Preliminary research on 1-(4-bromo-2-nitroimidazol-1-yl)-3-[18F]fluoropropan-2-ol as a novel brain hypoxia PET tracer in a rodent model of stroke

Article abstract of DOI:10.1016/j.ejmech.2015.06.023

Abstract The synthesis of the new radiotracer precursor 4-Br-NITTP and the radiolabeling of the new tracer 1-(4-bromo-2-nitroimidazol-1-yl)-3-[¹⁸F]fluoropropan-2-ol (4-Br-[¹⁸F]FMISO) is reported. The cyclic voltammetry behaviour, neu

Full text of DOI:10.1016/j.ejmech.2015.06.023

Accepted Manuscript
18
Preliminary research on 1-(4-bromo-2-nitroimidazol-1-yl)-3-[ F]fluoropropan-2-ol as
a novel brain hypoxia PET tracer in a rodent model of stroke
Elena Nieto, Mercedes Delgado, Mónica Sobrado, María L. de Ceballos, Ramón
Alajarín, Luis García-García, James Kelly, Ignacio Lizasoain, Miguel A. Pozo, Julio
Álvarez-Builla
PII:
S0223-5234(15)30099-4
10.1016/j.ejmech.2015.06.023
EJMECH 7950
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 18 December 2014
Revised Date: 10 April 2015
Accepted Date: 8 June 2015
Please cite this article as: E. Nieto, M. Delgado, M. Sobrado, M.L. de Ceballos, R. Alajarín, L. García-
García, J. Kelly, I. Lizasoain, M.A. Pozo, J. Álvarez-Builla, Preliminary research on 1-(4-bromo-2-
18
nitroimidazol-1-yl)-3-[ F]fluoropropan-2-ol as a novel brain hypoxia PET tracer in a rodent model of
stroke, European Journal of Medicinal Chemistry (2015), doi: 10.1016/j.ejmech.2015.06.023.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Graphical abstract
Br
N
NO2
O H
N
1
8
F

ACCEPTED MANUSCRIPT
Preliminary research on 1-(4-bromo-2-
1
8
nitroimidazol-1-yl)-3-[ F]fluoropropan-2-ol as a
novel brain hypoxia PET tracer in a rodent model
of stroke
a
b
c,1
d
Elena Nieto, Mercedes Delgado, Mónica Sobrado, María L. de Ceballos, Ramón
a
b
e,2
c
b,e,3
Alajarín, Luis García-García, James Kelly, Ignacio Lizasoain, Miguel A. Pozo,
Julio
a,3,*
Álvarez-Builla,
a
Departamento de Química Orgánica y Química Inorgánica, Universidad de Alcalá, 28871-
Alcalá de Henares, Madrid, Spain
b
CAI Cartografía Cerebral, Instituto Pluridisciplinar UCM, Paseo Juan XXIII, 1, Madrid
2
8040, Spain
c
Unidad de Investigación Neurovascular, Departamento de Farmacología, Facultad de
Medicina, Universidad Complutense, Avenida Complutense s/n, 28040 Madrid, Spain
d
Departamento de Neurobiología Molecular, Celular y del Desarrollo, Instituto Cajal, Avda.
Doctor Arce, 37, Madrid 28002, Spain
e
Instituto Tecnológico PET, Calle Manuel Bartolomé Cossío, 10, Madrid 28040, Spain
1

ACCEPTED MANUSCRIPT
ABSTRACT
The synthesis of the new radiotracer precursor 4-Br-NITTP and the radiolabeling of the new
1
8
18
tracer 1-(4-bromo-2-nitroimidazol-1-yl)-3-[ F]fluoropropan-2-ol (4-Br-[ F]FMISO) is
reported. The cyclic voltammetry behaviour, neuronal cell toxicity, transport through the
brain endothelial cell monolayer, in vivo PET imaging and preliminary calculations of the
tracer uptake for a rodent model of stroke were studied for the new compound and the results
18
were compared to those obtained with [ F]FMISO, the current gold standard PET hypoxia
tracer. The new PET brain hypoxia tracer is more easily reduced, has higher CLogP than
18
[
F]FMISO and it diffuses more rapidly through brain endothelial cells. The new compound
is non-toxic to neuronal cells and it allows the in vivo mapping of stroke in mice with higher
1
8
sensitivity. 4-Br-[ F]FMISO is a good candidate for further development in ischemic stroke.
Keywords
1
8
1
-(4-bromo-2-nitroimidazol-1-yl)-3-[ F]fluoropropan-2-ol, PET imaging, hypoxia, fluorine-
1
8, stroke, nitroimidazole
Abbreviations
BBB, blood-brain barrier; CBF, cerebral blood flow; CT, computed tomography; DHP, 3,4-
dihydro-2H-pyran; FDG, 2-fluoro-2-deoxy-D-glucose; FMISO, fluoromisonidazole, 1-fluoro-
3
-(2-nitroimidazol-1-yl)-propan-2-ol; FON, 1-(8-fluorooctyl)-2-nitro-1H-imidazole; FPN, 1-
(3-fluoropropyl)-2-nitro-1H-imidazole; LDH, lactate dehydrogenase; MRI, magnetic
resonance imaging; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide;
NBPht, N-bromophthalimide; NBS, N-bromosuccinimide; NITTP, 3-(2-nitro-imidazol-1-yl)-
2
-(tetrahydropyran-2-yloxy)-prop-1-yl tosylate; 4-Br-NITTP, 3-(4-bromo-2-nitroimidazol-1-
2

ACCEPTED MANUSCRIPT
yl)-2-(tetrahydropyran-2-yloxy)-prop-1-yl tosylate; OD, optical density; PET, positron
emission tomography; pMCAO, permanent occlusion of the right middle cerebral artery;
PPTS, pyridinium p-toluenesulfonate; SPECT, single photon emission computed
tomography; TBABF , tetrabutylammonium tetrafluoroborate; THP, tetrahydropyranyl.
4
*
Corresponding author: J.A.-B., Tel. +341918854606. Fax, +341918854686. E-mail,
1
julio.alvarez@uah.es; Present addresses:
M.S., Servicio de Neurología, Instituto de
Investigación Sanitaria, Hospital Universitario La Princesa, Calle Diego de León, 62, 28006
2
Madrid, Spain; J.K., Weill Cornell Medical College, Belfer Research Building, 413 E 69th
3
St, New York, NY 10021. Author contributions: J.A.-B. and M.A.P contributed equally to
this work. J.A.-B., M.A.P and I.L. conceived the project, E.N. performed the synthesis of
precursors and cold tracers, M.D. and L.G.-G. carried out micro-PET experiments, M.S.
prepared the pMCAO model and tested compounds on LDH and MTT assays, M.L.C.
performed BBB-permability assay, R.A. designed chemical experiments, J.K. performed the
radiosynthesis, E.N., M.D., M.S., M.L.C., R.A., L.G.-G. and J.K. wrote the manuscript. All
authors edited and approved the final version of the manuscript.
1
. Introduction
Stroke is an increasing common disease both in first-world and emergent countries as a
consequence of factors such as old age, high blood pressure, previous stroke, transient
ischemic attack, diabetes, high cholesterol, tobacco smoking and atrial fibrillation. As a
medical emergency, acute stroke requires quick diagnosis and treatment, i.e., within the first
minutes to hours after the stroke event. A rapid response is critical in order to initiate therapy
before the ischemic but still viable tissue becomes necrotic. One goal in the clinical treatment
3

ACCEPTED MANUSCRIPT
of stroke is to identify the presence and to map the extent, by neuroimaging techniques, of
hypoxic but still viable tissue – ischemic penumbra – that can be recovered through suitable
cerebral revascularization therapy. This possibility would facilitate the selection of patients
who may benefit from revascularization treatments, excluding those for whom the risk of
complications is very high or the brain damage by infarct is irreversible [1,2].
The identification and mapping of ischemic penumbra is the target for the current
diagnostic techniques used for the imaging of brain hypoxia, such as X-ray computed
tomography (CT), magnetic resonance imaging (MRI), positron emission tomography (PET)
and single photon emission computed tomography (SPECT) [3]. In the context of improving
the sensitivity and accessibility of current neuroimaging techniques, hypoxia radiotracers
based on the nitroimidazole core have received particular attention due to their ability to
accumulate in hypoxic but viable tissue and these may be good candidates for imaging in
ischemic stroke [4]. Once the radiotracer crosses the cell membrane, in the first step, the nitro
group is reduced by intracellular reductases to its radical nitro anion which is rapidly
reoxidized in normoxic cells and the tracer returns to blood flow. In hypoxic cells, the radical
nitro anion is then reduced in further stages to species that bond to cell components and the
modified tracer is unable to diffuse out from the cell, thus allowing the imaging of hypoxic
cells [5-9].
Fig.1. (2-column size)
18
18
18
[
F]Fluoromisonidazole ([ F]FMISO, [ F]1) is currently the gold standard for PET
imaging in oncology [10]. In acute stroke the use of this tracer is limited by poor brain
penetration, slow accumulation in the penumbral area and slow clearance from normoxic
4

ACCEPTED MANUSCRIPT
regions. As a consequence, long acquisition times are required late after the stroke event and
this compromises the early start of the therapy and the lifespan of recoverable tissue [4].
Researchers have focused on new hypoxia radiotracers with better pharmacokinetic
1
8
18
properties. [ F]FMISO ([ F]1) is a low lipophilicity compound (LogPoctanol/water: –0.40) [11]
and more lipophilic analogs have been prepared (Figure 1). The concentration of 4-
82
82
18
[
Br]bromomisonidazole (4-[ Br]BrMISO, [ F]2) increased in the hypoxic portion of
EMT-6/UW tumor at 2 hours, but high blood levels contributed to excessive background
1
8
18
[
12,13]. The initial uptake of 4-Br-[ F]FPN (4-bromo-1-(3-[ F]fluoropropyl)-2-nitro-1H-
18
18
18
imidazole, [ F]3) into rat brain was significantly higher than that for [ F]FMISO ([ F]1)
and this was followed by a rapid washout from the brain. The uptake in mice muscle tumor
18
18
was somewhat enhanced compared to the levels obtained with [ F]FMISO ([ F]1) and
1
8
18
18
18
[
[
F]FPN ([ F]4), but with lower tumor localization than [ F]FMISO ([ F]1) [14].
18
18
18
18
18
F]FON (1-(8-[ F]fluorooctyl)-2-nitro-1H-imidazole, [ F]5) and [ F]FPN ([ F]4) had
1
8
18
increased initial uptake in normal rat brain relative to [ F]FMISO ([ F]1), and this was
1
8
18
followed by a rapid washout from brain. Both [ F]4 and [ F]5 had significantly lower tumor
1
8
18
uptake and lower tumor-to-blood ratios than [ F]FMISO ([ F]1), a finding that suggests a
1
8
18
poor trapping mechanism within the tumor tissue. Neither [ F]4 nor [ F]5 offered improved
18
18
biological properties over [ F]FMISO ([ F]1) [11,14].
On the other hand, a newer generation of hypoxia radiotracers (Figure 1) includes 2-(2-
1
8
18
18
nitro-1H-imidazol-1-yl)-N-(2,2,3,3,3-[ F]pentafluoropropyl)-acetamide ([ F]EF5, [ F]6)
18
18
18
18
[
15], [ F]fluoroazomycin arabinoside ([ F]FAZA, [ F]7) [16,17], 3-[ F]fluoro-2-(4-((2-
18
18
nitro-1H-imidazol-1-yl)methyl)-1H-1,2,3,-triazol-1-yl)-propan-1-ol
([ F]HX4,
[ F]8)
6
2
4
[
18,19] and Cu-labeled bis(thiosemicarbazone) complexes such as Cu-diacetyl-bis(N -
6
2
18
64
4
methylthio-semicarbazone) ( Cu-ATSM, [ F]9) [20] and
Cu-diacetyl-bis(N -
6
4
18
ethylthiosemicarbazone) ( Cu-ATSE, [ F]10) [21], all of which have been studied for tumor
5

ACCEPTED MANUSCRIPT
hypoxia imaging without consensus on the best radiotracer [18,22]. It has been suggested that
1
8
18
tracer [ F]10 is a more sensitive marker of neurological hypoxia than analog [ F]9 as the
former is reduced at higher tissue oxygen concentrations [21]. To the best of our knowledge,
18
18
18
there is only one report in which compounds [ F]7, [ F]9 and [ F]10 have been compared
as PET tracers for brain hypoxia mapping in a rodent model of stroke. Unfortunately, none of
1
8
18
the examples surpassed or even equaled [ F]FMISO ([ F]1) for the early detection of acute
focal brain hypoxia and further development of these compounds in ischemic stroke is not
warranted [23].
In the field of stroke imaging there is still a need to find suitable tracers that can be
developed into radiopharmaceuticals for general use in acute stroke events in an effort to
allow patients to overcome the event with the lowest level of brain alteration. Important
issues that require improvement include: the amount of tracer that reaches the hypoxic cell
through cerebral blood flow (CBF) and blood flow across the blood-brain barrier (BBB), the
fraction that is further reduced from the radical nitro anion, the washout from normoxic cells,
the rate of trapping in hypoxic tissue which in turn yields better or worse imaging contrast
and the presence of cytosolic biomolecules that bind to the tracer.
We hypothesized, as it was also suggested before [12], that the modification of
18
18
[
F]FMISO ([ F]1) by the addition of bromo-substituents on the imidazole ring could have a
positive effect on the pharmacological properties of the radiotracer, such as redox potential
and BBB permeation. Thus, the tracer would provide better PET imaging at shorter times
after injection because of an easier reduction and higher lipophilicity. A nitroimidazole-based
82
18
tracer that bears a bromine on the imidazole ring (4-[ Br]BrMISO, [ F]2) was reported to
18
have increased concentration in the hypoxic portion of tumors at 2 hours while [ F]FMISO
18
18
(
[ F]1) showed to be useful at 4 hours. However, high blood levels of [ F]2 contributed to
6

ACCEPTED MANUSCRIPT
1
8
high background, incompatible with imaging [12]. Although [ F]2 could be prone to loose
the radiolabel since a bromine on the imidazole ring is less stable than a fluorine at the end of
18
an alkyl chain [12], the label remains on the imidazole ring, and [ F]2 is stable and appears
to retain its label in vivo [13]. Furthermore, the major metabolite corresponds to the
desmethyl derivative [12,24] Herein, we report the synthesis of 1-(4-bromo-2-nitroimidazol-
1
-yl)-3-fluoropropan-2-ol (4-Br-FMISO) and 1-(4,5-dibromo-2-nitroimidazol-1-yl)-3-
1
8
fluoropropan-2-ol, and the pharmacological study and radiolabeling of (4-Br-[ F]FMISO).
The imaging of ischemic penumbra in an experimental model of stroke in mouse is also
1
8
reported in order to evaluate 4-Br-[ F]FMISO as a more lipophilic and easier to reduce
nitroimidazole-based brain hypoxia PET tracer.
2
. Results and discussion
2
.1. Chemistry
The synthesis of precursors started from diol 11, which was prepared as reported in a
synthesis of NITTP (12) (Scheme 1) [25]. Bromination of 11 with N-bromophthalimide
NBPht) using 1.4 and 4.0 equivalents in DMF provided 13 (79%) and 14 (36%),
(
respectively. Subsequent selective tosylation yielded 15 (72%) and 16 (51%) as single
regioisomers [26]. Finally, the protection of the hydroxyl group as a tetrahydropyranyl (THP)
ketal provided the precursors 17 (4-Br-NITTP, 68%) and 18 (69%) as 1:1 and 44:56 mixtures
of diastereomers, respectively, for radiolabeling. Precursor 17 (69%) was also prepared as a
6
7:33 diastereomeric mixture by bromination of 12 using N-bromosuccinimide (NBS).
Compounds 12, 17 and 18 were used for radiolabeling. Compounds 19 (69%) and 20 (31%)
were prepared as 45:55 diastereomeric mixtures by fluoride nucleophilic displacement of
tosylate, using CsF in tert-amyl alcohol [27]. Subsequent removal of the THP protecting
7

ACCEPTED MANUSCRIPT
group by treatment with AcOH gave 21 (4-Br-FMISO, 65%) and 22 (69%). FMISO (1) was
also prepared by fluorination of 12 to give 23 (82%) as a 66:34 mixture of diastereomers
1
followed by hydrolysis of the ketal (55%). The diastereomeric ratios were measured by H
NMR. They were calculated by integration of the corresponding signals to the proton on the
ketal carbon at the THP ring. For 12, 17, 19 and 23 the signals for the protons on the
imidazole ring were also used.
Scheme 1. (2-column size)
Compounds 17 and 18 could not be prepared by reaction of 4(5)-bromoimidazole (24) or
4
,5-dibromoimidazole (25) with ditosylate 26 using the reported procedure for NITTP (12)
(Scheme 2) [28]. Firstly, the stoichiometric bromination of 2-nitroimidazole (27) has been
reported to give a mixture of 25 and 27 and only a trace amount of 24, while 25 (97%) was
prepared quantitatively when 2 equivalents of NBS were used [29]. Furthermore, the
reactions of 25 and 26 only gave trace amounts of 18 even under harsh conditions, probably
due to the low nucleophilicity of imidazolate anion 28.
Scheme 2. (1-colunm size)
2
.2. Radiosynthesis
1
8
The radiosynthesis of [ F]1 was performed as reported [28]. The labeling of the protected
18
tosyl precursor 17 with [ F]fluoride was carried out on an automated radiosynthesis platform
18
18
18
(
Scheme 3). In the synthesis of [ F]21, the [ F]fluoride was separated from the bulk H O
2
8

ACCEPTED MANUSCRIPT
solution by anion exchange chromatography following retention in a quaternary ammonium
cartridge and elution with K CO in an H O/MeCN solution containing kryptofix.
2
3
2
18
[
F]Fluoride was dried at 100 °C under reduced pressure and a solution of precursor 17 (9 ±
3
mg in 2.5 mL MeCN) was added. Labeling was performed at 130 °C over 5 minutes and
the reaction vessel was then cooled to 100 °C for hydrolysis of the THP ketal with 1N HCl.
The reaction mixture was transferred to a semi-preparative HPLC column for purification
18
and approximately 80% of the total activity was transferred. Free [ F]fluoride eluted after 3
minutes and a number of minor radiolabeled impurities were observed. A major impurity
eluted after 21 minutes and the major product eluted after 33 minutes. Both impurity and
product were collected and analyzed by HPLC and compared with the cold standard 21. The
peak for the product with a retention time of 33 minutes corresponded to the peak of the cold
standard. The impurity could not be identified by comparison with any known standard. The
18
18
total synthesis time was 55 minutes and [ F]21 (4-Br-[ F]FMISO) was synthesized in 18%
yield (decay corrected) with a specific activity of 150 ± 15 GBq/µmol (at the end of
radiosynthesis).
1
8
Radiosynthesis conditions were also applied to 18 but unfortunately evidence for [ F]22
was not observed. This result can be understood by considering the good leaving group
character of the very weakly basic imidazolate anion 28 and the presence of excess K CO3
2
under radiolabeling conditions, which could give rise to substitution or elimination reactions
1
8
18
involving the base. Further studies were only carried out with 1, [ F]FMISO ([ F]1), 21 and
1
8
18
4
-Br-[ F]FMISO ([ F]21).
Scheme 3. (1-column size)
9

ACCEPTED MANUSCRIPT
2
.3. Cyclic voltammetry
The in vivo reduction of nitroimidazole compounds is performed by cytosolic
nitroreductases through an initial single electron transfer to the nitro group to give the radical
nitro anion (Eq. 1), which under anaerobic conditions can be reduced to the nitroso,
hydroxylamino and amino compounds (Eq. 2–4) [30-32]. However, in aprotic media, such as
DMF, the radical nitro anion is reduced to the nitro dianion under a more negative potential
(Eq. 5) [33,34].
–
· –
R-NO + e
R-NO2
(Eq. 1)
2
·
–
–
+
R-NO2
+ e + 2H
RNO + H O (Eq. 2)
2
–
+
RNO + 2 e + 2H
RNHOH
(Eq. 3)
–
+
–
RNHOH + 2 e + 2H
RNH + H O (Eq. 4)
2
2
·
–
2–
R-NO2
+ e
RNO2
(Eq. 5)
In order to ascertain the influence that the 4-bromo-substituent on the nitroimidazole ring
has on the redox potential, compounds 21 and 1 were studied by cyclic voltammetry in
aprotic media (Bu NBF /DMF) and under anaerobic conditions (Table 1, Figure 2). A first
4
4
reversible peak (Ia) was observed for the reduction of the nitro group to the radical nitro
anion (Eq. 1; 1: –1.04 V; 21: –0.89 V), and a peak (Ib) was also observed for the oxidation of
the radical nitro anion to the nitro group (Eq. 1 reverse). A second irreversible peak (II) for
the reduction of the radical nitro anion to the nitro dianion (Eq. 5; 1: –1.97 V; 21: –1.64 V)
10

ACCEPTED MANUSCRIPT
was also observed. Compound 21 is more easily reduced than 1 to both the radical anion and
dianion, thus making 21 a better hypoxia probe.
Table 1.
Fig.2. (1-column size)
2
.4. Toxicity
Before carrying out animal experimentation in order to study 21 as in vivo tracer, its
toxicity was evaluated by viability/mortality in vitro assays performed in parallel with the
cold tracer 1 as reference (Figure 3). Assays were carried out on primary cultures of rat
cortical cells, both neuronal (see supplementary data) and mixed. Toxicity was determined by
LDH
(lactate
dehydrogenase)
and
MTT
(3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide) assays. Compound 21 showed a toxicity profile similar to that
of 1 at the concentrations assayed. It did not show toxicity when compared to vehicle (0 µM)
–
even at the highest concentration assayed (100 µM) – and did not show higher cell
mortality (P>0.05; one-way ANOVA and pos hoc Tukey’s test; Figure 4).
Fig. 3. (1.5-column size)
Fig. 4. (1-column size)
2
.5. Lipophilicity
11

ACCEPTED MANUSCRIPT
The ability to cross the BBB and reach the therapeutic target is the major challenge for
CNS drug candidates and it has been estimated that 98% fail to reach the target [35].
Important concepts in brain penetration such us unbound drug concentration, unbound brain-
to-plasma ratio, total brain-to-plasma ratio, fraction unbound, BBB passive permeability and
efflux ratio have to be considered together. A screening paradigm for the key brain
penetration properties has recently been developed to guide CNS compound optimization
[
36]. Lipophilicity, measured as LogP, is a parameter that is intimately related with BBB
passive transport and a mean CLogP value of 2.5 for the marketed CNS drugs has been
reported [37]. The LogPoct/water and CLogP values for previously reported tracers 1–10 and
compounds 21 and 22 are shown in Table 2. Compound 21 has a CLogP higher than most of
the nitroimidazole radiotracers in Figure 1, but specially higher than FMISO. Further in vitro
and in vivo experiments (see sections 2.6 and 2.7) showed that 21 crosses BBB.
Table 2.
Compound 21 showed a better reduction potential and higher CLogP than 1, it had a
1
8
comparable toxicity, and it was possible to radiolabel this compound. [ F]21 was therefore a
good candidate for further PET imaging in a rodent model of stroke.
2
.6. Blood-brain barrier permeability
As shown in Figure 5, FMISO (1) diffused across the bEnd-5 cell monolayer. More than
5
0% of the drug was present in the lower compartment at 15 min after the addition of the
12

ACCEPTED MANUSCRIPT
drug. Similar amounts of transported 1 were observed at 30 and 45 minutes and the level
increased slightly at 60 minutes. The diffusion of the bromo-derivative 21 was also rapid and
it tended to be higher compared to 1 at 15 min, although the difference did not reach
statistical significance. However, at longer times (30–60 minutes) compound 21 attained
significantly higher transport in comparison to 1 and almost all of it was present in the lower
compartment (99.4%).
The rapid initial diffusion of both compounds is compatible with a rapid access to the
brain. The limited permeability of 1 may be related to its lower LogP value, but the
possibility that the compound is retained in the cells and/or is metabolized cannot be ruled
out. In contrast, compound 21 completely diffused over time and its higher permeability is
consistent with its higher CLogP compared to FMISO.
Fig. 5. (1-column size)
1
8
2
.7. [ F]21 PET imaging of the hypoxic area
1
8
18
In order to evaluate 4-Br-[ F]FMISO ([ F]21) as an in vivo tracer, this compound, and
FMISO (1) as control, were assessed in an experimental model of brain ischemia by
permanent occlusion of middle cerebral artery (pMCAO) in male Fischer rats. The image
1
8
18
analysis revealed that 3 hours after pMCAO, the radiotracers [ F]21 and [ F]FMISO
18
(
[ F]1) were accumulated in the frontoparietal cortical area corresponding to middle cerebral
artery occlusion. The retention of both hypoxia tracers corresponds to the hypoxic region, as
1
8
18
revealed by [ F]fluorodeoxyglucose ([ F]FDG)-PET neuroimaging performed 24 hours
after the ischemic onset. The degree of lesion estimated as the reduction of glucose brain
13

ACCEPTED MANUSCRIPT
metabolism 24 hours after pMCAO was in the range 17–41% from the non-affected
1
8
18
18
contralateral area of [ F]FDG uptake. Representative [ F]21 and [ F]FDG-PET images are
shown in Figure 6.
Fig. 6. (1-column size)
1
8
18
18
To compare the labeling capability of [ F]21 and [ F]FMISO ([ F]1) to the hypoxic
area, accumulation in the injured region for both tracers were calculated. The uptake (see
Experimental section for a detailed description of how this parameter was obtained) for
1
8
18
[
(
F]21 corrected for lesion severity was 54% higher than that of the standard [ F]FMISO
1
8
18
[ F]1) (Table 3). Therefore, this result shows that [ F]21 yields an improved sensibility
1
8
18
compared to [ F]FMISO ([ F]1). This would allow an earlier and more sensitive detection
18
18
of the eventual brain damage generated by ischemia than [ F]FMISO ([ F]1), thus enabling
an earlier pharmacological treatment for preventing or minimizing the potential stroke-
induced consequences.
Table 3.
1
8
18
18
PET-CT rat brain coronal images of [ F]FMISO ([ F]1) and [ F]21, for which the
uptake at the hypoxic tissue 3 hours after pMCAO can be compared, are shown in Figure 7.
As observed, the areas visualized by both hypoxia tracers are fairly similar.
Fig. 7. (1-column size)
14

ACCEPTED MANUSCRIPT
In order to study the ability of the tracer to follow up the progression of the ischemic
18
lesion, [ F]21 was injected at 3 different time points after the onset of the pMCAO. This
tracer enabled the evolution of hypoxia to be evaluated. At 3 hours after ischemia onset, the
hypoxic area was clearly visualized (Figure 8, left panel). When evaluated at 9 hours, the
18
[
F]21 uptake was spread out to the neighbouring area (Figure 8, central panel), suggesting
1
8
that the penumbra area showed hypoxia signs. After 15 hours, a sharply reduced [ F]21
uptake in the center of the injury was detected, probably as a consequence of neuronal death
located at the level of the core. On the other hand, the surrounding penumbra area still
showed clear signs of hypoxia (Figure 8, right panel).
Fig. 8. (1-column size)
3
. Conclusions
1
8
18
The radiotracer 4-Br-[ F]FMISO ([ F]21) is an interesting probe with medicinal
chemical properties as a potential radiopharmaceutical for the imaging of stroke. In
1
8
18
comparison with the gold standard PET hypoxia tracer [ F]FMISO ([ F]1), compound 21
shows comparable in vitro low toxicity on neuronal cells, better reduction potential, higher
Clog P and faster and total permeability through an in vitro model of the BBB. The in vivo
1
8
PET molecular imaging of [ F]21 in a rodent model of stroke shows a relative accumulation
18
18
that is 54% higher than that obtained with the reference [ F]FMISO ([ F]1). In short,
18
18
18
[
F]21 is a new radiotracer with improved properties with respect to [ F]FMISO ([ F]1) for
the in vivo mapping of the brain ischemic area developed after an experimental stroke event
in rodents. This radiotracer could be a good candidate for further development in the study of
15

ACCEPTED MANUSCRIPT
ischemic stroke. The ongoing research is focusing on pharmacokinetic and imaging data and
results will be published elsewhere.
4
. Experimental section
4
.1. Chemistry
Solvents (HPLC quality, Scharlau) were dried in a Solvent Purification System (MBraun)
by passing through a pre-activated alumina column or were purchased as anhydrous quality.
Reagents were purchased from Sigma-Aldrich and were used as received. The reactions in
which moisture-sensitive compounds were handled were performed under an atmosphere of
dry argon. The reactions were monitored by thin-layer chromatography (TLC) on silica-
coated aluminum sheets (Alugram silica gel 60 F ). The compounds were visualized by UV
2
54
light (254 nm). Column chromatography was carried out with Merck silica gel (0.030–0.075
mm) and the solvents were used as received (Scharlau). Infrared spectra (IR, NaCl windows)
were recorded on a Perkin-Elmer FTIR 1725X instrument. The frequencies (ν) of the more
–
1
1
13
intense bands are given in cm . Nuclear magnetic resonance spectra ( H and C NMR) were
recorded using a Varian Gemini 200 (200 and 50 MHz, respectively), a Varian Mercury-VX-
PLUS
3
00 MHz (300 and 75 MHZ, respectively) and a Varian-UNITY
-500 (500 and 125 MHz,
respectively) instrument. Chemical shifts (
δ
) are given in ppm and are referenced to the
residual signal of the non-deuterated solvent. Coupling constants (J) are given in Hz.
Elemental analyses (C, H, N) were determined on a LECO CHNSO-932 elemental analyzer
and were within ±0.4% of the theoretical values. All compounds were analyzed by tandem
HPLC-MS. HPLC was performed on a C18, 3µm column (Luna, Phenomenex, 3 × 100 mm)
16

ACCEPTED MANUSCRIPT
–
1
using a gradient of MeOH/water/4% formic acid at a flow rate of 1.0 mL min . Products
were detected at = 254 nm. Compounds were ≥95% pure. Mass spectra were recorded on a
λ
Hewlett-Packard 5988A mass spectrometer. High-resolution mass spectra were recorded on
an Agilent 6210 LC/MS TOF mass spectrometer.
4
.2. General procedure for bromination of 11
A solution of diol 11 and N-bromophthalimide in anhydrous DMF was stirred under an
argon atmosphere. The solvent was removed under reduced pressure. Cold MeOH (20 mL)
was added and the phthalimide was filtered off. The solvent was removed under reduced
pressure and the residue was chromatographed on silica gel using hexane/AcOEt 2:1 as
eluent.
4
.2.1. 3-(4-Bromo-2-nitroimidazol-1-yl)propane-1,2-diol (13)
A mixture of diol 11 (2.25 g, 12.03 mmol) and N-bromophthalimide (3.80 g, 16.84 mmol)
in DMF (12 mL) was stirred at rt for 3 h. Compound 13 (2.51 g) was obtained as a yellowish
1
oil. Yield: 79%; Rf(hex/AcOEt 2:1): 0.32; H NMR (300 MHz, CD OD) δ 7.56 (s, 1H, H5-Im), 4.75
3
(
dd, 1H, J = 13.8 Hz, J = 3.3 Hz, CH -Im), 4.37 (dd, 1H, J = 13.8 Hz, J = 8.5 Hz, CH -Im),
2
2
1
3
4
.00–3.93 (m, 1H, CH-OH), 3.60 (m, 2H, CH -OH) ppm; C NMR (75 MHz, CD OD) δ
2 3
1
45.3 (C2 ), 128.6 (C5 ), 114.4 (C4 ), 71.3 (CH-OH), 64.6 (CH -OH), 54.0 (CH -Im)
Im
Im
Im
2
2
–1
ppm; IR (νmax, KBr): 3369 (OH), 15943 (C=N), 1469 (NO ), 1369 (NO ) cm ; HRMS (TOF-
2
2
+
APCI-POS) m/z calcd for C H BrN O [M] 265.9776, found 265.9777.
6
9
3
4
4
.2.2. 3-(4,5-Dibromo-2-nitroimidazol-1-yl)propane-1,2-diol (14)
17

ACCEPTED MANUSCRIPT
A mixture of diol 11 (1.56 g, 8.38 mmol) and N-bromophthalimide (7.57 g, 33.53 mmol)
in DMF (5 mL) were stirred at 10 ºC for 18 h. Compound 14 (1.03 g) was obtained as a
1
yellow solid. Yield: 36%; mp: 98–100 ºC; Rf(hex/AcOEt 2:1): 0.26; H NMR (300 MHz, CD OD)
3
δ 4.70 (d, 2H, J = 5.9 Hz, CH -Im), 3.95 (q, 1H, J = 5.9 Hz, CH-OH), 3.62 (dd, 2H, J = 5.9
2
1
3
Hz, J = 3.9 Hz, CH -OH) ppm; C NMR (75 MHz, CD OD) δ 117.5 (C5 ), 114.5 (C4 ),
2
3
Im
Im
7
1
3
1.3 (CH-OH), 65.1 (CH -OH), 53.8 (CH -Im); (ν KBr): 3392 (OH), 2933; 1543 (C=N),
2
2
max,
–
1
+
478 (NO ), 1363 (NO ) cm ; HRMS (TOF-APCI-POS) m/z calcd for C H Br N O [M]
2
2
6
8
2
3
4
43.8876, found 343.8847.
4
.3. General procedure for regioselective tosylation
A solution of diol 13 or 14, tosyl chloride, triethylamine and dibutyltin oxide in anhydrous
CH CN was stirred under an argon atmosphere at rt. The solvent was removed under reduced
3
pressure and the residue was treated as indicated below.
4
.3.1. Toluene-4-sulfonic acid 3-(4-bromo-2-nitroimidazol-1-yl)-2-hydroxypropyl ester (15)
A mixture of diol 13 (50.0 mg, 0.267 mmol), tosyl chloride (50.0 mg, 0.267 mmol),
triethylamine (0.037 mg, 0.267 mmol) and dibutyltin oxide (1.3 mg, 0.053 mmol) in CH CN
3
(4 mL) was stirred for 5 h. The residue was chromatographed on silica gel using
AcOEt/hexane 8:2. Compound 15 (76.0 mg) was obtained as a yellow solid. Yield: 83%; mp:
1
1
40.9–143.5 ºC; Rf(AcOEt/hex 8:2): 0.45; H NMR (300 MHz, CD OD)
δ
7.85 (d, 2H, J = 8.2 Hz,
3
Ho-Ts), 7.49 (d, 2H, J = 8.2 Hz, Hm-Ts), 7.41 (s, 1H, H5-Im), 7.13 (s, 1H, H4-Im), 4.66 (dd, 1H, J
13.8 Hz, J = 3.3 Hz, CH -Im), 4.36 (dd, 1H, J = 13.8 Hz, J = 8.2 Hz, CH -Im), 4.12–4.05
=
2
2
1
3
(
(
(
m, 3H, CH, CH -OTs), 2.50 (s, 3H, CH ) ppm; C NMR (50 MHz, DMSO-d )
δ
144.6
2
3
6
C4 ), 144.38 (C2 ), 131.4 (C1 ), 129.7 (2C, C3 ), 127.9 (C5 ), 127.2 (2C, C2 ), 126.8
Ts
Im
Ts
Ts
Im
Ts
C4 ), 71.0 (CH-OH), 66.2 (CH -OTs), 50.7 (CH -Im), 20.6 (CH ) ppm; IR (νmax, KBr):
Im
2
2
3
18

ACCEPTED MANUSCRIPT
–
1
3
204 (OH), 1539 (C= N), 1492 (NO ), 1365 (SO ), 1171 (SO ) cm ; HRMS (TOF-APCI-
2 2 2
+
POS) m/z calcd for C H N SO [M] 342.0706, found 342,0624; Anal. (C H N O S)
1
3
16
3
6
13 15
3
6
theoretical: C, 45.74; H, 4.42; N, 12.31; S, 9.39. Found: C, 45.58; H, 4.42; N, 12.11; S, 9.61.
4
.3.2. Toluene-4-sulfonic acid 3-(4,5-dibromo-2-nitroimidazol-1-yl)-2-hydroxypropyl ester
(16)
A mixture of diol 14 (0.905 g, 2.64 mmol), tosyl chloride (0.503 g, 2.64 mmol),
triethylamine (0.268 g, 2.64 mmol) and dibutyltin oxide (0.013 g, 0.052 mmol) in CH CN (40
3
mL) was stirred for 24 h. The residue was dissolved in CH Cl (20 mL), filtered and washed
2
2
with brine (2 × 10 mL). The aqueous phase was extracted with CH Cl (3 × 10 mL) and the
2
2
combined organic extracts were dried over anhydrous Na SO , filtered, and evaporated to
2
4
dryness. Finally, the residue was chromatographed on silica gel using a gradient of
hexane/AcOEt from 8:2 to 6:4. Compound 16 (0.67 g) was obtained as a yellow solid. Yield:
1
5
1%; mp: 44–46 ºC; Rf(Hex/AcOEt 6:4): 0.50; H NMR (300 MHz, CDCl ) δ 7.84 (d, 2H, J = 8.0
3
Hz, Ho-Ts), 7.48 (d, 2H, J = 8.0 Hz, Hm-Ts), 4.62 (m, 2H, CH -Im), 4.11 (s , 3H, CH-OH,
2
ap
1
3
CH -OTs), 2.50 (s, 3H, CH ) ppm; C NMR (75 MHz, CD OD) δ 146.8 (C4 ), 133.9
2
3
3
Ts
(
(
(
C1 ); 131.1 (2C, C3 ), 129.1 (2C, C2 ), 117.6 (C5 ), 114.6 (C4 ), 72.1 (CH-OH), 68.3
Ts Ts Ts Im Im
CH -OTs), 52.8 (CH -Im), 21.6 (CH ) ppm; IR (νmax, KBr): 3413 (OH), 1597 (C=N), 1474
2
2
3
–
1
NO ), 1362 (NO ), 1175 (SO ) cm ; HRMS (TOF-APCI-POS) m/z calcd for
2
2
2
+
C H Br N O S [M] 497.8965, found 497.8991; Anal. (C H Br N O S) theoretical: C,
13
14
2
3
6
13 13
2
3
6
3
1.28; H, 2.65; N, 8.42; S 6.42. Found: C, 31.39; H, 2.67; N, 8.61; S, 6.49).
4
.4. General procedure for tetrahydropyranyl ketal formation
A solution of 15 or 16, DHP and PPTS in anhydrous CH Cl was stirred at 0 ºC under an
2
2
argon atmosphere. The solvent was removed at rt under reduced pressure. The residue was
chromatographed on silica gel using AcOEt/hexane 6:4 as eluent.
19

ACCEPTED MANUSCRIPT
4
.4.1. Toluene-4-sulfonic acid 3-(4-bromo-2-nitroimidazol-1-yl)-2-(tetrahydropyran-2-
yloxy)propyl ester (17)
A mixture of alcohol 15 (1.2 g, 2.85 mmol), DHP (0.74 g, 8.83 mmol) and PPTS (0.89 g,
3
.54 mmol) in CH Cl (78 mL) was stirred for 22 h. Compound 17 (1.0 g) was obtained as a
2 2
1
yellow solid. Yield: 68%; mp: 120.0–122.3 ºC; Rf(Hex/AcOEt 6:4): 0.44; H NMR (500 MHz,
CDCl ) δ 7.79 (d, 2H, J = 8.1 Hz, Ho-Ts, dias. A), 7.78 (d, 2H, J = 8.1 Hz, Ho-Ts, dias. B), 7.36
3
(d, 2H, J = 8.1 Hz, Hm-Ts, dias. A+B), 7.09 (s, 1H, H5-Im, dias. B), 7.01 (s, 1H, H5-Im, dias. A),
4
.79 (dd, 1H, J = 14.0 Hz, J = 3.3 Hz, CH -Im, dias. A), 4.68 (dd, 1H, J = 14.0 Hz, J = 3.3
2
Hz, CH -Im, dias. B), 4.44 (m, 1H, H
, dias. B), 4.34 (m, 2H, CH -Im dias. A+B), 4.27
2-THP 2
2
(
m, 1H, H2-THP, dias. A), 4.21 (m, 2H, CH-OTHP, dias. B + CH -OTs, dias. A), 4.11 (dd, 1H,
2
J = 11.2 Hz, J = 4.5 Hz, CH -OTs, dias. B), 4.05–3.98 (m, 3H, CH-OTHP, dias. B + CH -
2
2
OTs, dias. A+B), 2.45 (s, 3H, CH , dias. A+B), 1.67–1.38 (m, 6H, H3,4,5-THP, dias. A+B) ppm;
3
1
3
C NMR (75 MHz, CDCl ) δ 145.8, 132.5, 130.1 (2C), 128.0 (2C), 126.3, 115.3, 100.9,
3
9
7.8, 74.1, 70.3, 68.5, 67.0, 63.2, 51.1, 50.5, 30.3, 30.1, 24.9, 24.8, 21.7, 19.7, 19.3 ppm; IR
–
1
(
(
(
νmax, KBr): 3449 (OH), 1535 (C=N), 1499 (NO ), 1350 (SO ), 1168 (SO ) cm ; HRMS
2 2 2
+
TOF-APCI-POS) m/z calcd for C H BrN O S [M] 504.0440, found 504.0400; Anal
1
8
23
3
7
C H BrN O S) theoretical: C, 42.86; H, 4.39; N, 8.33; S, 6.35. Found: C, 42.60; H, 4.46;
1
8
22
3
7
N, 8.24; S, 6.36.
4
.4.2. Toluene-4-sulfonic acid 3-(4,5-bromo-2-nitroimidazol-1-yl)-2-(tetrahydropyran-2-
yloxy)propyl ester (18)
A mixture of alcohol 16 (0.62 g, 1.25 mmol), DHP (0.34 g, 3.75 mmol) and PPTS (0.38 g,
1
.5 mmol) in CH Cl (35 mL) was stirred for 4 h. Compound 18 (0.5 g) was obtained as an
2 2
oil which was treated with Et O to give a yellow solid. Yield: 69%; mp: 51.0–53.0 ºC;
2
1
Rf(Hex/AcOEt 6:4): 0.54; H NMR (300 MHz, CDCl ) δ 7.78 (d, 2H, J = 8.3 Hz, Ho-Ts, dias.
3
20

ACCEPTED MANUSCRIPT
A+B), 7.36 (d, 2H, J = 8.3 Hz, Hm-Ts, dias. A+B), 4.66 (m, 2H, H2-THP, CH -Im, dias. A+B),
2
4
.36–3.95 (m, 4H, CH -Im + CH-OTHP + CH -OTs dias. A+B), 3.67–3.27 (m, 1H, H6-THP,
2 2
dias. A+B), 3.13 (m, 1H, H6-THP, dias. A+B), 2.45 (s, 3H, CH , dias. A+B), 1.54–1.18 (m, 6H,
3
1
3
H3,4,5-THP, dias. A+B) ppm; C NMR (75 MHz, CDCl ) δ 145.8, 145.4, 132.4, 132.2, 131.8,
3
1
2
30.0 (2C), 127.9 (2C), 101.3, 97.5, 74.7, 70.6, 67.9, 63.9, 62.9, 51.5, 50.5, 30.3, 24.7, 21.7,
0.2, 19.2 ppm; IR (νmax, KBr): 3435 (OH), 2945, 1597 (C=N), 1545, 1474 (NO ), 1363
2
–
1
+
(
NO ), 1176 (SO ) cm ; HRMS (TOF-APCI-POS) m/z calcd for C H Br N O S (M)
2 2 18 22 2 3 7
5
5
81.9540, found 581.9545; Anal (C H Br N O S) theoretical: C, 37.06; H, 3.62; N, 7.20; S,
18 21 2 3 7
.49. Found: C, 37.13; H, 3.63; N, 7.25; S, 5.67.
4
.5. General procedure for fluorination
A solution of sulfonate 17 (0.99 mmol) or 18 (0.35 mmol) and CsF (3 equiv) in tert-amyl
alcohol (3.3 mL per mmol of sulfonate) was heated at 120 ºC for 3 h. The solvent was
removed under reduced pressure and the residue was chromatographed on silica gel.
4
.5.1. 4-Bromo-1-[3-fluoro-2-(oxiran-2-yloxy)-propyl]-1H-2-nitroimidazole (19)
Compound 17 (0.50 g) gave 19 (0.24 g) as a yellowish oil after chromatography using
1
AcOEt/hexane 6:4 as eluent. Yield: 69%; Rf(AcOEt/Hex 6:4): 0.42; H NMR (300 MHz, CDCl ) δ
3
7
.17 (s, 1H, H5-Im, dias. B), 7.11 (s, 1H, H5-Im, dias. A), 4.81 (dd, 1H, J = 14.1 Hz, J = 3.6 Hz,
CH -Im, dias. A), 4.73 (dd, 1H, J = 14.1 Hz, J = 3.6 Hz, CH -Im, dias. B), 4.65–4.30 (m, 4H,
2
2
H2-THP+CH -Im+CH -F, dias. A+B), 4.28–4.16 (m, 1H, CH-OTHP, dias. A), 4.11–4.01 (m,
2
2
1
H, CH-OTHP, dias. B), 3.81–3.74 (m, 1H, H6-THP, dias. A), 3.45–3.38 (m, 1H, H6-THP, dias.
A), 3.25–3.21(m, 2H, H6-THP, dias. A+B), 1.70–1.37 (m, 6H, H3,4,5-THP, dias. A+B) ppm; IR
–
1
(
νmax, KBr): 3420 (OH), 1595 (C=N), 1462 (NO ), 1354 (NO ), 1358 (SO ), 1178 (SO ) cm ;
2 2 2 2
+
HRMS (TOF-APCI-POS) m/z calcd for C H BrFN O [M] 352.0303, found 352.0315.
1
1
16
3
4
21

ACCEPTED MANUSCRIPT
4
.5.2. 4,5-Dibromo-1-[3-fluoro-2-(oxiran-2-yloxy)propyl]-1H-2-nitroimidazole (20)
Compound 18 (0.20 g) gave 20 (0.045 g) as a yellowish oil after chromatography using
1
AcOEt/hexane 7:3 as eluent. Yield: 31%; Rf(AcOEt/Hex 7:3): 0.48; H NMR (300 MHz, CDCl ) δ
3
4
.81–4.40 (m, 5H, H2-THP+CH -Im+CH -F, dias. A+B), 4.24–4.16 (m, 2H, CH-OTHP, dias.
2 2
A+B), 3.80 (m, 1H, H6-THP, dias. A), 3.46 (m, 1H, H6-THP, dias. A), 3.20 (m, 2H, H6-THP, dias.
B), 1.64–1.22 (m, 6H, H3,4,5-THP, dias. A+B) ppm; HRMS (TOF-APCI-POS) m/z calcd for
+
C H Br FN O [M] 429.9408, found 429.9404.
11
15
2
3
4
4
.6. General procedure for tetrahydropyranyl ketal deprotection
The THP ketal 19 (0.59 mmol) or 20 (0.10 mmol) was dissolved in AcOH (4.1 mL per
mmol), THF (2.2 mL per mmol) and H O (1.1 mL per mmol) and the mixture was heated at
2
6
5 ºC. The solvent was removed under reduced pressure and the residue was
chromatographed on silica gel using hexane/AcOEt. The resulting thick oil was treated with
Et O to give a pale-yellow solid which was filtered off and dried.
2
4
.6.1. 1-(4-Bromo-1H-2-nitroimidazol-1-yl)-3-fluoropropan-2-ol (21)
Compound 19 (0.209 g) was heated for 6 h to give 21 (0.103 g) after chromatography
1
using AcOEt/hexane 4:6 as eluent. Yield: 65%; mp: 104–105 ºC; Rf(AcOEt/Hex 4:6): 0.44; H
NMR (500 MHz, CD OD: δ 7.57 (s, 1H, H5-Im), 4.75 (dd, 1H, J = 13.8 Hz, J = 3.3 Hz, CH2-
3
1
Im), 4.47 (dd, 2H, J = 47.1 Hz, J = 4.3 Hz, CH -F), 4.44 (dd, 1H, J = 13.8 Hz, J = 8.9 Hz,
H-F
2
2
13
CH -Im), 4.18 (dq, 1H, J = 19.8 Hz, J = 4.0 Hz, CH-OH) ppm; C NMR (75 MHz,
2
H-F
1
CD OD) δ 146.0 (C2 ), 128.7 (C5 ), 114.6 (C4 ), 85.3 (d, 1C, J = 171.2 Hz, CH -F),
3
Im
Im
Im
C-F
2
2
3
6
9.5 (d, 1C, J = 19.8 Hz, CH-OH), 53.1 (d, 1C, J = 7.4 Hz, CH -Im) ppm; IR (νmax,
C-F C-F 2
–1
KBr): 3398, 3103, 1541, 1510, 1470, 1400, 1377, 1261, 1012, 959, 845 cm ; HRMS (TOF-
22

ACCEPTED MANUSCRIPT
+
APCI-POS) m/z calcd for C H BrFN O [M] 267.9733, found 267.9719; Anal
6
8
3
3
(
C H BrFN O ) theoretical: C, 26.89; H, 2.63; N, 15.68. Found: C, 26.67; H, 2.82; N, 15.91.
6 7 3 3
4
.6.2. 1-(4,5-Dibromo-1H-2-nitroimidazol-1-yl)-3-fluoropropan-2-ol (22)
Compound 20 (0.045 g) was heated for 3 h to give 22 (0.025 g) as a yellowish oil after
1
chromatography using AcOEt/hexane 4:6 as eluent. Yield: 69%; Rf(AcOEt/Hex 4:6): 0.44; H
1
NMR (300 MHz, CD OD) δ 4.75 (dd, 2H, J = 6.6 Hz, J = 0.6 Hz, CH -Im), 4.50 (dd, 2H, J
3
2
H-
1
3
=
47.1 Hz, J = 4.4 Hz, CH -F), 4.24–4.12 (m, 1H, CH-OH) ppm; C NMR (125 MHz,
F
2
1
CD OD): δ 146.3 (C2 ), 117.7 (C5 ), 114.7 (C4 ), 85.3 (d, 1C, J = 170.7 Hz, CH -F),
3
Im
Im
Im
C-F
2
2
3
6
9.3 (d, 1C, J = 20.0 Hz, CH-OH), 52.5 (d, 1C, J = 7.6 Hz, CH -Im) ppm; IR (νmax,
C-F C-F 2
–
1
KBr): 3404, 1547, 1473, 1361, 1248 cm ; HRMS (TOF-APCI-POS) m/z calcd for
+
C H Br FN O [M] 345.8833, found 345.8879.
6
7
2
3
3
4
.7. Cyclic voltammetry
Compounds were dissolved in 0.1 M Bu NBF in DMF as SSE (solvent support
4
4
–
3
electrolyte) to a concentration of 4 × 10 M. Cyclic voltammetry was carried out in a
VoltaLab PGZ100 potentiostat at 25 ± 1 ºC under an argon atmosphere using glassy carbon
+
electrode, platinum as auxiliary electrode and Ag/satd Ag as reference electrode at a scan
speed of 100 mV/s. In the procedure, SSE (20 mL) was added to the cuvette, electrodes were
submerged into the solution, solvent was deoxygenated using a flow of argon for 15 min and
a scan of SSE was carried out. The compound (21 or 1) was added and the solvent was again
deoxygenated for 15 min. Finally, scans were performed using cycles from 0 to –2,5 V and
from –2,5 to 0 V recording cyclic voltammograms for each compound.
4
.8. bEnd5 cell line cultures
23

ACCEPTED MANUSCRIPT
Murine brain endothelial cells (b.End5), which are recognized as having brain
endothelium-like properties, were obtained from the European Collection of Cell Cultures
(
UK). This cell model is an appropriate choice to study blood-brain barrier function [38,39].
The cells were grown in Dulbeccos’s Modified Eagle’s Medium, with high glucose content
4.5 g/L; Lonza), supplemented with 10% heat inactivated fetal bovine serum (FBS, Gibco)
(
and gentamicin (40 µg/mL), and samples were maintained under standard cell culture
conditions at 37 °C and 5% CO in a humid atmosphere. The cells were trypsinized (Trypsin
2
0.5 mg/mL/EDTA 0.2 mg/mL, in Hank’s balanced salt solution, Sigma) and 50,000 cells in
100 µL of medium were seeded in inserts (culture treated polycarbonate membrane, pore size
0.4 ꢀm; 9 mm diameter; Costar). 700 µL of medium was added to the lower compartment.
Cultures were incubated for 24 h and the medium was changed to the same medium but without
phenol red and FBS prior to the assessment of drug transport. Sister cultures in P96 plates were
used to assess confluency. The drugs at the concentrations used did not alter the viability of the
cells at 24 h after the addition (not shown; MTT assay).
4
.9. Transport across an endothelial cell monolayer
The transport of the compounds under study was assessed in an in vitro system consisting
of a confluent endothelial cell monolayer on an insert (upper compartment), which was
placed onto a P24 well plate (lower compartment). Inserts placed in P24 transwell plates
consisted in 6.5 mm diameter inserts, with tissue culture treated polycarbonate membrane and
0
.4 µm pore size (Costar, Corning Inc., NY, USA). The compounds were added to the upper
compartment and the medium in the lower compartment was collected (600 µL) at 15 min
intervals up to 60 min after its addition. Duplicate cultures were used for each drug and for
each time. The media samples were freeze-dried and were stored at 4 ºC prior to
24

ACCEPTED MANUSCRIPT
quantification by HPLC analysis. Transport was expressed as percentage of the actual amount
used each day, which was also measured by HPLC.
4
.10. Primary cultures and assessment of neuronal injury
Primary mixed cultures were prepared from the cerebral cortex of 18-day-old Wistar rat
embryos and were grown in minimum essential medium (Invitrogen-Life Technologies,
Carslbad, CA, USA) as described previously [40]. Where indicated, primary cultures were
also grown in neurobasal medium (Invitrogen-Life Technologies) with serum-free B-27
supplement. Stock solutions (20 mM) in DMSO were prepared for 1 and 21. From these
solutions, new stock solutions (2 mM) in DMSO were made for each compound. The same
was carried out from the former solutions to prepare new stock solutions (0.2 mM) for both
compounds. Aliquots (10 µL) were taken from the 20, 2 and 0.2 mM solutions to prepare
1
00, 10 and 1 µM solutions in DMEM medium (2 mL), respectively. The final concentration
of DMSO in these solutions was 0.5 %. Cultures were then treated with different
concentrations (0, 1, 10, 100 µM) of compounds 1 and 2 at 7–9 days of in vitro culture. The
MTT and LDH reduction assays were used to measure cell viability after 24 h of incubation
with compounds. MTT (Aldrich, M2128) (0.5 mg/mL) was added to the medium and, after 2
h at 37 °C, the resulting formazan salts were dissolved in DMSO and were
spectrophotometrically quantified (OD, optical density) at 540 nm (Thermomax, Molecular
Devices). Viability percentage was calculated as % Viability = (OD treated cells / OD control
cells) × 100. Extracellular and intracellular LDH activities were spectrophotometrically
measured by following tetrazolium reduction at an absorbance wavelength of 490 nm
(Boehringer Mannheim kit). Total LDH activity was defined as the sum of intracellular and
extracellular LDH activity. Released LDH was defined as the percentage of extracellular
compared to total LDH activity.
25

ACCEPTED MANUSCRIPT
4
4
.11. Radiosynthesis
1
8 –
.11.1. Generation of F
1
8
–
18
18
F was produced by an O(p,n) F reaction in an IBA cyclotron following bombardment
1
8
of enriched H O (ROTEM Industries). When the theoretical activity was in the range 100
2
18
mCi to 3000 mCi (3.7 GBq to 111.0 GBq) the material was transferred in solution with H O
2
to the synthesizer.
1
8
4
.11.2. Radiolabeling of [ F]21
The radiosynthesis was undertaken on a Synthera® (IBA Molecular) automated
synthesizer using an Integrated Fluidic Processor (ABX). Reagents, including cryptand
solution, HCl, MeCN and citrate buffer were purchased from ABX. Cartridges for solid phase
1
8 –
purification were purchased from the Waters Corporation. The F was trapped on a pre-
conditioned Waters Sep-Pak QMA Light (quaternary methyl ammonium) anion exchange
cartridge (46 mg, ABX) and washed with H O before it was eluted with cryptand solution
2
(
2.2 mg kryptofix-222 and 7 mg K CO in 600 µL of a 1:1 mixture of H O and MeCN). The
2 3 2
eluate was dried under reduced pressure and the reactor was heated to 130 °C over 5 mintues.
Following addition of the precursor solution (6–12 mg dissolved in 2.5 mL MeCN), the
reactor was sealed and the mixture was heated for 5 min. The reaction mixture was cooled to
1
00 °C over 3 minutes before a solution of HCl (1M, 1 mL) was added to effect hydrolysis,
which was carried out at 100 °C over 5 minutes. The reaction mixture was buffered (pH =
®
18
5
.8) with a solution (4.5 mL, Synthera [ F]FLT reagent sets buffer solution, ABX)
containing 1M AcONa (0.5 mL), 1M NaOH (0.45 mL), EtOH (0.25 mL) and water (3.3 mL),
mixed, and transferred automatically to the HPLC system.
4
.11.3. Purification
26

ACCEPTED MANUSCRIPT
The crude reaction mixture was purified by automated injection onto a semi-preparative
C18 HPLC column (250 mm × 10 mm, 100 Å, Phenomenex) in a Synthera® HPLC system
(
ABX Molecular). Separated compounds were eluted with 95:5 H O/EtOH and a flow rate of
2
4
mL/min, and fractions were collected using the peak collection tool of the operating
1
8
software. Under these conditions the retention time of [ F]21 was 33 minutes and a major
radiochemical impurity eluted after 21 minutes. The total synthesis time was 55 minutes and
the radiochemical yield was 18% (decay corrected) with a specific activity of 150 ± 15
GBq/µmol (at the end of radiosynthesis).
4
.11.4. Formulation
Fractions were prepared for injection by dilution with a 0.9% saline solution in order to
reach the desired activity concentration (mCi/mL).
4
.11.5. Quality Control
Radiochemical purity was determined by analytical HPLC (Agilent 1200 series, Zorbax
Eclipse Plus C18 Column, 5 ꢀm, 4.6 × 100 mm; H O:EtOH = 95:5, flow rate = 1 mL/min)
2
and was greater than 99%. The identity of the compound was confirmed by comparison of the
radiolabeled peak with the UV peak contributed by the reference standard. Radionuclide
identity was confirmed by GENIE multichannel analysis, with the area of the peak at 511
keV in excess of 97%. In addition, the half-life of the sample was determined by taking
twenty-one activity measurements over a 20 minute interval and the use of the decay
equation. The half-life was calculated to be 108 minutes. The pH of the sample was measured
with colored pH strips and was found to be 6.9.
4
.12. Permanent middle cerebral artery occlusion (pMCAO)
27

ACCEPTED MANUSCRIPT
Male Fischer rats (250–300 g body weight; Harlan; Barcelona, Spain) (n = 7) were used
in the study. All experimental protocols adhered to the guidelines of the Animal Welfare
Committee of the Universidad Complutense (following EU directives 86/609/CEE and
2
003/65/CE). Animals were housed individually under standard conditions of temperature
and humidity and a 12 hours light/dark cycle (lights on at 08:00) with free access to food and
water. Rats were anesthetized (isoflurane 1.5–2% in 100% oxygen) and underwent analgesia
(metamizole 100 mg/kg ip). Body temperature was maintained at physiological levels with a
heating pad during the surgical procedure and anesthesia recovery. The permanent cerebral
ischemia model consisted of right pMCAO and of both common carotid artery occlusion
[
40]; after 75 min, the contralateral common carotid artery was de-occluded to reduce animal
death rate. In this model, the characteristic changes of ischemic necrosis are limited to the
parietal and sensory-motor cortex and large infarcts are reproducibly generated in the right
MCA territory after 24 h of surgery. The injured area – seen at 3 h after surgery – showed a
variable staining throughout the MCA territory combining: 1) areas with loss of cytoplasmic
Nissl staining and rounded with swollen neuronal cells and 2) areas showing a triangular
morphology with a reduced size of cell body and processes, indicating cellular damage, as
reported previously [40]. After the surgery, animals were returned to their cages to recover
from anaesthesia. The animals were then subjected to PET imaging study.
4
.13. Positron emission tomography
For in vivo comparison of the hypoxia-imaging compounds, brain PET scans were
performed in a dedicated small animal PET-CT hybrid tomograph (Albira ARS, Oncovision,
18
18
18
Valencia, Spain). Briefly, [ F]FMISO ([ F]1) or [ F]21 (55.5 MBq in approx. 0.2mL 0.9%
NaCl, i.v.) were injected into animals 30 minutes after surgery. After an uptake period of 150
min, the rats were anesthetized with isoflurane (4.5% for induction and 1.5% for
28