Combinatorial synthesis of labelled drugs and PET tracers: Synthesis of a focused library of 11C-carbonyl-labelled acrylamides as potential biomarkers of EGFR expression

Article abstract of DOI:10.1002/jlcr.2981

Combinatorial synthesis is extensively used in drug development and lead optimisation. However, this approach has rarely been used for positron emission tomography because of limitations in available technologies. [ ¹¹C]Carbon monoxide is amenable to combinatorial synthesis in transition-metal-catalysed reactions because it can react with a wide variety of electrophiles and nucleophiles, which opens up the possibilities for combinatorial radiochemistry. Herein, we exemplify the combinatorial approach by ¹¹C-labelling a library of epidermal growth factor receptor inhibitors. The selection of candidates was guided by molecular docking. Epidermal growth factor receptor is overexpressed in a variety of tumours, and it has become an important drug target. The ¹¹C-labelling reactions were performed using four substituted vinyl iodides and three different 4-anilino-6-aminoquinazolines using a palladium-mediated reaction with [ ¹¹C]carbon monoxide using a single set of reaction conditions. In total, 12 labelled acrylamide derivatives were radiolabelled and obtained in 24-61% decay-corrected radiochemical yield (from [¹¹C]carbon monoxide). Starting from 5.6 GBq [¹¹C]carbon monoxide, 0.85 GBq of formulated N-[4-(3-bromo-phenylamino)-quinazolin-6-yl]-acryl[ ¹¹C]amide [¹¹C]12da was obtained within 47 min from end of bombardment (specific activity of 60 GBq μmol^-1). This strategy is an example of how [¹¹C]carbon monoxide can be utilised in the labelling of libraries of drug candidates and positron emission tomography tracers for in vitro and in vivo testing. Copyright

Full text of DOI:10.1002/jlcr.2981

Research Article
Received 29 August 2012,
Revised 3 October 2012,
Accepted 5 October 2012
Published online 14 November 2012 in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/jlcr.2981
Combinatorial synthesis of labelled drugs and
PET tracers: synthesis of a focused library of
11C-carbonyl-labelled acrylamides as potential
biomarkers of EGFR expression
Ola Åberga, † and Bengt Långströma,b
*
Combinatorial synthesis is extensively used in drug development and lead optimisation. However, this approach has rarely been
used for positron emission tomography because of limitations in available technologies. [11C]Carbon monoxide is amenable
to combinatorial synthesis in transition-metal-catalysed reactions because it can react with a wide variety of electrophiles and
nucleophiles, which opens up the possibilities for combinatorial radiochemistry. Herein, we exemplify the combinatorial
approach by 11C-labelling a library of epidermal growth factor receptor inhibitors. The selection of candidates was guided by
molecular docking. Epidermal growth factor receptor is overexpressed in a variety of tumours, and it has become an important
drug target. The 11C-labelling reactions were performed using four substituted vinyl iodides and three different 4-anilino-6-
aminoquinazolines using a palladium-mediated reaction with [11C]carbon monoxide using a single set of reaction conditions.
In total, 12 labelled acrylamide derivatives were radiolabelled and obtained in 24–61% decay-corrected radiochemical yield (from
[11C]carbon monoxide). Starting from 5.6GBq [11C]carbon monoxide, 0.85 GBq of formulated N-[4-(3-bromo-phenylamino)-
quinazolin-6-yl]-acryl[11C]amide [11C]12da was obtained within 47 min from end of bombardment (specific activity of
60 GBq mmolꢀ1). This strategy is an example of how [11C]carbon monoxide can be utilised in the labelling of libraries of drug
candidates and positron emission tomography tracers for in vitro and in vivo testing.
Keywords: combinatorial chemistry; carbonylation; palladium; PET chemistry; radiolabelling
formed to the radionuclide, thus the chemical variation possible in
Introduction
a series of analogues is limited. The carbonyl group, on the other
Positron emission tomography (PET) is a non-invasive imaging
hand, is commonly found in biologically active molecules, and
technology that is based on the spatial detection of a positron-
the carbon binds two other atoms except for the doubly bound
emitting radionuclide incorporated into a biologically relevant
oxygen atom. Incorporation of a carbonyl group can be achieved
molecule, that is, a radiotracer. The use of high specific activity 11
C
via carbonylation. A unique and attractive feature of carbon monoxide
is that two bonds are formed in a typical cross-coupling reaction,
thus the possibility of using carbonylations for combinatorial
chemistry.2
Several chemical and technological methods to overcome
the low solubility and reactivity of [11C]carbon monoxide have
been evaluated including recirculation techniques,3 chemical com-
plexation with THFꢁBH3,3 or copper(I)scorpionates,4 high pressure
reactors,5 micro-fluidic reactors6 and xenon-based atmospheric
(t1/2 = 20.4min) allows a labelled molecule to be administered in
very small quantities (usually in the order of mg) and followed over
time in vivo using a PET scanner. The small mass administered
means that the regulatory burden is significantly lower when
compared with the administration of pharmacologically active
doses.1 Because of this reason, PET is increasingly used in drug
development for evaluation of new drug candidates and first
test in man following the micro-dosing concept.1 These phase 0
studies usually include biodistribution and dose occupancy studies
and often lead to a more rapid decision-making process on new
candidate drugs and save time and resources. Despite these benefits,
today, only the optimised lead in a typical drug development
programme is evaluated using this powerful technology. This is
aUppsala University, Department of Biochemistry and Organic Chemistry, BMC,
Box 576 SE-751 23, Uppsala, Sweden
to an extent associated with limitations in what type of chemical bDepartment of Nuclear Medicine, PET & Cyclotron Unit, Odense University
Hospital, Institute of Clinical Research, University of Southern Denmark, Odense,
Denmark
reactions that can be performed with the most commonly used
labelled precursors, that is, [11C]methyl iodide/triflate and [18F]
fluoride. Far from all drug candidates contain a methyl group or a
*Correspondence to: Ola Åberg, Uppsala University, Platform for preclinical PET,
fluorine suitable for labelling. 18F-labelling often involves harsh Box 574, SE-751 23 Uppsala, Sweden.
E-mail: ola.aberg@pet.medchem.uu.se
reaction conditions, thus lengthy multi-step reactions and prosthetic
† Current address: Uppsala, University Platform for preclinical PET, Box 574, SE-751
groups are needed to overcome this issue. Another shortcoming of
methylation and fluorination is that only one chemical bond is 23 Uppsala, Sweden.
J. Label Compd. Radiopharm 2012, 55 477–483
Copyright © 2012 John Wiley & Sons, Ltd.

O. Åberg and B. Långström
in a cold trap (ꢀ196 ꢃC). [11C]Carbon dioxide was reduced to [11C]carbon
monoxide over Zn granules at 400 ꢃC and concentrated silica in a
cold trap (ꢀ196 ꢃC). The carbonylation reactions were carried out in
a 200-mL stainless steel micro-autoclave using a previously described
technique.5a
Thin-layer chromatography was performed on pre-coated Merck silica
gel plates (60 F254) and visualised with UV light. 1H and 13C NMR spectra
were recorded on a Varian (Palo Alto, CA, USA) 400 or 500-MHz spectro-
meter, and chemicals shifts are given in ppm (d) and referenced to
residual solvent. J values are given in Hz. LC-MS analyses of reference com-
pressure reactors.7 These efforts have the potential to make [11C]
carbon monoxide available for any lab with an on-site cyclotron.
[11C]Carbon monoxide can be used to synthesise a plethora of
functional groups commonly found in drugs via transition-
metal-mediated carbonylative cross-coupling reactions. These
include but are not limited to amides, ureas, sulfonylureas,
ketones, carboxylic acids, carbonates, esters, thioesters, carbamates
and thiocarbamates.8
Radiolabelling and screening of libraries is a powerful optimi-
sation tool and has the potential to bring valuable knowledge pounds and precursors were performed on a Gilson reverse-phase HPLC
equipped with a Finnigan ESI mass spectrometer (MeCN/H2O and 0.1% for-
mic acid). Intermediates, precursors and references were synthesised
using known methods, and their spectroscopic characteristics
corresponded well with those reported; 3-chloro-4-fluoroaniline,16 3H-qui-
nazoline-4-one (1), 6-nitro-3H-quinazolin-4-one (2), 4-chloro-6-nitro-quina-
zoline (3),17 (3-bromo-phenyl)-(6-nitro-quinazolin-4-yl)-amine (4a),18 (3-
chloro-4-fluoro-phenyl)-(6-nitro-quinazolin-4-yl)-amine (4b),19 [3-chloro-
4-(3-fluoro-benzyloxy)-phenyl]-(6-nitro-quinazolin-4-yl)-amine (4c), 2-
chloro-1-(3-fluoro-benzyloxy)-4-nitro-benzene (6), 3-chloro-4-(3-fluoro-benzy-
loxy)-phenylamine (7),20 (E)-3-(tributylstannyl)-2-propen-1-ol (8),21 (E)-3-iodo-
about structure activity relationship for both tracer and
drug candidates as exemplified by 11C-labelled analogues of
WAY-100635,9 11C-labelled inhibitors of the vascular endothelial
growth factor receptor 2,10 18F-labelled peptides for the integrin
avb6,11 11C-labelled cyclooxygenase inhibitors12 and 11C-labelled
DAA1106 analogues.13
In this manuscript, we wish to emphasise the usefulness of
[
11C]carbon monoxide for synthesis of focused radiolabelled
compound libraries, here exemplified with a series of epidermal
growth factor receptor (EGFR) inhibitors labelled in the carbonyl 2-propen-1-ol (9),22 (E)-3-bromo-1-iodo-propene (10),23 (E)-4-morpholin-4-
yl-but-2-enoic acid [4-(3-chloro-4-fluoro-phenylamino)-quinazolin-6-yl]-amide
(12bb),24 N-[4-(3-bromo-phenylamino)-quinazolin-6-yl]-acrylamide (12da),19
N-[4-(3-chloro-4-fluoro-phenylamino)-quinazolin-6-yl]-acrylamide (12db)19
and N-{4-[3-chloro-4-(3-fluoro-benzyloxy)-phenylamino]-quinazolin-6-yl}-acryla-
mide (12dc).25
position. The development of a PET tracer has similarities
to drug development and radiolabelled drugs, but there are
also differences. Occasionally, radiolabelled drugs can serve as
tracers as exemplified with the EGFR tyrosine kinase inhibitor
(TKI) [11C]erlotinib.14 On the other hand, the EGFR TKI [18F]gefitinib
was inefficient for in vivo tumour imaging because of high non-
specific binding and non-saturable cellular uptake.15 From the
two examples mentioned, it is clear that even small molecular
differences will have a profound effect on a molecule’s chance
to work as a tracer. Our library approach enables important
molecular properties such as pKa, lipophilicity and reactivity to
be systematically varied.
N4-(3-Bromo-phenyl)-quinazoline-4,6-diamine (5a)26
To a slurry of 4a (0.538 g, 1.56mmol) in ethanol (20 mL) and water (10mL)
was added iron (0.190 g, 3.41 mmol) and acetic acid (1mL). The reaction
mixture was stirred under heating (95 ꢃC heating bath) for 90min. After
neutralisation with 2M NaOH, the mixture was filtered. The filtrate was
concentrated to approximately half volume and extracted with ethyl acetate,
and the combined organic phases were washed with brine. The solvent was
removed under reduced pressure to give the product as a yellow solid
1
(0.197 g, 25%). H NMR (400 MHz, DMSO-d6, 25 ꢃC) d = 9.49 (s, 1H), 8.36
Experimental
(s, 1H), 8.24 (s, 1H), 7.88 (m, 1H), 7.52 (m, 1H), 7.37 (s, 1H), 7.32–7.20 (m, 2H),
5.62 (br s, 2H, NH2) ppm. LC-MS, ESI + for C14H11BrN4 m/z [M+ H]: 315/317.
General
N4-(3-Chloro-4-fluoro-phenyl)-quinazoline-4,6-diamine (5b)27
Reagent grade THF was freshly distilled from sodium and benzophenone in
a nitrogen atmosphere. All chemicals were purchased from Aldrich/Fluka
(Stockholm, Sweden) or Fluorochem (Derbyshire, UK) and used as
received. The identities of the purified 11C-labelled compounds [11C]
12bb, [11C]12da, [11C]12db and [11C]12dc were determined with liquid
chromatography–massspectrometry(LC-MS) andanalyticalHPLCequipped
with a ultraviolet (UV) and radioactivity detector using authentic samples as
references. For the remaining compounds, LC-MS or LC-MS equipped with a
radioactivity detector was used. LC-MS analyses on 11C-labelled compounds
(decayed and/or radioactive) were performed on a Gilson (Middleton, WI,
USA) reverse-phase HPLC equipped with a Waters (Milford, MA, USA)
Quattro Premier triple quadrupole mass spectrometer with ESI, operating
in selective positive ion mode. Analytical HPLC was performed on a
Beckman (Bromma, Sweden) system, equipped with a Beckman 126 pump,
a Beckman 166 UV detector in series with a Bioscan b+-flow count detector and
a Beckman Ultrasphere ODS dp 5-m column (250 ꢂ 4.6 mm). A Gilson 231
XL was used as the auto-injector. Purification with semi-preparative
HPLC was performed on a similar Beckman system equipped with an
Alltech Alltima HP C18 5-m guard cartridge (7.5 ꢂ 4.6 mm) and a Grace
Synthesised according to the procedure for 5a. Yield 78% as a yellow solid.
1H NMR (500MHz, DMSO-d6, 25 ꢃC) d = 9.47 (br s, 1H, NH), 8.36 (s, 1H, H-2),
8.20 (dd, 4J 2.6, 4JH,F 6.9, 1H, H-20), 7.82 (ddd, 3JH,H 9.0, 4JH,H 2.6, 4JH,F 4.4, 1H,
H-60), 7.54 (d, 3JH,H 8.9, 1H, H-8), 7.40 (dd, 3JH,H 9.0, 3JH,F 9.0, 1H, H-50), 7.31
4
3
4
(d, JH,H 2.4, 1H, H-5), 7.24 (dd, JH,H 8.9, JH,H 2.4, 1H, H-7), 5.62 (br s, 2H,
NH2) ppm. LC-MS ESI+ for C14H10ClFN4 m/z [M+ H]: 289/291.
N4-[3-Chloro-4-(3-fluoro-benzyloxy)-phenyl]-quinazoline-4,6-diamine
(5c)28
Synthesised according to the procedure for 5a. Yield 70% as a yellow solid.
1H NMR (500MHz, DMSO-d6, 25 ꢃC) d = 9.33 (br s, 1H, NH), 8.31 (s, 1H, H-2),
8.03 (d, 4JH,H 2.6, 1H, H-5), 7.72 (dd, 4JH,H 2.6, 3JH,H 9.0, 1H, H-7), 7.51 (d, 3JH,H
4
3
8.8, 1H, H-50), 7.46 (m, 1H), 7.34–3.30 (m, 3H), 7.23 (dd, JH,H 2.5, JH,H 8.8,
1H, H-60), 7.22 (d, 3JH,H 9.0, 1H, H-8), 7.17 (m, 1H), 5.72 (br s, 2H, NH2), 5.23
(s, 2H, CH2) ppm. LC-MS, ESI + for C21H16ClFN4O m/z [M + H]: 395/397.
4-((E)-3-Iodo-allyl)-morpholine (11b)
Genesis C18 4-m column (250 ꢂ 10 mm). Flow 4 mL minꢀ1. Method A: To an ice-cold stirred slurry of morpholine (0.117, 1.35mmol) and K2CO3
30–70% acetonitrile over 10 min in 50-mM ammonium formate pH 3.5.
(0.186 g, 1.35 mmol) in MeCN (5mL) was added dropwise (E)-3-bromo-1-
Method B: 45–50% acetonitrile over 10min in 10-mM KH2PO4. Method iodo-propene 10 (0.302g, 1.22 mmol) as a solution in MeCN (5mL) over
C: 30–80% acetonitrile over 10 min in 50-mM ammonium formate pH 3.5. 10min. The reaction mixture was allowed to warm to room temperature
11C was prepared by the 14N(p,a)11C nuclear reaction using 17-MeV protons and stirred for 18h. The solids were filtered off, and the filtrate was concen-
produced by a Scanditronix MC-17 Cyclotron at Uppsala Imanet, GE Health- trated under reduced pressure to give an oil, which was purified using
care and obtained as [11C]carbon dioxide. The target gas used was nitrogen flash chromatography eluting with 0–50% EtOAc in pentane and a few
(AGA Nitrogen 6.0) containing 0.05% oxygen (AGA Oxygen 4.8). [11C]Carbon drops of triethylamine to give the product as an oil (0.146 g, 47%). 1H NMR
dioxide was transferred in a stainless steel tubing concentrated on silica
(400MHz, CDCl3, 25 ꢃC) d = 6.53 (dt, J 7.0, 14.4, 1H), 6.25 (dt, J 1.4, 14.4, 1H),
www.jlcr.org
Copyright © 2012 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2012, 55 477–483

O. Åberg and B. Långström
(E)-4-Morpholin-4-yl-but-2-enoic [carbonyl-11C]acid [4-(3-chloro-4-
fluoro-phenylamino)-quinazolin-6-yl]-amide (12bb)
3.68 (m, 2H), 2.93 (dd, J 1.4, 7.0, 1H), 2.41 (m, 2H) ppm. dC (100MHz, CDCl3,
25 ꢃC) 142.5, 79.0, 67.0, 63.1, 53.4 ppm.
1-((E)-3-Iodo-allyl)-piperidine (11a)
HPLC method B. Product tR =9.9–11.7 min, aniline precursor tR =11.8–14 min.
LC-MS ESI + for C22H21ClFN5O2 m/z: 221.5 (M+2H)2+
.
1
Synthesised according to the procedure for 11b. Yield 37% as an oil. H
NMR (400 MHz, CDCl3, 25 ꢃC) d = 6.58 (dt, J 7.0, 14.3, 1H), 6.19 (dt, J 1.3,
14.3, 1H), 2.91 (dd, J 1.3, 7.0, 2H), 2.35 (m, 4H), 1.57 (m, 4H), 1.42
(m, 2H) ppm. 13C NMR (100 MHz, CDCl3, 25 ꢃC) d = 143.4, 78.2, 63.6, 54.4,
26.0, 24.3 ppm.
(E)-4-Morpholin-4-yl-but-2-enoic [carbonyl-11C]acid {4-[3-chloro-4-(3-
fluoro-benzyloxy)-phenylamino]-quinazolin-6-yl}-amide (12bc)
HPLC method C. Product tR = 10.1–10.5 min, aniline precursor tR = 10.1–13.9 min.
Theproduct masspeakcouldnot be detectedbecauseofionsuppressionby
the precursor.
4-((E)-3-Iodo-allyl)-thiomorpholine (11c)
(E)-4-Thiomorpholin-4-yl-but-2-enoic [carbonyl-11C]acid [4-(3-bromo-
phenylamino)-quinazolin-6-yl]-amide (12ca)
Synthesised according to the procedure for 11b. Yield 40.4 mg, 58% as
1
an oil. H NMR (500 MHz, CDCl3, 25 ꢃC) d = 6.54 (dt, J 6.7, 14.5, 1H), 6.26
(dt, J 1.4, 14.5, 1H), 2.96 (dd, J 1.4, 6.7, 1H), 2.68 (m, 8H) ppm. 13C NMR
(125 MHz, CDCl3, 25 ꢃC) d = 142.9, 78.4, 63.6, 54.8, 28.2 ppm.
HPLC method C. Product tR = 7.2–8.5min, aniline precursor tR = 7.1–9.5min.
Theproduct masspeakcouldnot be detectedbecauseofionsuppressionby
the precursor.
General procedure for synthesis of [carbonyl-11C]acrylamides
[
11C]12aa–dc
(E)-4-Thiomorpholin-4-yl-but-2-enoic [carbonyl-11C]acid [4-(3-chloro-4-
fluoro-phenylamino)-quinazolin-6-yl]-amide (12cb)
Pd2(dba)3 (0.70 mg, 1.2mmol) and PPh3 (2.70 mg, 10 mmol) were placed in a
0.8-mLvialequippedwitharubberseptum. THF (600 mL)wasadded, andthe
resulting solution was degassed with argon. Vinyl halide 11a–c (6.0mmol)
was added, whereupon the colour of the solution changed from yellow to
pale yellow. Quinazoline precursor 5a–c (12.5 mmol) was added, and the
mixture was loaded through a filter (13-mm Teflon, 0.45-mm pore size) into
a 200-mL injection loop. THF was pumped through the loop using an HPLC
pump (30–40MPa), and the reagents were transferred into a pre-heated
200-mL, stainless steel micro-autoclave containing [11C]carbon monoxide
(10–100 nmol, 0.12–1Pa) in helium (40mmol, 500 kPa). The reagents were
kept in the micro-autoclave for 5 min at 110 ꢃC. The reaction mixture was
then transferred to a 1-mL, septum-equipped evacuated glass vial, and the
radioactivity was measured. The vial was heated at 80ꢃC for 1–2min during
the removal of volatiles under a stream of nitrogen gas, and radioactivity
was measured again. The crude product was redissolved in acetonitrile
(400 mL) and then water (400 mL) and purified with semi-preparative HPLC.
The radioactivity remaining in the vial was less than 2%. The radioactivity
of the purified product was measured, and analytical HPLC equipped with
a UV and radioactivity detector and LC-MS was used to assess its identity.
The radiochemical purity was >98% for all products. Formulation of the
purified product was performed by diluting the purified fraction (3–8 mL)
with 15-mL phosphate buffer (10mM, pH 7.4) and concentrated on a solid
phase extraction (SPE) column (Phenomenex (Torrance, CA, USA) Strata-
X, previously activated with EtOH and washed with H2O). The solid
phase extraction (SPE) column was washed with H2O (3 mL) before the
radioactive product was eluted using EtOH (0.5 mL) followed by
a solution of cyclodextrin (100 mL, 300 mg/mL) diluted in phosphate
buffer (950 mL, pH 7.4).
HPLC method B. Product tR = 13.9min, aniline precursor tR = 10.2–13.8 min.
The product mass peak could not be detected because of ion suppression
by the precursor.
(E)-4-Thiomorpholin-4-yl-but-2-enoic [carbonyl-11C]acid {4-[3-chloro-4-
(3-fluoro-benzyloxy)-phenylamino]-quinazolin-6-yl}-amide (12cc)
HPLC method B. Product tR =14.2–16.3 min, aniline precursor tR =12.1–15.0 min.
LC-MS ESI + for C29H27ClFN5O2S m/z: 564 (M+H)+.
N-[4-(3-Bromo-phenylamino)-quinazolin-6-yl]-[carbonyl-11C]acry-
lamide (12da)29
HPLC method A. Product tR =11.1–12.0 min, aniline precursor tR =7.5–10 min.
LC-MS ESI + for C17H13BrN4O m/z: 369 (M + H)+.
N-[4-(3-Chloro-4-fluoro-phenylamino)-quinazolin-6-yl]-[carbonyl-11C]
acrylamide (12db)
HPLC method A. Product tR = 10.2–11.4 min, aniline precursor tR =7.5–10.1 min.
LC-MS ESI + for C17H12ClFN4O m/z: 343 (M + H)+.
N-{4-[3-Chloro-4-(3-fluoro-benzyloxy)-phenylamino]-quinazolin-6-yl}-
[carbonyl-11C]acrylamide (12dc)
HPLC method B. Product tR =13.1–14.0 min, aniline precursor tR =10–12.3 min.
LC-MS ESI + for C24H18ClFN4O2 m/z: 449 (M+H)+.
Specific activity
The synthesis of [11C]12da was performed as described earlier, starting with
5.6 GBq of [11C]carbon monoxide. After 47min (from end of bombardment
(EOB)), the product was isolated with a radioactivity of 0.85GBq. The volume
of the product fraction was 3.1mL. The concentration was determined to be
4.50 mM using an HPLC UV (l = 254 nm) calibration curve, corresponding to
(E)-4-Piperidin-1-yl-but-2-enoic [carbonyl-11C]acid [4-(3-bromo-pheny-
lamino)-quinazolin-6-yl]-amide (12aa)
HPLC method B. Product tR =7.2–8.9 min, aniline precursor tR = 10.5–16.0 min.
LC-MS ESI + for C23H24BrN5O m/z: 466 (M+H)+.
14 nmol of product and a specific radioactivity of 60 GBqmmolꢀ1
.
(E)-4-Piperidin-1-yl-but-2-enoic [carbonyl-11C]acid [4-(3-chloro-4-
fluoro-phenylamino)-quinazolin-6-yl]-amide (12ab)
Computational docking studies
HPLC method B. Product tR =8.5–10.0 min, aniline precursor tR =10.9–13.9 min.
LC-MS ESI + for C23H23ClFN5O m/z: 440 (M+H)+.
The following X-ray structures representing the inactive conformation
of EGFR were downloaded from http://www.rcsb.org and prepared for com-
parisons; 1XKK, 2RGB, 3GT8, 2JIV, 2RFD and 2GS7. The docking calculations
for the inactive conformation were performed on 1XKK.30 The missing
amino acid side chains on the following residues K713, R748, E749, K754,
K757, E758, N808, E866, K867, V876, M945, R962, E985, R986, S995, R999,
E1004, V1010, V1011, E1015, L1017 and T1018 were modelled. None of
these residues was found within 11 Å from the ligand. The atom types of
the co-crystallised ligand GW-502016 were adjusted manually, and all water
molecules except three in the close proximity of the ligand were deleted
(E)-4-Piperidin-1-yl-but-2-enoic [carbonyl -11C]acid {4-[3-chloro-4-(3-
fluoro-benzyloxy)-phenylamino]-quinazolin-6-yl}-amide (12ac)
HPLC method B. Product tR =13.0–14.2 min, aniline precursor tR =11.8–14.5 min.
LC-MS ESI + for C30H29ClFN5O2 m/z: 273.5 (M+2H)2+
.
(E)-4-Morpholin-4-yl-but-2-enoic [carbonyl-11C]acid [4-(3-bromo-pheny-
lamino)-quinazolin-6-yl]-amide (12ba)
HPLC method B. Product tR = 10.0–11.8 min, aniline precursor tR = 11.8–16 min. as well as two phosphate ions, using the Schrödinger Maestro Interface.
LC-MS ESI + for C22H21ClFN5O2 m/z: 468 and 470 (M + H)+.
Orientation of hydroxyls in Ser, Thr, and Tyr and sulfhydryls in Cys chain in
J. Label Compd. Radiopharm 2012, 55 477–483
Copyright © 2012 John Wiley & Sons, Ltd.
www.jlcr.org

O. Åberg and B. Långström
a reactive acrylamide moiety that irreversibly binds to Cys797
attheentrance oftheadenosinetriphosphate(ATP) bindingpocket
were superior to reversible ones. This is likely due to fast washout of
reversible inhibitors because of ATP present in mM concentration
competing for the binding site.36 So far, the success of irreversible
binding EGFR TKI PET-tracers has been limited because of slow
tumour uptake, fast metabolism and high non-specific binding.
One cause of fast metabolism and unspecific binding may be
due to high reactivity of the acrylamide moiety towards other
the protein structure was made using the Protein Preparation script in
Schrodinger Glide. The same script was used for setting formal charges
and protonation of the protein.31 The X-ray structure of the inhibitor
GW-502016 bound to the activesite of the kinase domain of EGFR (PDBentry
1XKK) was used as template to construct the three-dimensional models of test
compounds as well as 12ac, 12bc, 12cc and 12dc. All the structures were
minimised in the absence of the protein using Macro Model32 and Merck
molecular force field (MMFFs)33 force field, with an implicit solvent model
of water. The Polak-Ribiere Conjugate Gradient (PRCG) method was used
witha gradientthreshold of 0.05, andminimisation was continueduntil con-
vergence (usually <3000 steps). Ionisable ligands (at physiological pH) were endogenous nucleophiles such as glutathione, a feature that
modelled both in their charged and neutral forms. The protein active site
was defined by a 10 ꢂ 10 ꢂ 10Å box around the ligand, and no constraints
were used. The docking model was tested by docking a set of 58 known
potent compounds (1–1900 nM),34 and the lowest energy poses were
inspected visually. The same method was used for the active conformer
1M17 and the remaining ligands. Docking and scoring of 12aa–dc were
performed using standard precision mode in the software Glide from
Schrodinger, and the distance between the sulfur atom of Cys797 and the
b-carbon of the acrylamide was measured. A covalent bond was manually
possibly can be altered by altering the substituents of the b-carbon
of the acrylamide (R0 in 12aa–dc). Another cause of non-specific
binding could be low selectivity over other kinases because
the ATP binding pocket is highly conserved in the over 500 kinases
known in the human genome.37 It has been postulated that
targeting the inactive conformation of kinases may be a way to
achieve higher selectivity because the inactive conformation does
need to fulfil the requirements to fit ATP, which could allow for
more specific ligand-receptor interactions.38 For the EGFR/HER-2
added between the b-carbon of the acrylamide to the sulfur atom of Cys797
.
Atoms more than 6 Å away from the ligand were frozen, whereas the atoms inhibitor lapatinib, it has been shown that by adding an aromatic
within 6 Å from the ligand were constrained with a force constant of 100.
The ligand, the side chain of Cys797 and the hydrogens of the water molecule
were allowed to move freely during the minimisation using Macro Model and
OPLS 2005 force field with an implicit solvent model of water. PCRG was used
with a gradient threshold of 0.05 kcal/mol and continued until converged.
moiety in the para position of the aniline, the inhibitor reaches a
hydrophobic pocked deep within the binding site and locks the
kinase in its inactive form, which resulted in a remarkable selectivity
profile that could be beneficial for imaging purposes.38,39
During the design of the focused library, we performed docking
studies on 12aa–12dc (Table 1). The quinazoline core and the
aromatic moieties at the 4-position were docked deep in the
hydrophobic binding pocket and were oriented in a similar way
as erlotinib (active) or lapatinib (inactive) in their X-ray structures.
The distance between the electrophilic b-carbon of the acrylamide
and the sulfur of Cys797 was measured (3.8–5.4 Å) to confirm that
irreversible binding would in theory be possible. The acrylamide
moiety of 12aa–12dc pointing towards the solvent interface
Results and discussion
During the last decade, several research groups have reported
radiolabelling and imaging evaluation of small molecule TKIs,
and their efforts have recently been reviewed.35 For EGFR TKIs,
the scaffold is the 4-(phenylamino)quinazoline or 4-(phenyla-
mino)-3-cyanoquinoline. It has been shown that PET tracers with
Table 1. Products obtained from aminocarbonylation of vinyl iodides with 6-aminoquinazoline derivatives
a
a
Vinyl iodide
Nucleophile
Product
cLogD7.4
Calculated pKa
Conversion of [11C]O (%)b
Isolated RCYc
11a
5a
5b
5c
5a
5b
5c
5a
5b
5c
5a
5b
5c
[11C]12aa
[11C]12ab
[11C]12ac
[11C]12ba
[11C]12bb
[11C]12bc
[11C]12ca
[11C]12cb
[11C]12cc
[11C]12da
[11C]12db
[11C]12dc
3.7
3.7
5.2
3.8
3.8
5.3
4.0
4.0
5.6
4.2
4.2
5.8
8.6
8.6
8.6
6.3
6.3
6.3
7.6
7.6
7.6
n.a.
n.a.
n.a.
92 ꢄ 1 (3)
86 ꢄ 4 (3)
91 ꢄ 4 (2)
89 ꢄ 3 (2)
94 ꢄ 1 (3)
92 ꢄ 0 (2)
90
36 ꢄ 2 (3)
31 ꢄ 6 (3)
24 ꢄ 9 (3)
39 ꢄ 0 (2)
35 ꢄ 1 (3)
36 ꢄ 2 (2)
31
11b
11c
11d
90
90
38
32
84 ꢄ 11(4)
89 ꢄ 7 (6)
95 ꢄ 5 (4)
47 ꢄ 15 (4)
48 ꢄ 9 (6)
61 ꢄ 13 (4)
n.a.. not applicable.
apKa values of the aliphatic amine calculated using Chemaxon Marwin online software.
bDecay-corrected conversion yield of [11C]carbon monoxide to non-volatile products remaining in the reaction mixture after
removal of solvent.
cThe number in parentheses is the number of experiments.
dDecay-corrected radiochemical yield based on the amount of radioactivity at the start of the synthesis and the radioactivity of the
isolated product. After the reaction mixture was transferred from the micro-autoclave to an evacuated 1-mL vial, the radioactivity
in the vial was measured. The radioactive residues left in the micro-autoclave were estimated to be less than 1%. Hence, the
amount of initial radioactivity could be determined.
eRadiochemical purity was >97% in all experiments.
www.jlcr.org
Copyright © 2012 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2012, 55 477–483

O. Åberg and B. Långström
O
O
N
O
N
appeared to have many low energy conformations and high
rotational freedom, which is in agreement with the less defined
electron densities reported for the flexible substituents at the
6-position and 7-position of the quinazoline cores of lapatinib and
erlotinib. Ligands 12ac, 12bc, 12cc and 12dc with the extra
phenyl–ether moiety targeting the inactive conformation were
not able to dock into the active conformation because of steric
clash. We also confirmed that b-substitution of the acrylamide
(piperidine, morpholine or thiomorpholine) did not affect the
binding mode of the quinazoline core, which promoted us to go
ahead with the synthesis of precursors and reference compounds.
The final binding modes exemplified with 12bb (active) and 12ac
(inactive) (Figure 1) were obtained by manual formation of the
covalent carbon–sulfur bond followed by a constrained minimisation.
The non-covalent ligand–protein interactions include a hydrogen
bond between N1 and the backbone N–H of Met793 in addition
to the water-mediated bond via the N3 of the inhibitor and the
hydroxyl of Thr790 (active) or hydroxyl Thr854 (inactive). In
the deep hydrophobic pocket that becomes accessible in the
inactive conformation, we observed a favourable p-interaction
between the 4-fluoro-benzyloxy substituent in an edge-to-face
manner to Phe856. The tertiary amine substituent on the acrylamide
interacts with Asp800 via hydrogen bonding. The docking results
are in good agreement with recently published X-ray structures
of structurally related irreversibly bound EGFR inhibitors.40
O2N
a
b
OH
NH2
NH
NH
1 (89%)
2 (76%)
c
R'''
HN
R''
Cl
O2N
O2N
d
N
N
N
N
4a-c (74-86%)
3 (94%)
e
R'''
R''
HN
N
a R'' = Br R''' = H
b R'' = Cl R''' = F
H2N
c R'' = Cl R''' = OCH2(3-F-Ph)
N
5a-c (25-74%)
Scheme 1. (a) Formamide, 130 ꢃC, 7 h. (b) HNO3/H2SO4, 95 ꢃC, 40 min. (c) SOCl2,
reflux, 3 h. (d) Ar–NH2, i-PrOH, reflux, 2 h. (e) Fe, HOAc, EtOH, H2O, reflux, 90 min.
Synthesis of precursors
b
a
HO
HO
HO
Br
Sn(nBu)3
I
The syntheses of the precursors were performed by previously
described procedures with minor modifications. Briefly, the
4-anilino-6-amino-quinazolines 5a–c were synthesised starting
from anthranilic acid and formamide to give the quinazolinone 1
(Scheme 1), which was nitrated in the 6-position to form 2.41
Deoxychlorination was accomplished using thionyl chloride, and
different anilines were reacted with 3 to give the 4-anilino-6-nitro-
quinazolines 4a–c. Finally, the nitro group was reduced using iron
and acetic acid to yield the nucleophile precursors 4-anilino-6-
amino-quinazolines 5a–c (Scheme 1). The b-position substituent
of the acrylic amide is pointing towards the solvent interface of
the binding site (Figure 1). These substituents may influence the
reactivity of the acrylamide as well as the overall solubility and
lipophilicity of the molecule (Scheme 3).17 Vinyliodide precursors
11a–c with different substituents were synthesised according to
8 (44%)
d
c
I
9 (86%)
10 (75%)
N
I
N
I
N
I
S
O
11a (37%)
11b (47%)
11c (58%)
Scheme 2. (a) Bu3SnH, AIBN, 80ꢃC, 2 h. (b) I2, dichloromethane, room temperature,
2 h. (c) Ph3P, CBr4, dichloromethane, 0 ꢃC, 10 min. (d) K2CO3, MeCN, a secondary amine,
0 ꢃC to room temperature, 18h.
Figure 1. (a) Plausible binding modes obtained using computational docking for 12bb bound to the adenosine triphosphate site of the active epidermal growth
factor receptor (EGFR) conformation followed by manually adding the covalent bond to Cys797 and minimisation of (b) 12ac docked and covalently bound to the inactive
conformation of EGFR.
J. Label Compd. Radiopharm 2012, 55 477–483
Copyright © 2012 John Wiley & Sons, Ltd.
www.jlcr.org

O. Åberg and B. Långström
R'''
R''
R'''
R''
HN
N
HN
N
a
H
R'
N
R'
H2N
*
N
I
+
N
O
[Carbonyl-11C]12aa-dc
11a R' = CH2-piperidine
5a R'' = Br R''' = H
5b R'' = Cl R''' = F
5c R'' = Cl R''' = OCH2(3-F-Ph)
11b R' = CH2-morpholino
11c R' = CH2-thiomorpholino
11d R' = H
* = [11C]
Scheme 3.
[
11C]O in a helium carrier, Pd2(dba)3, PPh3, THF, 110 ꢃC, 5 min.
Scheme 2. Addition of an excess of tributyltin hydride to propargyl reported IC50 value of 2 nM in an enzyme assay,24 was obtained
alcohol gave mainly the syn addition product 8.21 Iodination and in higher decay-corrected radiochemical yield and specific
subsequent bromination gave the intermediate 10, which was activity (47%, 60 GBq mmolꢀ1) than was reported previously
combined with secondary amines to obtain vinyl iodides 11a–c (12%, 10 GBq mmolꢀ1).29 In general, there was only one radio
used as substrates in the labelling reactions.23
peak with a retention time of more than 5 min. The hydrophilic
by-products formed were not analysed, but we believe these
could be the corresponding hydroxy carbonylation product from
a reaction where water acts as the nucleophile, as well as
Labelling chemistry
Compound libraries and combinatorial chemistry have since the labelled carbonates as seen in previous studies.44
early nineties been extensively used for structure activity relation-
ship (SAR) analyses and lead optimisation in drug discovery.42 On
Conclusion
a laboratory scale, carbon monoxide has been used to synthesise
The abundance of the carbonyl group in many drug candidate
libraries and other biologically active molecules and the generality
libraries using microwave techniques and in situ produced
carbon monoxide.43 In the current work, we use on-line pro-
duced [11C]carbon monoxide to radiolabel a focused library of
and functional group tolerance of the carbonylation labelling
chemistry make [11C]carbon monoxide versatile in labelling
EGFR inhibitors in a combinatorial fashion.
synthesis. A combinatorial approach to synthesise a library of
An active palladium complex was formed by mixing tris(dibenzy-
11C-labelled potential irreversible EGFR inhibitors based on
the 4-anilino-6-acrylamido-quinazoline core has successfully
been explored. The labelling was performed using various vinyl
lideneacetone)dipalladium(0) [Pd2(dba3)] and triphenylphosphane
in THF under argon to form a yellow solution. To the in situ formed
complex was added vinyl iodide 11a–d, whereupon the colour
iodides as substrates in palladium-mediated carbonylations with
of the solution changed to pale yellow. The amine nucleophile
[
11C]carbon monoxide and 4-anilino-6-amino-quinazolines as
5a–c was then introduced, and the mixture was injected into a
micro-autoclave pre-charged with no-carrier-added [11C]carbon
monoxide in helium obtained on-line via reduction of [11C]carbon
dioxide over zinc granules at 400 ꢃC. To ensure a high radiochemical
yield, the reaction time must be kept short because of the short
half-life of 11C, and thus, the rate of the reaction must be high.
To meet these criteria, all reagents were used in large excess
compared with [11C]carbon monoxide, and around 90% of the
[11C]carbon monoxide could be converted to non-volatile products
within 5 min. The stoichiometric relationship between the reagents
Pd/PPh3/vinyl iodide/amine nucleophile in the reaction mixture
was 1:4.2:2.5:5.2, whereas the amount of [11C]carbon monoxide
was much lower (0.004–0.04 equivalents). Unsubstituted acryla-
mides have been 11C-labelled previously via a Grignard reaction
followed by chlorination and finally by reaction with an amine
nucleophile.29 However, it is a clear advantage to use a single-step
reaction, especially in view of the short half-life of 11C as well as the
physical handling of the radioactivity. Another advantage using
[11C]carbon monoxide is the low abundance of carbon monoxide
in the atmosphere (0.05–0.2 ppm) compared with carbon dioxide
(380 ppm), which will decrease the risk of isotopic dilution while
handlingthe reagents. Isotopicdilution leads to lowspecific activity
that can lead to undesired pharmacological effects and saturation
of binding sites in biological systems.
the nucleophiles using the same conditions. In total, 12 analogues
were 11C-labelled at the carbonyl position and were obtained in
biologically useful radiochemical yields and low isotopic dilution.
Biological evaluation of several of these tracers will follow.
Acknowledgements
Uppsala Imanet and GE Healthcare are acknowledged for the
financial support of this work. Dr Niklas Tyrefors is acknowledged
for support with LC-MS analyses. We are grateful to Dr Torben
Rasmussen for computational advice, Prof Anders Karlén for
computational resources and Dr Tamara Church and Dr Graham
Smith for linguistic advice and commenting on this text.
Conflict of Interest
The authors did not report any conflict of interest.
References
[1] M. Bergström, A. Grahnen, B. Långström, Eur. J. Clin. Pharmacol.
2003, 59, 357.
[2] a) H. M. Colquhoun, D. J. Thompson, M. V. Twigg, Carbonylation
Direct Synthesis of Carbonyl Compounds, Plenum Press, N.Y., 1991;
b) L. Kollár, in Modern Carbonylation Methods (Ed.: L. Kollár), Wiley
VCH Verlag GmbH & Co. KGaA: Weinheim, 2008.
[3] P. Lidstrom, T. Kihlberg, B. Långström, J. Chem. Soc. Perkin. Trans. 1
1997, 2701.
The crude products were purified using semi-preparative
HPLC to give the labelled products [11C]12aa–dc (Scheme 3,
Table 1). The yields were in the range 24–61% (Table 1). As a
comparison, [11C]12da, a known irreversible inhibitor with a
[4] S. Kealey, P. W. Miller, N. J. Long, C. Plisson, L. Martarello, A. D. Gee,
Chem. Commun. 2009, 3696.
www.jlcr.org
Copyright © 2012 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2012, 55 477–483

O. Åberg and B. Långström
[5] a) Amersham PLC WO/2002/102711, 2002; b) E. D. Hostetler, H. D.
Burns, Nucl. Med. Biol. 2002, 29, 845; c) T. Kihlberg, B. Långstrom, J.
Org. Chem. 1999, 64, 9201.
[27] B. Marvania, P.-C. Lee, R. Chaniyara, H. Dong, S. Suman, R. Kakadiya,
T.-C. Chou, T.-C. Lee, A. Shah, T.-L. Su, Bioorg. Med. Chem. 2011, 19, 1987.
[28] Shanghai Allist Pharmaceutical EP1990337, 2008.
[6] P. W. Miller, N. J. Long, A. J. de Mello, R. Vilar, H. Audrain, D. Bender, [29] I. Ben-David, Y. Rozen, G. Ortu, E. Mishani, Appl. Radiat. Isot. 2003, 58, 209.
J. Passchier, A. Gee, Angew. Chem. Int. Ed. 2007, 46, 2875.
[7] J. Eriksson, J. van den Hoek, A. D. Windhorst, J. Label. Compd.
Radiopharm. 2012, 55, 223.
[8] B. Långström, O. Itsenko, O. Rahman, J. Label. Compd. Radiopharm.
2007, 50, 794.
[30] H. M. Berman, J. Westbrook, Z. Feng, G. Gilliland, T. N. Bhat, H. Weissig,
I. N. Shindyalov, P. E. Bourne, Nucleic Acids Res. 2000, 28, 235.
[31] R. A. Friesner, J. L. Banks, R. B. Murphy, T. A. Halgren, J. J. Klicic,
D. T. Mainz, M. P. Repasky, E. H. Knoll, M. Shelley, J. K. Perry, D. E. Shaw,
P. Francis, P. S. Shenkin, J. Med. Chem. 2004, 47, 1739.
[9] O. Itsenko, E. Blom, B. Långström, T. Kihlberg, Eur. J. Org. Chem. 2007, 4337. [32] F. Mohamadi, N. G. J. Richards, W. C. Guida, R. Liskamp, M. Lipton,
[10] O. Ilovich, O. Åberg, B. Långström, E. Mishani, J. Label. Compd.
Radiopharm. 2009, 52, 151.
[11] M. K. J. Gagnon, S. H. Hausner, J. Marik, C. K. Abbey, J. F. Marshall,
J. L. Sutcliffe, Proc. Natl. Acad. Sci. 2009, 106, 17904.
C. Caufield, G. Chang, T. Hendrickson, W. C. Still, J. Comput. Chem.
1990, 11, 440.
[33] T. A. Halgren, J. Comput. Chem. 1996, 17, 490.
[34] a) GlaxoSmithKline WO 98/02434, 1998; b) GlaxoSmithKline WO 98/
02438, 1998; c) GlaxoSmithKline WO 98/02437, 1998.
[12] M. Takashima-Hirano, M. Shukuri, T. Takashima, M. Goto, Y. Wada,
Y. Watanabe, H. Onoe, H. Doi, M. Suzuki, Chem. Eur. J. 2010, [35] a) E. Mishani, G. Abourbeh, M. Eiblmaier, C. J. Anderson, Curr. Pharm.
16, 4250.
Des. 2008, 14, 2983; b) J. G. Gelovani, Cancer Metastasis Rev. 2008,
27, 645; c) J. W. Hicks, H. F. VanBrocklin, A. A. Wilson, S. Houle, N.
Vasdev, Molecules 2010, 15, 8260; d) F. Pisaneschi, Q. D. Nguyen, E.
Shamsaei, M. Glaser, E. Robins, M. Kaliszczak, G. Smith, A. C. Spivey,
E. O. Aboagye, Bioorg. Med. Chem. 2010, 18, 6634; e) P. Slobbe, A.
J. Poot, A. D. Windhorst, G. A. M. S. van Dongen, Drug Discov. Today
2012. DOI: 10.1016/j.drudis.2012.06.016
[13] O. Rahman, B. Långström, J. Label. Compd. Radiopharm. 2007, 50, 1192.
[14] A. A. Memon, S. Jakobsen, F. Dagnaes-Hansen, B. S. Sorensen,
S. Keiding, E. Nexo, Cancer Res. 2009, 69, 873.
[15] H. Su, Y. Seimbille, G. Z. Ferl, C. Bodenstein, B. Fueger, K. J. Kim, Y. T. Hsu,
S. M. Dubinett, M. E. Phelps, J. Czernin, W. A. Weber,Eur. J. Nucl. Med. Mol.
Imaging 2008, 35, 1089.
[16] M. Tordeux, C. Wakselman, J. Fluorine Chem. 1995, 74, 251.
[36] A. Levitzki, Acc. Chem. Res. 2003, 36, 462.
[17] J. B. Smaill, G. W. Rewcastle, J. A. Loo, K. D. Greis, O. H. Chan, E. L. Reyner, [37] G. Manning, D. B. Whyte, R. Martinez, T. Hunter, S. Sudarsanam,
E. Lipka, H. D. H. Showalter, P. W. Vincent, W. L. Elliott, W. A. Denny,
J. Med. Chem. 2000, 43, 1380.
Science 2002, 298, 1912.
[38] Y. Liu, N. S. Gray, Nat. Chem. Biol. 2006, 2, 358.
[18] Z. Rachid, F. Brahimi, A. Katsoulas, N. Teoh, B. J. Jean-Claude, J. Med. [39] a) E. R. Wood, A. T. Truesdale, O. B. McDonald, D. Yuan, A. Hassell,
Chem. 2003, 46, 4313.
S. H. Dickerson, B. Ellis, C. Pennisi, E. Horne, K. Lackey, K. J. Alligood,
D. W. Rusnak, T. M. Gilmer, L. Shewchuk, Cancer Res. 2004, 64,
6652; b) M. A. Fabian, W. H. Biggs, D. K. Treiber, C. E. Atteridge,
M. D. Azimioara, M. G. Benedetti, T. A. Carter, P. Ciceri, P. T. Edeen,
M. Floyd, J. M. Ford, M. Galvin, J. L. Gerlach, R. M. Grotzfeld, S.
Herrgard, D. E. Insko, M. A. Insko, A. G. Lai, J. M. Lelias, S. A. Mehta,
Z. V. Milanov, A. M. Velasco, L. M. Wodicka, H. K. Patel, P. P. Zarrinkar,
D. J. Lockhart, Nat. Biotechnol. 2005, 23, 329.
[19] J. B. Smaill, B. D. Palmer, G. W. Rewcastle, W. A. Denny, D. J. McNamara,
E. M. Dobrusin, A. J. Bridges, H. R. Zhou, H. D. H. Showalter, R. T. Winters,
W. R. Leopold, D. W. Fry, J. M. Nelson, V. Slintak, W. L. Elliot, B. J. Roberts,
P. W. Vincent, S. J. Patmore, J. Med. Chem. 1999, 42, 1803.
[20] K. G. Petrov, Y. M. Zhang, M. Carter, G. S. Cockerill, S. Dickerson, C. A.
Gauthier, Y. Guo, R. A. Mook, D. W. Rusnak, A. L. Walker, E. R. Wood, K.
E. Lackey, Bioorg. Med. Chem. Lett. 2006, 16, 4686.
[21] M. E. Jung, L. A. Light, Tetrahedron Lett. 1982, 23, 3851.
[22] R. Rossi, F. Bellina, A. Catanese, L. Mannina, D. Valensin, Tetrahedron
2000, 56, 479.
[40] C. H. Yun, K. E. Mengwasser, A. V. Toms, M. S. Woo, H. Greulich,
K. K. Wong, M. Meyerson, M. J. Eck, Proc. Natl. Acad. Sci. U. S. A.
2008, 105, 2070.
[23] D. Brasseur, I. Marek, J. F. Normant, Tetrahedron 1996, 52, 7235.
[41] M. T. Bogert, G. A. Geiger, J. Am. Chem. Soc. 1912, 34, 524.
[24] H. R. Tsou, N. Mamuya, B. D. Johnson, M. F. Reich, B. C. Gruber, F. Ye, [42] E. M. Gordon, R. W. Barrett, W. J. Dower, S. P. A. Fodor, M. A. Gallop,
R. Nilakantan, R. Shen, C. Discafani, R. DeBlanc, R. Davis, F. E. Koehn, J. Med. Chem. 1994, 37, 1385.
L. M. Greenberger, Y. F. Wang, A. Wissner, J. Med. Chem. 2001, 44, 2719. [43] P. Nilsson, K. Olofsson, M. Larhed, in Microwave Methods in Organic
[25] Hanmi Parm. Co WO2006/71017 A1, 2006. Synthesis, Vol. 266, Springer, 2006, p. 103.
[26] G. W. Rewcastle, W. A. Denny, A. J. Bridges, H. R. Zhou, D. R. Cody, [44] O. Åberg, Ö. Lindhe, H. Hall, P. Hellman, T. Kihlberg, B. Långström,
A. McMichael, D. W. Fry, J. Med. Chem. 1995, 38, 3482. J. Label. Compd. Radiopharm. 2009, 52, 295.
J. Label Compd. Radiopharm 2012, 55 477–483
Copyright © 2012 John Wiley & Sons, Ltd.
www.jlcr.org