Rapid, one-step, high yielding 18F-labeling of an aryltrifluoroborate bioconjugate by isotope exchange at very high specific activity

Article abstract of DOI:10.1002/jlcr.2990

A rapid, single step, aqueous ¹⁸F-labeling method that proceeds under mild conditions to provide radiotracers in high radiochemical yield and at high specific activity represents a long-standing challenge. Arylboronates capture aqueous ¹⁸F-fluoride ion in buffered pH 2-3 at moderate temperature to provide a highly polar ¹⁸F-ArBF₃ ^- anion. Similarly, ¹⁹F-¹⁸F isotope exchange on a ¹⁹F-ArBF₃^- should create an ¹⁸F-ArBF₃^-. We hypothesized that this reaction would proceed in volumes that would be amenable to the high levels of ¹⁸F-activity used in clinical hospitals. In order to measure both radiochemical and chemical yields, along with specific activity, we linked an alkyne-¹⁹F-ArBF₃^- to rhodamine azide by standard click chemistry to afford a precursor Rh-¹⁹F-ArBF ₃^-. This precursor was aliquoted in portions of 50 nmol and lyophilized for on-demand use. Using robotic manipulators in a hot cell, we combined >29.6 GBq (800 mCi) and 50 nmol of the Rh-¹⁹F-ArBF ₃^- in aqueous dimethylformamide at buffered pH 2-3. Following mild heating (40 °C) for 10-15 min, the reaction was quenched and analyzed. We observed radiochemical yields of 50% and specific activities of nearly 555 GBq/μmol (15 Ci/μmol). Similar radiochemical yields and slightly lower specific activities were also obtained with ~400 mCi (n = 2). With radiochemical yields in the hundreds of millicuries and specific activities that are 3-10-fold higher than most radiotracers, this method is very attractive method for preparing clinically useful radiotracers. Moreover, the ability to produce tracers at extraordinarily high specific activities expands the distribution time window from production labs to distant positron emission tomography scanners. Copyright 2012 John Wiley & Sons, Ltd. The specific activity of most ¹⁸F-radiotracers is limited to ~6 Ci/μmol while labeling is often cumbersome due fluoride's lack of reactivity in water. With 800 mCi, an aryltrifluoroborate is ¹⁸F-labeled by isotope exchange at 30-40° C, buffered pH 2-3 in yields of 50% at specific activities of ~15 Ci/μmol in 10-15 minutes. To measure such high specific activity we took advantage of a pendant Rhodamine which now provides for ¹⁸F-labeled fluorophores at extraordinarily high specific activity in one step. Copyright

Full text of DOI:10.1002/jlcr.2990

Research Article
Received 2 October 2012,
Revised 31 October 2012,
Accepted 1 November 2012
Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/jlcr.2990
18
Rapid, one-step, high yielding F-labeling of
an aryltrifluoroborate bioconjugate by isotope
exchange at very high specific activity†
Zhibo Liu,a Ying Li,a Jerome Lozada,a Jinhe Pan,b Kuo-Shyan Lin,b
c
a
*
Paul Schaffer, and David M. Perrin
A rapid, single step, aqueous 18F-labeling method that proceeds under mild conditions to provide radiotracers in high radioche-
mical yield and at high specific activity represents a long-standing challenge. Arylboronates capture aqueous 18F-fluoride ion in
buffered pH 2–3 at moderate temperature to provide a highly polar 18F-ArBF3À anion. Similarly, 19F-18F isotope exchange on a
19F-ArBF3À should create an 18F-ArBFÀ3 . We hypothesized that this reaction would proceed in volumes that would be amenable
to the high levels of 18F-activity used in clinical hospitals. In order to measure both radiochemical and chemical yields, along
with specific activity, we linked an alkyne-19F-ArBF3À to rhodamine azide by standard click chemistry to afford a precursor
Rh-19F-ArBFÀ3 . This precursor was aliquoted in portions of 50nmol and lyophilized for on-demand use. Using robotic manipula-
tors in a hot cell, we combined >29.6GBq (800mCi) and 50nmol of the Rh-19F-ArBF3À in aqueous dimethylformamide at buffered
pH 2–3. Following mild heating (40 ꢀC) for 10-15 min, the reaction was quenched and analyzed. We observed radiochemical
yields of 50% and specific activities of nearly 555 GBq/mmol (15 Ci/mmol). Similar radiochemical yields and slightly lower specific
activities were also obtained with ~400 mCi (n = 2). With radiochemical yields in the hundreds of millicuries and specific activities
that are 3–10-fold higher than most radiotracers, this method is very attractive method for preparing clinically useful radiotra-
cers. Moreover, the ability to produce tracers at extraordinarily high specific activities expands the distribution time window
from production labs to distant positron emission tomography scanners.
Keywords: one-step 18F-labeling; fluorescent radiotracers; high specific activity; bioconjugates
specific activity of 37 GBq/mmol (1 Ci/mmol) that was also
Introduction
described as ‘high’.9 These reports underscore the need for
Positron emission tomography imaging provides some of the
radiosynthetic methods whereby labeling occurs rapidly, under
aqueous conditions, and in a single step.15,18–25 Toward these
highest spatio-temporal in vivo resolution of deep-tissue localiza-
tion of targets for validation, preclinical evaluation, and patient
ends, several new captors of aqueous fluoride ion based on alumi-
diagnosis.1 Although isotope choice is based on various considera-
num, silicon, and boron may expedite a one-step aqueous labeling,
as recently reviewed.22–25
tions, the main deciding factors are ease of labeling, isotope avail-
ability, specific activity, and half-life. Of various b+-radionuclides,
Previously, we proposed that arylboronates would capture
aqueous 18F-fluoride ion to provide an 18F-ArBF3À.26 ArBF-3 salts,
fluorine-18, with its high isotopic purity, single decay process,
low b+ emission energy, moderate half-life, and facile on-demand
which are considered to be non-toxic, are synthesized from
potassium 18F-fluoride under mildly acidic conditions according
production by hospital cyclotrons accounts for its widespread
to Figure 1.27–29 Labeling requires only microgram quantities
demand.
18F-fluoride ion is produced by bombardment 18O(p,n)18F
of arylboronate precursor and proceeds rapidly with mild heat-
ing under mildly acidic conditions, that is, pH 2–3. Only the
from 18OH2. Whereas carrier-free fluoride ion has a specific
activity of 63.6 TBq/mmol (1720 Ci/mmol), in practice, the speci-
fic activity of no carrier added (NCA) 18F-fluoride ion falls to
925–1320 GBq/mmol (25–40 Ci/mmol).2,3 Drying and anion
aChemistry Department, 2036 Main Mall, University of British Columbia, Vancouver,
B.C., V6T-1Z1, Canada
exchange trapping leads to further reduction of specific activ-
ity and hence small molecule tracers4,5 have been labeled at
bBC Cancer Agency - Vancouver Centre, Centre for Functional Imaging, 600 West
10th Avenue, Vancouver, B.C. V5Z-4E6, Canada
185–296 GBq/mmol (5–8 Ci/mmol), whereas most are labeled
at even lower specific activities.6–9 Similarly, the specific
activities of most 18F-labeled peptides fall below 74GBq/mmol
(2Ci/mmol).8,10–15 Nevertheless, a value of 37–74 GBq/mmol
(1–2 Ci/mmol), is commonly described as ‘high’ and therefore
is considered very useful for imaging.16,17 For instance, Ritter
and coworkers used 37 GBq (1 Ci) of NCA 18F-fluoride ion to
produce radiotracers in radiochemical yields of 6–30% at a
cTriumf 4004 Wesbrook Mall, Vancouver, B.C. V6T-2A3, Canada
* Correspondence to: David M. Perrin, Chemistry Department, 2036 Main Mall,
University of British Columbia, Vancouver, B.C., V6T-1Z1, Canada.
E-mail: dperrin@chem.ubc.ca
†Supporting information may be found in the online version of this article.
J. Label Compd. Radiopharm 2012, 55 491–496
Copyright © 2012 John Wiley & Sons, Ltd.

Z. Liu et al.
isotope exchange (either directly or through re-equilibration
in situ with the arylboronate) to provide an 18F-ArBF3À conjugate
in high radiochemical and chemical yield and at high specific
activity thereby greatly simplifying radiosynthesis. Such would
be analogous to 19F-di-tertbutylphenylfluorosilane-bioconjugates
designed to undergo isotope exchange.36 During the preparation
of this manuscript, a recent study showed demonstrated in vivo
stability of a phosphonium-stabilized 19F-ArBF3À would undergo
19F-18F isotope exchange although the specific activities reported
were 18.5GBq/mmol (0.5Ci/mmol) or lower.37
Here, we sought very high specific activity with good radioche-
mical yields in conjunction with production levels of 18F-activity in
fully shielded hot cells. In order to measure the very high specific
activity anticipated from using high levels of 18F-activity, we labeled
a conjugate containing a suitable chromophore that would enable
accurate mass measurement from the HPLC trace, from which
specific activity would be determined. Hence, we conjugated an
19F-ArBFÀ3 to rhodamine, (lmax = 569 nm, e = 100,000 MÀ1 cmÀ1).
In keeping with prior conjugation chemistries where we linked
the arylboron moiety to the RGD peptide for tumor visualization,
rhodamine was first conjugated to the nonradioactive alkyne-19F-
ArBFÀ3 via a copper mediated 2 + 3 cycloaddition ‘click’ reaction
to give the radiosynthetic precursor shown in Figure 3.
Figure 1. General labeling of an arylboronic acid (Y=O) or arylborimidine (Y=NH)
where L is a biomolecule or clickable handle such as an alkyne.
18F-ArBFÀ3 is stable at physiological pH; all other monofluorinated
and difluorinated species solvolyze within seconds once the
reaction is quenched at pH 7.5.30
Synthesis, fluorescent screening, 19F-NMR kinetics, HRMS,
pulse-chase-TLC, and serum stability assays verified physiological
stability of certain aryltrifluoroborates,30 whereas 18F-ArBFÀ3
labeled conjugates of marimastat,20,31 biotin, and32 Lympho-
seek™33 demonstrated in vivo stability with little if any signal in
bone, the appearance of which would have suggested solvoly-
tic/metabolic defluorination. Recently, we labeled peptidic RGD
conjugates either in a single step or in a one-pot-two-step click
labeling of an RGD azide with an alkyne-modified 18F-ArBF3À as
shown in Figure 2 (Li et al., manuscript in press). In both cases,
good tumor images were obtained despite the rather low
specific activity of the RGD-conjugates (2.2–5.9 GBq/mmol or
0.06–0.16 Ci/mmol).
This conjugate was then subjected to a 19F-18F isotope exchange
reaction, either directly or through re-equilibration, for the rapid
production of a fluorescent 18F-ArBFÀ3 -bioconjugate at very high
specific activity. The salient advantages embodied in this labeling
method are the following:(i) rapid synthesis time: <15 min; (ii)
excellent radiochemical yields ~50%; (iii) the use of aqueous condi-
tions; (iv) the use of high levels of radiation common to production
labs; (v) a kit-like approach that uses only 50 nmol of lyophilized
precursor; and (vi) the production of a dual-modal 18F-labeled
fluorophore at very high specific activity.38–40
Figure 2. Two 18F-labeled RGD species that were imaged; top—RGD-SuPi-18F-
ArBFÀ3 was directly labeled from the arylboronates precursor by treatment with
K
18F in acidic conditions; bottom—an alkyne-18F-ArBF3À was first synthesized
from the alkyne-arylborimidine and then conjugated to an RGD azide by standard
Cu+-catalyzed click conjugation.
Despite these in vivo images, a significant limitation is that
18F-ArBFÀ3 conjugates have been labeled at low specific activity
~7.4 GBq/mmol (~0.2 Ci/mmol). We and others have argued that
low specific activities reflect the use of low levels of starting
18F-activity, (370–1850 MBq or 10–50 mCi) and that use of high
levels of 18F-activity routinely produced by hospital cyclotrons
would provide much higher specific activities.9,34,35 Yet, the use
of high levels of 18F-activity requires robotic manipulation in
shielded hot cells, which might not accommodate the low reaction
volumes needed for good radiochemical yields. Encouraged by
in vivo tumor images gleaned from two differently linked
RGD-18F-ArBF3À conjugates, we hypothesized that a 19F-ArBF3À
bioconjugate, prepared in advance, would undergo facile 19F-18F
Figure 3. Synthetic scheme for click conjugation of the unlabeled alkyne-19F-ArB
FÀ3 to an azide-rhodamine to give the Rh-ArBFÀ3 precursor that undergoes isotope
exchange.
Experimental
Materials
Chemicals were purchased from Sigma-Aldrich unless stated otherwise.
The azidotriethyleneglycol-bis-sulfonyl rhodamine was purchased from
Click Chemistry Tools Inc. as predominantly one of two regio-isomers.
www.jlcr.org
Copyright © 2012 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2012, 55 491–496

Z. Liu et al.
Deuterated solvents were purchased from Cambridge Isotope Labora-
tories. TLC analysis and preparation were performed by Silica Gel
60 F254 glass TLC plates from EMD Chemicals.
(pH 2). This solution was transferred via syringe to the tube containing
the NCA 18F-fluoride ion. The tube was then placed in a heating block
set at 40 ꢀC. After 10–15 min, the reaction was removed from the heat-
ing block and quenched by the addition of 2 mL 5% NH4OH in 50%
MeCN/H2O via syringe to resuspend the contents of the labeling
reaction. From this resuspension, a small volume 50 mL was removed
from the hot cell, counted, and injected onto an RPC18 analytical HPLC
column for analysis.
HPLC methods
Unless otherwise stated, all samples were loaded onto a Phenomenex Jupi-
ter 10 m C18 300Ǻ 4.6 Â 250 mm column and material was eluted using a
step gradient (solvent A: 0.04M ammonium formate pH 6.8; solvent
B: MeCN); 0–5 min: 0–5% B, 5–10min: 5–35% B, 10–20min, 35–45% B,
20–22 min: 45–100% B, 22–28 min: 100–100% B, 28–30min: 100–20%
B, 30–33 min: 20–5% B; flow rate: 1 mL/min, column temperature: 19–21 ꢀC.
Synthesis of alkyne-19F-ArBF3À
Results and discussion
Briefly, 1,3,5-trifluoro-2-carboxyphenyl-4-boronic acid was prepared
according to previous reports.20,41 Instead of converting the boronate
to the tretaphenylpinacolate, a diaminonaphthalene (dan) group was
installed according to previous reports.42–44 The carboxylate was
activated in the presence of EDC according to previous reports and
condensed with propargyl amine and purified by flash chromatography.
Following previous reports for the production of the 1,3,5-trifluoro-4-
ArBFÀ3 ,32 (50 nmol) was dissolved in 5 mL DMF to which was added 2 mL
1 M pyridazine-HCl buffer in 50% DMF/H2O, and 1 mL 100 mM KHF2
aqueous solution were added to make a cocktail. The solvent was
removed under vacuum in speed-vac under 45 ꢀC. Generally, this proce-
dure took 15 min. And the alkyne-ArBFÀ3 has been previously reported in
which we have disclosed the 1H- and 19F-NMR spectra (Li et al. AJNMMI in
press), here, the residue was directly used without further purification.
The alkyne-19F-ArBFÀ3 was prepared as previously described
(Li et al., AJNMMI in press) and reacted with commercially
available azide-triethyleneglycolyl-rhodamine bis-sulfonate in
the presence of Cu2+ and ascorbate to provide the Rh-19F-
ArBFÀ3 bioconjugate shown in Figure 3. Following TLC and
HPLC purification, the conjugate was lyophilized in aliquots
of 50 nmol. For labeling, 816.5 mCi NCA 18F-fluoride ion was
concentrated at 100 ꢀC for 10 min in a polypropylene tube.
An aliquot of Rh-19F-ArBF3À (50 nmol) was resuspended in a
DMF-water solution (50 mL) containing a pyridazine-HCl buffer
(pH = 2). This solution was transferred by syringe to the plastic
tube containing the NCA 18F-fluoride ion. After 10–15 min at
40 ꢀC, the reaction was quenched by the addition of 2 mL 5%
NH4OH in 50% MeCN/H2O that served to resuspend the entire
reaction contents. About 2 h following quench, a small portion
of the crude reaction (4.1 mCi) was removed and injected onto
the HPLC.
Figure 4 shows the radioactive and corresponding UV-vis
HPLC traces of the crude labeling reaction and leads to sev-
eral important observations as follows: (i) there are essentially
only two radiopeaks, one corresponding to free 18F-fluoride
ion and one corresponding to the desired Rh-18F-ArBF3À; (ii)
the radiochemical yield, decay corrected to the end of
synthesis, appears to be 75% (63% not decay corrected),
which implies the production of >500 mCi of radiotracer
starting with 816 mCi although isolated radiochemical yields
are slightly lower, that is, 40–50%; (iii) there is a ~50% chemi-
cal radiochemical yield as the UV trace shows a second, well-
separated fluorescent product that elutes at 28.5 min and
which is the arylboronic acid whose formation either
competes with isotope exchange or occurs through an in situ
re-equilibration between the arylboronic acid, the ArBFÀ3 , and
free fluoride.
Synthesis of rhodamine-19F-ArBF3À
Approximately 250 nmol of alkyne-19F-ArBF3À was resuspended in 8 mL
5% NH4OH in 50% MeCN/H2O. A 5 mL MeCN solution that contained
500 nmol rhodamine azide, 250 nmol CuSO4, and 600 nmol sodium
ascorbate was added and the reaction was placed in a speed-vac at room
temperature. After ~10 min, a dark reddish solid remained. The residue
was resuspended with 20 mL 50% MeCN/H2O then applied to a TLC plate,
which was developed with 15% methanol in dichloromethane. The TLC
band, corresponding to the desired product was excised and extracted
into methanol. NOTE: If the plate was not completely dry prior to resolu-
tion, Rf values were quite variable. An aliquot of the TLC purified solution
was injected into HPLC to verify purity and elution time. Because of the
small amount of material in this case and the commercial availability of
the azido-rhodamine for use in copper catalyzed 2 + 3 cycloaddition,
the 1H-NMR and 19F-NMR spectra were not acquired following click
conjugation. Instead, TLC purification followed by HPLC purification along
with low-res MS and UV-vis spectra were considered sufficient to confirm
the purity of this conjugate: ESI-MS (M-H) = 1038; lmax = 568–569 nm.
NCA 18F-fluoride ion preparation
Encouraged by these high radiochemical yields, we next
sought to measure specific activity. To do this, we relied
on the excellent absorbance properties of the rhodamine
chromophore. Although peptides generally do not have chro-
mophores that afford reliable measurement of the low
concentrations characteristic of high specific activity, this
rhodamine bioconjugate has a reliable UV-vis signature and
a very high molar e. In measuring and reporting the specific
activity, we took into account two important considerations,
(i) there was an unnecessary lag between the time the reac-
tion was quenched and when it was injected on the HPLC that
was due to instrument availability in a shared hospital facility
and (ii) there was a 25-min lag between the time the sample
was loaded on the HPLC and when the radiotracer eluted.
About >29.6 GBq (>800 mCi) of 18F-fluoride ion was prepared by 18O-
water bombardment with 12–14 MeV protons. A QMA cartridge was
pretreated with (i) deionized (DI) water (6 mL); (ii) saturated brine
(6mL); and (iii) DI water (6mL) whereupon the irradiated water containing
the NCA 18F-fluoride ion was passed from the target over the cartridge. The
NCA 18F-fluoride ion was eluted with 0.25 mL, 8 mg/mL NaClO4 in water
was used to elute the radioactivity into a well-sealed plastic polypropylene
vial. Approximately 0.5mL of MeCN was added and the fluoride ion was
azeotropically concentrated at 100ꢀC for 10 min. This step served to simply
concentrate rather than dry the 18F-fluoride ion, which appeared to be a
pasty, damp, translucent residue in the tube.
Isotope exchange reaction
Just prior to labeling, 50 nmol 19F-Rhodamine-ArBFÀ3 was resuspended
in 20 mL DMF and 20 mL water and 10 mL containing 2 M pyridazine-HCl To report specific activity values in a meaningful way, we
J. Label Compd. Radiopharm 2012, 55 491–496
Copyright © 2012 John Wiley & Sons, Ltd.
www.jlcr.org

Z. Liu et al.
Time (min)
Time (min)
Figure 4. HPLC trace of the crude reaction; left—radiotrace where the y-axis is in mV; right—a UV-vis trace at 569 nm where the y-axis is in mAu.
corrected for the unnecessary decay between quench and
In order to measure specific activity, we conjugated an alky-
HPLC injection but did not correct for the HPLC purification ne-19F-ArBF3À to rhodamine azide in order to provide a surrogate
time that necessarily erodes specific activity. Hence, specific bioconjugate whose chromophore unequivocally afforded a mea-
activities are reported at time of collection (TOC) and not sure of low concentrations. These results now suggest that this
back-corrected to EOS, which would otherwise reflect even method will find use in labeling a fluorescent peptide to create
higher specific activities.
bimodal imaging agents in which there is considerable inter-
For the radiosynthesis featured in Figure 4, a small portion of est.33,41,45 Furthermore, because we previously labeled RGD, both
the crude reaction that contained 4.1 mCi was loaded on the by direct conversion of an RGD-boronate to an RGD-18F-ArBF3À
HPLC, from which was eluted 81.4 MBq (~2.2 mCi) (53% and by preparing an alkyne-18F-ArBF3À followed by click conjuga-
isolated), 25–26 min later. Integration of the visible peak at tion onto an RGD azide, chemical logic suggests this method of
569 nm that eluted at 25.5 min provided a quantitative mass isotope exchange is applicable to many other peptides.
value of 330 pmol and therefore gave a specific activity of
Unlike previously reported ArBFÀs syntheses, which require rela-
248 GBq/mmol (6.7 Ci/mmol) at TOC. Correcting for the 2 h lag tively high concentrations of total fluoride ion, necessitating low
time between quenching and loading, the real specific activity volumes and/or the addition of excess 19F-fluoride ion, the isotope
at TOC was an extraordinary 529 GBq/mmol (14.3 Ci/mmol). On exchange reaction herein uses no added carrier 19F-fluoride ion
the basis of the assumption that only one atom of fluoride is and only very low concentrations of carrier 19F-ArBF3À (1mM).
exchanged, the calculated specific activity of 592 GBq/mmol Because isotope exchange is likely to be pseudo-first order in
(16 Ci/mmol, that is, 0.8 Ci/50 nmol ArBFÀ3 ), corrected to begin- 19F-ArBFÀ3 concentration, radiochemical yields are independent
ning of synthesis is consistent with the specific activity that of the amount of activity used. This labeling has been repeated
was measured. Gratifyingly, this reaction was repeated using numerous times at low specific activity, and herein, three times
17.9 (485 mCi) and 15.2 GBq (412 mCi) in radiochemical yields with more than 14.8 GBq (400mCi) of 18F-activity. In each case,
of 50% and 45%, respectively, and specific activities in the range reaction in as little as 10–15min provided radiochemical yields of
of 111–259 GBq/mmol (3–7 Ci/mmol), corrected to TOC (see 50% or higher.
supporting information). These slightly lower specific activities
are consistent with the use of commensurately lower levels of most other radiotracers, including clickable radiosynthons on the
18F-activity. basis of 18F-SiFA conjugates and 18F-Al-NOTA chelates that have
Notably, the specific activities herein are higher than those of
Whereas separation herein gave radiochemically pure mate- been labeled at specific activities of ≤259 GBq/mmol (≤7 Ci/mmol)
rial at very high specific activity, the separation of labeled prior to click.46,47 Although we have argued that specific activ-
bioconjugates from other unlabeled products may not be ities as low as 3.7 GBq/mmol (0.1 Ci/mmol) are sufficient for
required for imaging. For instance, although bisRGD-18F-Al- imaging,48–51 it is likely that high specific activity will be crucial
NOTA could be prepared at ~6 Ci/mmol, it could not be sepa- for early cancer detection where target concentration is necessa-
rated from unlabeled bisRGD. Consequently, the effective rily low and for which regulatory agencies require that injected
specific activity was reduced to 18.5 GBq/mmol (0.5 Ci/mmol) radiotracer concentration also be low (<0.1 Â Kd) such that
although no time was “wasted” on HPLC separation.15 Had <10% target is bound in order to avoid pharmacological
we not separated labeled Rh-18F-ArBF3À, the effective specific effects.17 Although a recent review2 suggests that to date, there
activity (defined as GBq/mmol of all tracer components), is no consensus as to the minimum specific activity needed for
corrected to TOC, would have been 259 GBq/mmol (7Ci/mmol), a imaging, a value of 18.5 GBq/mmol (0.5 Ci/mmol) is a generally
value that is still on par with some of the highest values accepted minimum value required for imaging.15
obtained for any radiotracer. Thus, this work shows that high
As such, an interesting conclusion then arises. If this minimum
effective specific activities will prevail even sans separation. threshold is not raised in the future, the ability henceforth to
Finally, cartridge-based separation methods may eventually produce tracers at exceptionally high specific activity opens the
obviate the need for HPLC separation; for instance, a sandwich possibility of intercontinental distribution whereby radiotracers
of (i) silica to remove free fluoride ion; (ii) a diol resin to produced at 529 GBq/mmol (14.3 Ci/mmol) can be transported
remove the boronic acid side product; and (iii) a reverse phase 8.85 h before decaying to 0.5 Ci/mmol. We believe that this
resin to remove hydrophobic organics, may provide rapid feature may prove especially significant for the industrial produc-
purification.
tion, distribution, and sale of radiotracers.
www.jlcr.org
Copyright © 2012 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2012, 55 491–496

Z. Liu et al.
[10] M. Butters, J. N. Harvey, J. Jover, A. J. J. Lennox, G. C. Lloyd-Jones,
P. M. Murray, Angew. Chem. Int. Ed. 2010, 49, 5156.
[11] B. de Bruin, B. Kuhnast, F. Hinnen, L. Yaouancq, M. Amessou,
L. Johannes, A. Samson, R. Boisgard, B. Tavitian, F. Dolle, Bioconjug.
Chem. 2005, 16, 406.
[12] J. Becaud, L. J. Mu, M. Karramkam, P. A. Schubiger, S. M. Ametamey,
K. Graham, T. Stellfeld, L. Lehmann, S. Borkowski, D. Berndorff,
L. Dinkelborg, A. Srinivasan, R. Smits, B. Koksch, Bioconjug. Chem.
2009, 20, 2254.
[13] S. H. Hausner, J. Marik, M. K. J. Gagnon, J. L. Sutcliffe, J. Med. Chem.
2008, 51, 5901.
[14] D. E. Olberg, A. Cuthbertson, M. Solbakken, J. M. Arukwe, H. Qu,
A. Kristian, S. Bruheim, O. K. Hjelstuen, Bioconjug. Chem. 2010, 21,
2297.
[15] O. Jacobson, L. Zhu, Y. Ma, I. D. Weiss, X. Sun, G. Niu, D. O. Kiesewetter,
X. Y. Chen, Bionconj. Chem. 2011, 22, 422.
[16] M. C. Lasne, C. Perrio, J. Rouden, L. Barre, D. Roeda, F. Dolle, C. Crouzel,
Contrast Agents II Top. Curr. Chem. 2002, 222, 201.
[17] L. S. Cai, S. Y. Lu, V. W. Pike, Eur. J. Org. Chem. 2008, 2853.
[18] J. J. Agresti, B. T. Kelly, A. Jaschke, A. D. Griffiths, Proc. Nat’l. Acad. Sci.
U.S.A. 2005, 102, 16170.
Conclusion
Here, we have used click chemistry to prepare a fluorescent
bioconjugate 19F-ArBFÀ3 precursor that can be aliquoted and lyo-
philized for future use in kit-like fashion. For labeling, 50 nmol
(52 mg) is resuspended in 50 mL aqueous DMF (pH 2 buffer) and
combined with >29.6 GBq (>800 mCi) of NCA 18F-fluoride ion
via robotic manipulation in a shielded hot cell. After 10–15 min
at 40 ꢀC, the reaction was quenched and analyzed for radiochemi-
cal yield and specific activity. Radiochemical yields are ~50% and
specific activities are as high as 529GBq/mmol (14.3 Ci/mmol),
corrected to the TOC following HPLC purification. Two other
radiosyntheses each with ~14.8 GBq (~400 mCi) provided simi-
lar radiochemical yields at specific activities of approximately
111–222 GBq/mmol (3–7 Ci/mmol). In contrast to many other
methods, labeling herein has several advantages as follows: (i)
speed; (ii) use of aqueous conditions at mild temperatures in
buffer; (iii) high radiochemical yields; (iv) high radiochemical
purity; (v) extraordinarily high specific activity; (vi) the creation
[19] C. A. D’Souza, W. J. McBride, R. M. Sharkey, L. J. Todaro, D. M. Goldenberg,
Bioconjug. Chem. 2011, 22, 1793.
of a high polar anionic 18F-labeled prosthetic that is readily [20] Y. Li, R. Ting, C. W. Harwig, U. A. D. Keller, C. L. Bellac, P. F. Lange,
J. A. H. Inkster, P. Schaffer, M. J. Adam, T. J. Ruth, C. M. Overall,
separated by HPLC; and (vii) compatibility with both hot cell
robotic manipulation and high levels of 18F-activity. On the basis
D. M. Perrin, Medchemcomm 2011, 2, 942.
[21] A. P. Kostikov, J. Chin, K. Orchowski, S. Niedermoser, M. M. Kovacevic,
of successful tumor imaging with 18F-ArBF3À-bioconjugates that
A. Aliaga, K. Jurkschat, B. Wangler, C. Wangler, H. J. Wester,
were labeled at much lower specific activities, we believe this
work should be applicable to imaging the same bioconjugates
at specific activities ~100-fold higher than we previously
reported and 3–10-fold higher than what is normally achieved for
most other 18F-labeled radiotracers reported by others. This work
should empower others to label many other bioconjugate-19F-
ArBFÀ3 precursors by isotope exchange.
R. Schirrmacher, Bioconjug. Chem. 2012, 23, 106.
[22] R. Schirrmacher, C. Wangler, E. Schirrmacher, Mini-Rev. Org. Chem.
2007, 4, 317.
[23] H. F. Huang, K. Liang, Y. P. Liu, S. T. Huang, T. W. Chu, Prog. in Chem.
2011, 23, 1501.
[24] G. E. Smith, H. L. Sladen, S. C. G. Biagini, P. J. Blower, Dalton Trans.
2011, 40, 6196.
[25] B. Wangler, R. Schirrmacher, P. Bartenstein, C. Wangler, Mini Rev.
Med. Chem. 2011, 11, 968.
[26] R. Ting, M. J. Adam, T. J. Ruth, D. M. Perrin, J. Am. Chem. Soc. 2005,
127, 13094.
[27] G. A. Molander, R. Figueroa, Aldrichimica Acta 2005, 38, 49.
[28] G. A. Molander, N. Ellis, Acc. Chem. Res. 2007, 40, 275.
[29] S. Darses, J. P. Genet, Chem. Rev. 2008, 108, 288.
[30] R. Ting, C. Harwig, J. Lo, Y. Li, M. J. Adam, T. J. Ruth, D. M. Perrin,
J. Org. Chem. 2008, 73, 4662.
[31] U. A. D. Keller, C. L. Bellac, Y. Li, Y. M. Lou, P. F. Lange, R. Ting,
C. Harwig, R. Kappelhoff, S. Dedhar, M. J. Adam, T. J. Ruth, F. Benard,
D. M. Perrin, C. M. Overall, Cancer Res. 2010, 70, 7562.
[32] R. Ting, C. W. Harwig, U. auf dem Keller, S. McCormick, P. Austin,
C. M. Overall, M. J. Adam, T. J. Ruth, D. M. Perrin, J. Am. Chem.
Soc. 2008, 130, 12045.
Acknowledgements
This work was supported by a grant from the Canadian Cancer
Society #20071 and NSERC Operating funds.
Conflict of Interest
[33] R. Ting, T. A. Aguilera, J. L. Crisp, D. J. Hall, W. C. Eckelman, D. R. Vera,
R. Y. Tsien, Bioconjug. Chem. 2010, 21, 1811.
[34] M. Namavari, O. P. De Jesus, Z. Cheng, A. De, E. Kovacs, J. Levi,
R. Zhang, J. K. Hoerner, H. Grade, F. A. Syud, S. S. Gambhir, Mol. Imaging
Biol. 2008, 10, 177.
UBC has sought patent protection of various aryltrifluoroborates
for use in 18F-imaging.
[35] B. Kuhnast, F. Hinnen, R. Hamzavi, R. Boisgard, B. Tavitian, P. E. Nielsen,
F. Dolle, J. Label. Compd. Radiopharm. 2005, 48, 51.
[36] R. Schirrmacher, G. Bradtmoller, E. Schirrmacher, O. Thews,
J. Tillmanns, T. Siessmeier, H. G. Buchholz, P. Bartenstein, B.
Waengler, C. M. Niemeyer, K. Jurkschat, Angew. Chem. Int. Ed.
2006, 45, 6047.
[37] Z. Li, K. Chansaenpak, S. Liu, W. R. Casey, P. S. Conti, F. P. Gabbaï, Med.
Chem. Commun. 2012, 3, 1305.
[38] T. Poethko, M. Schottelius, G. Thumshim, U. Hersel, M. Herz,
G. Henriksen, H. Kessler, M. Schwaiger, H. J. Wester, J. Nucl. Med.
2004, 45, 892.
[39] D. Thonon, C. Kech, J. Paris, C. Lemaire, A. Luxen, Bioconjug. Chem.
2009, 20, 817.
[40] N. K. Devaraj, R. Weissleder, Acc. Chem. Res. 2011, 44, 816.
[41] R. Ting, J. Lo, M. J. Adam, T. J. Ruth, D. M. Perrin, J. Fluor. Chem. 2008,
129, 349.
[42] H. Ihara, M. Koyanagi, M. Suginome, Org. Lett. 2011, 13, 2662.
[43] N. Iwadate, M. Suginome, J. Organomet. Chem. 2009, 694, 1713.
[44] H. Noguchi, T. Shioda, C. M. Chou, M. Suginome, Org. Lett. 2008,
10, 377.
References
[1] R. Y. Tsien, Nat. Cell Biol. 2003, Ss16.
[2] S. E. Lapi, M. J. Welch, Nucl. Med. Biol. 2012, 39, 601.
[3] J. Bergman, O. Eskola, P. Lehikoinen, O. Solin, Appl. Radiat. Isot. 2001,
54, 927.
[4] B. C. Lee, J. Y. Paik, D. Y. Chi, K. H. Lee, Y. S. Choe, Bioconjug. Chem.
2004, 15, 104.
[5] E. Briard, S. S. Zoghbi, F. G. Simeon, M. Imaizumi, J. P. Gourley,
H. U. Shetty, S. Y. Lu, M. Fujita, R. B. Innis, V. W. Pike, J. Med. Chem.
2009, 52, 688.
[6] P. Carberry, B. P. Lieberman, K. Ploessl, S. R. Choi, D. N. Haase, H. F. Kung,
Bioconjug. Chem. 2011, 22, 642.
[7] F. Wuest, M. Berndt, R. Bergmann, J. van den Hoff, J. Pietzsch, Bioconjug.
Chem. 2008, 19, 1202.
[8] R. Haubner, B. Kuhnast, C. Mang, W. A. Weber, H. Kessler, H. J. Wester,
M. Schwaiger, Bioconjug. Chem. 2004, 15, 61.
[9] E. Lee, A. S. Kamlet, D. C. Powers, C. N. Neumann, G. B. Boursalian,
T. Furuya, D. C. Choi, J. M. Hooker, T. Ritter, Science 2011, 334,
639.
J. Label Compd. Radiopharm 2012, 55 491–496
Copyright © 2012 John Wiley & Sons, Ltd.
www.jlcr.org

Z. Liu et al.
[45] T. W. Hudnall, T. P. Lin, F. P. Gabbai, J. Fluor. Chem. 2010, 131, [49] S. R. Banerjee, M. Pullambhatla, Y. Byun, S. Nimmagadda, G. Green, J. J. Fox,
1182.
A. Horti, R. C. Mease, M. G. Pomper, J. Med. Chem. 2010, 53, 5333.
[50] Y. Chen, C. A. Foss, Y. Byun, S. Nimmagadda, M. Pullambahatla, J. J. Fox,
M. Castanares, S. E. Lupold, J. W. Babich, R. C. Mease, M. G. Pomper,
J. Med. Chem. 2008, 51, 7933.
[46] P. Laverman, W. J. McBride, R. M. Sharkey, A. Eek, L. Joosten, W. J. G.
Oyen, D. M. Goldenberg, O. C. Boerman, J. Nucl. Med. 2010, 51, 454.
[47] L. F. Tietze, K. Schmuck, Synlett 2011, 1697.
[48] T. Ujula, M. Huttunen, P. Luoto, H. Perakyla, I. Simpura, I. Wilson, [51] R. R. Flavell, P. Kothari, M. Bar-Dagan, M. Synan, S. Vallabhajosula,
M. Bergman, A. Roivainen, Bioconjug. Chem. 2010, 21, 1612. J. M. Friedman, T. W. Muir, G. Ceccarini, J. Am. Chem. Soc. 2008, 130, 9106.
www.jlcr.org
Copyright © 2012 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2012, 55 491–496