Synthesis of azide-fluoro-dehydrocoelenterazine analog as a photoaffinity-labeling probe and photolysis of azide-fluoro-coelenterazine

Article abstract of DOI:10.1016/j.tet.2005.04.025

A photosensitive azide-fluoro-dehydrocoelenterazine analog (Az-F-DCT) was synthesized, starting from 4-fluorophenylacetic acid, as a photoaffinity- labeling probe in order to analyze symplectin active site. To examine the photo-reactivity of Az-F-DCT, azi

Full text of DOI:10.1016/j.tet.2005.04.025

Tetrahedron 61 (2005) 5754–5762
Synthesis of azide-fluoro-dehydrocoelenterazine analog as a
photoaffinity-labeling probe and photolysis of
azide-fluoro-coelenterazine
Masaki Kuse,a Issei Doi,b Nobuhiro Kondo,b Yukari Kageyamab and Minoru Isobeb,
*
aChemical Instrument Division, RCMS, Nagoya University, Chikusa-ku, Furo-cho, 4648602 Nagoya, Japan
bLaboratory of Organic Chemistry, Graduate School of Bioagricultural Sciences, Nagoya University, Chikusa, 464-8601 Nagoya, Japan
Received 8 April 2005; revised 14 April 2005; accepted 15 April 2005
Available online 11 May 2005
Abstract—A photosensitive azide-fluoro-dehydrocoelenterazine analog (Az-F-DCT) was synthesized, starting from 4-fluorophenylacetic
acid, as a photoaffinity-labeling probe in order to analyze symplectin active site. To examine the photo-reactivity of Az-F-DCT, azide-fluoro-
coelenterazine analog (Az-F-CT) was used as a potent symplectin chromophore model. Photolysis of Az-F-CT in 2,2,2-trifluoroethanol
afforded nitrene intermediate to give an insertion product. The structure of this product was confirmed through spectroscopic analyses
particularly by using a proton/deuterium (H/D) exchange experiments with ESI-Q-TOF-MS and -MS/MS measurement.
q 2005 Elsevier Ltd. All rights reserved.
1. Introduction
bioluminescent process, especially on the dynamism of
symplectin active site surrounding the chromophore.
Photoaffinity labeling is the best method for such a research
purpose, and has been established with many successful
reports.7,8 Though, there are many photosensitive functional
groups such as diazirine and benzophenone,9 we selected
azide group as a photosensitive functional group, since, the
introduction of azide into aromatic ring is not so difficult
and the size of azide is relatively small to minimize a steric
repulsion in the active site. Therefore, we decided to
introduce an azide group into DCT.
Symplectin, the photoprotein of an oceanic luminous squid,
TOBIIKA (Symplectoteuthis oualaniensis, L.),1 has a
covalently bound chromophore, and its structure has been
known to be a conjugate addition product of dehydro-
coelenterazine (DCT; 1) to the protein through an active site
cysteine.2–6 In the presence of mono-valent ions (KC, NaC)
and molecular oxygen, the chromophore is oxidized at
pH 7.8 (optimum) to convert into an oxidized product while
emitting a blue light (470 nm) as shown in Scheme 1.
We now focus on the molecular mechanism of symplectin
Azide group always decomposes by light irradiation to give
Scheme 1. Postulated bioluminescent mechanism of symplectin photoprotein.
Keywords: Coelenterazine; Deuteriation MS; Exchangeable protons; Azide-fluoro-phenyl; Photoaffinity probe.
*
Corresponding author. Tel.: C81 52 789 4109; fax: C81 52 789 4111; e-mail: isobem@agr.nagoya-u.ac.jp
0040–4020/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2005.04.025

M. Kuse et al. / Tetrahedron 61 (2005) 5754–5762
5755
for the preparation of probe 2. For such a reason, Kita’s
method for the introduction of azide into an aromatic ring in
one-pot is very attractive for us.13 However, we could not
prepare Az-F-DCT (2) from 20-fluoro-DCT analog (3)
selectively by using Kita’s method (Scheme 2).
We then changed the strategy for the introduction of azide
into 20-fluoro-DCT analog (3); thus, we planned to prepare
the probe 2 through using the conventional synthetic method
established by Kishi, Inoue, and Goto et al.14–17 As shown in
Scheme 3, Az-F-DCT (2) would be derived from the
condensation of coelenteramine (4)15 and 3-(3-azide-4-
fluorophenyl)-2-oxopropanal (5). The ketoaldehyde (5)
would be prepared starting from 4-fluorophenylacetic acid
(6).
Figure 1. Structure of azide-fluoro-dehydrocoelenterazine analog (2).
Esterification of 4-fluorophenylacetic acid (6) in EtOH
afforded the ethyl ester (7) in 98% yield, followed by
nitration of 7 provided ethyl-4-fluoro-3-nitrophenylacetate
(8) in 79% yield as shown in Scheme 4 first row.
Hydrogenolysis of the nitro group of 8 was facilitated by
Pd/C catalyst to give ethyl-(3-amino-4-fluorophenyl)acetate
(9) and the subsequent azidation18 afforded ethyl-(3-azide-
4-fluorophenyl)acetate (10). Saponification of the ester (10)
afforded (3-azide-4-fluorophenyl)acetic acid (11) in 93%
yield, which was further, converted to acid chloride. It was
further, treated with diazomethane to give homologation
productin (12) 85% (2 steps) (Scheme 4, second row).
Bromination of this diazoketone (12) with HBr in acetic
acid provided bromoketone (13) in 93% yield. After
converted to its pyridinium salt, it was condensed with
N,N-dimethyl-nitrosoaniline to give N-(N,N-dimethyl-
aminophenyl)-3-(3-azide-4-fluorophenyl)-2-oxopropani-
mine oxide (14: nitrone) in 95% (2 steps) (Scheme 4, third
row). Hydrolysis of the nitrone (14) in aqueous H2SO4
afforded 3-(3-azide-4-fluorophenyl)-2-oxopropanal (5) in
74% yield. Condensation of the ketoaldehyde (5) with
coelenteramine (4) in a mixture of H2O and dioxane in the
presence of aqueous HCl provided azide-fluoro-coelenter-
azine analog (15: Az-F-CT) in 60% yield. Finally, oxidation
of Az-F-CT (15) with manganese(IV) oxide in a mixture of
ether and 2-PrOH afforded azide-fluoro-dehydrocoelenter-
azine analog (2: Az-F-DCT) in 64% yield (Scheme 4). Az-
F-DCT (2) is a powerful probe to analyze symplectin active
site by using photoaffinity-labeling, furthermore, Az-F-CT
(15) will be also useful for the photoaffinity-labeling of
aequorin photoprotein, which is the famous substance for
jellyfish (Aequoria aequoria) bioluminescence.19
Scheme 2. Trial for one-pot introduction of azide into 20-fluoro-DCT
analog (3) by using Kita’s method.
a nitrene intermediate, which after follows a ring expansion
to afford an electrophilic azepine intermediate. Actually,
Niwa and Ohashi et al. reported the isolation of azepine
derivative by photolysis of an azide-coelenteramide analog
in MeOH containing Et2NH.10 The ring expansion is also
useful if some nucleophilic residues exist near the azepine
intermediate. However, in order to label the active site of
symplectin protein efficiently even if there were no
nucleophilic residues, we considered it better to avoid the
ring expansion for the successful photoaffinity labeling.
Instead of using azepine intermediate, we planned to use
nitrene intermediate itself for labeling the symplectin active
site. Platz et al. reported that fluorine atom adjacent to
nitrene stabilized its singlet state.11 Furthermore, we have
demonstrated that the fluorinated DCT analogs are very
active substances for symplectin bioluminescence.12 There-
fore, we planned to synthesize the azide-fluoro-dehydroco-
elenterazine analog (Az-F-DCT; 2) as a photoaffinity-
labeling probe to analyze symplectin active site (Fig. 1).
In this report, we here, describe the 14-step synthesis of 2 by
starting from 4-fluorophenylacetic acid and photolysis of
azide-fluoro-coelenterazine analog (Az-F-CT; 15), which is
the final oxidation precursor for 2 having the same
oxidation state as the symplectin chromophore, in 2,2,2-
trifluoroethanol.
2. Results and discussion
2.2. Photolysis of Az-F-coelenterazine analog
2.1. Synthesis of azide-fluoro-dehydrocoelenterazine
analog (Az-F-DCT)
Although, we synthesized Az-F-DCT (2) as described in the
previous section, the structure of symplectin chromophore is
a reduced form of 2. Thus, Az-F-CT (15) has the same
One-pot introduction of azide into DCT is an ideal strategy
Scheme 3. Retro-synthesis of 2 from coelenteramine (4) and 3-(3-azide-4-fluorophenyl)-2-oxopropanal (5), and of 5 from 4-fluorophenylacetic acid (6).

5756
M. Kuse et al. / Tetrahedron 61 (2005) 5754–5762
Scheme 4. Synthetic route for Az-F-DCT (2) starting from 4-fluorophenylacetic acid (6).
oxidation state with symplectin chromophore. To test the
photo-reactivity of Az-F-DCT (2) in symplectin active site,
we examined the photo-reactivity of Az-F-CT (15) as a
symplectin chromophore model.
with anticipating the efficient production of a nitrene
intermediate. The photolysis with a high-pressure mercury
lamp was monitored with a nano-LC-Q-TOF-MS with our
pre-packed-gradient (PPG) program.22 The nano-LC
chromatogram was shown in Figure 2 monitoring at
254 nm absorption. Az-F-CT (15) was almost consumed
only remained 4% at 25.0 min after 15 min irradiation, and
three new peaks were observed at 20.2, 20.5, 24.6 min in the
chromatogram (Fig. 2). Then, these peaks were analyzed
with ESI-Q-TOF-MS and -MS/MS measurements. The
molecular weight of peak A (20.2 min, 33% yield) was m/z
441 (MCH)C, which was assigned as reduced amine (19)23
as shown in Figure 3. The peak at 20.5 min (24% yield) was
proved to be coelenteramine (4) with molecular ion at m/z
278 (MCH)C. The peak B at 24.6 min was a solvent
insertion product with molecular ion at m/z 539 (MCH)C.
There were three possible structures for B as shown in
Figure 3; O–H inserted product (20), C–H inserted product
(21), or ring-expanded azepine derivative (22).
During the analysis of chemiluminescent mechanism of
coelenterazine analog (16), we have reported that photo-
oxygenation of the analog (16) in 2,2,2-trifluoroethanol
(TFE)/methanol (7:3) afforded 2-peroxide (17) and
dioxetanone (18) intermediates effectively as shown in
Scheme 5.20,21
Since, luminous intermediates (17, 18) were accumulated
only when used (TFE) as solvent along with affording high
light yield of luminescence,20,21 we thought trifluoromethyl
group might have some stabilization effect on the excited
state of the coelenterazine analog (16) to give luminous
intermediates in high yield. Therefore, we selected TFE as
the solvent for light irradiation of 15 to decompose azide
Scheme 5. Photooxygenation of coelenterazine analog (16) afforded 2-peroxide (17) and dioxetanone (18) intermediates.20,21

M. Kuse et al. / Tetrahedron 61 (2005) 5754–5762
5757
Figure 2. PPG-nano-LC chromatogram of photo-irradiated products of 15 in TFE (detection at UV 254 nm absorption). Values in parentheses show yields
estimated by integration of peak areas.
Figure 3. Assignment of the peaks with ESI-Q-TOF-MS analysis, and three possible structures for the solvent insertion product (B) (20, 21, or 22).
For the characterization of structure of peak B, proton/
deuterium (H/D) exchange experiments were carried out
with ESI-Q-TOF-MS and -MS/MS measurement,24,25 on
the basis of different numbers in the estimated exchangeable
protons of 20 and 21. Comparing the respective MS
spectrum with both proton and deuterium measurement,
the structure of peak B would be assigned from the fact of
four exchangeable protons (20) or five (21).
The samples for proton/deuterium (H/D) exchange experi-
ments were prepared as such that photo-irradiated sample of
15 was completely dried in vacuo, then redissolved with
Figure 4. MS/MS spectra of solvent insertion product B: (I) MS/MS spectrum of deuteriated B precursor m/z 543; (II) MS/MS spectrum of natural abundant B
precursor m/z 539.

5758
M. Kuse et al. / Tetrahedron 61 (2005) 5754–5762
Figure 5. Assignment of precursor and fragment ions shown in MS/MS spectra (Fig. 4). *Dn: numbers of the exchangeable protons.
99% CH3OD:1% CH3COOD for ESI-Q-TOF-MS with
deuterium measurement or with 99% CH3OH:1% CH3-
COOH for MS with proton measurement. The resultant
spectra are shown in Figure 4 (I: deuterium measurement, II:
proton measurement). Pseudo molecular ion of peak B was
observed at 539 (MCH)C, which was employed as the
precursor ion for MS/MS measurement (sample cone 60 V,
collision 15–40 V), and the result is shown in Figure 4II.
The product ions from m/z 539 precursor are seen at m/z
439, 411, and 261 in protonic solvent (Fig. 4II), while those
from m/z 543 are also varied to appear at m/z 443, 415, and
262 in deuterium solvent as shown in Figure 4I.
symplectin by using Az-F-DCT (2), and analysis of the
photo-labeled site with nano-LC-Q-TOF-MS and -MS/MS
equipped with PPG program22 is now underway in our
group.
3. Conclusion
Synthesis of photosensitive azide-fluoro-dehydrocoelenter-
azine analog (Az-F-DCT; 2) has been accomplished starting
from 4-fluorophenylacetic acid (6). Az-F-DCT (2) must be a
powerful photoaffinity-labeling probe for the analysis of
symplectin active site. For the examination of the photo-
reactivity of Az-F-DCT (2), the azide-fluoro-coelenterazine
analog (Az-F-CT; 15) was used as a symplectin chromo-
phore model, due to the fact that chromophore structure in
symplectin bound form of Az-F-DCT (2) should have the
same chromophoric structure as Az-F-CT (15). Photo-
irradiation into a solution of Az-F-CT (15) in 2,2,2-
trifluoroethanol (TFE) afforded a nitrene intermediate to
give solvent O–H insertion product (20); this product (20)
has been characterized from proton/deuterium (H/D)
exchange data in ESI-Q-TOF-MS and -MS/MS spectra.
Figure 5 summarizes the assignment of these fragment
structures. For the validation of the above product ions
generated during the MS/MS collision process, deuteriation
of all the exchangeable protons, as shown in Figure 4I, made
mass increase of mass numbers m/z 539 to m/z 543. The
increased mass units (4, 4, 4, and 1) to the product ions also
provide strong support for their structures in detail as shown
in Figure 5. It has been concluded that the structure of
solvent insertion product (B) is now concluded to be O–H
inserted product (20) from the facts that four exchangeable
protons of B was not for 21 (five exchangeable protons) but
20.
4. Experimental
4.1. General
Concerning the ring-expanded azepine derivative (22), from
the results of MS and MS/MS measured in proton/
deuterium, we excluded the possibility that the solvent
adduct was azepine derivative (22).26
All melting points were measured on a Yanaco MP-S3 and
uncorrected. UV spectra were obtained on a JASCO U-best
50 spectrometer. IR spectra were recorded on a PERKIN
ELMER Paragon 1000 FT-IR spectrophotometer. Proton
NMR spectra were recorded on a JEOL GSX 270 for
270 MHz, or on a JEOL JNML-500 for 500 MHz, or on a
Bruker AMX-600 for 600 MHz. Chemical shifts (d) are
given in parts per million relative to tetramethylsilane (d
0.00), CD3OD (d 3.30), or DMSO-d6 (d 2.49) as internal
standard. Coupling constants (J) are given in Hz. Proton
decoupled carbon NMR spectra were recorded on a JEOL
GSX 270 for 67.8 Hz, or on a JEOL JNML-500 for
125.7 Hz, or on a Bruker AMX-600 for 150.9 MHz.
Chemical shifts are (d) given in parts per million relative
to CDCl3 (d 77.0), CD3OD (d 49.0), or DMSO-d6 (d 45.0) as
Through these experiments, we demonstrated that
photolysis of compound 15 in TFE produced a nitrene
intermediate to give the solvent O–H inserted adduct (20).
We designed Az-F-DCT (2) to suppress ring-expansion of
nitrene to give azepine intermediate up on photolysis. By
using Az-F-CT (15) as symplectin chromophore model, no
ring expansion was observed during photolysis as we
expected.23 These results enable us to utilize singlet nitrene
itselffor the photo-affinity labeling of symplectin active site.
We also demonstrated that proton/deuterium (H/D)
exchange experiments are very powerful method when
coupled with LC-Q-TOF-MS and -MS/MS to investigate
organic reaction. Photo-affinity labeling of semi-synthetic

M. Kuse et al. / Tetrahedron 61 (2005) 5754–5762
5759
internal standard. Coupling constants (J) are given in Hz.
(d, 3JC–FZ8 Hz), 154.7 (d, 1JC–FZ263 Hz), 170.1 ppm. 19
F
NMR (376 MHz, CDCl3) d K49.8 ppm. EI-MS m/z 227
(MC). HRMS (EI) calcd for C10H10O4NF: 227.0594, found
227.0619 (MC).
Fluorine NMR spectra were recorded on a JEOL A-400 for
376 MHz. Chemical shifts are (d) given in parts per million
relative to 1,1,1-trifluorotoluene (d 0.00) as external
standard. Low-resolution EI MS and FAB MS were
measured with a JEOL JMS-700. High-resolution (HR)
MS were measured with a JEOL JMS-700. ESI mass spectra
were measured with a Q-TOF mass spectrometer (Micro-
mass, Manchester, UK) equipped with a Z-spray type ESI
sources. Mr. S. Kitamura in Analytical Laboratory of this
school performed the combustion elemental analyses.
Fluorescence spectra were measured with a JASCO
FP-777 spectrometer. All the solvents were of reagent
grade. Analytical thin-layer chromatography (TLC) was
conducted on precoated TLC plates: silica gel 60 F-254
[E.Merck (Art 5715) Darmstadt, Germany], layer thickness
0.25 mm. Silica gel column chromatography utilized Silica
Gel 60 (spherical) 40–50 mm [KANTO CHEMICAL CO.,
INC].
4.1.3. Ethyl-(3-amino-4-fluorophenyl)acetate (9). To a
solution of 8 (3.5 g, 15.4 mmol) in 150 ml of EtOH was
added 10% Pd/C (175 mg) at rt in a round flask. The flask
was plugged with hydrogen at atmospheric pressure. After
stirred at rt for 2 days, Pd/C (175 mg) was further, added to
the mixture. Then the stirring was continued one more day
at rt under hydrogen atmosphere. The mixture was filtered
through a Celite pad. After removing solvent in vacuo, the
residue was purified by silica gel column chromatography
(2:3 ether/hexane) to give 2.87 g (94%) of Ethyl-(2-amino-
4-fluorophenyl)acetate (9) as orange oil. IR (KBr) nmax
3375, 1731, 1633, 1519, 1445, 1303, 1207, 1161 cmK1. 1H
NMR (500 MHz, CDCl3) d 1.25 (3H, t, JZ7.1 Hz, CH3),
3.47 (2H, s, ArCH2), 3.70 (2H, br s, NH2), 4.14 (2H, q, JZ
7.1 Hz, OCH2), 6.58 (1H, ddd, JZ8.3, 4.5, 2.0 Hz, Ar-6H),
6.70 (1H, dd, JZ8.6, 2.0 Hz, Ar-2H), 6.90 (1H, dd, JZ11.0,
8.3 Hz, Ar-5H) ppm. 13C NMR (125 MHz, CDCl3) d 14.1,
40.7, 60.8, 115.1 (d, 2JC–FZ18 Hz), 117.6 (d, 3JC–FZ3 Hz),
4.1.1. Ethyl-(4-fluorophenyl)acetate (7). A solution
(6.0 ml) of 4-fluorophenylacetic acid (6) (3.06 g,
19.9 mmol) in 6.0 ml of EtOH was refluxed at 80 8C for
2.5 h in the presence of 98% H2SO4 (0.84 ml) under Ar
atmosphere. The solution was then poured into water at
0 8C, and the resultant solution was extracted with ether.
The organic layer was washed with water and brine, and
then was passed through a Na2SO4–SiO2–Na2SO4 column.
Evaporation of the effluent afforded Ethyl-(4-fluorophenyl)-
acetate (1) in 98%. The product 7 was used for the following
reaction without any purification. For the spectroscopic
analyses, compound 7 was recrystallized with hexane to
afforded a colorless crystalline. Mp 29.0–30.0 8C. IR (KBr)
nmax 1736, 1511, 1224 cmK1. 1H NMR (270 MHz, CDCl3) d
1.23 (3H, t, JZ7.3 Hz, CH3), 3.57 (2H, s, ArCH2), 4.14 (2H,
q, JZ7.3 Hz, OCH2), 6.99 (2H, t, JZ8.8 Hz, Ar-3H), 7.24
(2H, dd, JZ8.8, 5.1 Hz, Ar-2H) ppm. 13C NMR (67.8 MHz,
3
4
119.3 (d, JC–FZ6 Hz), 130.3 (d, JC–FZ4 Hz), 134.4 (d,
2JC–FZ14 Hz), 150.9 (d, JC–FZ236 Hz), 171.6 ppm. 19F
1
NMR (376 MHz, CDCl3) d K75.1 ppm. EI-MS m/z 197
(MC), 124. HRMS (EI) calcd for C10H12O2NF: 197.0852,
found 197.0871 (MC).
4.1.4. Ethyl-(2-azide-4-fluorophenyl)acetate (10). To a
solution of 9 (2.87 g, 14.6 mmol) in 30 ml of trifluoroacetic
acid (TFA) at 0 8C was added NaNO2 (2.01 g, 29.2 ml)
followed by NaN3 (2.85 g, 43.8 mmol). After stirring for
10 min at 0 8C, the solution was diluted with ether at 0 8C,
then the resultant solution was poured into a cold water.
Water layer was extracted with ether. The combined organic
layer was washed with water, brine then dried over Na2SO4.
After removing solvent in vacuo, the residue was purified by
silica gel column chromatography (1:2 ether/hexane) to give
3.09 g (95%) of Ethyl-(2-azide-4-fluorophenyl)acetate (10)
as yellow oil. IR (KBr) nmax 2125, 1736, 1515, 1427, 1320,
1227, 1160 cmK1. 1H NMR (500 MHz, CDCl3) d 1.26 (3H,
t, JZ7.1 Hz, CH3), 3.56 (2H, s, ArCH2), 4.16 (2H, q, JZ
2
CDCl3) d 14.2, 40.5, 60.9, 115.2 (d, JC–FZ22 Hz), 129.7,
3
1
130.6 (d, JC–FZ8 Hz), 161.8 (d, JC–FZ244 Hz),
171.2 ppm. 19F NMR (376 MHz, CDCl3) d K53.1 ppm.
EI-MS m/z 182 (MC), 154, 109. HRMS (EI) calcd for
C10H11O2F: 182.0743, found 182.0719 (MC). Anal. Calcd
for C10H11O2F: C, 65.92; H, 6.09; N, 0.00. Found: C, 65.91;
H, 6.14; N, 0.27.
7.1 Hz, OCH2), 6.97–7.07 (3H, m, Ar-H) ppm. 13C NMR
2
(125 MHz, CDCl3) d 14.1, 40.3, 61.1, 116.6 (d, JC–F
Z
Z
3
2
19 Hz), 121.7, 126.6 (d, JC–FZ6 Hz), 127.8 (d, JC–F
4.1.2. Ethyl-(4-fluoro-3-nitrophenyl)acetate (8). To a
solution of 7 (3.54 g, 19.5 mmol) in 14 ml of H2SO4 at
0 8C was added 69% HNO3 (1.4 ml) dropwise over 5 min.
After stirring for 35 min at 0 8C, the solution was diluted
with AcOEt at 0 8C, then the resultant solution was poured
into a cold water. Water layer was extracted with AcOEt.
The combined organic layer was washed with water, brine
then dried over Na2SO4. After removing solvent in vacuo,
the residue was purified by silica gel column chromato-
graphy (1:2 ether/hexane) to give 3.5 g (79%) of Ethyl-(4-
fluoro-3-nitrophenyl)acetate (8) as pale yellow oil. IR (KBr)
1
12 Hz), 131.0, 154.0 (d, JC–FZ248 Hz), 171.0 ppm. 19F
NMR (376 MHz, CDCl3) d K66.0 ppm. EI-MS m/z 223
(MC). HRMS (EI) calcd for C10H10O2N3F: 223.0757, found
223.0771 (MC). Anal. Calcd for C10H10O2N3F: C, 53.81; H,
4.52; N, 18.83. Found: C, 53.81; H, 4.71; N, 18.53.
4.1.5. 2-Azide-4-fluorophenylacetic acid (11). To a
solution of 10 (3.09 g, 13.9 mmol) in 10 ml of ether was
added 10 ml of 2 N KOH/EtOH at 0 8C. After stirring at
0 8C for 15 min, the reaction was warmed to rt and stirred
for 1 h. Then the solution was cooled to 0 8C and was
acidified with 1 N aqueous HCl until pH 1–2. The resultant
solution was extracted with AcOEt. The organic layer was
washed with water then dried over Na2SO4. Concentration
in vacuo yielded 2.50 g (93%) of 2-Azide-4-fluorophenyl-
acetic acid (11). The product 11 was used for the following
reaction without any purification. For the spectroscopic
nmax 1737, 1540, 1352, 1254, 1178 cmK1 1H NMR
.
(500 MHz, CDCl3) d 1.28 (3H, t, JZ7.1 Hz, CH3), 3.68
(2H, s, ArCH2), 4.19 (2H, q, JZ7.1 Hz, OCH2), 7.26 (1H,
dd, JZ10.8, 8.6 Hz, Ar-4H), 7.57 (1H, ddd, JZ8.6, 4.2,
2.4 Hz, Ar-5H), 8.00 (1H, dd, JZ7.9, 2.3 Hz, Ar-2H) ppm.
13C NMR (125 MHz, CDCl3) d 14.1, 39.9, 61.4, 118.4 (d,
2JC–FZ21 Hz), 126.8, 131.1, 134.7 (d, 2JC–FZ7 Hz), 136.4

5760
M. Kuse et al. / Tetrahedron 61 (2005) 5754–5762
analyses, compound 11 was recrystallized with ether–
hexane to afforded an orange crystalline. Mp 48.5–
49.5 8C. IR (KBr) nmax 2130, 1718, 1707, 1518,
270.9757, found 270.9755 (MC). Anal. Calcd for C9H7-
ON3FBr: C, 39.73; H, 2.59; N, 15.44. Found: C, 39.84; H,
2.46; N, 15.56.
1
1239 cmK1. H NMR (500 MHz, CDCl3) d 3.60 (2H, s,
ArCH2), 6.97–7.07 (2H, m, Ar-H), 7.05 (1H, br t, JZ
4.1.8. N-(N,N-Dimethylaminophenyl)-3-(3-azide-4-
fluorophenyl)-2-oxopropanimine oxide (14). A solution
of 13 (2.44 g, 8.97 mmol) in 30 ml of CH2Cl2 was refluxed
at 70 8C in the presence of 3.0 ml of pyridine under Ar
atmosphere for 2 h. Concentration of the solution in vacuo
afforded a pyridinium salt as a foamy solid. To a solution of
the pyridinium salt in 90 ml of H2O at 0 8C was added N,N-
dimetyl-4-nitrosoaniline (1.4 g, 9.42 mmol) followed by
9 ml of 1 N aqueous NaOH. The resultant suspension was
warmed to rt and stirred for 2 h at rt with sonicating
occasionally. Filtration of the suspension yielded 2.19 g
(95%, two steps) of N-(N,N-dimethylaminophenyl)-3-(3-
azide-4-fluorophenyl)-2-oxopropanimine oxide (14). The
product 14 was used for the following reaction without any
9.6 Hz, Ar-H) ppm. 13C NMR (125 MHz, CDCl3) d 40.0,
2
3
116.8 (d, JC–FZ19 Hz), 122.0, 126.7 (d, JC–FZ7 Hz),
2
4
128.0 (d, JC–FZ12 Hz), 130.1 (d, JC–FZ4 Hz), 154.2 (d,
1JC–FZ248 Hz), 177.3 ppm. 19F NMR (376 MHz, CDCl3) d
K65.3 ppm. EI-MS m/z 195 (MC), 167 (MCKN2). HRMS
(EI) calcd for C8H6O2N3F: 195.0444, found 195.0449
(MC). Anal. Calcd for C8H6O2N3F: C, 49.24; H, 3.10; N,
21.53. Found: C, 49.25; H, 3.31; N, 21.35.
4.1.6. (3-Azide-4-fluorophenyl)methyl diazomethyl
ketone (12). A solution of 11 (2.50 g, 12.8 mmol) in
8.0 ml of thionylchloride was refluxed 1.5 h. Generating
vapors (SO2 and HCl) were trapped with saturated aqueous
NaHCO3. Concentration of the reaction mixture in vacuo
equipped a trap cold by liquid N2 gave a crude oil of acyl
chloride of 5. The acyl chloride was purified with Kugelrohr
(145 8C) in vacuo, and was immediately used for the
following reaction. To the solution of acyl chloride (2.54 g)
of 5 in 60 ml of ether was added a ether solution of
diazomethane until acylchloride was perfectly consumed at
0 8C. After removing solvent in vacuo, the residue was
purified by silica gel column chromatography (2:1 ether/
hexane) to give 2.23 g (85%, two steps) of (3-Azide-4-
fluorophenyl)methyl diazomethyl ketone (12) as a lemon
colored crystalline. Mp 34.8–35.5 8C. IR (KBr) nmax 2092,
1614, 1512, 1426, 1112 cmK1. 1H NMR (270 MHz, CDCl3)
d 3.56 (2H, s, ArCH2), 5.18 (1H, s, CHN2), 6.92–7.10 (3H,
m, Ar-H) ppm. 13C NMR (67.8 MHz, CDCl3) d 46.7, 55.1,
purification. IR (KBr) nmax 2121, 1605, 1490, 1170 cmK1
.
1H NMR (270 MHz, CDCl3) d 3.03 (6H, s, NMe), 5.45, (1H,
s, ArCH), 6.66 (2H, d, JZ9.2 Hz, ortho-aniline), 7.02 (1H,
dd, JZ10.6, 8.6 Hz, Ar-5H), 7.42 (1H, ddd, JZ8.6, 5.0,
2.3 Hz, Ar-6H), 7.50 (1H, s, CH]NO), 7.57 (2H, d, JZ
9.2 Hz, meta-aniline), 7.62 (1H, dd, JZ8.2, 2.3 Hz, Ar-2H),
13.77 (1H, br s, OH) ppm. 13C NMR (67.8 MHz, CDCl3) d
2
40.3, 110.7, 111.2, 116.5 (d, JC–FZ19 Hz), 120.9, 121.5,
122.3, 126.4 (d, 2JC–FZ7 Hz), 131.0, 133.3 (d, 3JC–FZ5 Hz),
3
1
134.1, 149.4 (d, JC–FZ2 Hz), 151.4, 153.1 (d, JC–F
Z
250 Hz) ppm. 19F NMR (376 MHz, CDCl3) d K64.7 ppm.
EI-MS m/z 341 (MC), 325 (MCK16). HRMS (EI) calcd for
C17H16O2N5F: 341.1288, found 341.1238 (MC).
4.1.9. 3-(3-Azide-4-fluorophenyl)-2-oxopropanal (5).
Compound 14 (700 mg, 2.05 mmol) was suspended into
10% aqueous H2SO4 (40 ml) at 0 8C. The suspension was
warmed to rt and stirred for 1 h with sonicating occasion-
ally. The suspension was extracted with ether. The organic
layer was washed with water and brine, and then was passed
through a Na2SO4–SiO2–Na2SO4 column. Removing
solvent in vacuo, yielded 314 mg (74%) of 3-(3-azide-4-
fluorophenyl)-2-oxopropanal (5) as a dark yellow amor-
phous. The product 5 was used for the following reaction
without any purification. For the spectroscopic analyses,
compound 5 was recrystallized with ether–hexane to
afforded a pale yellow amorphous. Mp 102.0–104.0 8C
(decomposed). IR (KBr) nmax 3329, 2137, 1672, 1650,
1409 cmK1. 1H NMR (270 MHz, CDCl3) d 6.09 (1H, d JZ
0.9 Hz, ArCH), 6.47 (1H, s, OH), 7.12 (1H, dd, JZ10.5,
8.4 Hz, Ar-5H), 7.51 (1H, ddd, JZ8.4, 4.6, 2.1 Hz, Ar-6H),
7.66 (1H, dd, JZ8.1, 2.1 Hz, Ar-2H), 9.25 (1H, d, JZ
0.9 Hz, COH) ppm. 13C NMR (67.8 MHz, CDCl3) d 116.9
2
3
116.9 (d, JC–FZ19 Hz), 121.8 (d, JC–HZ1 Hz), 126.5 (d,
3JC–FZ7 Hz), 128.1 (d, JC–FZ11 Hz), 131.3, 154.0 (d,
2
1JC–FZ248 Hz), 191.2 ppm. 19F NMR (376 MHz, CDCl3) d
K65.6 ppm. EI-MS m/z 219 (MC), 163 (MCK2N2).
HRMS (EI) calcd for C9H6ON5F: 219.0556, found
219.0540 (MC). Anal. Calcd for C9H6ON5F: C, 49.32; H,
2.76; N, 31.95. Found: C, 49.36; H, 2.67; N, 31.61.
4.1.7. (3-Azide-4-fluorophenyl)methyl bromomethyl
ketone (13). To a solution of 12 (2.10 g, 9.59 mmol) in
12 ml of AcOH at 0 8C was added 47% aqueous HBr
(2.6 ml, 14.9 mmol) dropwise slowly. The reaction was
warmed to rt and stirred for 1 h. Then the solution was
poured into water washing with H2O and ether. The mixture
was neutralized with Na2CO3 to pH 4, then with NaHCO3 to
pH 7. The solution was extracted with ether. The organic
layer was washed with water, brine, then dried over Na2SO4.
Concentration in vacuo afforded crude compound.
Recrystallization of the crude compound with ether yielded
2.44 g (93%) of (3-azide-4-fluorophenyl)methyl bromo-
methyl ketone (13) as a pale yellow needle. Mp 74.6–
2
3
(d, JC–FZ20 Hz), 120.2, 122.5, 127.9 (d, JC–FZ7 Hz),
3
1
128.2, 130.7, 148.6 (d, JC–FZ3 Hz), 154.7 (d, JC–F
Z
253 Hz), 187.8 ppm. 19F NMR (376 MHz, CDCl3) d
K60.8 ppm. EI-MS m/z 207 (MC), 179 (MCK28), 163
(MCKN2–CHO). HRMS (EI) calcd for C9H6O2N3F:
207.0444, found 207.0436 (MC). Anal. Calcd for
C9H6O2N3F: C, 52.18; H, 2.92; N, 20.28. Found: C,
52.17; H, 2.66; N, 20.37.
1
75.6 8C. IR (KBr) nmax 2128, 1736, 1513, 1240 cmK1. H
NMR (270 MHz, CDCl3) d 3.92 (4H, s, ArCH2, CH2Br),
6.90–6.96 (2H, m, Ar-H), 7.07 (1H, br t, JZ9.5 Hz,
Ar-H) ppm. 13C NMR (67.8 MHz, CDCl3) d 33.4, 45.4,
2
3
117.0 (d, JC–FZ28 Hz), 122.0 (d, JC–HZ2 Hz), 126.7
3
2
(d, JC–FZ7 Hz), 128.0, 129.9 (d, JC–FZ7 Hz), 154.1 (d,
1JC–FZ249 Hz), 198.5 ppm. 19F NMR (376 MHz, CDCl3) d
K65.1 ppm. EI-MS m/z 273 (MC), 271 (MC), 245 (MCK
N2), 243 (MCKN2). HRMS (EI) calcd for C9H7ON3F79Br:
4.1.10. 8-Benzyl-2-(3-azide-4-fluorophenylmethyl)-6-(4-
hydroxyphenyl)-3,7-dihydroimidazo[1,2-a]pyrazine-3-
one (15) (Az-F-CT). To a degassed solution of

M. Kuse et al. / Tetrahedron 61 (2005) 5754–5762
5761
coelenteramine (4) (214 mg, 0.77 mmol) and ketoaldehyde
(5) (223 mg, 1.08 mmol) in 6.0 ml of 20% water/dioxane
was added 1.0 ml of 10% aqueous HCl, and was stirred
under argon atmosphere at rt for 5 min. Then the reaction
was warmed to 80 8C and stirred for 3.5 h. After cooled, the
solution was poured into water at 0 8C. The resultant
mixture was extracted with AcOEt. The organic layer was
washed with water, brine, and dried over anhydrous
Na2SO4. After removing solvent in vacuo, the residue was
purified by silica gel column chromatography (4% MeOH in
CH2Cl2) to give 215 mg (60%) of 8-benzyl-2-(3-azide-4-
fluorophenylmethyl)-6-(4-hydroxyphenyl)-3,7-dihydroimi-
dazo[1,2-a]pyrazine-3-one (15) as a brown solid. For the
spectroscopic analyses, compound 15 was recrystallized
with ether–MeOH to afforded a yellow powder. Mp 140–
143 8C (decomposed). UV (MeOH) lmax (log 3) 438
(3.97) nm. FL (MeOH) Em. 526.5 nm (Ex. 350 nm). IR
(KBr) nmax 3100 (br), 2357, 2349, 2121, 1564, 1548,
4.1.12. Photolysis of Az-F-CT (15): general procedure.
To a solution of Az-F-CT (15) (2.0 mM) in 2,2,2-
trifluoroethanol in a NMR tubing was irradiated light with
a high-pressure mercury lamp (100-W high pressure Hg
lamp) for 15 min at rt with Ar bubbling. For nano-LC-Q-
TOF-MS measurement, 1 ml of the sample was injected. For
proton/deuterium (H/D) exchange measurement, 2 ml of the
sample was dried in vacuo, then redissolved with 99%
CH3OD:1% CH3COOD (200 ml) for ESI-Q-TOF-MS with
deuterium measurement or with 99% CH3OH:1% CH3-
COOH for MS (200 ml) with proton measurement. The
resultant solutions were injected into ion-source with a
mechanical syringe. TFE insertion product 20. HRMS
(FAB/NBA) calcd for C28H23N4O3F4: 539.1706, found
539.1710 (MHC).
Acknowledgements
1
1513 cmK1. H NMR (CD3OD, 600 MHz), d 4.14 (2H, s,
CH2Ph), 4.41 (2H, s, CH2Ph), 6.88 (3H, ddd, JZ9.6, 8.4,
4.8 Hz, Ar-H), 7.07 (1H, dd, JZ10.8, 8.4 Hz, F-Ar-H), 7.13
(1H, m, Ar-H), 7.19 (1H, br d, JZ8.4 Hz, Ar-H), 7.23 (1H,
t, JZ7.2 Hz, Ar-H), 7.30 (2H, t, JZ7.5 Hz, Ar-H), 7.38
(2H, d, JZ7.8 Hz, Ar-H), 7.45 (2H, br s, Ar-H) ppm. 13C
NMR (CD3OD, 150 MHz), d 49.9, 108.0, 116.8, 117.4 (d,
2JC–FZ20 Hz), 122.5, 127.4 (d, 3JC–FZ6 Hz), 128.2, 128.7
(d, 2JC–FZ11 Hz), 129.3, 129.6, 129.7, 137.1, 137.7, 154.8
The authors are grateful for financial support from MEXT
Grant-in-Aid for Specially Promoted Research (16002007),
JSPS Grant-in-Aid for Encouragement of Young Scientists
(15780085), and SUNBOR Grant.
1
(d, JC–FZ247 Hz), 160.1 ppm. 19F NMR (376 MHz,
References and notes
CD3OD) d K68.9 ppm. FAB-MS (NBA) m/z 467 (MHC).
HRMS (FAB/NBA) calcd for C26H20N6O2F: 467.1632,
found 467.1663 (MHC). Anal. Calcd for C26H19N6O2F: C,
66.95; H, 4.11; N, 18.02. Found: C, 66.95; H, 4.37; N, 17.72.
1. Tsuji, F. I.; Leisman, G. Proc. Natl. Acad. Sci. U.S.A. 1981, 78,
6719–6723.
2. Takahashi, H.; Isobe, M. Bioorg. Med. Chem. Lett. 1993, 3,
2647–2652.
3. Takahashi, H.; Isobe, M. Chem. Lett. 1994, 843–846.
4. Isobe, M.; Kuse, M.; Yasuda, Y.; Takahashi, H. Bioorg. Med.
Chem. Lett. 1998, 8, 2919–2924.
4.1.11. 8-Benzyl-2-(3-azide-4-fluorobenzylidene)-6-(4-
hydroxyphenyl)-2,3-dyhydroimidazo[1,2-a]pyrazine-3-
one (2) (Az-F-DCT). To a solution of coelenterazine analog
(15) (70 mg, 0.15 mmol) in a mixture of diethyl ether
(175 ml) and EtOH (35 ml) was added 85% manganese(II)
oxide (700 mg, 6.83 mmol) at 0 8C under Ar atmosphere.
After stirring for 1 h, the reaction was warmed to rt and
stirred for 1.5 h. The mixture was filtered through a Celite
pad, then was concentrated in vacuo. Recrystallization of
the residue with 2-PrOH/ether yielded 45 mg (64%) of
8-benzyl-2-(3-azide-4-fluorobenzylidene)-6-(4-hydroxy-
phenyl)-2,3-dyhydroimidazo[1,2-a]pyrazine-3-one (2) as a
purple amorphous. Mp 154–156 8C (decomposed). UV
(2-PrOH) lmax (log 3) 541 (3.55), 371 (3.93), 281
(4.11) nm. IR (KBr) nmax 2359, 2133, 1721, 1591,
5. Kuse, M.; Isobe, M. Tetrahedron 2000, 56, 2629–2639.
6. Fujii, T.; Ahn, J. Y.; Kuse, M.; Mori, H.; Matsuda, T.; Isobe,
M. Biochem. Biophys. Res. Commun. 2002, 293, 874–879.
7. Souto, M. L.; Um, J.; Borhan, B.; Nakanishi, K. Helv. Chim.
Acta 2000, 83, 2617–2628.
8. Dorman, G.; Prestwich, G. D. Trends Biotechnol. 2000, 18,
64–77.
9. Kurono, M.; Shimomura, A.; Isobe, M. Tetrahedron 2004, 60,
1773–1780.
10. Zheng, J. L.; Chen, F. Q.; Hirano, T.; Ohmiya, Y.; Maki, S.;
Niwa, H.; Ohashi, M. Bull. Chem. Soc. Jpn. 2000, 73,
465–469.
11. (a) Poe, R.; Scnapp, K.; Young, M. J. T.; Grayzar, J.; Platz,
M.S. J.Am. Chem. Soc. 1992, 114, 5054–5067. (b)Platz,M. S. In
Reactiveintermediatechemistry;Moss,R.A., Platz,M. S., Jones,
M., Jr., Eds.; Wiley: New Jersey, 2004; pp 501–559.
12. Isobe, M.; Fujii, T.; Kuse, M.; Miyamoto, K.; Koga, K.
Tetrahedron 2002, 58, 2117–2126.
1
1503 cmK1. H NMR (DMSO-d6, 600 MHz) d 4.29 (2H,
s, CH2Ph), 6.81 (2H, d, JZ8.8 Hz, Ar-H), 7.24 (1H, t, JZ
7.3 Hz, Ar-H), 7.35 (2H, t, JZ7.7 Hz, Ar-H), 7.50 (3H, dd,
JZ13.2, 5.4 Hz, Ar-H), 7.53 (1H, s, 5-H), 7.78 (2H, d, JZ
8.6 Hz, Ar-H), 7.94 (1H, s, Ar-H), 8.10–8.12 (1H, m, Ar-H),
8.62 (1H, d, JZ8.4 Hz, Ar-H), 9.66 (1H, br s, OH) ppm. 13C
NMR (DMSO-d6, 150 MHz) d 49.3, 108.0, 116.8, 117.4
(2JC–FZ20 Hz), 127.4 (3JC–FZ6 Hz), 128.1, 128.7
(2JC–FZ11 Hz), 129.3, 129.6, 129.7, 137.1, 137.7, 154.8
(1JC–FZ247 Hz), 160.1 ppm. 19F NMR (376 MHz,
CD3OD) d K58.4 ppm. FAB-MS (NBA) m/z 465 (MHC).
HRMS (FAB/NBA) calcd for C26H18N6O2F: 465.1475,
found 465.1501 (MHC). Anal. Calcd for C26H17N6O2F: C,
67.24; H, 3.69; N, 18.09. Found: C, 67.02; H, 3.80; N, 18.16.
13. Kita, Y.; Tohma, H.; Inagaki, M.; Hatanaka, K.; Yakura, T.
Tetrahedron 1991, 32, 4321–4324.
14. Kishi, Y.; Tanino, H.; Goto, T. Tetrahedron Lett. 1972, 13,
2747–2748.
15. Kakoi, H. Chem. Pharm. Bull. 2002, 50, 301–302.
16. Inoue, S.; Sugiura, S.; Kakoi, H.; Hashizume, K.; Goto, T.; Iio,
H. Chem. Lett. 1975, 141–144.
17. Recently, we reported the alternative synthetic route for
various 6-substituted coelenteramine analogs: Kuse, M.;

5762
M. Kuse et al. / Tetrahedron 61 (2005) 5754–5762
Kondo, N.; Ohyabu, Y.; Isobe, M. Tetrahedron 2004, 60,
835–840.
sion product to benzazepine type intermediate was observed.26
This may be responsible for the fluorine atom, which
prevents carbon migration to the nitrene due to its electro-
negativity.
18. Pinney, K. G.; Katzenellenbogen, J. A. J. Org. Chem. 1991,
56, 3125–3133.
.
24. Kuse, M.; Kanakubo, A.; Suwan, S.; Koga, K.; Isobe, M.;
19. Head, J.; Inoue, S.; Teranishi, K.; Shimomura, O. Nature 2000,
405, 372–376.
Shimomura, O. Bioorg. Med. Chem. Lett. 2001, 11,
1037–1040.
20. Usami, K.; Isobe, M. Tetrahedron Lett. 1995, 36, 8613–8616.
21. Usami, K.; Isobe, M. Tetrahedron 1996, 52, 12061–12090.
22. Kurahashi, T.; Miyazaki, A.; Suwan, S.; Isobe, M. J. Am.
Chem. Soc. 2001, 123, 9268–9278.
25. Kanakubo, A.; Isobe, M. J. Mass Spectrom. 2004, 39,
1260–1267.
26. We considered the possibility for the ring expansion of the
nitrene intermediate. If ring expansion happened, the resultant
azepine derivative (22) also has four exchangeable protons.
However, the initial fragmentation product observed at m/z
439 should have three exchangeable protons. If this is the case,
m/z 442 should be observed in deuterium solvent. Thus, the
peak at m/z 443 in Figure 4I was critical to exclude the
possibility of ring expansion product (22).
23. In deuterium ESI-MS measurement, the molecular weight of
19 (m/z 441) increased to plus 5 mass units to give molecular
ion at m/z 446. The exchangeable 5 protons strongly support
our assignment. We also suppose it possible that 19 might be
produced from C–H inserted product (21) by hydrolyzed
aminal moiety, or be directly produced from nitrene
intermediate through intersystem crossing.11 No ring expan-
.