (0.4−1.0
mL/min), but no separation of the diastereomers could be
achieved. The independent preparation of pure diastereomers
of 8 and 11, or determination of their absolute configuration at
C8 via stereoselective synthesis was considered, but given that
the compounds 8 and 11 demonstrated only moderate
radiosensitization in vitro, it was not deemed practical. No
further isolation of the diastereomers was pursued.
Table 2. Estimated Cytotoxicities (IC50, mM) of
Compounds 8, 11, 13, and IAZA in Selected Murine and
Human Cell Lines
a
cell line IC50 (mM)
compd
MOO6X
HeLa
KBALB
KBALB-STK
EMT-6
For the synthesis of 13, treatment of benzylated compound 4
with TBAF or CsF led to the formation of two olefinic
regioisomers in varied reagent-dependent ratios. Thus, the
reaction of 4 with TBAF led to the formation of both 12a and
12b (isomeric ratio 5:1) while the use of CsF afforded the
isomer 12b exclusively. The treatment of 4 with DBU−acetic
acid led to the regioselective formation of 12a, a C8−C9 olefin
product, along with 6 (ratio of 12a:6 = 6:4; Scheme 4). Because
the use of DBU afforded the C8−C9 olefin product 12a
selectively in mild conditions, this method was used for
converting 5 into the corresponding olefinic compound 12,
which was obtained as a mixture of α-12 and β-12 anomers (α/
β ratio, 13:1) in 28% yield (Scheme 4). The 1H NMR spectrum
of the isomeric mixture of 12 revealed that the olefin bond was
generated exclusively between C8 and C9, and the coupling
constants for H8 and H9 (J8,9 = 14 Hz) confirmed that they
were in the trans orientation. H8 appeared as a doublet of
IAZA (std)
>1
>1
0.5
0.5
>1
0.5
0.5
>1
>1
0.6
>1
>1
8
0.05
>1
0.5
>1
11
13
b
ND
ND
ND
ND
a
IC50s were estimated from survival plots, the maximum concentration
tested for each compound was 1 mM. ND = not done.
b
Radiosensitization. HeLa, EMT-6, and MOO6X cancer
cell lines were irradiated using radiation doses up to 18 Gy with
compounds 8, 11, and 13, at concentrations ranging from 0.5
to 1 mM, under air (normoxic) and nitrogen (hypoxic)
conditions. SERs (surviving fraction under nitrogen/surviving
fraction under nitrogen plus sensitizer) were calculated for a
radiation dose (18 Gy) which produced OERs (surviving
fraction under nitrogen/surviving fraction under air) ranging
from >2.0 to 4.3. Under these conditions, 13 was consistently
the most effective (SER 1.3−1.7) and 8 was the least effective
(SER 1−1.2) in the three cell lines used (Table 3). SERs for
well-characterized radiosensitizers such as IAZA and MISO
have been reported to be around 1.5, depending on cell line
and radiosensitizer concentrations used.39
triplet (dt) due to its coupling with H7 and H7′ protons (J8,7
=
J8,7′ = 5.6 Hz). Deacetylation of 12 using 0.1 M aqueous NaOH
followed by its treatment with Dowex 50W × 4−200 (H+)
resin gave a pure mixture of α-13 and β-13 (isomeric ratio,
71.4: 22.6 by HPLC). Besides shielding H1−H4 protons in 13,
H8 and H9 protons resonances moved downfield in
comparison to the acetylated product 13 and appeared at δ
6.28 (dd, J8,9 = 14 Hz) and at δ 7.62 (d, J8,9 = 14 Hz),
respectively. Proton−proton decoupling and NOE experiments
Transport Studies. In vitro 14C-glucose influx experiments
showed moderate inhibition of glucose uptake by Xenopus
oocytes when challenged by compounds 8, 11, or 13. It appears
that glucoazomycins most likely bind strongly to the trans-
porter and therefore are poorly transported just like 6-NBDG
that has high-affinity for GLUT-1, but it is not readily
translocated via GLUT-1.40 Of the three glucoazomycins, 13
was the most potent inhibitor of 14C-glucose accumulation by
Xenopus oocytes (Figure 4). Further analysis of data for 13
1
were performed to confirm the assignment of the H NMR
signals. NOE irradiation of H8 affected the signals for H7 and
H7′ (3.6%) and H9 protons (1.3%) and reaffirmed these
assignments. The C-13 NMR spectrum, elemental composition,
and mass spectral (ES+) analyses provided additional structural
confirmation for this product. A comparison of α/β ratio for
the final products to their acetylated derivatives indicated that a
6038
dx.doi.org/10.1021/jm2017336 | J. Med. Chem. 2012, 55, 6033−6046

Journal of Medicinal Chemistry
Article
Table 3. In Vitro Cell Survival Following a Single Radiation Dose of 18 Gy (Maximum Dose) Using 0.5 mM Concentrations of
Compounds 8, 11, and 13
in vitro radiosensitization
MOO6X
EMT-6
HeLa
a
a
a
compd (0.5 mM)
surviving fraction (N2)
OER
SER
surviving fraction (N2)
OER
SER
surviving fraction (N2)
OER
SER
8
0.15
0.15
0.08
2.8
3.3
2.4
1
0.05
0.2
3.8
3.8
4.3
1
0.06
0.05
0.05
3.6
3.0
2.8
1.2
1.3
1.7
11
13
1.3
1.30
1.2
1.4
0.2
a
OER and SER values were estimated from plots of surviving fraction as a function of radiation dose. OER and SER values quoted are based on the
survival fractions observed following a single 18 Gy X-ray dose.
(Figure 5) indicated an effective inhibitor dose (ED50) of
approximately 0.5 mM.
impact on either nitro reduction or on GLUT-1-facilitated
transmembrane transport. Consequently, only limited efforts to
separate these diastereomers were considered appropriate,
which did not lead to pure diastereomers.
ClogPs of these glucoazomycins (11 = −1.70, 13 = −2.20, 8
= −2.99) indicated that they are more hydrophilic than current
clinically used azomycin radiopharmaceuticals (ClogP, FAZA =
−1.23, IAZA = −0.36) (Figures 1 and 2). Preliminary in vitro
evaluations indicated that 13 demonstrates moderate hypoxia-
selective radiosensitization of MOO6X, EMT-6, and HeLa
cancer cells (SER 1.3−1.7, Table 3), while 8 and 11 were
weaker radiosensitizers than literature compounds like IAZA.
These differences in radiosensitization potency (SERs) were
small and most apparent at high μM to low mM concentrations
as is typical for other 2-nitroimidazoles. However, the
determination of the effects of the C8 substituent itself (i.e., I
vs F vs OH vs others) on bioreduction requires additional
study, including the determination of enzyme-catalyzed
Figure 4. Competitive inhibition of [14C]glucose uptake by oocytes
expressing GLUT-1 in the presence of 8, 11, and 13.
reduction kinetics and/or determination of their E1 and
7
correlation of such data with in vitro adduct formation
experiments. The cytotoxicities of these glucoazomycins toward
MOO6X, HeLa, EMT-6, K-Balb, and K-Balb-STK cancer cell
lines were similar to IAZA, a known radiosensitizer and
clinically used hypoxia-selective radiopharmaceutical for diag-
nosing and staging of a variety of solid tumors.
Glucose uptake inhibition studies were indicative of
interaction with GLUT-1 (Figure 4), but the only dose−
response relationship data that could be modeled was for the
inhibition of glucose uptake by 13 in the GLUT-2 expressing
oocyte model (Figure 5). Importantly, GLUTs affinity and
transport are not synonymous as reported for 6-NBDG. 6-
NBDG was initially thought to be transported by GLUT-1 but
has more recently been found to be a very strong binder that
impedes glucose uptake by blocking the transporter rather than
by competing for transport. Thus, 6-NBDG has high-affinity for
GLUT-1, but it is not readily transported via GLUT-1 because
of the large substituent at glucose C6.40 The glucose challenge
data for 8, 11, and 13 demonstrate rather weak inhibition of
glucose uptake in the oocytes model and low in vitro
cytotoxicity. In summary, the introduction of the glucose
substituent at N1 of azomycin appeared to have only small
impact on electron densities around the nitro substituent (the
bioreductive arm). Current GLUT-1 data indicate that the
glucoazomycins may bind to the transporter as is seen with 6-
NBDG, but its transport via GLUT-1 has not been ascertained.
It is possible that without facilitated transport, diffusion of these
very hydrophilic glucoazomycins across the phospholipid cell
membrane may be very slow, thereby precluding localized
intracellular concentrations required for effective oxygen-
mimetic radiosensitization.
Figure 5. Estimation of the median effective concentration (EC50) for
the inhibition of [14C]glucose uptake by uptake by oocytes expressing
GLUT-2 in the presence of 13.
CONCLUSIONS
■
Three representatives of the novel 6-O-glucoazomycin class of
bioreductive agents, 8, 11, and 13, were synthesized and
chemically characterized. The synthesis of 8 and 11 led to the
formation of four diastereomers due to α- and β-anomers at C1
and the chiralitiy of the C8 carbon. Consequently, α- and β-
anomers of 8R and 8S stereomers were obtained but not
isolated. Moderate radiosensitization and glucose inhibition
properties for 8R/S pairs in 8 and 11 (chiral compounds)
indicated that stereochemistry at C8 would exert little or no
6039
dx.doi.org/10.1021/jm2017336 | J. Med. Chem. 2012, 55, 6033−6046

Journal of Medicinal Chemistry
Article
oxirane H8 for both diastereomers, 3.31 (dd, J6′,6 = 11.6 Hz, J6,5 = 4.6
Hz, H6 for two diastereomers), 3.36 and 3.37 (two s, each for 3H,
OCH3 for two diastereomers), 3.42 (dd, J6,6′ = 11.6 Hz, J5,6′ = 5.6 Hz,
H6′ for two diastereomers), 3.54 (dd, J1,2 = 3.6 Hz, J3,2 = 9.2 Hz, 1H,
H2 for two diastereomers), 3.60 (dd, J3,4 = 9.2 Hz, J5,4 = 9.6 Hz, H4 for
two diastereomers), 3.66 (dd, J6,5 = 4.0 Hz, J4,5 = 10.4 Hz, 1H, H5, 8S
diastereomer), 3.68 (dd, J7′,7 = 11.2 Hz, J8,7 = 4.9 Hz,, H-7 for two
diastereomers), 3.72 (ddd, J4,5 = 9.6 Hz, J6,5 = 4.6 Hz, J6′,5 = 5.6 Hz,
1H, H5 for 8R diastereomer), 3.79 (dd, J8,7′ = 3.2 Hz, J7,7′ = 11.2 Hz,
H7′ for two diastereomers), 3.99 (dd, J2,3 = J4,3 = 9.2 Hz, H3 for two
diastereomers), 4.60 (d, J2,1 = 3.6 Hz, H1 for two diastereomers), 4.61
EXPERIMENTAL SECTION
■
Chemistry: General. All reagents were purchased from
commercial suppliers (Sigma-Aldrich, USA; Fluka, USA) and were
of reagent grade. Anhydrous solvents were used as obtained from the
suppliers. Sodium hydride (60% in mineral oil) was prewashed by the
hexanes prior to use. The progress of synthetic reactions was
monitored by thin layer chromatography (TLC) in a suitable solvent
system (system A, 5:95, v/v, CH3OH:CH2Cl2; system B, 10:90, v/v,
CH3OH:CH2Cl2; system C, hexanes:EtOAc, 2:1, v/v; system D,
hexanes:EtOAc, 1:2, v/v; and system E, hexanes:EtOAc 1:1, v/v).
Column chromatography was performed on Merck silica gel 60 (70−
200 and 230−400 mesh ASTM). Dry chromatography, wherever
needed, was performed using TLC grade silica gel 5−15 μm from
Sorbent Technologies Inc. Preparative HPLC purification used a
reverse phase ACE C18 (5 μm, 150 mm × 4.6 mm) column (Agilent)
and gradients of aqueous 0.1% H3PO4 solution and acetonitrile
(mobile phase A, gradient elution) as mobile phase at a flow rate of 1.0
mL/min that enabled the collection of pure diastereomeric mixtures of
the glucoazomycins. The attempts to separate the diastereomers
included reversed phase columns (Agilent, 150 mm × 4.6 mm) and
the columns with chiral stationary phases (ChiralpakAD 150 mm × 4.6
mm 10 μm containing amylase carbamate and ChiralcelOJ-H 150 mm
× 4.6 mm, cellulose tris [4-methyl benzoate] coated on 5 μm silica gel,
Daicel, USA) using hexanes:iso-PrOH (9:1, v/v) at a flow rate of 1.0
mL/ min or hexanes:iso-PrOH 1:1 (v/v) at a flow rate of 0.4 mL/min
as eluents. Synthesized compounds were isolated as diastereomeric
(d, J4a,b = 11.2 Hz, 1H, H4a of 8S diastereomer, Bn), 4.64 (d, J4a,b
=
11.2 Hz, 1H, H4a of 8R diastereomer, Bn), 4.66 (d, J3a,3b = 12.0. Hz,
1H, H3a Bn, both diastereomers), 4.79 (d, J3a,3b = 12.0 Hz, 1H, H3b
Bn, both diastereomers), 4.82 and 4.83 (d, J4a,b = 11.0 Hz, 1H, H4b
Bn, each d denotes one diastereomer), 4.88 and 4.90 (two d, J2a,2b
=
11.2 Hz, 1H, H2a, each d for one diastereomer), 4.98 (d, J2a,2b = 11.2
Hz, 1H, H2b Bn, both diastereomers), 7.28−7.37 (15H, 3 phenyl
moieties for both diastereomers). 13C NMR (CDCl3): δ 50.60 and
50.71 (oxirane C8, both diastereomers), 55.16 (Me), 69.78 and 69.87
(oxirane C9, both diastereomers), 69.99 and 70.04 (C5), 72.03 and
72.24 (C6), 72.30 and 72.56 (C7), 73.39 (CH2 at C4), 75.00 and
75.05 (CH2 at C2), 75.72 (CH2 at C3), 77.54 (C4), 79.72 and 79.84
(C2), 81.98 and 82.08 (C3), 98.21 (C1), 127.51−138.75 (phenyl Cs).
ESI-MS (C31H36NaO7): m/z 543.2348 (M + Na)+.
Synthesis of Methyl-1-α-D-(2,3,4-tri-O-benzyl)-6-O-(9-[2-
nitro-1H-imidazolyl]-8S-hydroxypropyl)glucopyranose and
Methyl-1-α-D-(2,3,4-tri-O-benzyl)-6-O-(9-[2-nitro-1H-imidazol-
yl]-8R-hydroxypropyl)glucopyranose (3a). 2-Nitroimidazole
(3.62 g, 32 mmol) and Cs2CO3 (1.04 g, 3.2 mmol) were added to a
solution of 2 (16.66 g, 32 mmol) in ethanol (310 mL) under stirring,
and the suspension was heated to reflux (80 °C) for 27 h. The
resulting clear yellow solution was cooled to 22 °C, the solvent was
removed under reduced pressure, and the yellow semisolid residue was
redissolved in EtOAc (300 mL) and then filtered through a pad of
silica gel (for TLC). The filtrate was collected and the solvent was
removed under reduced pressure to afford a yellow semisolid (19.7 g)
that was subjected to dry chromatography using EtOAc−hexanes
(20% → 80%) as the eluent to afford pure product 3a as a mixture of
1
mixtures, and therefore the melting points were not determined. H,
19F, and 13C NMR spectra, NOE studies were acquired using a Bruker
AM-300 spectrometer or a Varian Mercury 400 MHz NMR
spectrophotometer (Naeja) in deuterated chloroform (CDCl3) or
methanol (CD3OD), depending on the solubility of the product.
Chemical shifts are reported in δ ppm downfield with respect to
tetramethylsilane as an internal standard in H and 13C NMR spectra
1
and trichlorofluoromethane as an external standard for 19F NMR
spectra. Being sugar conjugates, the nomenclature of these molecules
uses glucose as the “base” moiety for denoting the positions of the
carbons, and this is extended to the linker; nitroimidazole protons are
represented by the prefix “Im”. The elemental composition of final
products was confirmed by C, H, and N elemental analyses and/or by
HRMS (ES+) spectra recorded using an AEI-MS-12 mass
spectrometer. Low resolution mass spectra were measured on a
Waters 2795 separation module using either ES+ or ES− ionization
modes. No attempt was made to resolve the diastereomers formed by
creation of an asymmetric center at C8 in the three-carbon linker that
couples the glucosyl and nitroimidazolyl moieties. HPLC analyses
confirmed >95% chemical purity for the final target compounds.
Synthesis of 1-α-D-O-Methyl-2,3,4-tri-O-benzyl-6-O-((oxiran-
2-yl)methyl)-glucopyranose (2). A precooled (0 to 5 °C) solution
of 1 (35.0 g, 0.0754 mol) in anhydrous DMF (60 mL) was added
dropwise to a stirred cooled (ca. 0 °C) suspension of NaH (4.53 g,
0.113 mol) in anhydrous DMF (75 mL). The resultant mixture was
allowed to warm up to 22 °C (1 h) and then cooled down to +5 °C.
Epibromohydrin (15.5 g, 0.113 mol) was added dropwise to this
reaction mixture under the stirring at +5 to 10 °C. The mixture was
allowed to warm to 22 °C, and stirring was continued for an additional
20 h to the complete reaction. EtOAc (300 mL) was added to the
reaction mixture, and the resultant mixture was washed with cold water
(3 × 150 mL), citric acid (0.5M, 2 × 100 mL), saturated aqueous
solution of NaHCO3 (100 mL), brine (200 mL), and dried over
anhydrous MgSO4. Filtration and removal of the solvent from the
filtrate under reduced pressure provided yellow oil (41.1 g). Dry
chromatography (5% EtOAc−hexanes → 30% EtOAc−hexanes) of
this material afforded a mixture of 8R and 8S diastereomers (due to the
chirality at C8, diastereomers ratio 1:1) of product 2 as a colorless oil,
1
two diastereomers with ratio 1:1. Combined yield: 15.9 g (78%). H
NMR (CDCl3): δ 2.74 and 2.83 (two d, J8‑OH = 5.2 Hz, OH for 8S and
8R diastereomers), 3.37 and 3.38 (two s, each for 3H, OCH3 for 8S
and 8R diastereomers), 3.43 and 3.44 (two dd, J3,4 = 9.6 Hz, J5,4 = 8.4
Hz, each representing 1H, H4 for 8S and 8R diastereomers), 3.49 and
3.51 (two dd, J3,2 = 9.6 Hz, J1,2 = 3.6 Hz, each denoting 1H, H2, 8S and
8R diastereomers), 3.54 (two dd, J5,6 = 4.4 Hz, J6′,6 = 12.2 Hz, H6 for
both diastereomers), 3.62 (dd, J5,6′ = 3.2 Hz, J6′,6 = 12.0 Hz, H6′, both
diastereomers), 3.69 (dd, J8,7 = 3.2 Hz, J7′,7 = 12.0 Hz, H7, both
diastereomers), 3.72 (ddd, J4,5 = 8.0 Hz, J6,5 = 4.4 Hz, J6′,5 = 3.2 Hz,
denoting H5, two diastereomers), 3.73 (dd, J8,7′ = 4.8 Hz, J7,7′ = 12.0
Hz, H7′, two diastereomers), 3.99 (dd, J2,3 = J4,3 = 8.8 Hz, H3, both
diastereomers), 4.08 (m, 1H, H8, both diastereomers), 4.33 (two dd,
J8,9 = 6.8 Hz, J9,9′ = 14.0 Hz, each representing one diastereomer, H9),
4.57 (dd, J2,1 = 4.0 Hz, H1, two diastereomers), 4.58 (d, J2a,b = 11.8 Hz,
H2a, Bn), 4.60 (broad, 1H, H9′, both diastereomers), 4.67 (d, J3a,3b
12.0 Hz, H3a, Bn), 4.79 (d, J3a,3b = 12.0 Hz, H3b, Bn), 4.81 (d, J4a,4b
11.0 Hz, H4a, Bn), 4.90 (d, J4a,b = 11.0 Hz, H4b, Bn), 4.99 (d, J2a,2 b
=
=
=
11.8 Hz, H2b, Bn), 7.09 and 7.10 (two s, representing 8S and 8R
diastereomer, H4, Im), 7.14 and 7.15 (two s, H5, 8S and 8R
diastereomers, Im), 7.28−7.38 (15H, 3 phenyls). 13C NMR (CDCl3):
δ 52.16 and 52.24 (C9 for 8S and 8R diastereomers), 55.29 (Me),
69.24 and 69.34 (C8, 8S and 8R diastereomers), 70.00 and 70.04 (C5),
70.55 (C6), 72.60 (C7), 73.34, (CH2 at C4), 74.87 and 74.97 (CH2 at
C2), 75.76 (CH2 at C2), 77.19 and 77.27 (C4), 79.90 (C2), 81.89
(C3), 98.05 and 98.11 (C1), 127.61−128.93 (phenyl Cs). ESI-MS
(C34H39N9O9): m/z 634 (M + 1)+.
Synthesis of Methyl-1-α-D-(2,3,4-tri-O-benzyl)-6-O-(9-[2-
nitro-1H-imidazolyl]-8S-O-(4-nitrobenzenesulfonyloxy)-
propyl)glucopyranose and Methyl-1-α-D-(2,3,4-tri-O-benzyl)-6-
O - ( 9 - [ 2 - n i t r o - 1 H - i m i d a z o l y l ] - 8 R - O - ( 4 -
1
which were not separated. Combined yield 31.3 g (85%). H NMR
(CDCl3): δ 2.56 (dd, J9′,9 = 9.2 Hz, J8,9 = 4.9 Hz for 8S and J8,9 = 2.8
Hz for 8R diastereomers, oxirane H9 for both diastereomers), 2.76
(dd, J8,9′ = 4.2 Hz, J9,9′ = 9.4 Hz, oxirane H9′ for both diastereomers),
3.13 (dddd, J9,8 = 4.9 Hz, J9′,8 = 4.2 Hz, J7,8 = 4.9 Hz, J7′,8 = 3.2 Hz,
6040
dx.doi.org/10.1021/jm2017336 | J. Med. Chem. 2012, 55, 6033−6046

Journal of Medicinal Chemistry
Article
nitrobenzenesulfonyloxy)propyl)glucopyranose (4). A solution
of 4-(dimethylamino)pyridine (2.69 g, 22 mmol) and triethylamine
(4.45 g, 44 mmol) in CH2Cl2 (60 mL) was added dropwise to a cold
stirred (ca. −10 °C) solution of compound 3a (13.93 g, 22 mmol) and
4-nitrobenzenesulfonyl chloride (5.85 g, 26.4 mmol) in anhydrous
CH2Cl2 (60 mL), and the reaction mixture was kept at this
temperature. The reaction was complete after 14 h. The mixture
was warmed to 22 °C and then washed sequentially with water (3 ×
100 mL) followed by 1N HCl (100 mL). The organic phase was
separated, dried over MgSO4, and then filtered through a pad of silica
gel (for TLC). The solvents were removed from the organic filtrate
under reduced pressure to afford pure 8R and 8S diastereomers (ratio
described in Chemistry: General, 11.81 min for all isomers. HPLC:
99.16% (combined chemical purity for all diastereomers).
α-5 (8R and 8S). 1H NMR (CDCl3): δ 2.03, 2.05, 2.13, 2.19 (4 s for
8S diastereomer) and 2.04, 2.05, 2.07 and 2.20 (4 s, for 8R
diastereomer, each for 3H, CH3), 3.50 (dd, J6′,6 = 11.2 Hz, J5′6′ = 4.0
Hz, 1H, H6 for 8S diastereomer), 3.61 (dd, J6′,6 = 11.2 Hz, J5,6′ = 2.0
Hz, 1H, H6′ for 8S diastereomer), 3.66 (dd, J5,6 = 2.0 Hz, J6′,6 = 8.8 Hz,
1H, H6, 8R diastereomer), 3.70 (dd, J5,6′ = 2.8 Hz, J6′,6 = 8.8 Hz, 1H,
H6′, 8R diastereomer), 3.73 (d, J7,7′ = 11.2 Hz, J8,7 = 3.2 Hz, 1H, H7,
8S diastereomer), 3.80 (dd, J7′,7 = 12.0 Hz, J8,7 = 3.6 Hz, 2H, H7 and
H7′ for 8R diastereomer), 3.88 (dd, J7′,8 = 2.8 Hz, J7′,7 = 12.0 Hz, 1H,
H7′, 8S diastereomer), 4.02 (m, J4,5 = 10.4 Hz, J6′5 = 2.0 Hz, J6,5 = 3.6
Hz, H5, both diastereomers), 4.55 (dd, J8,9 = 8.4 Hz, J9′,9 = 14.4 Hz,
1H, H9, 8S diastereomer), 4.60 (dd, J8,9′ = 3.2 Hz, J9′,9 = 14.4 Hz, 1H,
H9′, 8S diastereomer), 4.71 (dd, J8,9 = 8.4 Hz, J9′,9 = 14.4 Hz, 1H, H9,
8R diastereomer), 4.78 (dd, J8,9′ = 3.6 Hz, J9′,9 = 14.4 Hz, 1H, H9, 8R
diastereomer), 4.96 (dddd, J9,8 = 8.4 Hz, J9′,8 = 3.2 Hz, J7,8 = 3.6 Hz, J7′,8
1
1:1) of product 4 as a yellowish solid. Yield: 17.7 g (98%). H NMR
(CDCl3): δ 3.37 and 3.38 (two s, each for 3H, OCH3 for 8S and 8R
diastereomers), 3.41 (dd, J3,4 = 9.6 Hz, J5,4 = 9.2 Hz, 1H, H4 for 8S
diastereomer) and 3.50 (dd, J3,4 = 10.0 Hz, J5,4 = 9.2 Hz, 1H, H4 for
8R diastereomer), 3.51 (dd, J3,2 = 9.2 Hz, J1,2 = 3.2 Hz, representing
H2, 8S diastereomer), 3.58 (two dd, J5,6 = 2.0 Hz, J6′,6 = 11.2 Hz, H6,
8S and 8R diastereomers), 3.62 (dd, J5,6′ = 2.8 Hz, J6′,6 = 11.2 Hz,
representing H6′, 8S and 8R diastereomers), 3.68 (dd, J8,7 = 3.6 Hz, J7′,7
= 2.8 Hz, 1H, H8 of 8S diastereomer), 5.05 (dddd, J9′,8 = 8.4 Hz, J9,8
=
3.6 Hz, J7,8 = 3.6 Hz, J7′,8 = 2.8 Hz, 1H, H8 of 8R diastereomer), 5.08
(two dd, merged, J1,2 = 4.0 Hz, J3,2 = 10.0 Hz, H2, both diastereomers),
5.17 (dd, J5,4 = 10.4 Hz, J3,4 = 9.6 Hz, 1H, H4, 8S diastereomer), 5.26
(dd, J5,4 = 10.4 Hz, J3,4 = 9.6 Hz, 1H, H4, 8R diastereomer), 5.48 (dd,
= 11.2 Hz, H7, 8S and 8R diastereomers), 3.71 (dd, J4,5 = 9.2 Hz, J6,5
=
2.0 Hz, J6′,5 = 2.8 Hz, denoting H5 for two diastereomers), 3.74 (dd,
J8,7′ = 4.4 Hz, J7,7′ = 11.2 Hz, H7′ for two diastereomers), 3.99 (dd, J2,3
= 9.2 Hz, J4,3 = 9.6 Hz, H3 for 8S and 8R diastereomers), 4.13 and 4.46
(two dd, each for one diastereomer, J8,9 = 8.4 Hz, J9,9′ = 13.8 Hz, H9,
8S and 8R diastereomers), 4.55 (broad dd, J8,9′ = 8.4 Hz, J9,9′ = 13.8 Hz,
H9′, 8S and 8R diastereomers), 4.58 (dd, J2,1 = 4.0 Hz, representing
H1 for two diastereomers), 4.68 (d, J2a,b = 12.0 Hz, 1H, H2a, Bn), 4.73
(d, J3a,3b = 12.0 Hz, 1H, H3a, Bn), 4.81 (d, J4a,4b = 11.2 Hz, 2H, H4a
J2,3 = 10.0 Hz, J4,3 = 9.6 Hz, 1H, H3, 8S diastereomer), 5.49 (dd, J2,3
=
10.0 Hz, J3,4 = 9.6 Hz, 1H, H3, 8R diastereomer), 6.29 (d, J2,1 = 4.0 Hz,
1H, H1, 8S diastereomer), 6.31 (d, J2,1 = 4.0 Hz, 1H, H1, 8R
diastereomer), 7.05 (s, H4, Im, 8S diastereomer), 7.08 (s, H4, Im, 8R
diastereomer), 7.15 (s, H5, Im, 8S diastereomer) and 7.26 (s, H5, Im,
8R diastereomer), 7.85 and 7.95 (two dd, J6,5 = J2,3 = 9.2 Hz, 4H, H2,
H3, H5 and H6 of Ns, 8S diastereomer), 8.32 and 8.34 (two d, J5,6
=
J3,2 = 9.2 Hz, 4H, H2, H3, H5, and H6 of Ns, 8R diastereomer).
and H4b, Bn), 4.90 (d, J3a,3b = 12.0 Hz, 1H, H3b, Bn), 4.99 (d, J2a,2 b
=
1
β-5 (8R and 8S). H NMR (CDCl3): The signals for most of the
protons for this isomer appeared as buried under the corresponding
signals for the α-isomer. Only the signals that appeared separate from
α-5 and were identifiable in this complex mixture of diastereomers are
being provided. δ 2.03, 2.06, 2.11 and 2.12 (4 s, each representing 3H
of CH3), 3.80 (m, H7), 3.85 (dd, J8,7 = 4.0, J7,7′ = 12 Hz, 1H, H7′), 4.60
(H9), 4.78 (H9′), 5.65, and 5.68 (two d, J2,1 = 8.4 Hz, H1 for two
diastereomers).
11.2 Hz, 1H, H2b, Bn), 5.02 (m, J8,9 = 8.4 Hz, J8,9′ = 8.0 Hz, H8 for
two diastereomers), 6.99 and 7.01 (two s, each 1H, 8S and 8R
diastereomers, H4, Im), 7.04 (s, represents two diastereomers, H5,
Im), 7.26−7.39 (m, 3 phenyls from two diastereomers), 7.82 and 7.83
(two d, J = 8.8 Hz, for two diastereomers each denoting 1H, H3,
nosyl), 7.86 (two d, J = 8.8 Hz, each for 1H of diastereomers, H5,
nosyl), 8.21 (two d, J = 8.4 Hz, 1H of each diastereomer, H2, nosyl),
8.23 (two d, J = 8.4 Hz, 1H of each diastereomer H6, nosyl). 13C
NMR (CDCl3): δ 49.94 (C9), 55.35 (Me), 69.24 (C8), 70.44 (C5),
70.70 (C6), 72.65 (C7), 73.44 (CH2 at C4), 74.94 (CH2 at C2), 75.76
(CH2 at C2), 78.88 (C4), 79.95 (C2), 81.86 (C3), 98.15 (C1), 114.57
(C4, Im), 124.52 (C5, Im), 127.68−128.85 (phenyl Cs), C2 imidazole
and C4 nosyl (buried in the baseline). ESI-MS (C40H42N4O13S): m/z
841.2350 (M + Na)+.
Synthesis of 1-α/β-D-(1,2,3,4-Tetra-O-acetyl)-6-O-(9-[2-nitro-
1H-imidazolyl]-8S-O-(4-nitrobenzenesulfonyloxy)propyl)-
glucopyranose and 1-α/β-D-(1,2,3,4-Tetra-O-acetyl)-6-O-(9-[2-
nitro-1H-imidazolyl]-8R-O-(4-nitrobenzenesulfonyloxy)-
propyl)glucopyranose (5). Trimethylsilyl triflate (4.0 g, 18 mmol)
was added dropwise to a stirred cooled (ca. −40 °C) solution of
compound 4 (1.64 g, 2 mmol) in acetic anhydride (20 mL), allowed to
warm up to +14 °C, and stirred for an additional 4 h at this
temperature. The resulting dark-colored reaction solution was cooled
to −5 °C, diluted with cold EtOAc (100 mL), and then poured into a
cold (0 °C) saturated aqueous NaHCO3 solution (100 mL). The
organic phase was separated off and washed with saturated aqueous
NaHCO3 (5 × 100 mL), followed by brine (100 mL), and then dried
over MgSO4. Filtration of the contents through a pad of silica gel and
removal of the solvent under reduced pressure gave 2.2 g of red oil,
that was subjected to dry chromatography using EtOAc−hexanes
(20% → 70%) to afford product 5 (1.32 g) as a yellow solid. An
additional purification by dry chromatography using toluene followed
by EtOAc−toluene (40%) afforded 1.32 g of 5. A final purification of
this product using reversed phase preparative HPLC (eluent; MeCN−
gradient 0.1% CF3CO2H in H2O) eluted pure diastereomeric mixture
of α- and β-5 anomers at a retention time of 11.8 min. Combined yield
for the diastereomeric mixture: 0.512 g (36%). ESI-MS: m/z 703 (M +
1)+. Anal. Calcd (C26H30N4O17S): C, 44.45; H, 4.30; N, 7.97. Found:
C, 44.17; H, 4.39; N, 7.75. HPLC retention time using conditions as
Synthesis of Methyl-1-α-D-(2,3,4-tri-O-benzyl)-6-O-(9-[2-
nitro-1H-imidazolyl]-8S-O-(acetyl)propyl)glucopyranose and
Methyl-1-α-D-(2,3,4-tri-O-benzyl)-6-O-(9-[2-nitro-1H-imidazol-
yl]-8R-O-(acetyl)propyl)glucopyranose (6). Acetic anhydride
(3.75 g, 36.7 mmol) was added dropwise at +3° to +5 °C to a stirred
solution of 3a (15.51 g, 24.5 mmol) in anhydrous pyridine (60 mL).
The yellow reaction solution was allowed to warm up to 22 °C and
was stirred for 20 h. EtOAc (300 mL) was added to the reaction
solution. The resultant mixture was washed with water (3 × 100 mL),
2 N HCl (2 × 100 mL), saturated aqueous NaHCO3 (100 mL), brine
(150 mL), and dried over MgSO4. The solvent from the filtrate was
removed to provide a diastereomeric mixture of 6 as yellow semisolid.
Yield: 16.4 g (99%). 1H NMR (CDCl3): δ 1.91 (s, 3H of COCH3, 8S
diastereomer) and 1.96 (s, 3H of COCH3, 8R diastereomer), 3.39 (s,
3H, CH3, 8S diastereomer) and 3.40 (s, 3H of CH3, 8R diastereomer),
3.48 and 3.50 (two dd, J3,2 = 8.8 Hz, J1,2 = 3.5 Hz, 1H, H2, 8S and 8R
diastereomers), 3.53 (dd, J6,5 = 4.4 Hz, J6,6′ = 10.8 Hz, 1H, H6, both
diastereomers), 3.55 (dd, J6′,5 = 4.4 Hz, J6,6′ = 10.8 Hz, 1H, H6′ for
both diastereomers), 3.59 (dd, J3,4 = 9.6 Hz, J5,4 = 8.4 Hz, 1H, H4, 8S
diastereomer) and 3.61 (two dd, J3,4 = 9.6 Hz, J5,4 = 8.4 Hz, 1H, H4,
8R diastereomer), 3.62 (dd, J4,5 = 10.4 Hz, J6,5 = J6′,5 = 4.4 Hz, 1H, H5,
8S diastereomer) and 3.64 (dd, J4,5 = 10.4 Hz, J6,5 = J6′5 = 4.4 Hz, 1H,
H5, 8R diastereomer), 3.67 (dd, J7′,7 = 11.0 Hz, J8,7 = 7.6 Hz, H7, both
diastereomers), 3.75 (dd, J7,7′ = 11.0 Hz, J8,7′ = 7.6 Hz, H7′, both
diastereomers), 4.00 (dd, J2,3 = J4,3 = 8.8 Hz, 1H, H3), 4.45 (two dd,
each denoting one diastereomer, J9′,9 = 11.2 Hz, J8,9 = 8.4 Hz, H9, both
diastereomers), 4.59 (two dd, each for one diastereomer, J9,9′ = 11.2
Hz, J8,9′ = 8.0 Hz, H9′, both diastereomers), 4.61 (d, J2,1 = 3.6 Hz, H1,
both diastereomers), 4.68 (two d, each denoting one diastereomer, J2a,b
= 12.0 Hz, 1H, H2a, Bn), 4.82 (d, J3a,3b = 10.8 Hz, H3a, Bn, both
diastereomers), 4.83 (d, J4a,4b = 10.8 Hz, H4a, both diastereomers),
4.86 (two d, each for one diastereomer, J4a,4b = 10.8 Hz, 1H, H4b, Bn),
6041
dx.doi.org/10.1021/jm2017336 | J. Med. Chem. 2012, 55, 6033−6046

Journal of Medicinal Chemistry
Article
4.90 (two d, each representing one diastereomer, J3a,3b = 10.8 Hz, H3b,
Bn), 4.99 (d, J2a,2b = 10.0 Hz, H2b, both diastereomers, Bn), 5.02 (m,
J8,9 = 8.4 Hz, J8,9′ = 8.0 Hz, representing H8 for two diastereomers),
7.07 (d, J5,4 = 1.2 Hz, 8S diastereomer, H4, Im) and 7.08 (d, J5,4 = 1.2
Hz, 8R diastereomer, H4, Im), 7.08 (d, J5,4 = 1.2 Hz, 8S diastereomer,
H5, NI) and 7.09 (d, J5,4 = 1.2 Hz, 8R diastereomer, H5, NI), 7.27−
7.39 (15H for each diastereomer, 3 phenyls).
91.99 (C1), 128. 08 (C4, Im), 129.26 (C5 Im for two diastereomers),
other carbon signals were buried under the signals for α-anomer.
Synthesis of 1-α/β-D-(6-O-(9-[2-Nitro-1H-imidazolyl]-8S-
hydroxypropyl)glucopyranose and 1-α/β-D-(6-O-(9-[2-Nitro-
1H-imidazolyl]-8R-hydroxypropyl)glucopyranose (8). Com-
pound 7 (5.595 g, 10 mmol) was dissolved in a solution of 0.1 M
NaOH in methanol (600 mL). After 20 min of stirring at 22 °C, the
reaction solution was treated with Dowex 50W × 4−200 (H+, 18.6 g)
resin to adjust the pH of the solution to 5 and then filtered. The filtrate
was evaporated under reduced pressure to dryness to provide 2.46 g of
a yellow solid that was subjected to dry chromatography (5% MeOH−
CHCl3 → 30% MeOH−CHCl3) to afford 8 as a yellow solid. Yield:
2.18 g (62%). HPLC: 97.23% (combined chemical purity for all
diastereomers). ESI-MS: m/z 350 (M + 1)+. Anal. Calcd
(C12H19N3O9·1H2O): C, 39.24; H, 5.76; N, 11.44. Found: C, 38.90;
H, 5.85; N, 11.16.
Synthesis of 1-α/β-D-(1,2,3,4-Tetra-O-acetyl)-6-O-(9-[2-nitro-
1H-imidazolyl]-8S-O-(acetyl)propyl)glucopyranose and 1-α/β-
D-(1,2,3,4-Tetra-O-acetyl)-6-O-(9-[2-nitro-1H-imidazolyl]-8R-O-
(acetyl)propyl)glucopyranose (7). Trimethylsilyl triflate (59.3 g,
267 mmol) was added dropwise to a stirred cooled (ca. −40 °C)
solution of 6 (16.4 g, 24.3 mmol) in acetic anhydride (243 mL). The
reaction solution was allowed to warm up to 22 °C and stirred for an
additional 18 h. After cooling to −10 °C, the dark reaction solution
was diluted with cold EtOAc (700 mL) and poured in a cold (0 °C)
saturated aqueous NaHCO3 (500 mL). The resultant mixture was
stirred for 0.5 h at room temperature. The organic phase was separated
off and washed with saturated aqueous NaHCO3 (10 × 300 mL) and
brine (500 mL) and dried over MgSO4. Filtration through a pad of
silica gel (for TLC) and removal of the solvent from the filtrate under
reduced pressure provided 23.3 g of dark oil, which was subjected to
dry chromatography (30% EtOAc−hexanes → 80% EtOAc−hexanes)
to afford 11.09 g of product 7 as a yellow solid. An additional
purification of 7 by dry chromatography (5% MeCN−toluene → 35%
MeCN−toluene) provided 7, a yellowish solid, as an R and S
stereoisomeric mixture of α- and β-anomers (α/β anomeric ratio
∼83.9:14.0), which were not separated. Yield: 7.91 g (58%). ESI-MS:
m/z 560 (M + 1)+. Anal. (C22H29N3O14) CHN. HPLC: 97.9%
(combined chemical purity for all diastereomers).
α-8 (8R and 8S). Retention time 7.03 min. Isomeric composition
1
50.4% by HPLC). H NMR (CD3OD): δ 3.34 (dd, J5,4 = J3,4 = 10.0
Hz, 1H, H4), 3.38 (ddd, J1,2 = 3.6 Hz, J2,3 = 10.0 Hz, 1H, H2), 3.44
and 3.54 (two dd, merged, J7′,7 = 9.6 Hz, J8,7 = 4.8 Hz, 2H, H7 and
H7′), 3.67 (dd, J4,3 = J2,3 = 10.0 Hz, 1H, H3), 3.66 and 3.68 (two dd,
J6,5 = 2.8 Hz, J5,6′ = 2.0 Hz, J6,6′ = 11.2 Hz, 2H, H6 and H6′), 3.70 (dd,
J6′,5 = 3.6 Hz, J4,5 = 10.0 Hz, 1H, H5), 4.07 (m, 1H, H8), 4.48 (ddd, J9′,9
= 14.0 Hz, J8,9 = 4.0 Hz, JOH‑9 = 2.4 Hz, 1H, H9), 4.74 (ddd, J9′,9 = 14.0
Hz, J8,9′ = 4.0 Hz, JOH‑9′ = 2.4 Hz, 1H, H9′), 5.10 (d, J2,1 = 3.6 Hz, 1H,
H1), 7.12 (d, J5,4 = 1.2 Hz, 1H, H4 Im), 7.50 (d, J4,5 = 1.2 Hz, 1H, H5,
Im). 13C NMR (CDCl3): δ 52.14 and 52.19 (C7, two diastereomers),
68.78 (C9), 68.83 (C8), 70.33 and 70.38 (C6, dual resonances due to
diastereomers), 70.73 (C4), 72.62 (C5), 72.67 (C2), 73.62 and 73.65
(C3, dual resonances due to diastereomers), 92.87 (C1), 126.90 (C4
Im), 128.09 (C5 Im), 145.33 (C2, Im).
α-7 (8R and 8S). Retention time, 11.02 min. 1H NMR (CDCl3): δ
2.00, 2.01, 2.07 (s, 12H of CH3, 8S stereomer) and 2.03, and 2.04 (s,
12H, of 4 CH3, for 8R stereomer), 2.20 (s, 3H, COCH3 at C8), 3.44
(dd, J6,5 = 5.2 Hz, J6′,6 = 11.2 Hz, H6, both 8S and 8R diastereomers),
3.56 (dd, J6,6′ = 11.2 Hz, J6′,5 = 4.0 Hz, H6′, both diastereomers), 3.57
(dd, J8,7 = 5.6 Hz, J7′,7 = 10.8 Hz, H7, both diastereomers), 3.64 (dd,
β-8 (8R and 8S). Retention time 6.97 min. Isomeric composition
47.6% by HPLC. 1H NMR (CD3OD): δ 3.14 (dd, J1,2 = 7.6 Hz, J3,2
=
9.2 Hz, 1H, H2), 3.28−3.34 (m, J5,4 = J3,4 = 10.0 Hz, 2H, H3 and H4),
3.40 (m, 1H, H5), 3.54 (dd, J7′,7 = 9.6 Hz, J8,7 = 4.0 Hz, 1H, H7), 3.66
(two dd, merged, J6,5 = 3.2 Hz, J5,6′ = 2.4 Hz, J6′,6 = 9.6 Hz, 2H, H6 and
H6′), 3.71 (dd, J7,7′ = 9.6 Hz, J8,7′ = 4.0 Hz, 1H, H7′), 3.86 (m, 1H,
H8), 4.44 (ddd, JOH,9 = 3.2 Hz, J8,9 = 4.8 Hz, J9,9′ = 14.6 Hz, 1H, H9),
4.48 (dd, J2,1 = 8.0 Hz, 1H, H1), 4.70 (ddd, JOH,9′ = 3.2 Hz, J8,9′ = 5.6
Hz, J9,9′ = 14.6 Hz, 1H, H9′), 7.11 (d, J5,4 = 1.2 Hz, 1H, H4 Im), 7.46
(d, J4,5 = 1.2 Hz, 1H, H5, Im). 13C NMR (CD3OD): δ 52.24 and 52.26
(C7, two diastereomers), 68.78 (C9, dual resonances due to
diastereomers), 68.86 (C8), 70.54 and 70.63 (C6, two diastereomers),
70.78 (C4), 75.07 and 75.09 (C2, dual resonances due to
diastereomers), 75.58 (C3), 76.81 and 76.85 (C5), 97.09 (C1),
126.90 (C4 Im), 128.20 (C5 Im), 145.33 (C2, Im).
J7,7′ = 10.8 Hz, J8,7′ = 4.0 Hz, H7′, both diastereomers), 4.03 (dd, J6,5
=
5.2 Hz, J6′,5 = 4.0 Hz, J4,5 = 9.6 Hz, H5 for both diastereomers), 4.53
(dd, J8,9 = 6.8 Hz, J9′,9 = 14.2 Hz, 2H, H9 and H9′ for 8S and 8R
diastereomers), 4.90 (dd, J9′,9 = 14.2 Hz, J8,9′ = 3.6 Hz, 2H, H9′ and H9
for 8S and 8R diastereomers), 5.09 (two merged dd, J1,2 = 3.6 Hz, J3,2
=
10.0 Hz, each for 1H, H2 for both diastereomers), 5.19 (dd, J5,4 = 9.6
Hz, J3,4 = 9.2 Hz, 1H, H4, 8S diastereomer) and 5.22 (dd, J5,4 = 9.6 Hz,
J3,4 = 9.2 Hz, 1H, H4, 8R diastereomer), 5.28 (ddd, J9,8 = 6.8 Hz, J9′,8
=
3.6 Hz, J7′,8 = 4.0 Hz, J7,8 = 5.6 Hz, H8, both diastereomers), 5.48 (dd,
J2,3 = 10.0 Hz, J4,3 = 9.2 Hz, 1H, H3, 8S diastereomer) and 5.49 (dd,
Synthesis of Methyl-1-α-D-(2,3,4-tri-O-benzyl)-6-O-(9-[2-
nitro-1H-imidazolyl]-8S-fluoropropyl)glucopyranose, and
Methyl-1-α-D-(2,3,4-tri-O-benzyl)-6-O-(9-[2-nitro-1H-imidazol-
yl]-8R-fluoropropyl)glucopyranose (9). DAST (3.24 g, 20.1
mmol) was added dropwise to a cooled (ca. −20 °C) stirred solution
of 3 (8.50 g, 13.4 mmol) in anhydrous THF (250 mL). The reaction
mixture was allowed to warm up to 22 °C and stirred for an additional
20 h. After cooling to −10 °C, the reaction mixture was quenched by
MeOH (20 mL), the solvent was removed, and the residual red oil was
redissolved in EtOAc (250 mL). The resultant solution was washed
with saturated aqueous NaHCO3 (250 mL) and brine (250 mL) and
dried over MgSO4. Filtration of the solution through a pad of silica gel
(for TLC), followed by removal of the solvent from the filtrate under
reduced pressure afforded red oil (8.9 g) that upon dry silica gel
chromatography (EtOAc−hexanes 20%→ 50%) afforded pure
diastereomeric mixture of 9 as yellow oil. Yield: 6.62 g (78%).
HPLC: 97.89% (combined chemical purity for two diastereomers). 1H
NMR (CDCl3): δ 3.38 (s, 3H, OCH3, both diastereomers), 3.50 and
3.54 (two dd, J5,4 = 9.6 Hz, J3,4 = 9.6 Hz, 1H each, H4, 8S and 8R
diastereomers), 3.52 and 3.55 (two dd, J1,2 = 3.6 Hz, J3,2 = 10.0 Hz, 1H
each, H2, 8S and 8R diastereomers), 3.63 (dd, J5,6 = 4.4 Hz, J6,6′ = 11.2
Hz, H6, both diastereomers), 3.68 (dd, J5,6′ = 2.8 Hz, J6,6′ = 11.2 Hz,
J2,3 = 10.0 Hz, J4,3 = 9.2 Hz, 1H, H3, 8R diastereomer), 6.32 (d, J2,1
=
3.6 Hz, H1 for both diastereomers), 7.14 (s, 1H, H4, Im, 8S
diastereomer) and 7.19 (s, 1H, H4, Im, 8R diastereomer), 7.20 (s, 1H,
H5, Im, 8S diastereomer) and 7.27 (s, 1H, H5, Im, 8R diastereomer).
13C NMR (CDCl3): δ 20.71−21.13 (CH3S), 49.67 and 49.72 (C7, two
diastereomers), 68.18 and 68.24 (C6, two diastereomers), 69.40 and
69.42 (C9, for two diastereomers), 69.60 and 69.63 (C4, two
diastereomers), 69.77 (C8), 70.13 and 70.19 (C5 for two
diastereomers), 70.30 and 70.39 (C2, two diastereomers), 71.13 and
71.16 (C3 for two diastereomers), 89.30 (C1), 127.08 and 127.16 (C4
Im for two diastereomers), 128.40 and 128.45 (C5 Im for two
diastereomers), 145.67 (C2, Im), 169.14, 169.66, 169.73, 169.91,
169.96, 169.98, 170.43, 170.45 (8 s, representing 4 CO for two
diastereomers).
β-7 (8R and 8S). Retention time, 11.02 min. The presence of these
diastereomers was detected in ∼13% (calculated by NMR integration
and HPLC chromatographic analysis). Proton signals that appeared at
identifiable, and separate chemical shifts from the α-anomer are
1
reported. H NMR (CDCl3): δ 1.98, 2.03, 2.06 and 2.12 (four s, 4 ×
CH3), 5.14 (H2), 5.23 (H8), 5.68 (d, J2,1 = 8.0 Hz, H1), 7.11 and 7.15
(H4, Im), 7.23 and 7.25 (H5, Im). 13C NMR (CDCl3) δ: 21.04−21.13
(signals for CH3s), 49.55 (C6, two diastereomers), 73.04 and 73.10
(C3, two diastereomers), 74.09 and 74.16 (C5 for two diastereomers),
H6′, both diastereomers), 3.73 (ddd, J7′,7 = 11.4 Hz, J8,7 = 6.4 Hz, JF,7
=
31.2 Hz, H7, both diastereomers), 3.75 (ddd, J4,5 = 9.6 Hz, J6,5 = 4.4
6042
dx.doi.org/10.1021/jm2017336 | J. Med. Chem. 2012, 55, 6033−6046

Journal of Medicinal Chemistry
Article
Hz, J6′,5 = 2.8 Hz, H5, both diastereomers), 3.79 (ddd, J7′,7 = 11.4 Hz,
Im), 169.08−170.47 (COCH3). 19F NMR (CDCl3): δ 192.33 (dtt, JF,7
= 31.2 Hz, JF,9 = 16.2 Hz, JF,8 = 47.5 Hz) for 8R diastereomer, and
194.06 (dtt, JF,7 = 31.2 Hz, JF,9 = 16.2 Hz, JF,8 = 47.0 Hz) for 8S
diastereomer.
J8,7′ = 6.4 Hz, JF,7′ = 31.0 Hz, H7′, both diastereomers), 4.00 (dd, J4,3
9.2 Hz, J2,3 = 10.0 Hz, H3, both diastereomers), 4.53 and 4.58 (J9′,9
=
=
14.4 Hz, J8,9 = 6.8 Hz, J8,9′ = 8.4 Hz, JF,H = 20.0 Hz, H9 and H9′, both
diastereomers), 4.59 (dd, J2,1 = 3.6 Hz, H1, both diastereomers), 4.60
(d, J2a,b = 11.2 Hz, H2a, Bn, both diastereomers), 4.67 and 4.69 (d,
J3a,3b = 12.0 Hz, diastereomeric H3a, Bn, both diastereomers), 4.75−
4.88 (complex m, H4a, H4b of Bn, and H8, both diastereomers), 4.91
(d, J3a,3b = 12.0 Hz, H3b, Bn, both diastereomers), 4.99 (d, J2a,2 b = 11.2
Hz, 1H, H2b, Bn, both diastereomers), 7.11 (s, H4, Im, 8S
diastereomer) and 7.12 (s, H4, Im, 8R diastereomer), 7.13 (s, 1H,
H4, Im, 8S diastereomer) and 7.14 (s, 1H, H4, Im, 8R diastereomers)
and 7.27−7.38 (m, 30H, each diastereomer, 3 × Ph). 19F NMR
(CDCl3): δ −191.88 (m). ESI-MS (C34H38FN3O8): m/z 635 (M +
1)+.
β-10 (8R and 8S). Retention times for two diastereomers 9.19
1
(2.0%) and 9.34 (9.5%) min. H NMR (CDCl3): NMR signals for
both R and S diastereomers are observed in the spectrum, but being
very low in the composition most of the signals were of very low
intensity and buried in the baseline. Clearly identified signals are
indicated; δ 4.66 (H8), 4.84 (H9), 4.87 (H9′), 5.70 (d, J2,1 = 8.4 Hz,
1H, H1). 13C NMR (CDCl3): δ 20.79 and 21.03 (COCH3s), 68.02
(C7), 73.05 and 73.22 (C3), 74.05 and 74.42 (C5), 92.02 and 92.05
(C1), 128.42 (C4 Im), 129.5 (C5 Im). Other carbon signals appeared
to be buried in the baseline due to low amounts of this isomer. 19F
NMR (CDCl3): δ 192.33 ((dtt, JF,7 = 31.2 Hz, JF,9 = 16.2 Hz, JF,8
47.5 Hz) for 8R diastereomer, and −194.06 (dtt, JF,7 = 31.2 Hz, JF,9
16.2 Hz, JF,8 = 47.0 Hz) for 8S diastereomer.
=
=
Synthesis of 1-α/β-D-(1,2,3,4-Tetra-O-acetyl)-6-O-(9-[2-nitro-
1H-imidazolyl]-8S-fluoropropyl)glucopyranose and 1-α/β-D-
(1,2,3,4-Tetra-O-acetyl)-6-O-(9-[2-nitro-1H-imidazolyl]-8R-
fluoropropyl)glucopyranose (10; fluoroglucoazomycin ace-
tate). Trimethylsilyl triflate (25.5 g, 114.6 mmol) was added dropwise
to a stirred cooled (ca. −40 °C) solution of 9 (6.62 g, 10.4 mmol) in
acetic anhydride (117 mL). The reaction solution was allowed to
warm up to 22 °C and stirred for additional 14 h. After cooling to −10
°C, the dark reaction solution was diluted with cold EtOAc (400 mL)
and poured in a cold (0 °C) saturated aqueous NaHCO3 (500 mL).
The resultant mixture was stirred for 0.5 h at room temperature. The
organic phase was separated off and washed with saturated aqueous
NaHCO3 (10 × 200 mL) and brine (200 mL) and dried over MgSO4.
Filtration through a pad of silica gel (for TLC) and removal of the
solvent from the filtrate provided 8.9 g of dark oil, which was subjected
to dry chromatography (30% EtOAc−hexanes → 80% EtOAc−
hexanes) to afford 3.69 g of 7 as a yellow solid. Another purification of
this material by dry chromatography (5% MeCN−toluene → 30%
MeCN−toluene) provided diastereomeric (R and S) mixture of pure
Synthesis of 1-α/β-D-(6-O-(9-[2-Nitro-1H-imidazolyl]-8S-
fluoropropyl)glucopyranose and 1-α/β-D-(6-O-(9-[2-Nitro-1H-
imidazolyl]-8R-fluoropropyl)glucopyranose (11). Compound 10
(2.66 g, 5.12 mmol) was dissolved in 0.1 M NaOH in methanol (256
mL). After 20 min of stirring at room temperature, the reaction
solution was treated with Dowex 50W × 4−200 (H+, 20 g) to adjust
the pH of the solution to a value of 5. The resin was filtered off, and
the filtrate was evaporated under reduced pressure to dryness to
provide 1.43 g of a yellowish solid, which was subjected to dry
chromatography (2% MeOH−CHCl3 → 25% MeOH−CHCl3) to
afford pure isomeric mixture of 11 as a pale solid (α/β ratio, 10:9 by
HPLC and 1H NMR). Yield: 1.26 g (70%). HPLC: 98.4% (combined
chemical purity for all diastereomers). ESI-MS: m/z 352 (M + 1)+.
Anal. Calcd (C12H18FN3O8·0.5H2O): C, 41.03; H, 5.41; N, 11.96.
Found: C, 41.03; H, 5.16; N, 11.96.
α-11 (8R and 8S). Retention times 6.37 and 6.46 min (isomeric
composition by HPLC, 53.3% for αR and αS diastereomers; mobile
1
α/β-isomers of 10. α/β ratio 91:9 by H NMR analyses. Yield: 2.92 g
1
phase A). H NMR (CD3OD): δ 3.31 (dd, J5,4 = J3,4 = 10.0 Hz, 1H,
(54%). HPLC: 95.9% (combined chemical purity for all diaster-
eomers). MS (ES+): m/z 520 (M + 1)+. Anal. (C20H26FN3O12) CHN.
α-10 (8R and 8S). Retention times for two diastereomers 11.08
H4), 3.38 (dd, J1,2 = 3.6 Hz, J2,3 = 8.4 Hz, 1H, H2), 3.67 (dd, J4,3 = J2,3
= 8.4 Hz, 1H, H3), 3.73 (dd, J6,5 = 2.8 Hz, J6,6′ = 11.2 Hz, 1H, H6),
3.76 (dd, J6,5 = 2.8 Hz, J6,6′ = 11.2 Hz, 1H, H6′), 3.78 (ddd, J7′,7 = 10.0
Hz, J8,7 = 8.0 Hz, JF,H = 29.7 Hz, 1H, H7), 3.81 (ddd, J7,7′ = 10.0 Hz,
1
(74.5%) and 12.57 (12%) min. H NMR (CDCl3): δ 2.02, 2.06, 2.06,
and 2.19 (4 s, each for 3H of 4 CH3s, 8S diastereomer) and 2.03, 2.04,
2.08 and 2.20 (4 s, each for 3H of 4 CH3s, 8R diastereomer), 3.55 (dd,
J8,7′ = 8.0 Hz, JF,H = 29.7 Hz, 1H, H7′), 3.89 (ddd, J6,5 = 2.8 Hz, J4,5
=
10.0 Hz, 1H, H5), 4.74 (ddd, J8,9 = 7.6 Hz, J9,9′ = 11.2 Hz, JF,9 = 16.0
Hz, 1H, H9, both diastereomers), 4.86 (ddd, J9′,9 = 11.2 Hz, J8,9′ = 3.6
Hz, JF‑9′ = 16.0 Hz, 1H, H9′, both diastereomers), 5.05 (dddd, J8,7 = J8,7′
= 8.0 Hz, J8,9 = 7.6 Hz, J8,9′ = 3.6 Hz, JF‑8 = 48.0 Hz, H8, both
diastereomers), 5.10 (d, J2,1 = 3.6 Hz, H1, both diastereomers), 7.16 (s,
H4 Im, both diastereomers), 7.52 (s, H5 Im, both diastereomers). 13C
J6′,6 = 11.6 Hz, J5,6 = 4.4 Hz, 1H, H6, 8S diastereomer), 3.64 (dd, J6′,6
=
11.6 Hz, J5,6 = 3.6 Hz, 1H, H6, 8R diastereomer), 3.66 (dd, J5,6 = 2.0
Hz, J6,6′ = 8.8 Hz, 1H, H6′, 8S diastereomer), 3.72 (dd, J5,6′ = 3.2 Hz,
J6′,6 = 8.8 Hz, 1H, H6′, 8R diastereomer), 3.80 (ddd, J7,7′ = 11.2 Hz, J8,7
= 3.6 Hz, JF,H = 30.4 Hz, 1H, H7, both diastereomers), 3.86 (ddd, J7′,7
= 11.2 Hz, J8,7′ = 2.8 Hz, JF,H = 31.2 Hz, 1H, H7′, both diastereomers),
4.04 (complex ddd, J4,5 = 10.4 Hz, H5, both diastereomers), 4.63 (ddd,
J8,9 = 5.6 Hz, J9′,9 = 14.4 Hz, JF,9 = 16.2 Hz, 1H, H9, 8S diastereomer),
4.67 (ddd, J8,9 = 8.0 Hz, J9′,9 = 14.4 Hz, JF,9 = 16.2 Hz, 1H, H9, 8R
diastereomer), 4.81 (ddd, J8,9′ = 7.6 Hz, J9,9′ = 14.4 Hz, JF,9′ = 16.2 Hz,
1H, H9′,8S diastereomer), 4.88 (ddd, J8,9′ = 4.4 Hz, J9′,9 = 14.4 Hz, JF‑9′
= 16.2 Hz, 1H, H9′, 8R diastereomers), 4.92 (dm, JF,9 = 47.0 Hz, H8,
both diastereomers), 5.09 (dd, J1,2 = 3.6 Hz, J3,2 = 10.0 Hz, 1H, H2, 8S
diastereomer) and 5.11 (dd, J1,2 = 3.2 Hz, J3,2 = 10.0 Hz, 1H, H2, 8R
diastereomer), 5.19 (dd, J2,3 = J3,4 = 9.6 Hz, 1H, H3, 8S diastereomer),
5.26 (dd, J2,3 = 10.0 Hz, J3,4 = 9.6 Hz, 1H, H3, 8R diastereomer), 5.49
(dd, J3,4 = J5,4 = 10.0 Hz, 1H, H4, 8S diastereomer), 5.50 (dd, J3,4 = J5,4
= 10.0 Hz, 1H, H4, 8R diastereomer), 6.32 (two d merged, J2,1 = 3.2
Hz, 1H, H1, 8S diastereomer), 6.34 (d, J2,1 = 4.0 Hz, 1H, H1, 8R
diastereomer), 7.17 (s, H4 Im, 8S diastereomer) and 7.18 (s, H4 Im,
8R diastereomer), 7.24 (s, H5 of Im, 8S diastereomer) and 7.35 (s, H5
of Im, 8R diastereomer). 13C NMR (CDCl3): δ 20.69, 20.91 and 21.11
(COCH3s), 50.67 and 50.73 (JF,7 = 23 Hz, C7 for two diastereomers),
68.17 and 68.36 (C6 for two diastereomers), 69.40 and 69.44 (C4),
69.95 and 70.09 (C5), 70.21 and 70.26 (C2, dual resonances), 70.46
and 70.61 (C9, two d, JF,9 = 20 Hz, dual resonances), 71.17 and 71.24
(C3, dual resonances), 89.29 and 89.35 (C1, dual resonances), 90.31
and 90.88 (C8, two d, JF,8 = 177 Hz), 127.60 and 127.89 (C4 Im, two
resonances), 128.68 and 128.72 (C5 Im, dual resonances), 145.00 (C2
NMR (CD3OD): δ 54.09 (C7, d, JF‑C = 23.5 Hz), 74.15 (C9, d, JF‑C
=
21.4 Hz), 74.46 and 74.53 (C6, two diastereomers), 74.97 (C4), 76.55
(C5), 77.64 (C2), 79.00 (C3), 94.61 (C8, d, JF,C = 176.1 Hz), 96.76
(C1), 131.25 (C4 Im), 131.93 (C5 Im), 149.06 (C2, Im). 19F NMR
(CD3OD): δ −194.26 (dtt, JF,8 = 47.0 Hz, J7,F = 29.7 Hz, JF,9 = 16.0
Hz).
β-11 (8R and 8S). Retention time 5.33 min (for both βR and βS
1
diastereomers; mobile phase A). H NMR (CD3OD): δ 3.14 (dd, J1,2
= 6.0 Hz, J3,2 = 8.4 Hz, 1H, H2), 3.33−3.37 (m, 2H, H3 and H4), 3.36
(ddd, J6,5 = 4.0 Hz, J6′,5 = 1.2 Hz, J4,5 = 9.2 Hz, 1H, H5), 3.68 (dd, J6,5
=
4.0. Hz, J6′,6 = 9.2 Hz, 1H, H6), 3.71 (dd, J5,6′ = 2.4 Hz, J6′,6 = 9.2 Hz,
1H, H6′), 3.78 (ddd, merged, J7′,7 = 9.6 Hz, J8,7 = 8.0 Hz, J7,F = 16.0 Hz,
1H, H7), 3.87 (ddd, J7′,7 = 9.6 Hz, J8,7′ = 8.0 Hz, J7′,F = 27.5 Hz, 1H,
H7′), 4.47 (d, J2,1 = 6.0 Hz, 1H, H1), 4.75 (ddd, J8,9 = 7.6 Hz, J9,9′
11.2 Hz, JF,9 = 22.8 Hz, 1H, H9, both diastereomers), 4.83 (ddd, J9′,9
=
=
11.2 Hz, J8,9′ = 3.6 Hz, JF‑9′ = 16.0 Hz, 1H, H9′, both diastereomers),
4.90 (dddd, J8,7 = J8,7′ = 8.0 Hz, J8,9 = 7.6 Hz, J8,9′ = 3.6 Hz, JF‑8 = 48.0
Hz, 1H, H8 for both diastereomers), 7.15 (d, J5,4 = 1.2 Hz, 1H, H4
Im), 7.50 (d, J4,5 = 1.2 Hz, 1H, H5, Im). 13C NMR (CD3OD): δ 54.18
(C7, d, JF‑C = 22.7 Hz), 74.27 (C9, d, JF‑C = 23.5 Hz), 74.53 and 74.89
(C6, two diastereomers), 75.13 (C4), 79.63 (C2), 79.69 (C3), 80.85
(C5), 94.57 (C8, d, JF,C = 176.9 Hz), 100.99 (C1), 131.25 (C4 Im),
132.04 (C5 Im), 149.06 (C2, Im). 19F NMR (CD3OD): δ −194.26
6043
dx.doi.org/10.1021/jm2017336 | J. Med. Chem. 2012, 55, 6033−6046

Journal of Medicinal Chemistry
Article
(dtt, for both diastereomers, J7,F = 27.5 Hz, JF,8 = 47.0 Hz, JF,9 = 16.0
Hz).
= 10.0 Hz, 1H, H4), 3.42 (ddd, J6,5 = 5.5 Hz, J6′,5 = 2.9 Hz, J4,5 = 10.0
Hz, 1H, H5), 3.84 (dd, J6,5 = 5.5 Hz, J6′,6 = 10.9 Hz, 1H, H6), 3.92 (dd,
J6′,5 = 2.9 Hz, J6,6′ = 10.9 Hz, 1H, H6′), 4.26 (m, 2H, H7 and H7′), 4.46
(d, J2,1 = 8.8 Hz, 1H, H1), 6.24 (dd, J7,8 = 5.6 Hz, J9,8 = 14.0 Hz. 1H,
H8), 7.17 (s, 1H, H4, Im), 7.58 (d, J8,9 = 14.0 Hz, 1H, H9), 7.72 (s,
1H, H5, Im). 13C NMR (CD3OD): δ 68.39 (C6), 69.96 (C7), 70.68
(C4), 75.05 (C2), 75.82 (C3), 76.93 (C5), 97.07 (C1), 123.78 (C8),
124.13 (C9), 125.70 (C4, Im), 127.98 (C5, Im).
Lipophilicity. Estimated lipophilicity for the compounds, ex-
pressed as ClogP, was calculated using ChemDraw (CambridgeSoft/
Perkin-Elmer).
Biological Evaluations. Human mammary MOO6X, cervical
HeLa, and murine mammary EMT6, KBALB, and KBALB-STK cancer
cells were used for the biological studies. DMEM/F12 medium and
Garth medium (depending on the study) were used for cell work. Test
compounds were dissolved in sterile water and were serially diluted to
generate appropriate test concentrations. High purity gases were used
for radiosensitization studies.
Synthesis of 1-α-D-(1,2,3,4-Tetra-O-acetyl)-6-(((8,9E)-9-(2-
nitro-1H-imidazol-1-yl)allyloxy)methyl)glucopyranose and 1-
β-D-(1,2,3,4-Tetra-O-acetyl)-6-(((8,9E)-9-(2-nitro-1H-imidazol-1-
yl)allyloxy)methyl)glucopyranose (12; allylglucoazomyin ac-
etate). DBU (3.14 g, 20.6 mmol) was added dropwise to a stirred
solution of nosylate 5 (7.25 g 10.3 mmol) in anhydrous acetonitrile
(50 mL) at 22 °C. The dark reaction solution was stirred for 22.5 h
and then concentrated under reduced pressure. The dark oily residue
was treated with EtOAc (200 mL) and water (100 mL). The organic
phase was separated off and washed with water (2 × 100 mL), 1N HCl
(50 mL), saturated aqueous NaHCO3 (100 mL), and brine (100 mL).
After drying over MgSO4, the resultant solution was filtered through a
pad of silica gel (for TLC) and the solvent was removed under
reduced pressure to provide 3.8 g of a yellowish solid, which was
subjected to dry chromatography (5% MeCN−toluene → 30%
MeCN−toluene) to afford a mixture of α- and β-isomers (α/β
1
isomeric ratio 13:1 by H NMR spectral analysis) of 12 (87%) as
Cytotoxicity. The MTT-assay was used to estimate the toxicity of
8, 11, and 13 to human HeLa, MOO6X, K-BALB, and K-BALB-STK
cancer cells and murine EMT-6 cell lines in cell culture.21
Exponentially growing cells were trypsinized, centrifuged, and
suspended in growth medium and the cell counts were readjusted to
8 × 103 cells/mL. The cells were seeded into 96-well plates at a count
of 8 × 102 cells/well and incubated in 37 °C, 5% CO2 for 24 h. The
test compounds (stock solution 2 mM in 95% ethanol) were added to
the growth medium and exposed to the cells in 96-well plates at a final
volume of 300 μL to produce the required dilution of the experimental
design. Control wells were filled with 100 μL of medium. These plate
groups were incubated for 3 days at 37 °C in a humidified atmosphere
containing 95% air and 5% CO2. At the end of incubation, 50 μL
solution of 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H tetrazolium
bromide (MTT), predissolved in PBS (5 mg/mL), filtered through a
0.45 μm membrane, and diluted 1:5 in prewarmed medium, was added
to each well. The plates were incubated at 37 °C for 4 h. The medium
was then removed from the wells and 150 μL of dimethylsulfoxide was
added to each well. The plates were placed on a shaker for 15 min to
dissolve the formazan crystal and then read immediately at 540 nm on
a scanning multiwall spectrophotometer. Data are summarized in
Table 2. Data were plotted on Microsoft Excel.
yellow solid. Yield: 1.425 g (28%).
α-12. H NMR (CDCl3): δ 2.02, 2.03, 2.04 and 2.18 (s, 12H, 4 ×
CH3), 3.57 (dd, J6,5 = 4.4 Hz, Jgem = 10.8 Hz, 1H, H6), 3.66 (dd, J6′,5
1
=
2.4 Hz, Jgem = 10.8 Hz, 1H, H6′), 4.07 (ddd, J6,5 = 4.4 Hz, J6′,5 = 2.2 Hz,
J4,5 = 10 Hz, 1H, H5), 4.16 (dd, J7′,7 = 13.2 Hz, J8,7 = 6.0 Hz, 1H, H7),
4.23 (dd, J7,7′ = 13.2 Hz, J8,7′ = 6.0 Hz, 1H, H7′), 5.09 (dd, J1,2 = 3.2 Hz,
J3,2 = 10.0 Hz, 1H, H2), 5.22 (dd, J2,3 = 10 Hz, J4,3 = 9.2 Hz, 1H, H4),
5.49 (dd, J3,4 = 9.2 Hz, J5,4 = 10.0 Hz, 1H, H3), 6.02 (dt, J7,8 = J7′,8 = 6.0
Hz, J9,8 = 14.0 Hz, 1H, H8), 6.33 (d, J2,1 = 3.2 Hz, 1H, H1), 7.19 (s,
1H, H4, Im), 7.33 (s, 1H, H5, Im), 7.58 (d, Jvic = 14.0 Hz, 1H, H9).
β-12. 1H NMR (CDCl3): Only the specific proton signals visible in
the spectrum of the isomeric mixture are being described, the signals
for remaining carbons in the molecule were buried under the signals
belonging to the α-12; δ 2.03, 2.11 and 2.36 (s, 12H, 4 × CH3), H6
(buried along with the signal for α-isomer), 3.7 (dd, unresolved, H6′),
5.14 (dd, J1,2 = 8.0 Hz, J3,2 = 9.2 Hz, 1H, H2), 5.28 (dd, J2,3 = J4,3 = 9.2
Hz, 1H, H3), H4 (buried under the signal for α-isomer), 5.70 (d, J2,1
=
8.0 Hz, 1H, H1), 6.04 (dt, J9,8 = 14.0 Hz, J7,8 = 6.0 Hz, 1H, H8), 7.17
(s, 1H, H4, Im), 7.24 (s, 1H, H5 Im), 7.56 (d, 1H, J9,8 = 14.0 Hz, 1H,
H9).
1-α-D-(6-(((8,9E)-9-(2-Nitro-1H-imidazol-1-yl)allyloxy)-
methyl)glucopyranose and 1-β-D-(6-(((8,9E)-9-(2-Nitro-1H-imi-
dazol-1-yl)allyloxy)methyl)glucopyranose (13). Compound 12
(1.358 g, 2.72 mmol) was dissolved in 0.1 M NaOH in methanol (136
mL). After 20 min of stirring at 22 °C, the reaction solution was
treated with Dowex 50W × 4−200 (H+, 5 g) to adjust the pH of the
solution to 5. The resin was filtered off, and the filtrate was evaporated
under reduced pressure to dryness to provide 0.716 g of a yellowish
solid that was subjected to column chromatography (2% MeOH−
CHCl3 → 30% MeOH−CHCl3) to afford pure isomeric mixture of 13
Radiosensitization. Cell radiosensitization was determined using
a
60Co X-ray source together with a clonogenic survival assay22 using
HeLa, EMT-6, and MOO6X cancer cell lines. Cells (300000 in 3 mL
of DMEM/F12 medium per T60 Petri dish) were individually
incubated under 5% CO2 in air at 37 °C for 24 h. The dishes were
assigned to either the control (normoxic) or hypoxic groups, the test
substance (stock solution 10 mM in 95% ethanol) was added to these
groups to achieve a concentration of 100 μM, and the incubation was
allowed for 30 min. Those in the hypoxic group were degassed to
hypoxia by six consecutive vacuum and nitrogen fill cycles in a vacuum
chamber. The Petri dishes (hypoxic and normoxic controls) were
incubated for 30 min on an oscillating shaker at R/T × 60 cycles per
min and then irradiated in a 60Co γ-irradiator at 0 (control), 4, 8, 12,
16, and 20 Gy in N2 (hypoxic subgroup) and air chambers (normoxic
subgroup). The cells were then recovered from each dish by two
consecutive washes with PBS followed by the addition of trypsin (500
μL) and quenching with fresh medium (4.5 mL). Cells were then
plated at densities from 100 to 15000 cells/5 mL of medium for
normoxic conditions and 100 and 5000 cells/5 mL of medium for
hypoxic conditions. The cells were incubated for 10−14 days at 37 °C
under 5% CO2 and then stained with methylene blue or crystal violet
in ethanol, clones counted, and surviving fractions calculated. Tests
were done in triplicate. SERs, presented in Table 3, were derived from
cell survival plots on Microsoft Excel.
1
as a pale solid (α and β, isomeric ratio 71.4:22.6 determined by H
NMR). HPLC purity: 95.4% (combined chemical purity for all
diastereomers). Yield: 0.55 g (61%); mp (anomeric mixture) 156−158
°C; λmax 324 nm. ESI-MS: m/z 331.00 (M + 1)+. Anal. Calcd
(C12H17N3O8·4/5H2O): C 41.69, H 5.42, N 12.16. Found: C 41.71, H
5.12, N 12.02.
α-13. Retention time, 8.33 min (mobile phase A). Isomeric
1
composition 71.4% by HPLC. H NMR (CD3OD): δ 3.32 (dd, J5,4
=
J3,4 = 10.0 Hz, 1H, H4), 3.34 (dd, J2,3 = 9.8 Hz, J1,2 = 3.7 Hz, 1H, H2),
3.67 (dd, J4,3 = J2,3 = 9.8 Hz, 1H, H3), 3.77 (dd, J4,5 = 10.0 Hz, J6,5
=
5.5 Hz, 1H, H5), 3.73 (dd, J6,5 = 2.4 Hz, J6′,6 = 10.9 Hz, 1H, H6), 3.76
(dd, J6′,5 = 2.7 Hz, J6,6′ = 10.9 Hz, 1H, H6′), 4.26 (m, 2H, H7 and H7′),
5.10 (d, J2,1 = 3.7 Hz, 1H, H1), 6.28 (dd, J7,8 = 3.6 Hz, J9,8 = 14.0 Hz,
1H, H8), 7.18 (s, 1H, H4, Im), 7.62 (d, J8,9 = 14.0 Hz 1H, H9), 7.71
(s, 1H, H5, Im). 13C NMR (CD3OD): δ 68.31 (C6), 69.96 (C7),
70.47 (C4), 70.88 (C5), 72.64 (C2), 73.69 (C3), 92.84 (C1), 123.69
(C8), 124.06 (C9), 125.68 (C4, Im), 127.98 (C5, Im) and 143.97
(C2, Im).
Transmembrane Transport Studies. Xenopus oocytes expressing
recombinant GLUT-1 or GLUT-2 were prepared as reported
elsewhere.41 Oocytes were injected with either water (20 nL),
containing RNA transcript (1 ng/nL) encoding the required GLUT,
or water (10 nL) alone (control). Injected oocytes were then
incubated (4 days; 18 °C) in modified Barth’s medium (MBM)
β-13. Retention time, 7.94 min (mobile phase A). Isomeric
1
composition 22.6% by HPLC. H NMR (CD3OD): δ 3.14 (dd, J1,2
= 8.8 Hz. J3,2 = 10 Hz. 1H, H2), 3.30 (m, 1H, H3), 3.32 (dd, J5,4 = J3,4
6044
dx.doi.org/10.1021/jm2017336 | J. Med. Chem. 2012, 55, 6033−6046

Journal of Medicinal Chemistry
Article
(changed daily) (88 mM NaCl, 1 mM KCl, 0.33 mM Ca(NO3)2, 0.41
mM CaCl2, 0.82 mM MgSO4, 2.4 mM NaHCO3, 10 mM Hepes, 2.5
mM sodium pyruvate, 0.1 mg/mL penicillin, and 0.05 mg/mL
gentamycin sulfate; pH 7.5) prior to the assay. Radioisotope influx and
electrophysiology experiments were performed in Na+-containing
transport medium (100 mM NaCl, 2 mM KCl, 1 mM CaCl2, 1 mM
MGAZl2, 10 mM Hepes; pH 7.5). For radioisotope influx studies,
uptake in Xenopus oocytes was traced using radiolabeled permeant
([14C]glucose; 1 μCi/mL), each assay using an incubation period of 20
min at 20 °C on groups of 10−12 oocytes in transport medium. At the
end of the incubation period, extracellular label was removed by six
rapid washes in ice-cold transport medium, and individual oocytes
were dissolved in 1% (w/v) SDS for quantifying oocyte-associated
radioactivity by liquid scintillation counting (LS 6000 IC, Beckman,
Fullerton, CA, USA). Data plots, showing the inhibition of
[14C]glucose uptake as a function of inhibitor concentration for
GLUT-1, are shown (Figure 4). The median effective concentration
(EC50) for F-GAZ, defined as the concentration of inhibitor that can
be expected to cause a 50% reduction in [14C]glucose uptake in
oocytes expressing GLUT-2, was estimated using the four-parameter
logistic analysis and plotted on a log[conc] axis (Figure 5).
ment ratio; PBS, phosphate buffer saline; s, singlet; SER,
sensitization enhancement ratio; TMS-triflate, trimethylsilyltri-
fluoromethanesulfonate; t, triplet; tlc, thin layer chromatog-
raphy
REFERENCES
■
(1) Dowdy, A. H.; Bennett, L. R.; Chastain, S. M. Protective action of
anoxic anoxia against total body roentgen irradiation of mammals.
Radiology 1950, 55, 879−885.
(2) Gray, L. H.; Conger, A. D.; Ebert, M.; Hornsey, S.; Scott, O. C.
The concentration of oxygen dissolved in tissues at the time of
irradiation as a factor in radiotherapy. Br. J. Radiol. 1953, 26 (312),
638−648.
(3) Thomlinson, R. H.; Gray, L. H. The histological structure of
some human lung cancers and the possible implications for
radiotherapy. Br. J. Cancer 1955, 9, 539−549.
(4) Gray, L. H. Oxygenation in radiotherapy. I. Radiobiological
considerations. Br. J. Radiol. 1957, 30 (356), 403−406.
(5) Minchinton, A. I.; Tannock, I. F. Drug penetration in solid
tumours. Nature Rev. Cancer 2006, 6 (8), 583−592.
(6) Vaupel, P. Hypoxia and aggressive tumor phenotype: implications
for therapy and prognosis. Oncologist 2008, 13 (3), 21−26.
(7) Bache, M.; Kappler, M.; Said, H. M.; Staab, A.; Vordermark., D.
Detection and Specific Targeting of Hypoxic Regions within Solid
Tumors: Current Preclinical and Clinical Strategies. Curr. Med. Chem.
2008, 15, 322−338.
ASSOCIATED CONTENT
■
S
* Supporting Information
Data pertaining to the elemental analysis, H-, 13C-, 19F-NMR,
1
and mass spectra of the final and other relevant compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(8) Wardman, P. Chemical Radiosensitizers for Use in Radiotherapy.
Clin. Oncol. 2007, 19 (6), 397−417.
(9) Adams, G. E.; Dewey, D. L. Hydrated electrons and
radiobiological sensitization. Biochem. Biophys. Res. Commun. 1963,
12, 473−477.
(10) Fowler, J. F.; Denekamp, J. A review of hypoxic cell
radiosensitization in experimental tumors. Pharmacol. Ther. 1979, 7
(3), 413−444.
(11) Tanabe, K.; Zhang, Z.; Ito, T.; Hatta, H.; Nishimoto, S. Current
molecular design of intelligent drugs and imaging probes targeting
tumor-specific microenvironments. Org. Biomol. Chem. 2007, 5 (23),
3745−3757.
(12) Adams, G. E.; Stratford, I. J.; Wallace, R. G.; Wardman, P.;
Watts, M. E. Toxicity of nitro compounds toward hypoxic mammalian
cells in vitro: dependence on reduction potential. J. Natl. Cancer Inst.
1980, 64 (3), 555−560.
(13) Adams, G. E.; Flockhart, I. R.; Smithen, C. E.; Stratford, I. J.;
Wardman, P.; Watts, M. E. A correlation between structure, one
electron reduction potential and efficiency of nitroimidazoles as
hypoxic cell radiosensitizers. Radiat. Res. 1976, 67, 9−20.
(14) Brown, J. M.; Workman, P. Partition coefficient as a guide to the
development of radiosensitizers which are less toxic than misonidazole.
Radiat. Res. 1980, 82, 171−190.
AUTHOR INFORMATION
Corresponding Author
*Phone: +1.780.989.4313. Fax: +1 780.432.8483. E-mail:
Piyush.Kumar@albertahealthservices.ca; pkumar@ualberta.ca.
Address: Oncologic Imaging, Department of Oncology,
University of Alberta, Cross Cancer Institute, 11560 University
Avenue, Edmonton, Alberta T6G 1Z2, Canada.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This research was supported by grants from the Canadian
Institutes of Health Research (CIHR POP program), CIHR-
ACF Partner Bridge grant, the Canadian Association of
Radiation Oncologists (CARO), and the Alberta Cancer
Foundation (ACF Pilot). Additional equipment and resource
support by Naeja Inc., Edmonton, to facilitate chiral
chromatography experiments are sincerely appreciated.
(15) Asquith, J. C.; Watts, M. E.; Patel, K.; Smithen, C. E.; Adams, G.
E. Electron-Affinic Sensitisation V. Radiosensitisation of Hypoxic
Bacteria and Mammalian Cells in Vitro by Some Nitroimidazoles and
Nitropyrazoles. Radiat. Res. 1974, 60, 108.
(16) Grigsby, P. W.; Winter, K.; Wasserman, T. H.; Marcial, V.;
Rotman, M.; Cooper, J.; Keys, H.; Asbell, S. O.; Phillips, T. L.
Irradiation with or without misonidazole for patients with stages IIIB
and IVA carcinoma of the cervix: final results of RTOG 80−05.
Radiation Therapy Oncology Group. Int. J. Radiat. Oncol. Biol. Phys.
1999, 44 (3), 513−517.
ABBREVIATIONS USED
■
A-GAZ, allylglucoazomycin; DAST, diethylaminosulfurtrifluor-
ide; DBU, 1,8-diazabicyclo(5,4,9)-undec-7-ene; d, doublet; dd,
doublet of doublet; DMAP, dimethylaminopyridine; DMF,
dimethylformamide; dt, doublet of triplet; EtOH, ethanol;
EtOAc, ethyl acetate; E1 , single electron reduction potential; F-
7
GAZ, fluoroglucoazomycin; FAZA, 1-α-D-5-fluoro-5-deoxyar-
abinofuranose-2-nitroimidazole; FMISO, fluoromisonidazole;
GAZ, glucoazomycin; GLUTs, glucose transport proteins;
HPLC, high performance liquid chromatography; IAZA, 1-α-
D-5-iodo-5-deoxyarabinofuranose-2-nitroimidazole; iso-PrOH,
isopropyl alcohol; m, multiplet; MeCN, acetonitrile; MBM,
modified Barth’s medium; MTT, 3-(4,5-dimethyl-2-thiazolyl)-
2,5-diphenyl-2H-tetrazolium bromide; NaOH, sodium hydrox-
ide; NaHCO3, sodium bicarbonate; Na2CO3, sodium carbo-
nate; NOE, nuclear Overhauser effect; OER, oxygen enhance-
(17) Wiebe, L. I.; McEwan, A. J. B. Scintigraphic Imaging of Focal
Hypoxic Tissue: Development and Clinical Applications of 123I-IAZ.
Braz. Arch. Biol. Technol. 2002, 45, S89−102.
(18) Stypinski, D.; Wiebe, L. I.; McEwan, A. J.; Schmidt, R. P.; Tam,
Y. K.; Mercer, J. M. Clinical pharmacokinetics of 123IAZA in healthy
volunteers. Nucl. Med. Commun. 1999, 20, 559−567.
(19) Krohn, K. A.; Link, J. M.; Mason, R. P. Molecular imaging of
hypoxia. J. Nucl. Med. 2008, 49 (Suppl 2), 129S−48S.
6045
dx.doi.org/10.1021/jm2017336 | J. Med. Chem. 2012, 55, 6033−6046

Journal of Medicinal Chemistry
Article
(20) Postema, E. J.; McEwan, A. J.; Riauka, T. A.; Kumar, P.;
Richmond, D. A.; Abrams, D. N.; Wiebe, L. I. Hypoxia imaging using
1-α-D-(5-deoxy-5-[18F]-fluoroarabino-furanosyl)-2-nitroimidazole
(18F-FAZA): the initial results of a Phase I/II study. Eur. J. Nucl. Med.
Mol. Imaging 2009, 36, 1565−1573.
(37) Nakae, T.; Uto, Y.; Tanaka, M.; Shibata, H.; Nakata, E.;
Tominaga, M.; Maezawa, H.; Hashimoto, T.; Kirk, K. L.; Nagasawa,
H.; Hori, H. Design, synthesis, and radiosensitizing activities of sugar-
hybrid hypoxic cell radiosensitizers. Bioorg. Med. Chem. 2008, 16, 675−
682.
(38) Adams, G. E.; Flockhart, I. R.; Smithen, C. E.; Stratford, I. J.;
Wardman, P.; Watts, M. E. Radiat. Res. 1976, 67, 9.
(21) Kumar, P.; Stypinski, D.; Xia, H.; McEwan, A. J. B.; Machulla,
2
H-J.; Wiebe, L. I. Fluoroazomycin arabinoside (FAZA): synthesis, H,
3H-labelling and preliminary biological evaluation of a novel 2-
nitroimidazole marker of tissue hypoxia. J. Labelled Compd. Radio-
pharm. 1999, 42, 3−16.
(39) Mannan, R. H.; Somayaji, V. V.; Lee, J.; Mercer, J. R.; Chapman,
J. D.; Wiebe, L. I. Radioiodinated 1-(5-iodo-5-deoxy-β-D-arabinofur-
ansoyl)-2-nitroimidazole (iodoazomycin arabinoside: IAZA): a novel
marker of tissue hypoxia. J. Nucl. Med. 1991, 32, 1764−1770.
(40) Barros, L. F.; Bittner, C. X.; Loaiza, A.; Ruminat, I.; Larenas, V.;
Moldenhauer, H.; Ovarzun, M. Kinetic validation of 6-NBDG as a
probe for the glucose transporter GLUT1 in astrocytes. Neurochemistry
2009, 109, 94−99.
(22) Shi, K.; Souvatzoglou, M.; Astner, S. T.; Vaupel, P.; Nusslin, F.;
̈
Wilkens, J. J.; Ziegler, S. I. Quantitative assessment of hypoxia kinetic
models by a cross-study of dynamic 18F-FAZA and 15O-H2O in
patients with head and neck tumors. J. Nucl. Med. 2010, 51 (9), 1386−
1394.
(41) Yao, S. Y. M.; Cass, C. E.; Young, J. D. Membrane Transport: A
Practical Approach; Baldwin, S. A., Ed.; Oxford University Press:
London, 2000; pp 47−48.
(23) Arabi, M.; Piert, M. Hypoxia PET/CT imaging: implications for
radiation oncology. Q. J. Nucl. Med. Mol. Imaging 2010, 54 (5), 500−
509.
(24) Busk, M.; Horsman, M. R.; Jacobson, S.; Hansen, K. V.; Bussink,
J.; Van der Kogel, A.; Overgaard, J. Can hypoxia-PET map hypoxic cell
density heterogeneity accurately in an animal tumor model at a
clinically obtainable image contrast? Radiother. Oncol. 2009, 92, 429−
436.
(25) Schuetz, M.; Schmid, M. P.; Potter, R.; Kommata, S.; George,
̈
D.; Lukic, D.; Dudczak, R.; Kletter, K.; Dimopoulos, J.; Karanikas, G.;
Bachtiary, B. Evaluating repetitive 18F-fluoroazomycin-arabinoside
(18FAZA) PET in the setting of MRI guided adaptive radiotherapy in
cervical cancer. Acta Oncol. 2010, 49 (7), 941−947.
(26) Adams, G. E. Chemical radiosensitization of hypoxic cells. Br.
Med. Bull. 1973, 29 (1), 48−53.
(27) Adams, G. E.; Cooke, M. S. Electron-affinic sensitization. A
structural basis for chemical radiosensitizers in bacteria. Int. J. Radiat.
Biol. Relat. Stud. Phys. Chem. Med. 1969, 15, 457−471.
(28) Adams, G. E.; Clarke, E. D.; Flockhart, I. R.; Jacobs, R. S.;
Sehmi, D. S.; Stratford, I. J.; Wardman, P.; Watts, M. E.; Parrick, J.;
Wallace, R. G.; Smithen, C. E. Structure−activity relationships in the
development of hypoxic cell radiosensitizers. I. Sensitization efficiency.
Int. J. Radiat. Biol. 1979, 35, 133−150.
(29) Wuest, M.; Kumar, P.; Jans, H.-J.; Murray, D.; Roa, W.; Wiebe,
L. Radiopharmacological evaluation of 18F-labeled GAZ-F: A glucose-
nitroimidazole hybrid radiosensitizer. Canadian Association of Radiation
Oncologists Annual Scientific Meeting 2010, 188.
(30) Roa, W.; Yang, X.; Kumar, P.; Murray, D.; Wiebe, L.
Cytotoxicity and in vivo toxicity of GAZ-F, A glucose−azomycin
construct for radiosensitization of hypoxic tumors. Canadian
Association of Radiation Oncologists Annual Scientific Meeting 2010, 125.
(31) Loike, J. D.; Cao, L.; Brett, J.; Ogawa, S.; Silverstein, S. C.; Stern,
D. Hypoxia induces glucose transporter expression in endothelial cells.
Am. J. Physiol. 1992, 263 (2 Pt 1), C326−333.
(32) Patt, M.; Sorger, D.; Scheunemann, M.; Stocklin, G. Adduct of
2-[18F]FDG and 2-nitroimidazole as a putative radiotracer for the
detection of hypoxia with PET: synthesis, in vitro and in vivo
characterization. Appl. Radiat. Isotopes 2002, 57 (2), 705−712.
(33) Kumar, P.; Zheng, W.; McQuarrie, S. A.; Jhamandas, J. H.;
Wiebe, L. I. 18F-FESB: synthesis and automated radiofluorination of a
novel 18F-labeled PET tracer for β-amyloid plaques. J. Labelled Compd.
Radiopharm. 2005, 48, 983−996.
(34) Kumar, P.; Emami, S.; McEwan, A. J. B.; Wiebe, L. I.
Development of an economical, single step synthesis of FAZA, a
clinical hypoxia marker, and potential synthons to prepare its
positional analogs. Lett. Drug Des. Discovery 2009, 6, 82−85.
(35) Silverstein, R. M.; Bassler, G. C.; Morrill T. C. Spectrometric
Identification of Organic Compounds. 4th ed.; John Wiley & Sons: New
York, 1981; pp 181−303.
(36) Husstedt, W. S.; Wiehle, S.; Stillig, C.; Bergander, K.; Grimme,
S.; Haufe, G. Synthesis and preferred conformations of all region-and
diastereomeric methyl 2,3-fluorohydroxyalkanoates. Eur. J. Org. Chem.
2011, 355−363, DOI: 10.1002/ejoc.201001224.
6046
dx.doi.org/10.1021/jm2017336 | J. Med. Chem. 2012, 55, 6033−6046