Synthesis and in vitro evaluation of 18F- and 19F-labeled insulin: A new radiotracer for PET-based molecular imaging studies

Article abstract of DOI:10.1021/jm0509344

A new and regioselective strategy was developed for the preparation of fluorine-18-labeled insulin as a novel positron emission tomography (PET) tracer. [¹⁸F]-4-Fluorobenzoic acid (4-¹⁸FBA), which was produced in 83 ± 8% yield (n = 1

Full text of DOI:10.1021/jm0509344

1
466
J. Med. Chem. 2006, 49, 1466-1474
18
19
Synthesis and in Vitro Evaluation of F- and F-Labeled Insulin: A New Radiotracer for
PET-based Molecular Imaging Studies
†
†
‡
‡
‡
§
Katharina J. Guenther, Sabesan Yoganathan, Robert Garofalo, Thomas Kawabata, Thomas Strack, Ren e´ e Labiris,
§
|
,†
Myrna Dolovich, Raman Chirakal, and John F. Valliant*
Departments of Chemistry and Medical Physics & Applied Radiation Sciences, McMaster UniVersity, 1280 Main St. West,
Hamilton, ON, Canada L8S 4M1, Pfizer Inc., 235 East 42nd Street 3/58, New York, New York 10017, Department of Medicine,
McMaster UniVersity, 1200 Main St. West, Hamilton, ON, Canada L8M 3Z5, and Department of Radiology,
Hamilton Health Sciences, 1200 Main St. West, Hamilton, ON, Canada L8M 3Z5
ReceiVed September 20, 2005
A new and regioselective strategy was developed for the preparation of fluorine-18-labeled insulin as a
1
8
18
novel positron emission tomography (PET) tracer. [ F]-4-Fluorobenzoic acid (4- FBA), which was produced
18
18
in 83 ( 8% yield (n ) 10), through the use of succinimidyl [ F]-4-fluorobenzoate (4- FSB), was conjugated
through a short spacer (6-aminohexanoic acid, AHx) to the PheB1 residue of a protected form of insulin.
FB-AHx-insulin (8b) was repeatedly prepared in practical quantities (10-20 mCi, 370-740 MBq) in
good radiochemical yield (9 ( 5%, n ) 9) and in a specific activity of 7.8 mCi/µmol. The final product was
1
8
1
9
19
characterized by comparing the radioHPLC and radioTLC of 8b with that of the F-analogue ( FB-AHx-
insulin, 8a) and by analyzing a carrier-added synthesis by mass spectrometry. Dithiothreitol and endoproteinase
Glu-C digestion experiments on 8a confirmed that the prosthetic group was in fact conjugated to the PheB1
residue. An insulin receptor (IR) phosphorylation assay using CHO-hIR cells overexpressing recombinant
human insulin receptors indicated no statistical difference in the extent of autophosphorylation stimulated
by 8a as compared to that for human insulin (EC50 values of 0.82 nM and 1.0 nM, respectively). The
stimulation of 2-deoxyglucose uptake in 3T3-L1 mouse adipocytes utilizing 8a versus unmodified human
insulin gave similar EC50 values of 0.68 nM and 0.41 nM, respectively. The IC50 values for 8a versus native
insulin for the displacement of 1 I-insulin from HEK-293 cells were also the same within experimental
error (2.6 nM for 8a versus 2.4 nM for unmodified human insulin). These results support the use of the
25
1
8
F-insulin analogue as a PET tracer for imaging the distribution of insulin in vivo.
Introduction
done using in vitro assays. They are also needed as well
characterized reference standards for the radiolabeled analogues.
The availability of dedicated scanners for imaging small
animals is revolutionizing the drug discovery process and the
study of biochemical processes.1,2 Radioimaging is particularly
suited for these studies because its sensitivity provides the
opportunity to directly visualize the distribution and/or metabo-
lism of specific compounds in living subjects noninvasively.
This allows compounds to be studied in a more realistic
environment compared to traditional methods, and it also allows
longitudinal studies to be performed in the same animals, thereby
Human insulin is a polypeptide-based hormone involved in
3
the regulation of energy metabolite homeostasis. Specifically,
the secretion of insulin in response to high blood glucose levels
stimulates the storage of metabolites as glycogen, lipids, and
protein for future use when exogenous nutrients are not
3
available. Decreased insulin secretion alone or in combination
with peripheral tissue resistance to insulin action leads to
hyperglycemia and lipidemia both of which contribute to
significantly greater cardiovascular morbidity and mortality in
2
avoiding subject-to-subject variability. An advantage of using
4
radioimaging over other molecular imaging methods, such as
optical imaging, is that the radiotracers can be used for human
studies and are not limited to experiments in small animals.
The two main radioimaging modalities are single photon
emission computed tomography (SPECT) and positron emission
computed tomography (PET). To employ these techniques to
evaluate the distribution of drug candidates requires the synthesis
and characterization of radiolabeled versions of the compounds
under study. Not only must a method be developed to produce
the tracer in good yield and purity but nonradioactive analogues
must also be prepared and fully characterized. The “cold”
compounds are needed to evaluate the impact of introducing
the label on the biochemistry of the substrate, which is typically
affected individuals. Thus, ameliorating insulin resistance and
increasing insulin levels appropriately to restore normal glucose
and lipid levels are cornerstones of therapeutic strategies and
have resulted in significant improvements in patient’s health
5,6
and quality of life. It should also be noted that abnormal
increases in insulin utilization are associated with a number of
4
diseases, including diabetes, obesity, hypertension, heart fail-
7
8
9
ure, neurodegenerative diseases, and cancer.
A number of radiolabeled analogues of insulin have been
reported as tracers for studying the biochemistry of the hormone
and its role in disease progression and for evaluating different
1
25
insulin delivery strategies. I-Labeled insulin, in which the
isotope is bound to the aromatic ring of A14 tyrosine, is perhaps
the best-known radiolabeled insulin derivative, but it is used
*
To whom correspondence should be addressed. Tel: (905) 525-9140,
1
0-14
solely for in vitro studies.
There are few reports of the
ext 22840. Fax: (905) 522-2509. E-mail: valliant@mcmaster.ca.
†
Departments of Chemistry and Medical Physics & Applied Radiation
synthesis of radiolabeled insulin analogues for in vivo molecular
imaging studies using PET. Potential limitations of the studies
that have been reported (which involved labeling insulin with
Sciences, McMaster University.
‡
Pfizer Inc.
§
Department of Medicine, McMaster University.
Hamilton Health Sciences.
|
18 15,16
124 17,18
F
and
I
) include products that are of questionable
1
0.1021/jm0509344 CCC: $33.50 © 2006 American Chemical Society
Published on Web 01/26/2006

EValuation of 18F- and 19F-Labeled Insulin
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 4 1467
Scheme 2. Preparation of
1
Scheme 1. Preparation of B -(4-Fluorobenzoyl)insulin,
19
1
FB-insulin (4)
B -([4-Fluorobenzoyl]-6-aminohexanoyl)insulin, FB-AHx-insulin
8a, 8b)
(
stability, a lack of chemical and biological characterization data,
and/or low overall production yields.
Herein a new method for the preparation of 18F-labeled insulin
is described. The reported approach overcomes several of the
limitations of existing strategies and is accompanied by detailed
chemical and in vitro characterization of the nonradioactive (i.e.
19
F) analogue. The product of this research can be used to study
the distribution and, indirectly, the biochemistry of insulin in
vivo using PET.
Boc groups on the hormone, while the acid was converted to
the active ester to facilitate bioconjugation (Scheme 2). Reaction
of 5 with DBI (1) was done in a 10:1 molar ratio and was
complete in 30 min as shown by HPLC. The resulting conjugate,
which was precipitated from solution by the addition of ether,
was subsequently treated with 20% (v/v) piperidine in DMF
buffer to remove the Fmoc group. The reaction mixture was
subsequently passed through a size-exclusion column as a means
of removing dibenzofulvene and related derivatives using 100
mM NH HCO as the eluent. The purity of the free amine 6
Results and Discussion
1
9
Synthesis and Characterization of 4- FB-AHx-insulin.
The first objective was to prepare a 19F-labeled insulin analogue
that would be used to determine if the addition of the label
would have an impact on the biological activity of the hormone.
The approach involved derivatizing the phenylalanine B1
(
PheB1) amino acid of insulin with the N-hydroxysuccinimidyl
19
19
active ester of 4- F-fluorobenzoic acid (4- FSB). The amine
of the PheB1 residue was deemed the most attractive site
4
3
was confirmed by HPLC and the electrospray MS, which
because this region of the hormone has been shown to not be
showed the expected molecular ion at an m/z value of 6121.4.
Derivatization of 6 with 4- FSB was accomplished by adding
involved in receptor binding.1
9-22
Unfortunately, it has been
19
well-documented that the reactivity of the PheB1 amino residue
toward derivatization by electrophiles is less than that of the
GlyA1 and LysB29 amino acids.23 To circumvent this issue,
Shai et al. and Baud y_s et al. developed methods to selectively
2a to the substrate in DMSO-containing TEA. The reaction was
followed by HPLC and the consumption of the amine was
completed in 20 min, whereby the MS of 7a showed a m/z value
of 6243.4, which is consistent with the addition of a single FSB
group. The Boc protecting groups were subsequently removed
by treatment of 7a with TFA in the presence of 5% anisole,25
and the product was isolated by preparative HPLC. After
lyophilization, the resulting precipitate was desalted using either
dialysis against 100 mM NH HCO or size exclusion chroma-
1
5,24
protect the GlyA1 and LysB29 amines with Boc groups,
thus providing a suitable precursor (DBI, 1) for selective
derivatization at PheB1.
Compound 1, which was prepared in 70% yield, was reacted
19
with succinimidyl 4-fluorobenzoate (4- FSB, 2a) in the pres-
ence of triethylamine (TEA) (Scheme 1). The progress of the
reaction between 1 and 2a was followed by HPLC and it was
apparent that conjugation was complete after 20 min. Compound
4
3
tography using 100 mM NH HCO as the eluent. The overall
4
3
yield of synthesis was approximately 15% and the purity of 8a
was greater than 95% by RP-HPLC.
3
was purified by semipreparative HPLC, and the Boc groups
Liquid chromatography mass spectrometry (LCMS) (elec-
trospray) analysis of 8a showed a single peak in the HPLC at
6.56 min (Figure 1a), which resulted in three sets of peaks in
the mass spectrum at m/z values of 1209.9, 1512.1, and 2015.4
(calcd 1209.6, 1511.7, and 2015.3, respectively) (Figure 1b).
were removed using a mixture of trifluoroacetic acid (TFA)/
anisole according to the method described by Lundt et al.
2
5
Following purification by HPLC, the product (4) was lyophi-
lized, yielding a white powder, which showed a single peak in
the HPLC (tR ) 12.9 min) and exhibited the expected mass-
to-charge ratio in the electrospray mass spectrum (MS) (5928.0
m/z).
1
9
These three sets of peaks correspond to FB-AHx-insulin (8a)
bearing +5, +4, and +3 charges, respectively. Additional peaks
in the mass spectrum appeared that correlate with TFA adducts
of the parent ion. The MALDI-TOF mass spectrum showed the
parent ion at a m/z value of 6043.8, which is consistent with
the calculated mass of 8a.
It was later discovered (vide infra) that the PheB1 residue
18
does not react with 4- FSB when it is prepared in high specific
activity. As a result, a short spacer group was attached to the
PheB1 residue to increase the reactivity of the amine and to
extend the site of conjugation away from the globular structure
of insulin. The spacer chain used in this study was 6-amino-
hexanoic acid (AHx). The amine was protected with an Fmoc
group to allow for orthogonal deprotection with respect to the
The 19F NMR spectrum of 8a (Figure 2) shows a single
multiplet containing broad peaks centered at -109.60 ppm. The
3
4
JFH and JFH couplings were 8.9 and 5.6 Hz, respectively. As
1
9
19
a reference, the F NMR spectra of [ F]-4-fluorobenzoic acid
(4- FBA) and 2a were acquired, which each show single
1
9

1
468 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 4
Guenther et al.
Figure 1. (a) RP-HPLC chromatogram of 19FB-AHx-insulin (8a). (b) ES(+) MS of FB-AHx-insulin (8a) from LCMS analysis scanned from 1100
to 2200 m/z.
multiplets (nonets) centered at -106.67 and -101.42 ppm,
To confirm the site of derivatization, compound 8a was
3
4
19
respectively. The JFH and JFH coupling constants for 4- FBA
and compound 2a are identical to that in 8a. The broad lines in
the 19F NMR spectrum of 8a and the increased complexity of
the observed multiplet are likely the result of the existence of
multiple conformations of the hormone derivative, which is
consistent with the NMR studies on insulin reported by Weiss
et al.26
initially treated with dithiothreitol to disrupt the interchain
2
7
disulfide bonds. LCMS analysis of the sample (Table 1)
showed a peak at 88.36 min with an m/z value of 1190.1, which
corresponds to the molecular ion of the unmodified A-chain. A
peak at 104.99 min with an m/z value of 1222.0 corresponds to
the B-chain plus the FB-AHx pendant group attached ([M +
+
H ]).

EValuation of 18F- and 19F-Labeled Insulin
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 4 1469
Figure 2. 19F NMR spectra (DMSO-d , 470.6 MHz) of 4-fluorobenzoic acid (left), 19FSB (2a) (middle), and 19FB-AHx-insulin (8a) (right).
6
Table 1. LCMS (ES+) Studies of 19FB-AHx-insulin (8a) Fragments
CGSHLVE, which is consistent with the prosthetic group being
attached to the PheB1 residue and not the GlyA1 or LysB29
sites.
after Dithiothreitol Treatment
molecular
HPLC tRa
8.36
04.99
MS found (calcd)
ion
fragment
A-chain
FB-AHx-B-chain
In Vitro Screening of 8a. The bioactivity of 8a was
determined in vitro in direct comparison to human insulin using
a number of assays. The first assay was an intact cell insulin
receptor autophosphorylation assay in CHO cells, which over-
express recombinant human insulin receptors (CHO-hIR). There
was no statistically significant difference observed in the extent
of autophosphorylation induced by FB-AHx-insulin as compared
to unmodified human insulin (Figure 3). The calculated EC50
for the autophosphorylation was 0.82 nM for FB-AHx-insulin
and 1.0 nM for unmodified human insulin.
+
8
1
1190.1 (1192.9)
1222.0 (1222.8)
[M + 2H ]/2
+
[M + 3H ]/3
a
HPLC column and conditions: C18 Beckmann Ultrasphere (150 × 4.6
mm, porosity 5 µm); mobile phase A, H2O with 0.05% TFA; mobile phase
B, 90/10 (v/v) H2O/CH3CN with 0.05% TFA; gradient, 99:1 (A/B) for 10
min, 99:1 (A/B) to 11:89 (A/B) for 140 min, 11:89 (A/B) to 99:1 (A/B) for
3
min, 99:1 (A/B) for 17 min; flow rate ) 0.1 mL/min.
Table 2. LCMS (ES+) Studies of 19FB-AHx-insulin (8a) Fragments
after Endoproteinase Glu-C Digestion
HPLC
tR
molecular
ion
The second assay was performed to determine if compound
a
MS found (calcd)
fragment
GIVE (A-chain)
RGFFYTPKT (B-chain)
8a could stimulate glucose uptake. The extent of 2-deoxyglucose
+
6
7
5.41 417.5 (417.5)
[M + H ]
uptake in 3T3-L1 (mouse) adipocyte cells induced by 8a was
not significantly different from that induced by unmodified
human insulin (Figure 4). The calculated EC50 values were 0.41
nM for human insulin and 0.68 nM for 8a.
+
8.99 1116.6 (1117.3) [M + H ]
+
1
03.87 1145.1 (1146.3) [2M + 3H ]/3 FB-AHx-FVNQHLCGSHLVE
(B-chain)
a
HPLC column and conditions: C18 Beckmann Ultrasphere (150 × 4.6
The ability of 8a to bind to the insulin receptor was
mm, porosity 5 µm); mobile phase A, H2O with 0.05% TFA; mobile phase
B, 90/10 (v/v) H2O/CH3CN with 0.05% TFA; gradient, 99:1 (A/B) for 10
min, 99:1 (A/B) to 11:89 (A/B) for 140 min, 11:89 (A/B) to 99:1 (A/B) for
determined using a displacement assay involving recombinant
125
I-insulin and HEK-293 cells expressing the recombinant
3
min, 99:1 (A/B) for 17 min; flow rate ) 0.1 mL/min.
human insulin receptor. The displacement curves for human
1
9
insulin and FB-AHx-insulin (8a) were, within experimental
error, equivalent (Figure 4), with 50% displacement achieved
at concentrations of 2.6 nM for 8a and 2.4 nM for unmodified
The sample was subsequently treated with endoproteinase
Glu-C, an enzyme that facilitates cleavage of peptides at the
carbonyl side of glutamic acid residues.27 As demonstrated in
Table 2, the peak at 65.41 min resulted in a m/z of 417.5, which
insulin, respectively.
Radiochemistry. 18FB-AHx-insulin (8b) was synthesized in
+
is consistent with the [M + H ] fragment: GIVE from the
fives steps from compound 9 (Schemes 2 and 3), which was
A-chain, confirming that derivatization did not occur at the
GlyA1 residue. The peak at 78.99 min exhibited an m/z value
of 1116.6, which matches the fragment RGFFYTPKT ([M +
28 18
-
prepared following literature procedures.
F[F ] was added
to 9 and Kryptofix222 (4,7,13,16,21,24-hexaoxa-1,10-diaza-
bicyclo[8.8.8]hexacosane) in anhydrous CH3CN and the mixture
heated to 130 °C for 10 min (Scheme 3). Compound 10 was
purified using a silica cartridge and was obtained in 76 ( 9%
radiochemical yield (n ) 10, decay-corrected). TFA was added
to compound 10, cooled in an ice bath, to remove the
pentamethylbenzyl-protecting group. After 5 min, the residual
TFA was evaporated, which resulted in a significant loss of
activity (between 50% and 80%).
+
H ]) from the B-chain. There was no peak that corresponded
to the attachment of the pendant group 1 FB-AHx to the lysine
residue of RGFFYTPKT from the B-chain. Interestingly, the
fragment FB-AHx-FVNQHLCGSHLVE was also not observed;
however, the HPLC peak at 103.87 min gave rise to two m/z
values of 859.1 and 1145.1. These correspond to a proton-bound
9
+
dimer ([2M + 3H ]/3) of the fragment FB-AHx-FVNQHL-

1
470 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 4
Guenther et al.
Figure 3. Comparisons of Aventis (human) insulin and 19FB-AHx-insulin (8a): insulin receptor autophosphorylation ELISA with 50 mL of CHO-
hIR lysate/well (24-well plate); n ) 3 wells per treatment.
Figure 4. Comparisons of Aventis (human) insulin and 19FB-AHx-insulin (8a): insulin dose-response for 2-deoxyglucose uptake in 3T3-L1
1
25
adipocytes (left); displacement of bound I-insulin in HEK-293 cells expressing recombinant human insulin receptor (right).
Scheme 3. Preparation of Succinimidyl 4- F-Fluorobenzoate (4- FSB, 18F-2b)
1
8
18
To determine the nature of the volatile species, a trapping
experiment was performed where the TFA was removed by
blowing N2 over the sample while passing the gas through a
separate vial cooled to -5 °C. The radiochromatogram (HPLC)
of the volatile compounds showed that the collected radioactive
product was, as expected, compound 11. The extent of sublima-
tion, even under fairly mild evaporation conditions, was
significant.
To prepare the active ester (Scheme 3), N-hydroxysuccinimide
and 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochlo-
ride in anhydrous CH3CN were added to the CH3CN solution
of 11. The reaction was stirred at room temperature for 10 min
and the product subsequently isolated by semipreparative HPLC
in 83 ( 8% yield (n ) 10, decay-corrected). HPLC purification
of 2b was necessary to ensure reproducibility in the subsequent
conjugation reaction.
To alleviate the loss, it is possible to add NMe4OH to the
solution prior to the evaporation step.29 We found this approach
problematic in terms of the base’s impact on the reproducibility
in the subsequent step involving formation of the active ester.
As an alternative approach, following acid deprotection, the
resulting mixture was diluted with H2O (5 mL) and passed
through an activated C18 cartridge. The cartridge was washed
with H2O (5 mL) to remove any residual TFA and then dried
for 1 min using a stream of nitrogen. Compound 11 was
subsequently eluted from the cartridge using anhydrous CH3-
CN (3 mL) in 88 ( 6% yield (n ) 10, decay-corrected), and it
can be directly converted to the active ester 2b without having
to evaporate the solvent. This approach prevented loss due to
sublimation and is easily automated.
Initially, attempts were made to conjugate 2b to the PheB1
residue of diBoc-insulin 1, on the basis of the successful
synthesis of the nonradioactive analogue. However, formation
of the bioconjugate was not observed. Varying the solvent
(DMSO, DMF, PBS, or borate buffers), temperature (room
temperature to 50 °C), and the amount of 1 (2-18 mg) had no
affect on the reaction with the exception of increasing the
amount of 11, which formed as a result of hydrolysis of the
active ester.
In contrast to the direct reaction of 2b with 1, successful
conjugation with 6 was achieved in 20 min (Scheme 2). The
18
product, FB-AHx-DBI (7b), was isolated from low molecular
weight compounds using a size-exclusion cartridge in 43 ( 10%
yield (n ) 9, decay-corrected). The last step involved the

EValuation of 18F- and 19F-Labeled Insulin
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 4 1471
removal of the Boc protecting groups by treating 7b with neat
TFA in the presence of anisole, which was followed by
purification using a size-exclusion cartridge. Residual AHx-
insulin in the sample can be removed by preparative HPLC
of silica cartridges (Waters) was completed immediately prior to
use using Et
activated with EtOH (10 mL) followed by H
2
O (10 mL), while RP C18 cartridges (Waters) were
2
O (10 mL).
For nonradioactive materials, analytical and semipreparative
HPLC experiments were performed using a Varian ProStar system,
fitted with a 330 PDA multiwavelength detector, a 230 solvent
delivery module, and a Beckmann Ultrasphere C18 column (4.6 ×
(method B, see Experimental Procedures). However, because
the effective specific activity was already 7.8 mCi/µmol and
because our future imaging studies involve mixing the tracer
with a large excess of insulin, further purification was deemed
unnecessary. The HPLC retention times of the final product
matched that of the 1 F reference standard. Typically, 37-74
1
50 mm, 300 Å-5 µm), or a semipreparative Microsorb Dynamax
C
18 column (10 × 250 mm, 300 Å-5 µm). The elution protocol
method A) consisted of a gradient over 20 min from 75/25 A/B
(v/v) to 20/80 A/B which was followed by 10/90 A/B for 10 min
(mobile phase A ) 0.1% TFA in H O, mobile phase B ) 0.1%
TFA in CH CN) with the flow rates set at 1.0 and 4.0 mL/min for
(
9
1
18
MBq (1-2 mCi) of B -([4- F-fluorobenzoyl]-6-aminohexanoyl)-
insulin (8b) was obtained in 180-200 min starting from a 2.78
2
3
1
8
-
18
-
analytical and semipreparative runs, respectively. The UV detector
was set at λ ) 254 and 210 nm.
GBq (75 mCi) of F [F ], giving a total yield, based on F ,
of 9 ( 5% yield (n ) 9, decay-corrected). This approach was
also amenable to producing significant quantities of the product
tracer. Higher production runs starting with 25.2 GBq (680 mCi)
yielded 470 MBq (12.7 mCi) of 8b (decay-corrected yield of
For preparative HPLC runs, a Varian ProStar preparative HPLC
system, which consisted of a model 320 uniwavelength detector, a
model 215 solvent delivery system, and a Microsorb Dynamax C18
column (21.4 × 250 mm, 300 Å-5 µm), was used. The elution
protocol (method B) consisted of a gradient from 75/25 A/B (v/v)
to 20/80 A/B over 20 min followed by 20/80 A/B (v/v) to 100% B
2
6
%).
To further verify that 18FB-AHx-insulin (8b) was in fact the
product, carrier-added experiments were performed by adding
over 5 min (mobile phase A ) 0.1% TFA in H O, mobile phase B
1
9 -
0
.1 mg of F to the Kryptofix solution prior to the addition
3
) 0.1% TFA in CH CN) at a flow rate of 30 mL/min. The
1
8 -
of F . The radioactivity was allowed to decay to background
and the reaction mixture characterized by MALDI-TOF mass
spectrometry. A peak was observed at a m/z value of 6043.8,
which is consistent with the mass spectrum of 8a prepared from
absorbance was monitored at λ ) 254 nm.
For the analysis of radioactive compounds, a Hewlett-Packard
1090 analytical system fitted with a diode array multiwavelength
detector, a model PV5 solvent delivery system, a NaI detector
connected to a Carroll and Ramsey Model 105S ratemeter, and a
Beckmann Ultrasphere C18 column (4.6 × 150 mm, 300 Å-5 µm)
were employed. The elution protocol (method C) consisted of a
gradient over 20 min consisting of 75/25 A/B (v/v) to 20/80 A/B
and then 10/90 A/B for 10 min (mobile phase A ) 0.1% TFA in
2
a.
Conclusion
A reproducible method for the preparation of a 18F-labeled
19
analogue of insulin was developed. The F version was screened
in a series of in vitro assays where it was indistinguishable from
2 3
H O, mobile phase B ) 0.1% TFA in CH CN) at a flow rate of
1.0 mL/min. The absorbance was monitored at λ ) 254 nm.
For purification of radioactive materials, a Waters 6000A solvent
delivery system fitted with Beckman UV detector (fixed wavelength
at 280 nm) and a Bicron GM tube (SWGM B980C) attached to a
Bicron ratemeter was employed. Using a semipreparative C18 Vydac
column (10 × 250 mm, 300 Å-5 µm), the elution protocol (method
native insulin. These studies indicate that 18FB-AHx-insulin is
a viable PET tracer for studying the distribution and biochem-
istry of insulin in vivo. The compound is currently being
evaluated in a series of PET studies using both small and large
animal models, which will be reported in due course.
D) was isocratic 50/50 A/B (v/v) (solvent A ) H O; solvent B )
2
3
CH CN) at a flow rate of 3.0 mL/min. The absorbance was
Experimental Procedures
monitored at λ ) 254 nm.
Materials and Instrumentation. Chemicals were purchased
from Aldrich Inc., NovaBiochem Inc., or Fluka Inc. and were used
without further purification. Human insulin was obtained from
LCMS experiments were run on a Waters 2690 separations
module, a model 996 photodiode array multiwavelength detector
and a Micromass Quatro Ultima electrospray ionization instrument.
For the digestion studies, a C18 Beckmann Ultrasphere (150 × 4.6
mm, porosity 5 µm) column was employed using a mobile phase
3
Aventis Inc. Anhydrous CH CN (Aldrich, 99.8%), anhydrous
DMSO (Aldrich, 99.9%), Kryptofix222 (K222: 4,7,13,16,21,24-
hexaoxa-1,10-diazabicyclo[8.8.8]hexacosane, Aldrich, 98%), KH-
2 2 3
of A: H O with 0.05% TFA and B: 90/10 (v/v) H O/CH CN with
18
18
CO
3
(British Drug Houses, 99.5%), and [ O]H
2
O (Isotec, 98% O)
0.05% TFA and a gradient of 99/1 A/B (v/v) for 10 min, 99/1 A/B
to 11/89 A/B for 140 min, 11/89 A/B to 99/1 A/B for 3 min and
99/1 A/B for 17 min at a flow rate of 0.1 mL/min.
were used where indicated. Saline solution (0.9%, USP grade) was
obtained from Baxter Inc. Dibasic sodium phosphate (J. T. Baker,
USP grade) solution (0.9%) was prepared immediately prior to use
MALDI-TOF experiments were carried out on a Micromass TOF
Spec2E instrument. Prior to each analysis, a calibration standard
was run that consisted of a mixture of 1 Fmol/µL substance P, 2
pmol/µL renin substrate tetradecapeptide, 2 pmol/µL adrenocorti-
cotropic hormone fragment 18-39, and 10 pmol/µL cytochrome
c. This was done in the positive ion reflection mode at 20 kV.
NMR spectra were predominately recorded using a Bruker
Avance (AV 200 MHz) instrument and spectra were referenced to
2
using a sterile container and sterile H O designated for human use.
Pentamethylbenzyl 4-(N,N,N-trimethylammonium trifluoromethane-
sulfonate)benzoate (9) was synthesized following literature proce-
dures.28
TLC was run on plates of silica gel 60-F254 (Merck). Nonradio-
active compounds were visualized by UV (λ ) 254 nm), while
radioactive compounds were detected using a BioScan TLC
analyzer. Chromatatron plates were made from a slurry of silica
the residual proton peaks in deuterated solvents (DMSO-d
2.50 ppm; CD Cl , δ ) 5.32 ppm; CD OD, δ ) 3.30 and 4.78
ppm) or TMS. Characterization of 8a was completed using a Bruker
6
, δ )
gel (60-PF254) containing gypsum (120 g) in deionized H
2
O (230
2
2
3
mL) and dried for 24 h at 110 °C.
1
9
Size-exclusion chromatography (SEC) was run using Sephadex
G-25 resin (Aldrich, Inc.), which was swollen in 100 mM NH
HCO for 24 h prior to use. The void volume was determined by
AV Ultra Shield Plus (700 MHz) spectrometer at 37 °C. F NMR
4
-
experiments were recorded using a Bruker Avance (DRX 500 MHz)
instrument and the chemical shifts referenced to CFCl . Abbrevia-
3
3
measuring the volume required to elute a 1 mL solution of Dextran
Blue. SEC HiTrap desalting cartridges (GE Healthcare) were
tions of multiplicity: q ) quartet, qu ) quintet.
Radiochemistry. Radiosyntheses were performed in a shielded
hot cell (7.5 cm lead) using remote manipulators for high level
experiments (Caution: 18F is radioactive and should be handled
in a licensed facility using the appropriate shielding). Specific
radioactivity was determined as follows: the area of the absorbance
activated with 20 mL of 100 mM NH
mediately prior to use. Following use, the cartridges were washed
with 20 mL of buffer, 20 mL of H O, and 20 mL of 80/20 (v/v)
O/EtOH. Resins were stored in 80/20 (v/v) H O/EtOH. Activation
4 3
HCO (dropwise) im-
2
H
2
2

1
472 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 4
Guenther et al.
peak corresponding to the radiolabeled product was measured on
the HPLC chromatogram and compared to a standard curve relating
mass to absorbance.
solution, and reisolated by centrifugation. The residue was collected
by filtration through a fritted funnel. Anal. HPLC (method A): t
R
+
) 12.2 min. MS m/z ES(+): found, 1586.8 [M + 4H ]/4, 2116.2
Radioisotope Production. No-carrier-added 18F-fluoride was
produced using a Siemens 11 MeV proton-only cyclotron (RDS
+
[M + 3H ]/3; calcd, 1586.9, 2115.5, respectively.
The precipitate was transferred to a vial and DMF (600 µL)
containing 20% piperidine (v/v) added. The reaction was allowed
to stir for 20 min, at which time the mixture was transferred to a
1
8
18
18
1
12) using an O(p,n) F reaction. The [ O]H
2
O was separated
1
8 -
from F by passing the mixture through an anion exchange column
-
(
BioRad, AG 11 A8 resin, 50-500 mesh, converted to HCO
3
centrifuge vial containing cold Et
induced by the addition of CH CN (1.5 mL). Isolation of the solid
followed the same procedure described above. The solid was
dissolved in 75/25 (v/v) H O/CH CN containing 0.1% TFA and
2
O (2 mL) and precipitation
1
8 -
form). The
F
was subsequently eluted from the column using
3
of 95/5 (v/v) CH
and 2 mg KHCO
was evaporated at 130 °C. The residue was redissolved in anhydrous
CH CN (0.5 mL) and evaporated to dryness. The drying procedure
was repeated a second time before use.
3 2
CN/H O (1 mL) solution containing 8 mg of K222
3
and collected in a 10-mL vial, and the solvent
2
3
the desired product purified using preparative HPLC (method B).
The collected fractions were concentrated by rotary evaporation
3
(water bath at 30°C) to remove the majority of the CH
by lyophilization overnight, which yielded a white solid. The solid
was subsequently desalted by SEC using a 100 mM NH HCO
3
CN followed
1
29
Synthetic Procedures. A ,B -Di(tert-butyloxycarbonyl)insulin
DBI, 1). A modification of the literature procedure reported by
Shai et al. was employed. A solution of di-tert-butyl dicarbonate
(
4
3
15
buffer as the eluent. Following lyophilization, compound 6 (29 mg,
69%) was obtained as a colorless solid. Anal. HPLC (method A):
(
0.188 g, 0.86 mmol), TEA (200 µL), and N-hydroxysuccinimide
+
(0.207 g, 1.80 mmol) was stirred in DMSO (500 µL) for 1 h. The
t
R
) 8.91 min. MS m/z ES(+): found, 1531.1 [M + 4H ]/4, 2041.7
+
solution was subsequently added dropwise over 2 h (20 gauge
needle, 1 drop every 30 s) via a 1-mL syringe to a rapidly stirring
solution of human insulin (2.0 g, 0.344 mmol) dissolved in DMSO
[M + 3H ]/3; calcd, 1531.3, 2041.3, respectively.
1
29
1
A ,B -Di(tert-butyloxycarbonyl)-B -([4-fluorobenzoyl]-6-ami-
19
nohexanoyl)insulin ( FB-AHx-DBI, 7). A solution of 6 (10.0 mg,
1.63 µmol) and succinimidyl 4-fluorobenzoate (2a) (3.87 mg, 16.3
µmol) in DMSO (100 µL) containing 5% TEA (v/v) was allowed
to stir at room temperature for 20 min. The reaction mixture was
(
40 mL) containing 5% TEA (v/v). The reaction mixture was
allowed to stir for 1 h and subsequently divided equally among
eight centrifuge vials, which contained 30 mL of cold Et O.
Precipitation was induced through the addition of 5-8 mL CH
CN. The solid was isolated by centrifugation (3200 rpm) at 5 °C.
The residue was washed twice with 50/50 (v/v) Et O/CH CN and
isolated by centrifugation. The solid was dissolved in a 75/25 (v/
v) H O/CH CN containing 0.1% TFA and the desired product
2
3
-
transferred to a centrifuge vial containing cold Et
whereupon precipitation was induced by the addition of CH
2
O (1 mL),
CN
3
(0.5 mL). Isolation of the solid was accomplished by centrifugation
at 5 °C (3200 rpm). Compound 6 was isolated by semipreparative
HPLC (method A) which, after lyophilization, resulted in a colorless
2
3
2
3
purified using preparative HPLC (method B). Fractions collected
were concentrated by rotary evaporation (water bath at 30°C) to
R
solid (5.7 mg, 56%). Anal. HPLC (method A): t ) 11.4 min. MS
+
+
m/z ES(+): found, 1561.6 [M + 4H ]/4, 2082.7 [M + 3H ]/3;
remove the majority of the CH
overnight, which yielded a white solid. The solid was subsequently
desalted by SEC using a 100 mM NH HCO buffer as the eluent.
Following lyophilization, compound 1 (1.49 g, 72%) was obtained
as a colorless solid. Anal. HPLC (method A): t ) 7.80 min. MS
3
CN followed by lyophilization
calcd, 1562.0, 2082.4, respectively.
1
19
B -([4-Fluorobenzoyl]-6-aminohexanoyl)insulin ( FB-AHx-
insulin, 8a). To a vial containing 7 (2.1 mg, 0.34 µmol) were added
anisole (5 µL, 46.0 µmol) and TFA (100 µL), and the reaction was
allowed to stir at room temperature for 10 min. The reaction mixture
4
3
R
+
+
m/z ES(+): found, 1202.4 [M + 5H ]/5, 1503.0 [M + 4H ]/4,
2
was transferred to a centrifuge vial containing cold Et O (1 mL)
+
2
004.0 [M + 3H ]/3; calcd, 1202.8, 1503.0, 2003.6, respectively.
and the resulting precipitate isolated by centrifugation at 5 °C (3200
rpm). The target compound was isolated by semipreparative HPLC
(method A), the organic solvent removed by gentle rotary evapora-
tion, and the resulting solution lyophilized overnight. The solid
Succinimidyl 9-Fluorenylmethoxycarbonyl-6-aminohexanoate
(5). To a solution of 9-fluorenylmethoxycarbonyl-6-aminohexanoic
acid (1.0 g, 2.83 mmol) in CH CN (500 mL) were added
3
N-hydroxysuccinimide (0.411 g, 3.57 mmol) and 1-(3-dimethyl-
aminopropyl)-3-ethylcarbodiimide hydrochloride (0.698 g, 3.57
mmol). The reaction was allowed to stir at room temperature for 2
h, at which time it was concentrated to dryness using a rotary
residue was desalted using SEC and a 100 mM NH
producing a colorless solid (1.5 mg, 74%) after lyophilization.
NMR (DMSO-d ): δ -109.60 (tt, 1F, J ) 8.9, 5.6 Hz); -109.58
4 3
HCO buffer,
1
9
F
6
(t, J ) 8.9 Hz). Anal. HPLC (method A): t ) 9.15 min. MS m/z
R
+
+
evaporator. The residue was dissolved in CH
extracted with H
O (2 × 20 mL). The organic layer was dried with
MgSO , filtered, and then concentrated to dryness using a rotary
evaporator. The residue was redissolved in CH Cl (1-2 mL) and
purified using a chromatotron and a solvent gradient consisting of
00% CH Cl to 90/10 (v/v) CH Cl /CH OH. The product (0.83
g, 65%) was a colorless solid. R (DCM/MeOH 95/5, v/v): 0.78.
Mp ) 110-113 °C. H NMR (CD Cl ): δ 7.78 (d, 2H, J ) 7.5
Hz); 7.62 (d, 2H, J ) 7.4 Hz); 7.44 (dd, 2H, J ) 7.5, 7.4 Hz); 7.32
dd, 2H, J ) 7.4, 7.4 Hz); 4.98 (br s, 1H); 4.37 (d, 2H, J ) 6.7
Hz); 4.22 (t, 1H, J ) 6.5 Hz); 3.17 (d, 2H, J ) 6.2 Hz); 2.79 (s,
2
Cl
2
(25 mL) and
ES(+): found, 1209.9 [M + 5H ]/5, 1512.1 [M + 4H ]/4, 2015.4
+
[M + 3H ]/3; calcd, 1209.6, 1511.7, 2015.3, respectively.
2
4
4 3
Digestion Studies. A total of 25 µL of 0.4 M NH HCO (aq)
2
2
was added to 250 µL of 8a (2.4 mg/mL) in PBS in a 1.5 mL conical
vial and the mixture agitated gently. To this was added 5 µL of 45
mM aqueous dithiothreitol and the mixture incubated for 15 min
at 50 °C. After cooling to room temperature, 5 µL of 100 mM
aqueous iodoacetamide was added and the solution agitated
periodically over 15 min. An aliquot (56 µL) was mixed with 28
µL of 0.4 M NH HCO in 8 M aqueous urea and the analysis
4 3
performed by LCMS (ES+) over 160 min.
The remaining solution from the dithiothreitol experiment was
transferred to a glass vial. To this was added 2.5 µL of endopro-
teinase-Glu-C in MilliQ water and the mixture incubated for 16 h
1
2
2
2
2
3
f
1
2
2
(
4
(
H); 2.61 (t, 2H, J ) 7.2 Hz); 1.76 (qu, 2H); 1.53 (qu, 2H); 1.44
m, 4H). 13C NMR (CD
Cl ): δ 169.87, 169.24, 156.80, 144.77,
41.85, 128.16, 127.55, 125.64, 120.45, 66.83, 47.93, 41.15, 31.39,
2
2
1
2
1
-
1
9.89, 26.26, 26.19, 24.81. IR (KBr, cm ): ν 3366.10 (N-H),
725.87 (CdO), 1693.77 (CdO), 1528.69 (CdC). MS m/z ES-
at 37 °C. An aliquot (56 µL) was mixed with 28 µL of 0.4 M NH
HCO in 8 M aqueous urea. Analysis was preformed by LCMS
(ES+).
4
-
3
+
(+): found, 451.3 [M + H ]; calcd, 451.5.
1
29
1
A ,B -Di(tert-butyloxycarbonyl)-B -(6-aminohexanoyl)insu-
Receptor Binding Assay. A human embryonic kidney (HEK-
lin (AHx-DBI, 6). A solution of DBI (1) (40.7 mg, 6.77 µmol)
and 5 (30.5 mg, 67.7 µmol) in DMSO (400 µL) containing 5%
TEA (v/v) was allowed to stir at room temperature for 30 min.
The reaction mixture was transferred to a centrifuge vial containing
293) cell line that that has been stably transfected to express human
-
10
insulin receptors was incubated in the presence of 1.4 × 10
M
125
of
I-insulin and varying concentrations of unlabeled human
insulin or FB-AHx-insulin. Cells were incubated in 50 mM PBS
with 0.5% BSA for 120 min at 4 °C and then washed three times.
The pelleted cells were counted on a γ-counter and results were
reported as percent binding (cpm sample/cpm added × 100).
cold Et
CN (1.5 mL). The resulting solid was isolated by centrifugation
3200 rpm) at 5 °C, washed twice with a 50/50 (v/v) Et O/CH CN
2 3
O (2 mL) and precipitation induced by the addition of CH -
(
2
3

EValuation of 18F- and 19F-Labeled Insulin
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 4 1473
For cellular autophosphorylation assays, CHO-hIR cells were
grown in R-MEM, 10% FBS with 10 µg/mL gentamicin, serum
deprived for 1 h, and then incubated with either human insulin
CN (3 mL) containing 0.1% TFA (v/v) through the cartridge. Anal.
HPLC (method C): t
The CH CN solution of 7b was concentrated to dryness by
R
) 12.1 min.
3
(
1
Aventis) or FB-AHx-insulin at the indicated concentrations for
0 min. Cells were then lysed in a buffer consisting of 50 mM
HEPES, pH 7.4, 150 mM NaCl, 1% NP-40, 0.1% deoxycholate,
.1% SDS, 1 mM Na VO , 0.1 mM EGTA, and 1 Roche EDTA-
passing a stream of nitrogen over the sample, which was heated to
30 °C for approximately 15-20 min. TFA (400 µL) containing
5% anisole (v/v) was then added and the mixture allowed to stir
for 5 min, at which time, the TFA was removed under a stream of
nitrogen (2-3 min). The residue was redissolved in EtOH (200
0
3
4
free Protease Inhibitor Cocktail Tablet (per 50 mL volume). Lysates
were clarified by centrifugation at 100 000g for 5 min at 4 °C and
applied to 96-well plates coated with anti-insulin receptor mono-
clonal antibodies (29B4 antibody). The extent of autophosphory-
lation was quantified using a second antibody directed against
phosphotyrosine residues (HRP-conjugated PY20, Oncogene Re-
search Products) and a coupled horseradish peroxidase reaction
µL) and 50/50 NaCl/Na
mixture was transferred to an activated SEC cartridge, which was
eluted with 2 mL 50/50 NaCl/Na HPO (0.9%, USP grade), and
2 4
HPO (800 µL, 0.9%, USP grade). The
2
4
the 3 mL of buffer was collected in one vial. The solution containing
8b (31-77% yield) was filtered through a Millipore frit prior to
use. Anal. HPLC (method C): t ) 10.7 min.
R
(
ELISA). 2-Deoxyglucose uptake in 3T3-L1 adipocytes in response
Acknowledgment. The authors would like to thank Xiuling
Guo, Deborah Finco-Kent, Stephen J. Orena and Anne M.
Brodeur for performing the biological assays, Pfizer Inc. and
the Ontario Research and Development Fund (ORDCF) for their
financial support of this work, and the Ministry of Human
Resources, Training and Universities of Ontario for an Ontario
Graduate Scholarship for K.J.G. We would also like to
acknowledge the Department of Nuclear Medicine, Hamilton
to human insulin or FB-AHx-insulin at the indicated concentrations
was determined as previously described.30
1
18
Radiosynthesis. B -([4- F-Fluorobenzoyl]-6-aminohexanoyl)-
18
18
3
insulin ( FB-AHx-insulin, F-8b). To an anhydrous CH CN (500
µL) solution of 1 F-fluoride containing 8 mg of K222 was added
8
pentamethylbenzyl 4-(N,N,N-trimethylammonium trifluoromethane-
sulfonate)benzoate (9)28 (3.0 mg, 6.13 µmol) dissolved in anhydrous
3
CH CN (500 µL). The vial was capped tightly and placed in a
18
heating block at 130 °C for 10 min with stirring. The reaction vessel
was cooled using an ice-water bath and the activity measured.
Health Sciences, for providing F and access to radiochemistry
facilities.
2
The resulting light yellow solution was diluted with Et O (3 mL)
and passed through an activated silica cartridge collecting 1 F-10
in a 10-mL vial. The cartridge was washed with Et O (3 mL) and
the activity combined with the initial eluent to give the desired
product in 67-81% yield. Anal. HPLC (method C): t ) 25.7 min.
SiO -TLC: CH CN/H O 95/5 (v/v), R ) 0.97.
The Et O solvent was removed by heating the solution to 90 °C
while passing a gentle stream of nitrogen over the reaction vial for
-6 min. The dried sample was subsequently placed in a water-
ice bath, neat TFA (200 µL) added, and the mixture stirred for 5
min. The solution was diluted with H O (5 mL) and passed through
an activated C18 cartridge. The cartridge was washed with H O (5
mL) and then dried for 1 min by applying a stream of nitrogen.
8
Supporting Information Available: HPLC data and NMR
spectra. This material is available free of charge via the Internet at
http://pubs.acs.org.
2
R
References
2
3
2
f
(
1) Doubrovin, M.; Serganova, I.; Mayer-Kuckuk, P.; Ponomarev, V.;
Blasberg, R. G. Multimodality in Vivo Molecular-Genetic Imaging.
Bioconjugate Chem. 2004, 15, 1376-1388.
2
4
(2) Cherry, S. R. In Vivo Molecular and Genomic Imaging: New
Challenges for Imaging Physics. Phys. Med. Biol. 2004, 49, R13-
R48.
2
(
3) Saltiel, A. R.; Kahn, C. R. Insulin Signalling and the Regulation of
Glucose and Lipid Metabolism. Nature 2001, 414, 799-806.
4) DeFronzo, R. A. Insulin Resistance: A Multifaceted Syndrome
Responsible for NIDDM, Obesity, Hypertension, Dyslipidaemia and
Atherosclerosis. Neth. J. Med. 1997, 50, 191-197.
2
(
18
18
18
4
- F-Fluorobenzoic acid ( FBA) ( F-11) was eluted from the Sep-
Pak using anhydrous CH CN (3 mL), which was collected in a
0-mL vial in 77-95% yield. Anal. HPLC (method C): t ) 7.89
min. SiO -TLC: CH CN/H O 95/5 (v/v), R ) 0.34.
-(3-Dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride
34 mg, 0.17 mmol) and N-hydroxysuccinimide (20 mg, 0.17 mmol)
was added to 11 in anhydrous CH CN (500 µL). The reaction was
3
(
5) Stratton, I. M.; Adler, A. I.; Neil, H. A. W.; Matthews, D. R.; Manley,
S. E.; Cull, C. A.; Hadden, D.; Turner, R.; Holman, R. R. Association
of Glycaemia With Macrovascular and Microvascular Complications
of Type 2 Diabetes (UKPDS 35): Prospective Observational Study.
Br. Med. J. 2000, 321, 405-412.
1
R
2
3
2
f
1
(
(
6) DCCT: The Effect of Intensive Treatment of Diabetes on the
Development and Progression of Long-Term Complications in
Insulin-Dependent Diabetes Mellitus. New Engl. J. Med. 1993, 329,
3
stirred at room temperature for 10 min, whereupon it was
concentrated to approximately a 0.5-1 mL volume by heating the
solution at 90 °C for 6-8 min under a gentle stream of nitrogen.
9
77-986.
(
(
(
7) Paternostro, G.; Pagano, D.; Gnecchi-Ruscone, T.; Bonser, R. S.;
Camici, P. G. Insulin Resistance in Patients with Cardiac Hypertro-
phy. CardioVacular Res. 1999, 42, 246-253.
8) Hoyer, S. Is sporadic Alzheimer Disease the Brain Type of Non-
Insulin Dependent Diabetes Mellitus? A Challenging Hypothesis. J.
Neural Transm. 1998, 105, 415-422.
9) Kaaks, R.; Toniolo, P.; Akhmedkhanov, A.; Lukanova, A.; Biessy,
C.; Dechaud, H.; Rinaldi, S.; Zeleniuch-Jacquotte, A.; Shore, R. E.;
Riboli, E. Serum C-Peptide, Insulin-like Growth Factor (IGF)-I, IGF-
Binding Proteins and Colorectal Cancer Risk in Women. J. Natl.
Cancer Inst. 2000, 92, 1592-1600.
To this sample was added enough H
.5 mL. The solution was purified by semipreparative HPLC
method D), collecting the peak corresponding to the succinimidyl
2
O to make a total volume of
1
(
4
18
18
- F-fluorobenzoate (4- FSB). The HPLC eluent was transferred
to a 20-mL syringe containing 14 mL of H
was passed through an activated C18 cartridge. After drying the
cartridge under a stream of nitrogen (1 min), pure 2b was collected
2
O, and the entire volume
(
71-91% yield) in a 10-mL vial by flushing CH
the cartridge. Anal. HPLC (method C): t ) 10.53 min. SiO
TLC: CH CN/H O 95/5 (v/v), R ) 0.88.
3
CN (3 mL) through
R
2
-
(
10) Sodoyez, J. C.; Sodoyez-Goffaux, F.; Goff, M. M.; Zimmerman, A.
E.; Arquilla, E. R. Preparation, Physical, Immunological and Biologi-
cal Properties and Susceptibility to “Insulinase” Degradation. J. Biol.
Chem. 1975, 250, 4268-4277.
11) Wajchenberg, B. L.; Pinto, H.; de Toledo e Souza, I. T.; Ler a´ rio, A.
C.; Pieroni, R. R. Preparation of Iodine-125-Labeled Insulin for
Radioimmunoassay: Comparison of Lactoperoxidase and Chloram-
ine-T Iodination. J. Nucl. Med. 1978, 19, 900-905.
3
2
f
The solution containing 2b was concentrated to dryness at 90
C under a slow stream of nitrogen (6-8 min). To the dried sample
°
(
was added 6 (2 mg, 0.16 µmol) dissolved in DMSO (200 µL)
containing 10% TEA (v/v). The reaction was allowed to stir at room
temperature for 20 min and subsequently diluted with 100 mM NH
HCO
SEC Sep-Pak, which was washed with 2 mL of 100 mM NH
HCO
mL of 100 mM NH
This fraction was loaded onto a C18 cartridge, which was washed
with H O (5 mL) and dried for 1 min using a stream of nitrogen.
The desired product was isolated in 32-53% yield by passing CH
4
-
3
(1 mL). The reaction mixture was loaded onto an activated
(12) Jørgensen, K. H.; Larsen, U. D. Homogeneous Mono-[125I]-Insulins:
Preparation and Characterization of Mono-[125I]-(Tyr A14)- and
Mono-[125I]-(Tyr A19)-Insulin. Diabetologia 1980, 19, 546-554.
4
-
3
. The product was subsequently eluted using an additional 3
HCO , which was collected in a 10-mL vial.
(13) Rissler, K.; Engelmann, P. Labeling of Insulin with Non-Radioactive
4
3
1
27I and Application to Incorporation of Radioactive 125I for Use
in Receptor-Binding Experiments by High-Performance Liquid
2
Chromatography. J. Chromatogr. BsBiomed. Appl. 1996, 679, 21-
29.
3
-

1
474 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 4
Guenther et al.
(
14) Owens, D. R.; Coates, P. A.; Luzio, S. D.; Tinbergen, J. P.; Kurzhals,
R. Pharmacokinetics of 125I-Labeled Insulin Glargine (HOE 901)
in Healthy Men: Comparison With NPH Insulin and the Influence
of Different Subcutaneous Injection Sites. Diabetes Care 2000, 23,
(23) Blundell, T. L.; Dodson, G. G.; Hodgkin, D. C.; Mercola, D.
Insulin: The Structure in the Crystal and Its Reflection in Chemistry
and Biology. AdV. Protein Chem. 1972, 26, 279-402.
(24) Baudy sˇ , M.; Uchio, T.; Mix, D.; Wilson, D.; Kim, S. W. Physical
813-819.
Stabilization of Insulin by Glycosylation. J. Pharm. Sci. 1995, 84,
(
(
(
15) Shai, Y.; Kirk, K. L.; Channing, M. A.; Dunn, B. B.; Lesniak, M.
28-33.
A.; Eastman, R. C.; Finn, R. D.; Roth, J.; Jacobson, K. A.
(
25) Lundt, B. F.; Johansen, N. L.; Voelund, A.; Markussen, J. Removal
of tert-Butyl and t-Butoxycarbonyl Protecting Groups with Trifluo-
roacetic Acid. Mechanisms, Byproduct Formation and Evaluation of
Scavengers. Int. J. Pept. Protein Res. 1978, 12, 258-268.
1
8F-Labeled Insulin: A Prosthetic Group Methodology for Incor-
poration of a Positron Emitter into Peptides and Proteins. Biochem-
istry 1989, 28, 4801-4806.
16) Eastman, R. C.; Carson, R. E.; Jacobson, K. A.; Shai, Y.; Channing,
M. A.; Dunn, B. B.; Bacher, J. D.; Baas, E.; Jones, E.; Kirk, K. L.;
Lesniak, M. A.; Roth, J. In Vivo Imaging of Insulin Receptors in
Monkeys Using 18F-Labeled Insulin and Positron Emission Tomo-
raphy. Diabetes 1992, 41, 855-860.
17) Glaser, M.; Brown, D. J.; Law, M. P.; Iozzo, P.; Waters, S. L.; Poole,
K.; Knickmeier, M.; Camici, P. G.; Pike, V. W. Preparation of No-
Carrier-Added [124I]A14-Iodoinsulin as a Radiotracer for Positron
Emission Tomography. J. Labelled Compd. Radiopharm. 2001, 44,
(26) Weiss, M. A.; Nguyen, D. T.; Khait, I.; Inouye, K.; Frank, B. H.;
Beckage, M.; O’Shea, E.; Shoelson, S. E.; Karplus, M.; Neuringer,
L. J. Two-Dimensional NMR and Photo-CIDNP Studies of the Insulin
Monomer: Assignment of Aromatic Resonances with Application
to Protein Folding, Structure, and Dynamics. Biochemistry 1989, 28,
9855-9873.
(
27) Matsudaira, P. T. A Practical Guide to Protein and Peptide
Purification for Microsequencing; 2nd ed.; Academic Press: San
Diego, CA, 1993; pp 55-59.
465-480.
(
18) Iozzo, P.; Osman, S.; Glaser, M.; Knickmeier, M.; Ferrannini, E.;
Pike, V. W.; Camici, P. G.; Law, M. P. In Vivo Imaging of Insulin
Receptors by PET: Preclinical Evaluation of Iodine-125 and Iodine-
(28) Lang, L.; Jagoda, E. M.; Schmall, B.; Vuong, B.-K.; Adams, R. H.;
Nelson, D. L.; Carlson, R., E.; Eckelman, W. C. Development of
Fluorine-18-Labeled 5-HT1A Antagonists. J. Med. Chem. 1999, 42,
1
24 Labelled Human Insulin. Nucl. Med. Biol. 2002, 29, 73-82.
(
19) Pullen, R. A.; Lindsay, D. G.; Wood, S. P.; Tickle, I. J.; Blundell,
T. L.; Wollmer, A.; Krail, G.; Brandenburg, D.; Zahn, H.; Gliemann,
J.; Gammeltoft, S. Receptor-Binding Region of Insulin. Nature 1976,
1576-1586.
(
29) Fredriksson, A.; Johnstr o¨ m, P.; Stone-Elander, S.; Jonasson, P.;
Nygren, P.- A° .; Ekberg, K.; Johansson, B.-L.; Wahren, J. Labeling
of Human C-Peptide by Conjugation With N-Succinimidyl-4-[18F]-
fluorobenzoate. J. Labelled Compd. Radiopharm. 2001, 44, 509-
259, 369-373.
(20) Gliemann, J.; Gammeltoft, S. The Biological Activity and the Binding
Affinity of Modified Insulins Determined on Isolated Rat Fat Cells.
Diabetologia 1974, 10, 105-113.
(
(
519.
21) Hashimoto, M.; Takada, K.; Kiso, Y.; Muranishi, S. Synthesis of
Palmitoyl Derivatives of Insulin and Their Biological Activities.
Pharm. Res. 1989, 6, 171-176.
22) Murray-Rust, J.; McLeod, A. N.; Blundell, T. L.; Wood, S. P.
Structure and Evolution of Insulins: Implications for Receptor
Binding. BioEssays 1992, 14, 325-331.
(30) Orena, S. J.; Torchia, A. J.; Garofalo, R. S. Inhibition of Glycogen
Synthase Kinase-3 Stimulates Glycogen Synthase and Glucose
Transport by Distinct Mechanisms in 3T3-L1 Adipocytes. J. Biol.
Chem. 2000, 275, 15765-15772.
JM0509344