Arylfluoroborates and alkylfluorosilicates as potential PET imaging agents: High-yielding aqueous biomolecular18F-labeling

Article abstract of DOI:10.1021/ja053293a

Organometalloid compositions of silicon and boron permit rapid, high-yielding, one-step radiolabeling of a covalently linked protein ligand (biotin) under aqueous conditions to give the corresponding alkyltetrafluorosilicates and aryltrifluoroborate salts

Full text of DOI:10.1021/ja053293a

Published on Web 09/01/2005
Arylfluoroborates and Alkylfluorosilicates as Potential PET Imaging Agents:
High-Yielding Aqueous Biomolecular 18F-Labeling
Richard Ting,† Michael J. Adam,†,‡ Thomas J. Ruth,‡ and David M. Perrin*,†
Chemistry Department, 2036 Main Mall, UniVersity of British Columbia, and TRIUMF, VancouVer,
B.C. V6T-1Z1, Canada
Received May 19, 2005; E-mail: dperrin@chem.ubc.ca
Positron (â+) emission tomography (PET) permits high-resolution
localization of biological phenomena, provided that suitably labeled
ligands can specifically recognize molecular markers of the diseased
state.1 Larger biomolecules such as peptides and oligonucleotides
will be valuable for addressing this.
unmodified biotin that had been dissolved in 5 µL of DMF, MeOH,
MeCN, DMSO, or H2O. For reaction at pH 7.5, 200 mM HEPES
was used as a buffer in water. After 1 h at room temperature, reac-
tions were quenched by addition of 200 µL of 300 mM NaHCO3,
pH 7.5. The reaction was transferred to 15 µL of washed, poly-
disperse avidin magnetic particles (AMPs) with an experimentally
verified capacity for binding 525 pmol of biotin (see Supporting
Information) in 300 µL total volume. After 20 min, the AMPs were
magnetized, washed twice with 200 µL of 300 mM carbonate and
thrice with 200 µL of 1 M NaCl, 10 mM TrisHCl, pH 7.5, and 1
mM EDTA (total wash time ∼40 min), resuspended in 5 µL of
water, and transferred to a silica TLC plate, which was affixed with
clear tape and subjected to autoradiography as shown in Figure 2.
18F, with a half-life of 110 min, provides high specific radioactiv-
ity when introduced efficiently. Nevertheless, the chemistry of fluo-
rine limits labeling strategies involving C-F bond formation. Reac-
tive F+ reagents (e.g. F2, DAST) are difficult to prepare with high
specific activity and moreover are too reactive for direct labeling of
biomolecules. Anionic F-, while readily prepared with high specific
activity, is a poor nucleophile in aqueous media in which most
large biomolecules must be stored. Current labeling strategies in-
volve the generation of small radioactive precursors that are pre-
pared in polar aprotic solvents at 70-140 °C via nucleophilic dis-
placement by 18F- under scrupulously dry conditions. Coupling these
precursors to biomolecules requires additional transformations. In all
cases to date, 18F is introduced in the first of several chemical steps.2
Following synthesis, rapid decay and safety concerns limit the
ability to correlate radiochemical yields with the imaging agent’s
biological activity, which might be compromised by improper
coupling or by contaminants carried over from prior steps.3 Ideally,
a large-molecule imaging agent would be synthesized as a shelf-
stable precursor that could be rapidly labeled in a single high-
yielding step under aqueous conditions via simple addition of 18F-.
Indeed, the advantage of such a strategy is recognized in the case
of DOTA conjugates that are labeled by 64Cu wash-in.4 To date,
no such strategy has been described for 18F labeling.5
Figure 2. Autoradiogram of the washed AMPs: first row, boron 1; second
row, silicon 2; third row, unmodified biotin control. Spots 1-6 (left to
right): 50% MeOH, 50% DMF, 50% DMSO, 50% MeCN, 100% H2O, at
pH 4.5, then 100% H2O at pH 7.5. Bottom: Standard curve representing
percentage of total starting radioactivity. With a fluoridation efficiency of
90%, the density on the AMPs should be 0.15% (vide infra). Contrast is
increased to show radiographic location of control samples in third row.
Here we describe two new classes of biomolecule precursors
that acquire 18F highly efficiently in one step in aqueous or mixed-
solvent media: a pinacol phenylboronate diester (1) and an
alkyltriethoxysilane (2) (Figure 1). Each metalloid was conjugated
to biotin via an amide bond13 (see Supporting Information for
synthetic characterization) to investigate protein targeting to avidin.
Fluoridation is seen only for compounds 1 and 2, whereas negli-
gible fluoride carry-over is found on the AMPs that bound unmod-
ified biotin treated in the same way. Yields varied little with various
cosolvents. If one assumes that all avidin is bound (Kd < 10-12 M)
and no biotin dissociates during washing (kd < 10-5 min-1), the
ratio of AMP-bound counts to total counts must equal the ratio of
biotin binding capacity (525 pmol) to the total biotin (3 × 105 pmol)
multiplied by the fluoridation efficiency. To correlate autoradio-
graphic density with the fraction of F- spotted or incorporated,
serially diluted quantities of 18F- used in fluoridation were applied
to the plate. Fluoridation efficiencies were variable but approached
100% for 2 and 80% for 1 (see Supporting Information), consistent
with yields for such “ate” complexes obtained at higher fluoride
and organometalloid concentrations.6,9 19F NMR and HRMS spectra
Figure 1. Biotinylated p-aminophenylboronylpinacolate (1) and biotinylated
(aminopropyl)triethoxysilane (2).
Fluoride treatment of 1 and 2 gave trifluoroborate and tetrafluo-
rosilicate “ate” salts that were captured by the protein avidin. Thus,
this labeling should apply to other ligands (e.g. peptides, aptamers,
antibodies) that recognize various protein targets.
An aqueous volume of 5 µL (440 µCi from unfixed target water)
of 100 mM or 130 mM KHF2 (3.3 or 4.4 equiv of 19F-) in 200
mM NaOAc, pH 4.5, was added to 60 mM solutions of 1, 2, or
verified a F:B ratio of 3:1 for the “ate” salt of 1. For that of 2, 19
F
NMR shifts are given and low-resolution ESI indicated a F:Si ratio
of 4:1 (see Supporting Information).
Whereas the Si-F and B-F bonds are among the strongest bonds
known, and although alkyltetrafluorosilicates and aryltrifluorobo-
rates can be precipitated from water, to the best of our knowledge,
no study has examined the hydrolytic decomposition rate of either
an alkyltetrafluorosilicate or an aryltrifluoroborate. The only relevant
work was an investigation of the hydrolysis of tetrafluoroborate
† Chemistry Department, UBC.
‡ TRIUMF.
9
13094
J. AM. CHEM. SOC. 2005, 127, 13094-13095
10.1021/ja053293a CCC: $30.25 © 2005 American Chemical Society

C O M M U N I C A T I O N S
which proceeds very slowly (t1/2 ≈ 168 h, 100 °C).14 To that end,
we sought to measure the hydrolytic stability of the “ate” complexes
here at ∼21 °C. To do this we first verified that significantly less
fluoridation occurs at dilute conditions: 1.5-2 mM KHF2 (3-4
equiv of F-) and 1 mM 1 or 2 (∼3-10-fold over background, data
not shown). Therefore, if the “ate” were to hydrolyze at high dilu-
tion, no significant back reaction (i.e. refluoridation) would occur.
18F-labeled 1 and 2 were diluted to 0.3 mM into carbonate buffer
for periods of time prior to avidin capture. In addition, the same
were diluted 100-fold into 200 mM KH19F2. As fluoridation is both
kinetically and thermodynamically favorable at 200 mM KHF2, this
isotope exchange experiment ensured that any dissociated 18F would
be replaced by 19F. The autoradiogram of AMP capture of 1 and 2
after dilution is shown in Figure 3. Quantitative analysis followed
as above and as detailed further in the Supporting Information.
was studied because of its extensive use in biochips. It reasons
that biomolecules affixed to solid supports via a siloxy linkage can
be simultaneously released from the chip and labeled by 18F.
Due to safety concerns in this work, carrier 19F was added to
trace 18F (<100 pmol) to show high fluoridation efficiency. Never-
theless, this approach should deliver specific activities of >1 Ci/
µmol suitable for PET imaging without adding carrier: typically,
“no carrier added” 18F preparations have specific activities of ∼5
Ci/µmol, which contain significant quantities of carrier 19F. Thus,
reaction of 1 Ci (200 nmol total 18/19F) with 0.25 or 0.33 equiv of
metalloid will yield “ates” with specific activities of 16-20 Ci/
µmol. Since B or Si acquires three or four atoms of F- respectively,
this method, like no other, increases the effective specific activity
of the source 18F by 3 or 4 times (see Supporting Information). Of
note, other boron derivatives used in F- sensing may improve
radiochemical yields,10 particularly in cases where large molecules
may not be soluble at 30 mM. Smaller 18F-“ate” precursors could
be labeled first and then coupled to biomolecules as with traditional
precursors. Finally, as avidin itself has been used in imaging,15
biotins 1 and 2 should be of use with avidin-fusion proteins.
Acknowledgment. The authors thank the TRIUMF PET group
and Dr. Nick Burlinson for NMR expertise and support from West.
Econ. Dev. Off. & Can. Inst. Health Research grants. R.T. holds
Gladys Estella Laird and Michael Smith graduate student trainee-
ships. D.M.P. holds a Michael Smith Junior Career Award.
Supporting Information Available: Full references list; experi-
mental and characterization data. This material is available free of
charge via the Internet at http://pubs.acs.org.
Figure 3. Stability assay of “ate” salts of 1 and 2: first column, samples
are diluted into KHF2 at pH 7.5; second column, samples are simply diluted
-
into HCO3 buffer; first row, boron; second row, silicon; third row,
unmodified biotin. Incubation times prior to AMP capture are noted.
In the case of the tetrafluorosilicate, signal intensity dissipates
over time. Data were fit to a first-order function that returned a
rate constant of 0.01 min-1 in the presence of KHF2 and 0.008
min-1 in the absence, suggesting modest stability (see Supporting
Information for data fits). For the trifluoroborate, no decomposition
was observed, suggesting considerable stability in aqueous media.
For any labeling method to be used in imaging, the radioisotope
must be readily incorporated onto the imaging agent, which in turn
must be kinetically stable upon humoral dilution. To test this, “ates”
of 1 and 2 were incubated in either serum or whole blood for 1 h
prior to AMP capture. No time-dependent loss of 18F was observed
(see Supporting Information).
Conclusions and Discussion. Use of a short-lived isotope requires
that radiopharmaceutical synthesis be kinetically and thermody-
namically favorable at the time of preparation, and that the product
be at least kinetically stable following injection. To that end, we
contemplated the varied chemistries of organosilicon and organo-
boron6 that have found use respectively in protecting groups and
biochip fabrication,7 Suzuki chemistries,8,9 and fluoride sensors.10
Both compositions form stable linkages to alcohols yet react rapidly
and quantitatively with F- to give tetrafluorosilicates11 and tri-
fluoroborates that are sufficiently stable to be precipitated from
aqueous media.6,12 Organoboron and organosilicon bioconjugates
fluoridate in a single, rapid, and high-yielding step at pH 4-7 in
aqueous solvents and at temperatures that are unlikely to denature
the biomolecule. Their use should obviate multistep synthetic
transformations that normally follow radioisotope incorporation and
which require cumbersome robotic shielding. The aryltrifluoroborate
is appreciably more stable than the alkyltetrafluorosilicate and
should thus prove useful in developing stable biomolecule precur-
sors for imaging. The alkyltetrafluorosilicate was moderately stable
in aqueous media and decomposed with a rate that is on par with
that of 18F decay. Thus, the utility of an alkyltetrafluorosilicate
would be limited to ligands that associate very quickly with
physiological targets. Despite its limited stability, the triethoxysilane
References
(1) (a) Adam, M. J.; Wilbur, D. S. Chem. Soc. ReV. 2005, 34, 153-63. (b)
Britz-Cunningham, S. H.; Adelstein, S. J. J. Nucl. Med. 2003, 44, 1945-
61. (c) Jaffer, F. A.; Weissleder, R. JAMA 2005, 293, 855-62.
(2) (a) Dolle, F.; et al. Bioconj. Chem. 2005, 16, 406-20. (b) Grierson, J. R.;
Yagle, K. J.; Eary, J. F.; Tait, J. F.; Gibson, D. F.; Lewellen, B.; Link, J.
M.; Krohn, K. A. Bioconj. Chem. 2004, 15, 373-9. (c) Vaidyanathan,
G.; Zalutsky, M. R. Bioconj. Chem. 1994, 5, 352-6. (d) Glaser, M.;
Karlsen, H.; Solbakken, M.; Arukwe, J.; Brady, F.; Luthra, S. K.;
Cuthbertson, A. Bioconj. Chem. 2004, 15, 1447-53.
(3) Shai, Y.; Kirk, K. L.; Channing, M. A.; Dunn, B. B.; Lesniak, M. A.;
Eastman, R. C.; Finn, R. D.; Roth, J.; Jacobson, K. A. Biochemistry 1989,
28, 4801-6.
(4) (a) Chen, X. Y.; Park, R.; Tohme, M.; Shahinian, A. H.; Bading, J. R.;
Conti, P. S. Bioconj. Chem. 2004, 15, 41-9. (b) Wu, A. M.; et al. Proc.
Natl. Acad. Sci. U.S.A. 2000, 97, 8495-500.
(5) Lasne, M. C.; Perrio, C.; Rouden, J.; Barre, L.; Roeda, D.; Dolle, F.;
Crouzel, C. Top. Curr. Chem. 2002, 222, 201-58.
(6) Vedejs, E.; Chapman, R. W.; Fields, S. C.; Lin, S.; Schrimpf, M. R. J.
Org. Chem. 1995, 60, 3020-7.
(7) (a) Keana, J. F. W.; Shimizu, M.; Jernstedt, K. K. J. Org. Chem. 1986,
51, 1641-4. (b) Bartzoka, V.; Brook, M. A.; McDermott, M. R. Langmuir
1998, 14, 1887-91.
(8) (a) Molander, G. A.; Biolatto, B. J. Org. Chem. 2003, 68, 4302-14. (b)
Wright, S. W.; Hageman, D. L.; McClure, L. D. J. Org. Chem. 1994, 59,
6095-7.
(9) Darses, S.; Michaud, G.; Genet, J. P. Eur. J. Org. Chem. 1999, 1875-83.
(10) (a) Sole, S.; Gabbai, F. P. Chem. Commun. 2004, 1284-5. (b) Yamaguchi,
S.; Akiyama, S.; Tamao, K. J. Am. Chem. Soc. 2001, 123, 11372-5. (c)
Cooper, C. R.; Spencer, N.; James, T. D. Chem. Commun. 1998, 1365-
6. (d) DiCesare, N.; Lakowicz, J. R. Anal. Biochem. 2002, 301, 111-6.
(e) Wang, W.; Gao, X. M.; Wang, B. H. Curr. Org. Chem. 2002, 6, 1285-
317.
(11) (a) Tacke, R.; Becht, J.; Lopezmras, A.; Sheldrick, W. S.; Sebald, A. Inorg.
Chem. 1993, 32, 2761-6. (b) Tacke, R.; Pfrommer, B.; Lunkenheimer,
K.; Hirte, R. Organometallics 1998, 17, 3670-6.
(12) Mesmer, R. E.; Palen, K. M.; Baes, C. F. Inorg. Chem. 1973, 12, 89-95.
(13) (a) Crapatureanu, S.; Serbanescu, R.; Brevitt, S. B.; Kluger, R. Biocon-
jugate Chem. 1999, 10, 1058-67.
(14) Anbar, M.; Guttmann, S. J. Phys. Chem. 1960, 64, 1896-9.
(15) (a) Hnatowich, D. J.; Virzi, F.; Rusckowski, M. J. Nucl. Med. 1987, 28,
1294-302. (b) Najafi, A.; Peterson, A. Nucl. Med. Biol. 1993, 20, 401-5.
(c) Mariani, G.; et al. Eur. J. Nucl. Med. 1999, 26, 606-14. (d) Yla-
Herttuala, S.; et al. Gene Ther. 2003, 10, 2090-7. (e) Nakamoto, Y.; et
al. Nucl. Med. Biol. 1998, 25, 95-9.
JA053293A
9
J. AM. CHEM. SOC. VOL. 127, NO. 38, 2005 13095