Bioconjugate Chem. 2004, 15, 373−379
373
Production of [F-18]Fluoroannexin for Imaging Apoptosis with PET John R. Grierson,*,† Kevin J. Yagle,† Janet F. Eary,† Jonathan F. Tait,‡ Don F. Gibson,‡ Barbara Lewellen,† Jeanne M. Link,† and Kenneth A. Krohn† Departments of Radiology and Laboratory Medicine, University of Washington, Seattle, Washington 91895. Received September 19, 2003; Revised Manuscript Received January 27, 2004
Recombinant human-annexin-V was conjugated with 4-[F-18]fluorobenzoic acid (FBA) via its reaction with the N-hydroxysuccinimidyl ester (FBA-OSu) at pH 8.5. A series of reactions using varying amounts of annexin-V, unlabeled FBA-OSu, and time produced products with different conjugation levels. Products were characterized by mass spectrometry and a cell-binding assay to assess the effect of conjugation. In each case, the conjugated protein was a mixture of proteins with a range of conjugation. Annexin-V could be conjugated with an average of two FBA mole equivalents without decreasing its affinity for red blood cells (Kd 6-10 nM) with exposed phosphatidylserine. An average conjugation of 7.7 (range 3-13) diminished the binding 3-fold. Large-scale production and purification of [F-18]FBA-OSu from [F-18]fluoride was accomplished within 90 min and in 77% radiochemical yield (decaycorrected to the end of cyclotron bombardment). The conjugation reaction of annexin with [F-18]FBAOSu was studied with respect to activity level, protein mass, and concentration. Under the most favorable conditions, >25 mCi [F-18]fluoroannexin (FAN) was isolated in 64% yield (decay-corrected for a 22 min conjugation process) from labeling 1.1 mg of annexin-V. A pilot PET imaging study of [F-18]fluoroannexin in normal rats showed high uptake in the renal excretory system and demonstrated sufficient clearance from most other internal organs within 1 h. [F-18]Fluoroannexin should prove useful in imaging targeted apoptosis.
INTRODUCTION
Apoptosis is a normal physiological process of programmed cell death that plays a key role in maintaining constant cell density in tissues undergoing cell proliferation. It also provides a defense mechanism for eliminating abnormal cells with damaged DNA. As such, apoptosis plays an important role in cancer biology and cancer treatment. Cancer can result from a failure of apoptosis to balance cell proliferation, either by under-expressing genes that promote apoptosis or by overexpressing genes that inhibit it. In addition to genetic mutations that lead to unregulated cell proliferation, other mutations that affect the cell’s ability to undergo apoptosis can also be part of the progress toward uncontrolled growth. For example, the bcl-2 oncogene appears to contribute to the development of lymphoma by inhibiting apoptosis, without affecting cell proliferation (1). Furthermore, chemotherapy and radiation therapy generally kill cells by activating apoptosis. In this context, a sensitive method for imaging apoptosis in vivo should prove useful for monitoring cancer treatment. In contrast to normal cells, high levels of phosphatidylserine are exposed on the outer membrane leaflet of apoptotic cells. This provides a selective and accessible target for molecular imaging. Annexin-V (36 kDa) is an endogenous protein ligand with a high affinity (Kd ∼10 nM) for phosphatidylserine (PS) (2). Consequently, it has been adapted for use as a radiopharmaceutical after labeling it with technetium-99m (3-5). Studies with [Tc99m]annexin-V have demonstrated that apoptosis imaging is feasible in animal models of induced apoptosis and * Corresponding author. E-mail:
[email protected]. Ph: 206-598-6247. Fax: 206-598-4192. † Department of Radiology. ‡ Department of Laboratory Medicine.
with human subjects (6-8). This success prompted us to develop a fluorine-18-labeled annexin-V analogue for quantitative 3-D imaging with PET. We were also attracted by the potential application of combining [F-18]fluoroannexin-V and [C-11]thymidine in the same imaging session to investigate the balance between cell death and growth rates in tumors. In this report we show that annexin-V can be effectively labeled using the N-hydroxysuccinimidyl ester of [F-18]fluorobenzoic acid (Figure 1) and that the biological activity of the acylated product, [F-18]fluoroannexin (FAN), is equivalent to that of annexin-V. Exploratory PET imaging of [F-18]FAN in normal rats showed rapid blood clearance and low background activity in organs other than the kidneys and bladder, which are involved in clearance of the protein. EXPERIMENTAL PROCEDURES
Materials and Methods. (a) Chemicals. Reagent chemicals were obtained from Aldrich Chemical Co. (Milwaukee, WI) and were used as received. The labeling precursor for [F-18]FBA-OSu:ethyl 4-trimethylammonium benzoate trifluoromethanesulfonate salt was prepared according to Haka et al. (9) and stored in the cold. Unlabeled 4-fluorobenzoic acid N-hydroxysuccinimide ester (FBA-OSu) was prepared using standard peptide methods with 4-fluorobenzoic acid, N-hydroxysuccinamide, and dicyclohexylcarbodiimide in ether. The isolated solid was recrystallized twice from ether (mp 110-112 °C). Anhydrous dimethyl sulfoxide (DMSO), contained in break-seal ampules, was used for labeling experiments. Enriched (95%-atom) [O-18]water was obtained from Isonics Corp. (Golden, CO). AG 1-X8 ion-exchange resin, for separating [F-18]fluoride from [O-18]water, was obtained from Rio-Rad (Hercules, CA). C18-Plus SepPak
10.1021/bc0300394 CCC: $27.50 © 2004 American Chemical Society Published on Web 02/27/2004
374 Bioconjugate Chem., Vol. 15, No. 2, 2004
Grierson et al.
Figure 1. Radiolabeling scheme and timeline for the synthesis of [F-18]fluoroannexin. EOB (end of cyclotron bombardment); A (end of label incorporation rxn); B (end of FBA-OSu purification); C (isolation of FBA-OSu in dry ether); D (solvent-free FBA-OSu isolated). Table 1. Mass Spectrometry (ES+) Study of Annexin-V Acylation by Cold FBA-NHS stoichiometry: FBA-OSu/annexin
rxn time (min)
(0)
1:1 2:1 4:1 8:1 16:1 32:1
15 15 15 15 15 15
97 89 79 69 42 12
FBA conjugations: annexin:(x) (% distribution) (1) (2) (3) (4) 3 11 21 26 34 26
5 17 27
cartridges were obtained from the Waters Corporation (Milford, MA). (b) Annexin-V Protein. Laboratory grade recombinant (E. coli) human annexin-V (rh-annexin-V) was purified as described by Wood et al. (10). Stock protein solutions (1.0, 7.6, or 8.7 mg/mL) were prepared with 10 mM NaH2PO4 pH 7.4 buffer containing 140 mM sodium chloride and stored frozen (-15 °C) until used. Clinical grade rh-annexin-V was provided (1.0 mg/mL, 10 mM NaH2PO4 pH 7.4 buffer, 140 mM NaCl) as a gift from the Theseus Imaging Corporation (Worcester, MA) and stored at -80 °C until used. (c) Buffers. HNKGB buffer (10 mM HEPES pH 7.4, 136 mM NaCl, 2.7 mM KCl, 5 mM glucose, 1 mg/mL fatty acid free bovine serum albumin) was prepared as needed. Sodium chloride (0.9%, USP) was obtained from Baxter (Deerfield, IL) and used for the size exclusion HPLC purification of [F-18]fluoroannexin. For animal work, pH 7.4 PBS was used. Borate buffers (0.025 M and 0.50 M, pH 8.5) were prepared from sodium tetraborate decahydrate (Borax, 99.999%) and hydrochloric acid. The more concentrated buffer solution (9.52 g Borax + 2.53 mL concentrated HCl/50 mL) must be prepared with heating to dissolve the salt. The solution (rt) was used before recrystallization initiated (∼30 min). (d) Chromatography Systems and Detectors. Purifications were performed using a Thermo-Separation Products (Riviera Beach, FL) Constametric 4100 HPLC pump, Rheodyne (Cotati, CA) 7125 and 7725i injectors and flow-through radiation detector systems from Beckman (Fullerton, CA; model 170) or Carroll & Ramsey Associates (Berkeley, CA; model 105-S). For studies with unlabeled annexin a variable wavelength V4-UV absorbance detector from ISCO (Lincoln, NE) was used. HPLC columns were obtained from Phenomenex (Torrance, CA): column-A (LUNA-5 µm C18(2)) 250 × 10 mm i.d.; column-B (BioSep SEC-S 2000) 300 × 7.8 mm i.d.; column-C (LUNA-5 µm C18(2)) 250 × 4.6 mm i.d.. Radioactivity >100 µCi was measured using an ion
7 23
9
(5)
average: FBA/annexin
IC50 (nM)
5
0.03 0.11 0.21 0.37 0.89 2.06
12 ( 1 9(1 10 ( 2 8(2 10 ( 2 8(2
chamber Capintec CRC-12 Radioisotope Calibrator (Pittsburgh, PA) and 30%) if the active ester is not concentrated to a small area prior to its reaction. To minimize the problem we evaporated the stock aliquot (∼4 mL) to dryness and then used a small volume of ether (400 µL) to wash down the tube wall, followed by gentle evaporation to a final residual film. (h) Cell Binding Assay for Labeled Annexin. A previously developed cell-binding assay (5) was used to determine the affinity of labeled annexin samples to a standard preparation of human red blood cells with externalized (PS+) phosphatidylserine. The assay was conducted in the absence and presence of calcium, which is essential for annexin binding. Briefly, labeled annexin at a final concentration of 10 nM (the approximate Kd of annexin-V) was incubated for 15 min at rt in HNKGB buffer (total volume 1 mL) containing either 2.5 mM Ca2+ or 5 mM EDTA, and 100 µL of PS+ cells (∼4 × 108 erythrocytes). The specific binding (1-[cpm (Ca2+)/cpm total]) was measured after pelleting cells by centrifugation at 10 000g (Eppendorf model 5415D; Hamburg, Germany) for 3 min. Typically, the specific binding ranged from 80 to 90%, and this value was used to correct for nonbinding annexin in the determinations of Kd (section-j). (i) Determining Unlabeled Fluoroannexin-V Inhibition Concentration (IC50) Relative to FITC-
376 Bioconjugate Chem., Vol. 15, No. 2, 2004
Grierson et al.
Table 2. Radiochemical Yields and Product Analysis for [F-18]Fluoroannexin-V Production labeling reaction variables
[F-18]fluoroannexin production
entry
reaction activity, mCi
FBA-OSu batch
protein mass, µg
buffer volume, µL
reaction volume, µL
F-annexin-V activity (EOS), mCi
yield (SOS) vs FBA-OSu, %
specific activity (EOS), µCi/µg
protein recovery, %
1 2 3 4 5 6 7 8
24.5 50.7 18.3 138.4 48.1 41.3 44.7 44.2
A B A C D E E E
150 150 1140 1140 1140 1305 150 300
150 150 150 150 15 15 15 15
300 300 300 300 165 165 165 315
1.39 2.82 4.63 33.9 26.7 22.8 6.0 6.7
7 6 29 28 64 64 17 19
16 31 4 36 29 23 76 29
59 59 86 84 83 76 53 79
Figure 3. Fluoroannexin IC50 values vs reaction conditions for fluoroannexin syntheses. Product protein IC50 values were determined using a competitive binding assay with FITC-annexin as described in section-i; error bars (2 × SD).
Annexin-V Binding to PS+ (phosphatidylserine) Red Blood Cells. The concentration of purified fluoroannexin protein samples was first assayed by the Bradford method (13) using reagents from BioRad (Hercules, CA) with BSA fraction-V (Pierce, Rockford, IL) as a standard. Fluoroannexin samples were added at concentrations from 2 to 100 nM to a reaction tube containing 5 nM fluorescein-labeled annexin-V (FITC-annexin) (5) and 8 × 106 PS+ erythrocytes in HNKGB buffer, in a final volume of 1 mL. After incubating for 15 min at room temperature, cells were pelleted by centrifugation at 10 000g for 3 min. The supernatant was removed to a separate tube, and the FITC-annexin-V bound to the pellet was released by adding 950 µL of 5 mM EDTA/ HNKGB. The sample was recentrifuged, and 800 µL of supernatant was assayed for FITC-annexin-V content by fluorometric analysis using an Abbott TDX analyzer (Abbott Park, IL). The results from these inhibition experiments are shown in Figure 3. (j) Determining the Apparent Kd for [F-18]Fluoroannexin Bound to PS+ Red Blood Cells. Quantitative titrations were used to determine the dissociation constant, as described previously (5). Briefly, labeled fluoroannexin (2-300 nM) was added to 107 PS+ erythrocytes in HNKGB buffer (250 µL total volume) with either 2.5 mM Ca2+ or 5 mM EDTA. After incubation at room temperature for 15-30 min, bound and free fluoroannexin were separated by centrifugation through 40 µL of DC550 silicone oil (Dow Chemical, Midland, MI) in 0.3 mL small-tipped microtubes (Sarstedt; Nuembrecht, Germany) at 10 000g for 3 min. After cutting off the tube tip to isolate the pellet, the bound fluoroannexin in the cell pellet was counted. Counts were converted to molecules bound/cell using the known specific activity of the protein. The concentration of free annexin was
Figure 4. PS+-red blood cell binding data for [F-18]fluoroannexin.
determined by measuring activity remaining in the supernatant and correcting for nonbindable fluoroannexin, as determined using the method described in section-h. A typical data set is shown in Figure 4. Kd and Bmax values were calculated using a nonlinear leastsquares analysis (Microsoft Excel) to fit data to a model of noncooperative binding to homogeneous sites. (k) Mass Spectrometry. A Micromass (Manchester, UK) Quattro II Tandem Quadrupole Mass Spectrometer, that has unit resolution over a 4000 m/z range, was used with electrospray ionization (ESI+) for analysis of fluoroannexin samples containing mixed levels of protein conjugation with 4-fluorobenzoic acid. Samples were desalted by HPLC prior to their infusion into the electrospray source. The chromatographic method used a
[F-18]Fluoroannexin for Imaging Apoptosis
Poros RP1 packed column (150 × 2.1 mm, wide pore cross-linked polystyrene media supplied by Applied Biosystems Inc, Foster City, CA) and a Shimadzu (Columbia, MD) LC-10AD pump. The RP1 column was eluted at 300 µL/min using mixtures of mobile phase: A ) water0.05% trifluoroacetic acid (TFA) and B ) MeCN0.05%TFA. The mobile phase profile was 90:10 (A/B) between 0 and 5 min, followed by a linear gradient to 5:95 (A/B) between 5 and 20 min, and a final isocratic elution to 30 min. Micromass MassLynxNT 3.2 software was used to control the HPLC, the spectrometer, and data acquisition. The column eluate flow was diverted from the electrospray source for 2 min with a VICI Valco Instruments (Houston, TX) six-port injection valve. After 15 min of programmed elution, the fluoroannexin proteins eluted unseparated over a 4-min period. The spectrometer scan-time was 3.5 s between 500 and 2000 Da, with an interscan period of 0.1 s. MaxEnt, a statistical software package within the MassLynx operating system, was used to deconvolute complex multiplecharge-state mass spectral envelopes. A uniform Gaussian width-at-half-height of 2 Da was applied for mass peak recognition. An example data set and analysis is provided as Supporting Information. (m) PET Imaging. Normal male Sprague-Dawley rats (260-280 g) were purchased from Animal Technologies Ltd. (Kent, WA). Animals were housed according to institutional guidelines, and all procedures were approved by the University of Washington Institutional Animal Care and Use Committee (IACUC). Prior to imaging, rats were anesthetized with ketamine (68 mg/ kg) and xylazine (4.4 mg/kg) by iv injection. Anesthesia was maintained by booster injections of ∼7 mg/kg ketamine every 20 min. PET images were obtained using a GE (Waukesha, WI) Advance PET system. Image data was collected for 3 h, in 10 min time bins, starting 30 s after tail vein injection of [F-18]fluoroannexin (1-1.5 mCi). The specific activity was 13 µCi/µg at the time of injection. Data were collected in 2D-mode and reconstructed using the attenuated weighted OSEM algorithm (2 iterations, 28 subsets) with a 4 mm Gaussian postfilter. The reconstructed field of view was 20 cm. RESULTS AND DISCUSSION
[F-18]Fluoroannexin (FAN) was prepared by adapting a general protein labeling protocol developed by Wester et al. (11). The method uses the N-hydroxysuccinimidyl ester of 4-fluorobenzoic acid (FBA-OSu). Our choice was based on initial success with labeling FBA-OSu and the previous claim that, in vivo, protein-(fluoro)benzamide conjugation was more stable than protein-(fluoro)alkylamide conjugation (11). We have further investigated the production of [F-18]FBA-OSu and evaluated its potential for labeling annexin-V. The basic scheme is illustrated in Figure 1. The issue of fluorobenzoic acid (FBA) conjugate load was addressed with experiments using unlabeled FBA-OSu and annexin-V. Study of Annexin-V Conjugation. Annexin-V has 22 lysine residues and a free amino terminus that provide potential sites for fluorobenzoic acid (FBA) conjugation. Prior to this work, no information was available on how many FBA conjugations could be tolerated before the characteristic PS-binding of the parent protein (Kd 10 nM) was compromised. This issue was addressed by a series of experiments that varied the stoichiometry (1:1 to 32:1 FBA-OSu/annexin) of 15 min-pH 8.5-room temperature reactions. Fluoroannexin reaction products were purified by size exclusion HPLC and analyzed by
Bioconjugate Chem., Vol. 15, No. 2, 2004 377
Figure 5. Radiochromatographic study of [F-18]FBA-OSu stability. Injection-A and -B: C18 column eluted with 60% MeCN-water; Injection-C: size exclusion column eluted with 0.9% saline.
mass spectrometry to determine the distribution of annexin molecules with different numbers of FBA groups. Table 1 lists annexin-xFBA distributions within each sample and the associated IC50 values (50% inhibitory concentration) that measure the ability of protein samples to competitively displace FITC-annexin (Kd 10 nM) bound to PS+ cell membranes. The data showed that even a 32:1 molar ratio yields quite a low (2.1) average FBA substitution level and that the reaction protocol could yield adequate specific activity without loss of phosphatidylserine binding. The average level of FBA conjugation with annexin-V was linearly related (r2 ) 0.99) to the offering ratio up to the highest ratio tested, 32:1. Extended reaction times (30 and 60 min) with a 32:1 mole ratio were also examined to test the limits of conjugation. These experiments are summarized in Figure 3, which indicated that only an extended reaction time and a high FBA-OSu:annexin molar ratio (60 min, 32:1) yields a protein conjugate with significantly compromised binding to PS+ membranes. The two control samples were stock annexin solution and a stock sample that was purified by SEC. The product from the most heavily treated reaction was analyzed by mass spectrometry to determine how many FBA/annexin were present. The range of conjugation was 3-13, and the average was 7.7 (Supporting Information). Collectively, these data showed that it should be possible to prepare [F-18]fluoroannexin with high protein specific activity and maintain competitive binding. Radiolabeling FBA-OSu and Annexin-V. The production of [F-18]FBA-OSu was accomplished in 77% radiochemical yield (EOB). We chose to include a HPLC purification of FBA-OSu to circumvent potential problems with protein labeling and to eliminate unlabeled or labeled impurities that might interfere with the [F-18]FAN bioassay. Using the pure ester offered better quality control for evaluating the protein-conjugation reaction. The extra time taken to purify and isolate the ester in dry ether may be optional, but we have demonstrated that large quantities of the ester can be made. A typical timeline for the process is shown in Figure 1. Time points B and C mark, respectively, the collection of the ester from the HPLC column (91 min) and its subsequent
378 Bioconjugate Chem., Vol. 15, No. 2, 2004
Grierson et al.
Figure 6. PET image of a male rat injected with [F-18]fluoroannexin and tissue TAC’s. PET image set (top); tissue TAC: full scale (left); expanded scale (right).
reformulation in dry ether (108 min, 35% radiochemical yield, not decay corrected). Isolation of the ester in dry ether minimized its decomposition and provided a convenient solvent for rapidly evaporating aliquots. Figure 2 illustrates a set of HPLC radiochromatographic profiles for FBA-OSu and FAN purifications. Initial attempts with FBA-OSu production gave a 25% (not decay corrected) radiochemical yield (55%, decay corrected to the end of cyclotron bombardment (EOB)) on a >350 mCi scale. With practice, however, the yields steadily increased to 44% (not decay corrected) for collecting FBAOSu off the HPLC column, requiring 90 min overall. Figure 5 is a series of radiochromatograms that demonstrate the stability of FBA-OSu. Chromatogram-A shows a sampling of FBA-OSu in dry ether after the solution was allowed to stand for 2 h (initial activity 80 mCi/6 mL). Clearly the ester has good stability in ether. Another stability test of FBA-OSu at pH 8.5 mimicked the labeling conditions with annexin-V. An aliquot of ester stock solution was evaporated and treated with a 16% solution of acetonitrile in 0.025 M borate buffer (150 µL). Chromatogram-B represents a sample after 15 min at room temperature. Assuming FBA-OSu decomposes by hydrolysis or attack by other ambient species at the active ester site, a crude estimate of FBA-OSu decomposition gives a first order kinetic half-time of 32 min. This value acceptable compared to the 15 min labeling protocol. The identity of the FBA-OSu decomposition
byproducts in chromatogram-B was partially clarified using the size exclusion HPLC method involved with FAN purification. Chromatogram-C identified 4-fluorobenzoic acid (verified by coinjection) and an unidentified component “x”. Considering the composition of the test mixture, “x” is likely derived from acetonitrile decomposition byproducts that become labeled. This implicated fluorobenzamide or N-fluorobenzoylacetamide, but the issue was not pursued further. Table 2 lists selected fluoroannexin labeling results using [F-18]FBA-OSu. [F-18]FAN activity values are referenced to the end-of-synthesis (EOS) to reflect the quantities typically encountered with the 22 min reaction and purification process. In some cases (1, 2, and 7) radiochemical yields were influenced by low protein recovery. The poor yields (6%) seen for reactions 1 and 2 showed that a 1 mg/mL protein concentration diluted (1:1) with pH 8.5 buffer was impractical. By comparison, the higher protein stock concentration (7.4 mg/mL) used in reactions 3 and 4 improved the yields through gains in protein recovery and labeling efficiency. Protein recovery showed a practical limit of ∼85% when the protein content was above 1.1 mg (reactions 3-6). In these cases, unrecovered activity was accounted for by the reaction vessel residue and transfer loss to the HPLC column. Further improvement in yield (2.3-fold) was realized (e.g. reaction 5 vs 4) when the buffer volume was reduced 10-fold (15 µL vs
[F-18]Fluoroannexin for Imaging Apoptosis
150 µL). A supersaturated borate buffer (0.5 M) was used. Less concentrated buffers may be as effective but were not tested. The 64% yield for reactions 5 and 6 appears to be the limit, considering transfer losses in the process and the competing FBA-OSu decomposition. Overall, the final protein concentration in the reaction was the most important factor that determined the percent conjugation of FBA with annexin-V. Reactions 6-8 used a common batch of FBA-OSu and comparable amounts of activity and protein. This series also evaluated clinical grade (1.0 mg/mL) rh-annexin-V. Reaction 6 used the same research grade rh-annexin-V as reactions 1-5. In comparison to the yield for reaction 1, the yield for reaction 7 is higher than expected. This could be ascribed to differences in FBA-OSu batches A and E at the start of the labeling reaction. The specific activities in FBA-OSu batches A-D were not measured. However, fluoride activity levels for all batches were comparable (average 478 mCi, range 335-535 mCi). FBA-OSu specific activity for batch E (2 Ci/µmol) was determined indirectly by performing mass spectrometry on the FAN product from reaction 7. The value was calculated by dividing the protein specific activity (88 µCi/ µg decay corrected to SOS) by the average number of FBA conjugations per FAN protein molecule (1.6). The FANxFBA distribution frequency was 0-FBA(13%), 1-FBA(37%), 2-FBA(32%), 3-FBA(12%), 4-FBA(3%), and 5-FBA(3%), and the FBA-OSu/annexin ratio for reaction 7 was 5.3:1. Figure 4 shows rbc-binding profile for FAN from reaction 7. The apparent Kd and Bmax values ((2SD) for the protein were 10.8 ( 5.0 nM and 3.04 + 0.35 × 105/ cell, respectively, as calculated from a nonlinear leastsquares fit. The apparent Kd values compare favorably with the reported Kd for annexin V, which is between 6 and 10 nM at this Ca2+ concentration (2.5 mM) (5). From a collection of similarly prepared samples (n ) 8) the average Kd; Bmax values were 9.6 nM (( 20%) and 3.1 × 105/cell (( 13%), respectively. PET Imaging. Our goal is to use [F-18]fluoroannexin to image apoptosis in humans. As a preliminary step, normal rats were injected with [F-18]fluoroannexin and imaged to study which organs had the highest uptake and how rapidly the activity was cleared from tissues. Figure 6 shows the first PET images obtained with FAN in a normal rat. This series of 10-min frames centered on 45 min postinjection demonstrates the high uptake in the renal excretory system. The heart and liver are discernible above the kidneys, and the ureters are even visible. This figure also illustrates tissue time-activity curves (TAC) for heart, liver, kidney, and bladder. The graph on the right-hand side uses an expanded scale for activity values in order to better display the lower valued curves. It is clear from these curves and the image set that there is sufficiently rapid clearance of [F-18]fluoroannexin so that only modest activity remains in most internal organs after 1 h. The clearance characteristics of [F-18]fluoroannexin should prove useful for early imaging of apoptosis. We are currently investigating the use of [F-18]fluoroannexin in imaging apoptosis in animal models. ACKNOWLEDGMENT
Financial support was provided by the National Institutes of Health (CA42045). We thank William Howald
Bioconjugate Chem., Vol. 15, No. 2, 2004 379
and Ross Lawrence for their assistance with protein mass measurements. We also thank Theseus Imaging Corporation for the gift of clinical grade rh-annexin-V Supporting Information Available: Mass spectral ion data and its deconvolution for a fluoroannexin sample. This material is available free of charge via the Internet at http://pubs.acs.org/BC. LITERATURE CITED (1) Rantanen, S., Monni, O., Joensuu, H., Franssila, K., and Knuutila, S. (2001) Causes and consequences of BCL2 overexpression in diffuse large B-cell lymphoma. Leuk. Lymphoma 42, 1089-1098. (2) Thiagarajan, P., and Tait, J. F. (1990) Binding of annexin V/placental anticoagulant protein I to platelets. Evidence for phosphatidylserine exposure in the procoagulant response of activated platelets. J. Biol. Chem. 265, 17420-3. (3) Tait, J. F., and Gibson, D. (1994) Measurement of membrane phospholipid asymmetry in normal and sickle- cell erythrocytes by means of annexin V binding. J. Lab. Clin. Med. 123, 741-8. (4) Stratton, J. R., Dewhurst, T. A., Kasina, S., Reno, J. M., Cerqueira, M. D., Baskin, D. G., and Tait, J. F. (1995) Selective uptake of radiolabeled annexin V on acute porcine left atrial thrombi. Circulation 92, 3113-21. (5) Tait, J. F., Brown, D. F., Gibson, D. F., Blankenberg, F. G., and Strauss, H. W. (2000) Development and characterization of annexin-V mutants with endogenous chelation sites for 99mTc. Bioconjugate Chem. 11, 918-925. (6) Blankenberg, F. G., Katsikis, P., Tait, J. F., Davis, R. E., Naumovski, L., Ohtsuki, K., Kopiwoda, S., Abrams, M. J., Darkes. M., Robbins, R. C., Maecker, H. T., and Strauss, H. W. (1998) In vivo detection and imaging of phosphatidylerine expression during programmed cell death. Proc. Natl. Acad. Sci. U.S.A. 95, 6349-6354. (7) Blankenberg, F. G., Katsikis, P. D., Tait, J. F., Davis, R. E., Naumovski, L., Ohtsuki, K., Kopiwoda, S., Abrams, M. J., and Strauss, H. W. (1999) Imaging of apoptosis (programmed cell death) with 99mTc annexin V. J. Nucl. Med. 40, 184-191. (8) Belhocine, T., Steinmetz, N., and Hustinx, R., et al. (2002) Increased uptake of the apoptosis-imaging agent 99mTc recombinant human annexin-V in human tumors after one course of chemotherapy as a predictor of tumor response and patient prognosis. Clin. Cancer Res. 8, 2766-2774. (9) Haka, M. S., Kilbourn, M. R., Watkins, G. L., and Toorongain S. A. (1989 Aryltrimethylammonium trifluoromethanesulfonates as precursors to aryl[F-18]fluorides: Improved synthesis of [F-18]GBR-13119. J. Label. Comput. Radiopharm. 27, 823. (10) Wood, B. L., Gibson, D. F., and Tait, J. F. (1996) Increased erythrocyte phosphatidylserine exposure in sickle cell disease: flow-cytometric measurement and clinical associations. Blood 88, 1873-1880. (11) Wester, H. J., Hamacher, K., and Stocklin, G. (1996) A comparative study of the nca fluorine-18 labeling of proteins via acylation and photochemical conjugation. Nucl. Med. Biol. 23, 365-372. (12) Schlyer, D. J., Batos, M., and Wolf, A. P. (1987) A rapid quantitative separation of fluorine-18 from oxygen-18 water. J. Nucl. Med. 40, 25P. (13) Bradford, M. M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248-254.
BC0300394
