Synthesis and evaluation of ¹⁸F-labeled carbonic anhydrase IX inhibitors for imaging with positron emission tomography

Abstract

Two carbonic anhydrase IX (CA IX) inhibitors were radiolabeled with ¹⁸F, and evaluated for imaging CA IX expression. Despite good affinity for CA IX and excellent plasma stability, uptake of both tracers in CA IX-expressing HT-29 tumor xenografts in mice was low. ¹⁸F-FEC accumulated predominately in the liver and nasal cavity, whereas a significant amount of ¹⁸F-U-104 was retained in blood. Due to minimal uptake in HT-29 tumors compared to other organs/tissues, these two tracers are not suitable for use for CA IX-targeted imaging.

• Benzenesulfonamide
• coumarin
• cytochrome P450

Introduction

Carbonic anhydrase IX (CA IX) is a trans-membrane protein that catalyzes the hydration of carbon dioxide (CO₂ + H₂O ↔ H⁺+ )1. CA IX is not expressed in most normal tissues2 but is highly up-regulated in a variety of solid tumors including bladder3, cervical4, head and neck5, breast6,7, lung8 and kidney9 cancers. The expression of CA IX is induced by the transcription factor HIF-1α, which is stabilized and activated in response to hypoxia10. In clear cell renal cell carcinoma (CCRCC) the VHL gene is often mutated, leading to constitutive HIF-1α activation and a strong up-regulation of CA IX expression11,12. CA IX expression is associated with poor prognosis and tumor progression13–15. CA IX helps cancer cells survive under hypoxia by transporting into the cell to maintain a neutral intracellular pH16. The remaining H⁺ acidifies the extracellular microenvironment, activates metalloproteinases and facilitates invasion and metastasis17–20.

Due to its expression on the cell surface, CA IX is an attractive and accessible target for cancer therapy. There are currently three categories of anti-CA IX therapeutic agents under various stages of clinical trials: (1) sulfonamide derivative Indisulam (E7070) for the treatment of solid tumors21,22; (2) the humanized anti-CA IX monoclonal antibody WX-G250 (Rencarex®)23,24 and (3) anti-CA IX monoclonal antibodies (cG250) radiolabeled with cytotoxic radionuclides (¹³¹I and ¹⁷⁷Lu) for the treatment of CCRCC25. With promising therapeutic agents available, a suitable CA IX-targeted imaging agent is needed to quantify the expression of CA IX in tumors. This will help identify potential responders to CA IX-targeted therapies, and monitor the treatment efficacy. Since CA IX expression in most tumors is generally correlated with hypoxia, CA IX-targeted imaging agents could also be used as a noninvasive imaging tool to visualize hypoxic regions in tumors, and select patients for hypoxia-targeted treatment.

Since its development in the mid-1980s2, anti-CA IX monoclonal antibody G250 and derivatives have been labeled with ¹³¹I, ¹¹¹In, ¹²⁴I and ⁸⁹Zr to image CA IX expression in tumors with single photon emission computed tomography26,27 or positron emission tomography (PET)28,29. Divgi et al.28 reported that CCRCC could be distinguished accurately from other renal masses by ¹²⁴I-cG250 PET (94% sensitivity, 100% specificity), supporting the potential clinical use of ¹²⁴I-cG250 PET in renal masses as an alternative to biopsy for characterizing such lesions. However, the application of radiolabeled anti-CA IX antibodies for imaging hypoxia-associated CA IX expression is limited as the hypoxic tumor regions are usually less accessible to antibodies due to low perfusion caused by aberrant vasculature and increased interstitial pressure. Radiolabeled CA IX-targeted probes derived from small molecule inhibitors will be more suitable because of their ability to diffuse more freely in hypoxic tumor regions. Small molecules are also less expensive to prepare and easier to handle than antibodies, and are not immunogenic.

A large number of small molecule CA IX inhibitors have been developed and several attempts have been made to radiolabel them for imaging. Apte et al.30 reported the synthesis of an ¹⁸F-labeled sulfonamide derivative (Figure 1A), but no biological data were presented. The preparation of a ^99mTc-labeled sulfonamide conjugate (Figure 1B) was reported by Akurathi et al.31. Its accumulation in CA IX-expressing HT-29 tumors in mice was minimal with only 0.1% ID/g at 0.5 h post-injection (p.i.). Asakawa et al.32 reported the radiosynthesis of three potent 11C-ureido-substituted benzenesulfonamides (Figure 1C) as PET probes for imaging CA IX in tumors, but no biological evaluation data were presented. ¹⁸F-VM4-037 (Figure 1D), developed by Siemens33, showed intense background activity in a human study34, which precluded its use as a cancer imaging agent. Recently, Lu et al. reported the synthesis of a series of ^99mTc/Re-labeled benzenesulfonamide derivative35. Among them, ^99mTc-3d (Figure 1E) exhibited high binding affinity (IC₅₀ = 9 nM) to hypoxic CA IX-expressing HeLa cells, but no further evaluation in animal imaging study was presented. Here, we report our attempts on the synthesis, 18F labeling, and evaluation of two CA IX inhibitors, 7-(2-fluoroethoxy)coumarin (FEC) and U-104 (Figure 2), as potential tracers for imaging CA IX expression with PET.

Figure 1. Radiolabeled sulfonamide derivatives for imaging CA IX expression.

Figure 2. Structures of two potential CA IX inhibitors amenable for radiolabeling with ¹⁸F for PET imaging.

Materials and methods

Chemicals and instrumentation

U-10436 and its radiolabeling precursor 4-nitrophenyl 4-sulfamoylcarbanilate37 were prepared according to previously published procedures. All other chemicals and solvents were obtained from commercial sources, and used without further purification. Proton NMR spectra were obtained using a Bruker (Billerica, MA) Avance 400inv Spectrometer, and were reported in parts per million downfield from internal tetramethylsilane. Mass analyses were performed using a Bruker Esquire-LC/MS system with ESI ion source. Purification and quality control of ¹⁸F-labeled CA IX inhibitors were performed on an Agilent (Santa Clara, CA) HPLC System equipped with a model 1200 quaternary pump, a model 1200 UV absorbance detector and a Bioscan (Washington, DC) NaI scintillation detector. The radio-detector was connected to a Bioscan B-FC-1000 Flow-count System, and the output from the Bioscan Flow-count system was fed into an Agilent 35900E Interface which converted the analog signal to digital signal. The operation of the Agilent HPLC system was controlled using the Agilent ChemStation software. The HPLC columns used were a semipreparative column (Phenomenex C18, 5 µ, 250 × 10 mm) and an analytical column (Eclipse XOB-C18, 5 µ, 150 × 4 mm). ¹⁸F-Fluoride was produced by the 18O(p, n)¹⁸F reaction using an Advanced Cyclotron Systems Inc. (Richmond, BC, Canada) TR19 cyclotron. Radioactivity of ¹⁸F-labeled tracers were measured using a Capintec (Ramsey, NJ) CRC®-25R/W dose calibrator, and the radioactivity of mouse tissues collected from biodistribution studies were counted using a Packard (Meriden, CT) Cobra II 5000 Series auto-gamma counter. PET imaging experiments were conducted using a Siemens (Erlangen, Germany) Inveon microPET/CT scanner.

Synthesis of FEC

A mixture of 7-hydroxycoumarin (0.81 g, 5 mmol), 1-fluoro-2-tosyloxyethane (1.56 g, 7.1 mmol) and K₂CO₃ (3.45 g, 25 mmol) in DMF (15 mL) was heated at 70 °C for 24 h. After cooling to room temperature, CH₂Cl₂ (100 mL) was added to the mixture, and the resulted solution was washed with water (100 mL × 3). The CH₂Cl₂ phase was dried with anhydrous MgSO₄, concentrated under reduced pressure, and purified by flash column chromatography on silica gel using 4:6 EtOAc/hexanes to obtain the desired product FEC as a white solid (0.918 g, 84 %). ¹H NMR (CDCl₃) δ 4.28 (dt, J = 27.6, 4.0 Hz, 2H), 4.80 (dt, J = 47.2, 4.0 Hz, 2H), 6.27 (d, J = 9.2 Hz, 1H), 6.83 (d, J = 2.4 Hz, 1H), 6.89 (dd, J = 8.4, 2.4 Hz, 1H), 7.40 (d, J = 8.4 Hz, 1H), 7.65 (d, J = 9.2 Hz, 1H); MS (ESI) calculated for C₁₁H₉FO₃ (M + H)⁺ 209.1, found 209.0.

Synthesis of 7-(2-tosyloxyethoxy)coumarin

A solution of 7-hydroxycoumarin (0.81 g, 5 mmol), 1,2-bis(tosyloxy)ethane (2.78 g, 7.5 mmol) and K₂CO₃ (3.45 g, 25 mmol) in DMF (15 mL) was heated at 70 °C for 18 h. After cooling to room temperature, CH₂Cl₂ (100 mL) was added, and the resulted solution was washed with water (100 mL × 3). The CH₂Cl₂ layer was dried with anhydrous MgSO₄, concentrated under reduced pressure, and chromatographed on silica gel using 1:99 CH₃CN/CH₂Cl₂ to obtain the desired product as a white solid (0.282 g, 16%). ¹H NMR (CDCl₃) δ 2.47 (s, 3H), 4.19–4.25 (m, 2H), 4.39–4.45 (m, 2H), 6.28 (d, J = 9.2 Hz, 1H), 6.68 (d, J = 2.4 Hz, 1H), 6.76 (dd, J = 8.4, 2.4 Hz, 1H), 7.34–7.39 (m, 3H), 7.64 (d, J = 9.2 Hz, 1H), 7.83 (d, J = 8.4 Hz, 1H); MS (ESI) calculated for C₁₈H₁₆O₆S (M + H)⁺ 361.1, found 361.1.

Affinity measurement

The inhibition constants (K_i) of FEC to four human CA isoenzymes I, II, IX and XII were determined by CA catalyzed CO₂ hydration assays following previously published procedures38.

Radiosynthesis of ¹⁸F-FEC

The proton-bombarded H₂[¹⁸O]O was transferred via He gas push from the cyclotron target station to a Waters (Milford, MA) QMA light Sep-Pak cartridge set up in the hot cell. The ¹⁸F-fluoride was trapped in the QMA cartridge, and eluted out into a 4-mL V-shaped reaction vial with a mixture of water (0.3 mL) and CH₃CN (0.3 mL) containing 7 mg of K₂CO₃ and 22 mg of kryptofix 222 (K₂₂₂). The uncapped reaction vial was placed in a heating block, and the solution was heated at 120 °C. After most of the solvents evaporated, CH₃CN (1 mL × 2) was added to the reaction vial to facilitate complete removal of water. A solution of 7-(2-tosyloxyethoxy)coumarin (3 mg) in DMF (0.5 mL) was added to the reaction vial containing dry K[¹⁸F]F/K₂₂₂. The vial was capped, and the mixture was heated at 70 °C for 30 min. At the end of heating, the mixture was diluted with water (1 mL) and purified by HPLC using the semipreparative column eluted with 35% CH₃CN/65% H₂O at a flow rate of 4.5 mL/min. The retention time of ¹⁸F-FEC was 17.4 min. The eluting fraction containing ¹⁸F-FEC was collected, diluted with water (50 mL), and trapped on a Waters tC18 light Sep-Pak cartridge. After washing the tC18 light Sep-Pak cartridge with water (10 mL), ¹⁸F-FEC was eluted out with ethanol (0.4 mL), and formulated with saline (4 mL) for plasma stability study and µPET/CT imaging study. The quality control was performed by HPLC on the analytical column eluted with 25% MeCN/75% H₂O at a flow rate of 2 mL/min. The retention time of ¹⁸F-FEC was 8.5 min. The specific activity of ¹⁸F-FEC was measured using the analytical HPLC system. It was calculated by dividing the injected radioactivity in 0.2 mL of final ¹⁸F-FEC solution by the mass of FEC in the injected solution. The mass of FEC was estimated by comparing the UV absorbance obtained from the injection with a previously prepared standard curve.

Radiosynthesis of ¹⁸F-U-104

(1) Radiosynthesis of 1-[¹⁸F]fluoro-4-nitrobenzene: a solution of 1,4-dinitrobenzene (4 mg) in DMSO (0.5 mL) was added to a 4 mL reaction vial containing dry K[¹⁸F]F/K₂₂₂. The vial was capped, and the mixture was heated at 125 °C for 10 min. At the end of the heating, the mixture was diluted with water (10 mL). The resulted solution was passed through a Waters tC18 plus Sep-Pak cartridge, and the cartridge was washed with water (10 mL). The trapped 1-[¹⁸F]fluoro-4-nitrobenzene was eluted out of Sep-Pak cartridge with methanol (1.5 mL). (2) Radiosynthesis of 4-[¹⁸F]fluoroaniline. The above solution of 1-[¹⁸F]fluoro-4-nitrobenzene in methanol was added to a vial containing 10% palladium on carbon (4 mg) and sodium borohydride (10 mg). The resulted mixture was incubated at room temperature for 5 min. The reaction was quenched by adding 0.1 mL of concentrated HCl and diluted with 1 M NaOH (20 mL). The resulted solution was passed through a Lichrolut EN column (500 mg). The trapped 4-[¹⁸F]fluoroaniline in the Lichrolut EN column was eluted out with THF (2 mL) and dried by passing the THF solution through a pre-packed column containing celite (125 mg) and anhydrous MgSO₄ (125 mg). (3) Radiosynthesis of ¹⁸F-U-104: The above solution of 4-[¹⁸F]fluoroaniline in THF was added to a 4-mL reaction vial containing 4-nitrophenyl 4-sulfamoylcarbanilate (8 mg) and DIEA (10 µL) in DMF (0.5 mL). The resulted mixture was heated at 125 °C for 15 min, diluted with water (1 mL), and purified by HPLC using the semipreparative column eluted with 40% MeOH/60% H₂O at a flow rate of 4.5 mL/min. The retention time of ¹⁸F-U-104 was 27.4 min. The eluting fraction containing ¹⁸F-U-104 was collected, diluted with water (50 mL), and passed through a Waters tC18 light Sep-Pak cartridge. The trapped ¹⁸F-U-104 on the Sep-Pak cartridge was eluted out with ethanol (0.4 mL), and formulated with saline (4 mL) for plasma stability and biodistribution studies. The quality control of ¹⁸F-U-104 was performed on HPLC using the analytical column eluted with 25% MeCN/75% H₂O at a flow rate of 2 mL/min. The retention time of ¹⁸F-U-104 was 7.7 min. The specific activity of ¹⁸F-U-104 was measured following same procedures as described for the calculation of specific activity of ¹⁸F-FEC.

Stability in mouse plasma

Aliquots (100 µL) of the ¹⁸F-labeled tracer (¹⁸F-FEC or ¹⁸F-U-104) were incubated with 400 µL of balb/c mouse plasma (available from Innovative Research; Cat#: IMS-BCN-N-25 mL) for 5, 15, 30, 60 and 120 min at 37 °C. At the end of each incubation period, samples were passed through a 0.45 micron filter. The filtered sample was loaded onto the analytical HPLC to check the presence of metabolites, and analysis was conducted with Agilent ChemStation software.

In vivo experiments

Mice were maintained and the experiments were conducted in accordance with the guidelines established by the Canadian Council on Animal Care and approved by the Animal Ethics Committee of the University of British Columbia.

Tumor implantation. All experiments were performed using NODSCID IL2RKO mice bred in-house at the Animal Research Centre, British Columbia Cancer Research Centre, Vancouver, Canada. Mice were anesthetized briefly with 2.5% isoflurane in 2.0 L/min of oxygen during cells implantation. After wiping skin surrounding the injection site with an alcohol prep pad, a 31-Gauge needle was used to subcutaneously implant 5 × 10⁶ HT-29 cells (in 100 µL of 1xPBS and BD Matrigel Matrix at 1:1 ratio) under the right shoulder. Biodistribution studies and PET/CT imaging were performed when tumors reached 5–7 mm in diameter.

Biodistribution studies. After HT-29 tumor bearing mice (n = 3) were anesthetized using isofluorane, ¹⁸F-U-104 was administered intravenously through the caudal vein at a dose of 100 µCi in a volume of 100–200 µL. At 1 h p.i., mice were euthanized by CO₂ asphyxiation followed by cervical dislocation. During dissection, tissues of interest (blood, stomach, intestine, spleen, liver, pancreas, kidney, lung, heart, tumor, muscle, bone and brain) were harvested. Tissues of interest were rinsed with saline (exception of blood), blotted dry, weighed and counted to determine the percentage of injected dose per gram (%ID/g) of tissue. 

PET imaging and data analysis. Under anesthesia, 100-200 μCi of [¹⁸F]FEC was administered intravenously through the caudal vein into HT-29 tumor bearing mice (n = 4). CT scan was performed before a PET dynamic image sequence of 55 min. PET data were acquired in list mode acquisition. At 1 h p.i., mice were euthanized by CO₂ asphyxiation followed by cervical dislocation. The tissues of interest were harvested, weighed, and counted as described in the above Biodistribution Studies section. The PET data were reconstructed using the 3d-OSEM-MAP algorithm with CT-based attenuation correction. Three-dimensional regions of interest were placed on the reconstructed images to determine the %ID/g of tissue using the Inveon Acquisition Workplace software.

Immunohistochemistry

For validation of CA IX expression in HT-29 tumors, histologic tissue analysis was performed. Tumors were harvested and fixed in 4% paraformaldehyde in PBS for 48 h at room temperature. After dehydration, tissues were embedded in paraffin and 4 µm sections were mounted onto poly-L-lysine slides. Immunostaining was performed by the Centre of Translational Applied Genomics at the BC Cancer Agency using the Ventana Discovery XT instrument. Slides were incubated with a goat anti-human CA IX antibody (R&D AF2188) for 1 h without heat at a 1:200 dilution in CC1 antigen retrieval buffer. A rabbit anti-goat linker (1:500) was applied for 32 min followed by a 16 min incubation with Ultramap anti-rabbit HRP detection kit. The stained sections were examined and photographed with a Leica EC3 microscope.

Results and discussion

FEC is a coumarin derivative, whereas U-104 is an ureido-substituted benzenesulfonamide (Figure 2). Both coumarins39 and benzenesulfonamides36 have been shown to be potent inhibitors of CA IX. In addition, U-104 has been previously reported to inhibit tumor growth and metastasis in spontaneous and experimental models of metastasis, without inhibitory effects on CA IX-negative tumors13. As shown in Scheme 1(A), FEC was synthesized in 84% yield by coupling 7-hydroxycoumarin with 1-fluoro-2-tosyloxyethane40.

For the development of CA IX-targeted tracers, one concern is the binding to other major CA isoforms including I, II and XII. Cytosolic CA I and CA II are expressed in red blood cells41,42 and the binding of CA IX-targeted radiotracers to CA I and CA II would increase background signal and reduce the image contrast. On the contrary, binding of these CA IX-targeted radiotracers to CA XII could be beneficial as CA XII is expressed on membranes and is also up-regulated in hypoxic tumors43. The inhibition constants (K_i) for human CA isoenzymes I, II, IX and XII were 4622, >100 000, 70, and 88  nM for FEC, respectively. The K_i values for U-104 have been previously reported to be 5080, 95, 45 and 4.5  nM for CA isoenzymes I, II, IX and XII, respectively13,36.

The radiosynthesis of ¹⁸F-FEC was prepared via the ¹⁸F aliphatic nucleophilic substitution (Scheme 2A). The radiolabeling precursor, 7 -(2-tosyloxyethoxy)coumarin, was prepared by coupling 7-hydroxycoumarin with 1,2-bis(tosyloxy)ethane in 16% yield (Scheme 1B). After HPLC purification, ¹⁸F-FEC was obtained in 11–24 % decay-corrected yield in 1.6 h synthesis time with 1.7–5.7 Ci/µmol specific activity at the end of synthesis (EOS), and >99% radiochemistry purity. The preparation of ¹⁸F-U-104 was depicted in Scheme 2(B). The radiosynthesis involves three reactions. First, 1-[¹⁸F]fluoro-4-nitrobenzene was prepared in 40–60% radiochemical yield by direct aromatic nucleophilic substitution reaction using ¹⁸F-fluoride and dinitrobenzene. The isolated 1-[¹⁸F]fluoro-4-nitrobenzene was reduced to 4-[¹⁸F]fluoroaniline in >90 % yield following previously published procedure44 using NaBH₄ and 10% Pd/C. At the final step, ¹⁸F-U-104 was obtained in >50% radiochemical yield by reacting 4-[¹⁸F]fluoroaniline with 4-nitrophenyl 4-sulfamoylcarbanilate in DMF. However, due to multiple purification steps, ¹⁸8F-U-104 isolated in only 3–9% overall decay-corrected yield in 2.5 h synthesis time with >98% radiochemical purity, and 15.1–19.8 Ci/μmol specific activity at EOS. The reason that ¹⁸F-U-104 had much higher specific activity than ¹⁸F-FEC was because we used Teflon tubing as the H[¹⁸F]F/H₂[¹⁸O]O transfer line while preparing ¹⁸F-FEC. Before working on ¹⁸F-U-104, we replaced the Teflon transfer line with peek tubing. This change reduced the amount of fluoride leaching out from the transfer line, and significantly increased the specific activities of ¹⁸F-labeled tracers including ¹⁸F-U-104 prepared thereafter.

Scheme 1. Synthesis of (A) FEC and (B) its radiolabeling precursor 7-(2-tosyloxyethoxy)coumarin.

Scheme 2. Radiosynthesis of (A) ¹⁸F-FEC and (B) ¹⁸F-U-104.

Both ¹⁸F-U-104 and ¹⁸F-FEC were stable in mouse plasma with >99% of the tracers remaining intact after 2 h incubation at 37 °C. For the imaging/biodistribution studies, we used HT-29 colorectal tumor xenografts as our CA IX-expressing tumor model, and the expression of CA IX in the tumors was confirmed by immunohistochemistry (Figure 3). Due to its constitutive expression of CA IX, HT-29 cells have also been used by other investigators31,45 as a CA IX-expressing model for the development of CA IX-targeted tracers.

Figure 3. Immunohistochemical staining of CA IX expression in a HT-29 tumor.

The biodistribution data of ¹⁸F-U-104 (Table 1) indicated that the radioactivity was excreted via both renal and hepatobiliary pathways. The uptake in intestines and kidneys at 1 h p.i. were 13.66 ± 1.23 and 9.71 ± 1.68 %ID/g, respectively. Tumor uptake (%ID/g) of ¹⁸F-U-104 at 1 h p.i. was 0.83 ± 0.06 which was lower than blood (13.92 ± 3.07), muscle (1.19 ± 0.20) and major organs except brain (0.16 ± 0.01). High blood uptake of ¹⁸F-U-104 is likely due to the binding to CA II in erythrocytes. U-104 has good affinity (95 nM) for CA II and it has been shown that erythrocytes express high level of CA II41,42. Due to minimal uptake of ¹⁸F-U-104 in HT-29 tumors compared to normal tissues/organs, ¹⁸F-U-104 is not suitable for use for CA IX-targeted imaging.

Table 1. Biodistribution of ¹⁸F-FEC and ¹⁸F-U-104 in HT-29 tumor bearing NODSCID IL2RKO mice at 1 h post injection.

	^18F-FEC	^18F-U-104
Blood	1.54 ± 0.44	13.92 ± 3.07
Stomach	0.89 ± 0.46	2.94 ± 2.28
Intestine	3.14 ± 1.24	13.66 ± 1.23
Spleen	1.89 ± 0.88	6.48 ± 0.74
Liver	33.76 ± 8.31	6.76 ± 0.22
Pancreas	0.77 ± 0.10	3.33 ± 0.11
Kidney	6.97 ± 3.34	9.71 ± 1.68
Lungs	1.18 ± 0.24	5.71 ± 0.42
Heart	1.16 ± 0.27	3.46 ± 0.57
Tumor	1.16 ± 0.19	0.83 ± 0.06
Muscle	0.73 ± 0.13	1.19 ± 0.20
Bone	2.09 ± 0.50	2.52 ± 0.81
Brain	0.88 ± 0.16	0.16 ± 0.01

Data are expressed as %ID/g ± S.D.

Biodistribution and PET imaging studies of ¹⁸F-FEC showed that the radioactivity was excreted via both renal and hepatobiliary pathways (Table 1, Figure 4). In vivo defluorination of ¹⁸F-FEC was likely as uptake in bone (2.09 ± 0.50 %ID/g) was higher than the uptake in both blood (1.54 ± 0.44 %ID/g) and muscle (0.73 ± 0.13 %ID/g) at 1 h p.i. Tumor uptake of ¹⁸F-FEC was 1.16 ± 0.19 %ID/g at 1 h p.i. The tumors were not visualized from PET image (Figure 4) due to low tumor uptake and very high liver uptake (33.76 ± 8.31 %ID/g at 1 h p.i.). The high uptake in the liver was not simply due to the hepatobiliary excretion of ¹⁸F-FEC as the uptake in intestines was low (3.14 ± 1.24 %ID/g) and the uptake in the liver did not decrease over time (Figure 5).

Figure 4. A typical PET image of ¹⁸F-FEC in HT-29 tumor bearing NODSCID IL2RKO mice at 55 min post-injection. White arrows indicate tumor.

Figure 5. Time-activity curve of ¹⁸F-FEC in tumor and selected organs/tissues.

One possible explanation for the high uptake of ¹⁸F-FEC in the liver is due the action of 7-ethoxycoumarin O-deethylase (ECOD). ECOD is a family of cytochrome P450 enzymes that metabolize 7-ethoxycoumarin into 7-hydroxycoumarin and acetaldehyde46 as depicted in Scheme 3(A). ECOD is highly expressed in the liver, and one of the major ECOD is CYP1A246,47. Due to the similarity of 7-ethoxycoumarin and FEC in their structures, it is likely that ECOD also metabolizes ¹⁸F-FEC into 7-hydroxycoumarin and 2-[¹⁸F]fluoroacetaldehyde (Scheme 3B). The radioactive metabolite 2-[¹⁸F]fluoroacetaldehyde can be further metabolized into 2-[¹⁸F]fluoroacetate which in turn forms 2-[¹⁸]fluoroacetyl CoA and becomes trapped within the cell48. This hypothesis was supported by the high uptake of ¹⁸F-FEC observed in the nasal cavity (Figures 4 and 5) as olfactory mucosa is the only extrahepatic tissue expressing high level of CPY1A249. However, more studies are needed to confirm the trapping mechanism of ¹⁸F-FEC in the liver and nasal cavity.

Scheme 3. (A) Metabolism of 7-ethoxycoumarin and (B) proposed metabolism of ¹⁸F-FEC by ECOD.

Conclusions

We have synthesized two ¹⁸F-labeled CA IX inhibitors, ¹⁸F-FEC and ¹⁸F-U-104, and evaluated their potential as CA IX-targeted PET tracers. Both compounds showed good affinity for CA IX and excellent stability in mouse plasma. However, their uptake in CA IX-expressing HT-29 tumors was minimal which precludes their application as CA IX imaging agents. The unexpected trapping of ¹⁸F-FEC in the liver and nasal cavity could be due to the metabolism of ¹⁸F-FEC by ECOD. Once this is confirmed, ¹⁸F-FEC may be potentially used for imaging the expression/activity of ECOD with PET.

Declaration of interest

This work was supported by the Canadian Institutes of Health Research (Grant MOP 102685), and the postdoctoral position held by Jinhe Pan was funded by the Canadian Breast Cancer Foundation. The authors report no declarations of interest.

Acknowledgements

We are thankful to Jean-Pierre Appiah and Milan Vuckovic for operating the cyclotron and providing ¹⁸F-fluoride for radiolabeling experiments.