1974
R. Schirrmacher et al. / Tetrahedron Letters 52 (2011) 1973–1976
Table 1
Optimized conditions for the opening of epoxides with 18FÀ in CH3CN and t-amyl-OH
Epoxide/labeled compound
Solvent/temperature (°C)/time (min)a
Base system
RCY% (%) n = 10
1a/2a
1a/2a
1a/2a
1b/2b
1b/2b
1b/2b
1c/2c
1c/2c
1c/2c
1d/2db
CH3CN/100/30
CH3CN/100/30
t-Amyl-OH/110/30
CH3CN/100/30
CH3CN/100/30
t-Amyl-OH/110
CH3CN/100/30
CH3CN/100/30
t-Amyl-OH/100/30
t-Amyl-OH/100/30
K2CO3
24
30
80
22
28
86
7
10
90
87
2
3
3
4
2
3
3
4
5
7
K2(COO)2
K2(COO)2
K2CO3
K2(COO)2
K2(COO)2
K2CO3
K2(COO)2
K2(COO)2
K2(COO)2
a
The maximum RCY for all reactions was reached at approx. 30 min.
b
The reaction of 1d with 18FÀ was only performed in t-amyl-OH.
by 18FÀ.4 It is likely that the N-terminal epoxide moiety reacts
unselectively with side chain functionalities of the peptide.
To finally obtain [18F]5, the new prosthetic labeling precursor
HO
R
F
non-radioactive
standard compounds
[
18F]FPB ([18F]2d) was synthesized from 1d in high RCYs between
3a-d
160°
DMF
80% and 95% (based on HPLC) and conjugated to 6 in aqueous solu-
tion at pH 4 to form [18F]5 via chemoselective oxime formation in
reliable RCYs of 60–70% as determined by analytical HPLC. The 18F-
labeled product could be isolated in preparative RCYs of 28–32%
after 65–70 min total preparation time (Figs. 1 and 2). The prepara-
tive HPLC purification of [18F]5 was particularly easy since no
radioactive side products were formed during the conjugation
reaction (Fig. 3).
CsF
R
2
2
2
K
/
F
K
HO 18F
18F-/K222/K+
t-amyl-alcohol
T=100-110°
O
R
18
1a-d
]2a-d
[
F
77-95%
The preparation of the secondary labeling precursor [18F]FPB as
well as its conjugation efficiency to aminooxy-derivatized peptides
are thus comparable to the currently, mainly used aldehyde-com-
prising secondary labeling precursor 18F-fluorobenzaldehyde13,14
and identify [18F]FPB as a well-suited secondary labeling precursor
for the radiolabeling of complex aminooxy-derivatized molecules.
a
c
b
d
H
O
R a-d:
N
N
O
NO2
O
In all investigated reactions, the opening was 100% a-selective
with the 18FÀ reacting at the sterically less hindered position which
is in accordance to the findings of Park et al.15 During the epoxide
ring opening reactions, two different enantiomers are formed
which should in general be of no concern for an in vivo application
of labeled biomolecules obtained by this reaction type. For
Figure 1.
a-Selective epoxide opening of model compounds 1a and 1b, the labeling
precursor 1c for the synthesis of [18F]FMISO and 1d yielding [18F]2d for peptide
labeling via oxime formation. Non-radioactive standard compounds 3a–d were
synthesized reacting 1a–d with CsF in DMF at 160 °C.
[
18F]MISO, the formation of enantiomers was postulated not to
Interestingly and in contrast to the reactions performed in dipo-
lar aprotic solvents, in t-amyl-OH no radioactive side products
were detected either by radio-TLC or radio-HPLC. The same reac-
tion conditions were applied to the synthesis of [18F]MISO (2c)
from its epoxide precursor 2-nitro-1-(oxiran-2-ylmethyl)-1H-
imidazole (1c). [18F]MISO was obtained in RCYs of 80–95% in pure
t-amyl-alcohol after 30 min reaction time at 100 °C using K2(COO)2
as base. With this labeling reaction, the [18F]MISO formation
reached a plateau of about 90( 5)% RCY after 30 min. After 5, 10,
and 20 min, the RCYs were about 45( 5)%, 61( 5)%, and 78%( 5),
respectively. The corresponding reaction in CH3CN gave 7–10%
RCY only (Table 1). The preparative RCYs of [18F]MISO after HPLC
purification were 55% after a total synthesis time of 60 min.
In order to validate if those conditions are applicable to the 18F-
labeling of more complex biomolecules such as peptides, Tyr3-oct-
reotate was N-terminally modified with aminooxy-acetic acid11,12
(6, Fig. 2) and conjugated to racemic 4-(oxiran-2-ylmethoxy)benz-
aldehyde (1d) by chemoselective oxime formation yielding peptide
4 (Fig. 2).12 The direct 18F-fluorination of 4 under various condi-
tions did not yield the desired labeled peptide [18F]5. Neither dipo-
lar aprotic solvents nor t-amyl-OH resulted in the formation of
radiolabeled [18F]5. Besides unreacted 18FÀ, different non-radioac-
tive side products could be detected by HPLC (UV channel) which
were not further analyzed. This is in contrast to the results re-
ported for the one-step aziridine opening of complex biomolecules
negatively influence the pharmacologic properties of the radio-
tracer as well.16 Nevertheless, the formation of enantiomers pro-
duced during the epoxide opening may have negative effects on
the pharmacologic properties of small molecules and thus has to
be kept in mind.
Interestingly, the non-radioactive 19F-standard compounds 3a–
d synthesized for analytical purposes could not be obtained using
the corresponding 18F-labeling conditions employing macroscopic
amounts of KF/K222 in t-amyl-OH.
The 19F-fluorinated compounds 3a and 3b could only be synthe-
sized from the epoxide precursors using CsF in DMF at 160 °C for
24 h and analytical data (1H, 13C NMR, mass) were compared with
literature data.17,18 Compound 3c, which was used as non-radioac-
tive standard compound, was purchased from ABX (Radeberg, Ger-
many). The non-radioactive standard compound 3d was obtained
as follows: The racemic epoxide 1d (0.35 g, 1.96 mmol) and CsF
(0.89 g, 5.88 mmol) were dissolved in DMF (7 mL) and reacted at
160 °C for 24 h. Compound 3d was purified via column chromatog-
raphy. Compound 3d: 1H NMR (400 MHz, CDCl3) d (9.84 (s, 1H), 7.8
(d, 2H, J = 8.85 Hz), 6.98 (d, 2H, J = 8.74 Hz), 4.57 (ddd, 2H,
J = 46.97 Hz, 4.63 Hz, 2.31 Hz); 13C NMR (400 MHz, CDCl3)
d
190.80, 132.15, 115.08, 76.80, 69.19, 50.00, 44.72, 29.87, ESI-MS:
m/z 199 ([M+H]+).
The synthesis of aminooxy-derivatized TATE (6) was performed
as previously described.19 The non-radioactive peptide 5 and the