Skip to main content
NCBI home page
ACS Pharmacology & Translational Science logoLink to ACS Pharmacology & Translational Science
. 2023 Oct 20;6(11):1745–1757. doi: 10.1021/acsptsci.3c00187

Design, Synthesis, and Evaluation of 18F-Labeled Tracers Targeting Fibroblast Activation Protein for Brain Imaging

Ziyue Yu †, Yong Huang ‡, Hualong Chen †, Zeng Jiang †, Chengze Li ‡, Yi Xie †, Zhongjing Li ‡, Xuebo Cheng †, Yajing Liu §, Shengli Li ∥, Ying Liang ‡,*, Zehui Wu †,*
PMCID: PMC10644484  PMID: 37974629

Abstract

graphic file with name pt3c00187_0010.jpg

Fibroblast activation protein (FAP) is closely related to central nervous system diseases such as stroke and brain tumors, but PET tracers that can be used for brain imaging have not been reported. Here, we designed, synthesized, and evaluated 18F-labeled UAMC1110 derivatives suitable for brain imaging targeting FAP. By substituting the F atom for the H atom on the aromatic ring of compound UAMC1110, 1a–c were designed and prepared. 1a–c were confirmed to have a high affinity for FAP through molecular docking and enzyme assay. [18F]1a–c were successfully prepared and confirmed to have high affinity. The stability in vivo indicates that no obvious metabolites of [18F]1a,b were found in the plasma 1 h after injection, which is beneficial for brain imaging. In vitro cell uptake experiments showed that [18F]1a,b and [68Ga]FAPI04 exhibited similar uptake and internalization rates. PET imaging of U87MG subcutaneous tumor showed that [18F]1a,b could penetrate the blood–brain barrier with higher uptake and longer retention time than [68Ga]FAPI04 (uptake at 62.5 min, 1.06 ± 0.23, 1.09 ± 0.25% ID/g vs 0.21 ± 0.10% ID/g, respectively). The brain-to-blood ratios of [18F]1a,b were better than [68Ga]FAPI04. Biodistribution and PET imaging showed that [18F]1a had better uptake on tumors and a higher tumor-to-muscle ratio than [18F]1b and [68Ga]FAPI04. Further imaging of U87MG intracranial glioma showed that [18F]1a outlined high-contrast gliomas in a short period of time compared to [18F]1b. Therefore, [18F]1a is expected to be useful in the diagnosis of FAP-related brain diseases.

Keywords: fibroblast activation protein, brain imaging, PET, tracer, UAMC1110

The tumor microenvironment is the set of cellular and molecular components that contribute to tumor development and progression.1 Noncancerous stromal cells in the tumor microenvironment promote tumor invasion and metastasis by producing various growth factors, chemokines, and cytokines to promote extracellular matrix remodeling, angiogenesis, cell migration, drug resistance, and evading immune surveillance.2 Therefore, monitoring or regulating the expression of effective targets in the tumor microenvironment is expected to achieve the purpose of the diagnosis or treatment of tumors. Fibroblast activation protein (FAP) is a type II transmembrane serine protease, belonging to the prolyl-specific serine oligopeptidase family,3 which is widely expressed on the surface of cancer related fibroblasts and tumor-associated macrophage in the tumor microenvironment, such as liver cancer, colorectal cancer, pancreatic cancer and ovarian cancer.4 FAP is also highly expressed in lesions characterized by interstitial tissue activation, such as infectious, inflammatory, fibrotic diseases and wound healing,5,6 but not expressed or low expressed in normal tissues and benign tumor stroma. Therefore, FAP is a reliable biomarker for tumor and other diseases diagnosis, treatment, and predicting efficacy.

The first FAP radiotracer was the 131I-labeled monoclonal antibody F19 targeting FAP,7 which can accurately locate colon cancer recurrence and liver metastases. With the development of FAP inhibitors, the small molecular structure of (4-quinolinyl) glycyl-2-cyanoprolidine scaffold with high FAP affinity was discovered in 2013.8,9 Based on this structure, Haberkorn et al. innovatively designed a series of PET molecular tracers with nanomolar levels of FAP affinity and high selectivity in 2018.10 Among them, [68Ga]FAPI04 is one of the tracers with good imaging characteristics, according to the report. [68Ga]FAPI04 can be internalized (with values over 90%) into cells expressing FAP faster than antibodies or antibody fragments, with good pharmacokinetic characteristics and FAP affinity and no uptake in normal tissues.11,12 [68Ga]FAPI04 can visualize more than 30 types of tumors,12 which has caused a huge sensation. In addition, radiation therapy drugs in clinical applications have achieved great success in recent years.13−15 Therefore, the development of radiopharmaceuticals targeting FAP for diagnosis and treatment has become a hot topic in the field of radiopharmaceuticals today. A series of novel FAP tracers aimed at improving the contrast between tumors and nontargets have been reported (Figure 1), such as FAPI46,11 ONCOFAP-DOTAGA,16,17 FAPI-2286,18 etc.,19−23 and have good applications in peripheral tumors such as breast cancer, lung cancer, etc., and arthritis.24 Abnormal expression of FAP is closely associated with central nervous system diseases such as stroke, inflammation, and brain tumors.25−27 However, FAP PET imaging studies for related diseases are rare, mainly because the ability of the existing reported tracers to penetrate the blood-brain barrier still needs further optimization. Therefore, FAP tracers that can be applied to the diagnosis of central nervous system diseases still need to be studied. Herein, we report the chemical design, synthesis, and biological evaluation of 18F-labeled FAP-targeting PET tracers suitable for brain imaging.

Figure 1.

Figure 1

Open in a new tab

Chemical structures of UAMC1110, FAPI04, FAPI46, ONCOFAP-DOTAGA, and 1a–c designed in this paper.

Methods

General Information

All reagents were commercial products used without further purification, unless otherwise indicated. Deionized water used in our experiments was obtained from a Milli-Q water system. NMR spectra were recorded at 300 MHz at ambient temperature. Chemical shifts are reported in parts per million downfield from TMS (tetramethylsilane). Coupling constants in 1H NMR are expressed in Hertz. High-resolution mass spectrometry (HRMS) data was obtained with an AB Sciex X500R QTof, ExionLC system (Massachusetts, USA). Agilent 1260 VWD (Santa Clara, USA) and Eckert & Ziegler Diode Detector B-FC-3400-A (Valencia, USA) were selected as UV and Radio- High performance liquid chromatography (HPLC) detectors, respectively. PerkinElmer WIZARD22480 automatic gamma counter (Waltham, USA) was used to measure the radioactivity of compounds. Thin-layer chromatography (TLC) analyses were performed using Merck (Darmstadt, Germany) silica gel 60 F254 plates. Crude compounds were generally purified by flash column chromatography (FC) packed with Teledyne ISCO (Nebraska, USA). In this report, no unexpected or unusually high safety hazards were encountered. UAMC1110 and (S)-4,4-difluoro-1-glycylpyrrolidine-2-carbonitrile were purchased from Bide pharm Company (Shanghai, China). The precursors of [68Ga]FAPI04 and FAPI04 were purchased from Tanzhenbio Company (Nanchang, China).

General Procedures of 3a–c

Compounds 4a, 4b, or 4c (0.26 g, 1 mmol) and (S)-4,4-difluoro-1-glycylpyrrolidine-2-carbonitrile (0.13g, 0.69 mmol) were dissolved in DCM (20 mL), triethylamine (3 mL), HOBt (0.05 g, 0.34 mmol), and EDCI (0.4 g, 2.1 mmol) subsequently. The mixture was stirred overnight at room temperature and then extracted with water three times. The combined organic layers were dried over MgSO4, filtered, concentrated, and purified by FC to give 3a, 3b, or 3c.

3a

Yield: 31.2%. Rf: 0.38 (100% ethyl acetate); FC conditions: (100% ethyl acetate). 1H NMR (300 MHz, DMSO-d6) δ: 9.21 (brs, 1H), 9.05 (d, J = 4.2 Hz, 1H), 8.58 (s, 1H), 8.01 (dd, J = 16.0 Hz, 2H), 7.65 (d, J = 4.3 Hz, 1H), 5.19 (d, J = 9.0 Hz, 1H), 4.33–4.25 (m, 4H), 2.94–2.82 (m, 2H).13C NMR (75 MHz, DMSO-d6) δ: 168.22, 167.18, 151.40, 146.99, 141.30, 133.51, 132.02, 128.24, 125.85, 121.16, 120.52, 118.29, 51.69, 44.72, 41.88, 36.89. HRMS calcd for C18H15BrF2N3O2+, 422.0310[M + H]+; found, 422.0313.

3b

Yield: 31.5%. Rf: 0.36 (100% ethyl acetate); FC conditions: (100% ethyl acetate). 1H NMR (300 MHz, CDCl3) δ: 8.68 (d, J = 4.0 Hz, 1H), 8.07 (d, J = 1.5 Hz, 1H), 7.96 (d, J = 9.0 Hz, 1H), 7.71 (brs, 1H), 7.50 (dd, J = 9.0, 1.8 Hz, 1H), 7.34 (d, J = 4.1 Hz, 1H), 4.93–4.89 (m, 1H), 4.34–4.26 (m, 1H), 4.09–3.84 (m, 3H), 2.78–2.66 (m, 2H). 13C NMR (75 MHz, CDCl3) δ 171.21, 167.71, 167.20, 150.65, 148.75, 140.52, 131.76, 131.13, 128.82, 128.67, 126.57, 125.33, 124.32, 122.79, 122.02, 119.02, 116.51, 60.38, 52.38, 51.95, 51.53, 44.38, 42.11, 41.31, 37.44, 37.10, 36.76, 36.04, 34.64, 34.48, 31.55, 29.03, 27.63, 26.88, 25.25, 22.62, 22.59, 21.00, 20.67, 20.41. HRMS calcd for C18H15BrF2N3O2+, 422.0310[M + H]+; found, 422.0314.

3c

Yield: 35.6%. Rf: 0.43 (100% ethyl acetate); FC conditions: (100% ethyl acetate). 1H NMR (300 MHz, CD3OD) δ: 8.98 (d, J = 4.3 Hz, 1H), 8.32 (d, J = 8.3 Hz, 1H), 8.12 (d, J = 7.4 Hz, 1H), 7.67 (d, J = 4.3 Hz, 1H), 7.50 (t, J = 8.0 Hz, 1H), 5.17 (dd, J = 9.0, 2.9 Hz, 1H), 4.34–4.07 (m, 3H), 2.99–2.77 (m, 2H).13C NMR (75 MHz, CDCl3) δ: 167.50, 167.11, 150.51, 145.14, 141.40, 133.99, 128.17, 125.24, 119.71, 116.25, 60.39, 44.35, 42.21, 21.04. HRMS calcd for C18H15BrF2N3O2+, 422.0310[M + H]+; found, 422.0311.

General Procedures of 2a–c

To a stirred solution of 3a, 3b, or 3c (0.2 g, 0.47 mmol) in toluene (10 mL) at room temperature were added subsequently bis(triphenylphosphine)palladium(II) chloride (80 mg, 0.07 mmol), lithium chloride (30 mg, 0.71 mmol), and hexamethyldistannane (0.78 g, 2.4 mmol) were added subsequently. The solution was stirred at 110 °C for 1.5 h. The reaction solution was concentrated by vacuum and purified by FC to give 2a, 2b, or 2c.

2a

Yield: 21.5%. Rf: 0.54 (100% ethyl acetate); FC conditions: (100% ethyl acetate). 1H NMR (300 MHz, CDCl3) δ: 8.71 (d, J = 3.8 Hz, 1H), 8.32 (s, 1H), 7.97 (d, J = 8.2 Hz, 1H), 7.80 (d, J = 8.2 Hz, 1H), 7.48 (s, 1H), 7.34 (d, J = 4.0 Hz, 1H), 4.86 (t, J = 6.2 Hz, 1H), 4.34–4.26 (m, 1H), 4.12–3.83 (m, 3H), 2.73–2.65 (m, 2H), 0.42–0.24 (m, 9H). 13C NMR (75 MHz, CDCl3) δ 171.20, 167.83, 167.47, 149.58, 148.40, 143.54, 140.28, 136.74, 132.82, 128.54, 125.28, 123.79, 118.73, 116.27, 60.37, 52.39, 51.97, 51.54, 44.29, 42.12, 37.38, 36.87, 21.01, 14.15, −7.01, −9.31, −11.61. HRMS calcd for C21H24F2N3O2Sn+, 508.0853 [M + H]+; found, 508.0856.

2b

Yield: 33.4%. Rf: 0.53 (100% ethyl acetate); FC conditions: (100% ethyl acetate). 1H NMR (300 MHz, CDCl3) δ: 8.88 (d, J = 3.9 Hz, 1H), 8.39–8.10 (m, 2H), 7.72 (d, J = 8.1 Hz, 1H), 7.46 (d, J = 3.7 Hz, 1H), 7.28 (s, 1H), 4.94 (t, J = 6.1 Hz, 1H), 4.42–4.34 (m, 1H), 4.05–3.90 (m, 3H), 2.79–2.71 (m, 2H), 0.48–0.29 (m, 9H). 13C NMR (75 MHz, CDCl3) δ: 171.19, 167.55, 149.64, 147.82, 146.15, 140.55, 137.76, 134.52, 123.94, 118.79, 116.14, 60.39, 52.45, 52.02, 51.59, 44.29, 42.20, 37.44, 36.93, 29.68, 21.04, 14.18, 7.16, −9.46, −11.77. HRMS calcd for C21H24F2N3O2Sn+, 508.0853 [M + H]+; found, 508.0851.

2c

Yield: 30.1%. Rf: 0.23 (petroleum ether/ethyl acetate = 1/2); FC conditions: (petroleum ether/ethyl acetate = 1/2). 1H NMR (300 MHz, CDCl3) δ: 8.87 (d, J = 3.0 Hz, 1H), 8.20 (d, J = 7.3 Hz, 1H), 7.93 (d, J = 5.5 Hz, 1H), 7.49–7.44 (m, 2H), 7.16 (s, 1H), 4.94 (s, 1H), 4.42–4.36 (m, 1H), 4.13–4.02 (m, 3H), 2.77–2.75 (d, J = 6.2 Hz, 2H), 0.47–0.29 (m, 9H).13C NMR (75 MHz, CDCl3) δ: 167.99, 167.69, 153.26, 148.69, 148.38, 140.78, 137.71, 127.36, 125.12, 123.91, 118.44, 116.33, 60.42, 52.39, 51.96, 51.53, 44.29, 42.18, 37.49, 36.99, 21.06, 14.19, −5.86, −8.27, −10.69. HRMS calcd for C21H24F2N3O2Sn+, 508.0853 [M + H]+; found, 508.0849.

General Procedures of 1a–c

In a nitrogen atmosphere, 2a, 2b, or 2c (50 mg, 0.10 mmol), Cu(OTf)2 (71 mg, 0.19 mmol), KF (44 mg, 0.76 mmol), and pyridine (0.12 mL, 1.47 mmol) were weighed into a flask equipped with a magnetic stir bar. DMA (10 mL) was added. The reaction mixture was stirred at 140 °C for 4 h. The resulting solution was cooled to room temperature, diluted with diethyl ether, and washed with water (2 × 20 mL) and brine (20 mL). The organic extracts were dried, concentrated, and purified by FC.

1a

Yield: 29.6%. Rf: 0.52 (100% ethyl acetate); FC conditions: (100% ethyl acetate); purity > 95% (detected by HPLC). 1H NMR (300 MHz, CDCl3) δ: 8.85 (d, J = 4.2 Hz, 1H), 8.11 (dd, J = 9.1, 5.5 Hz, 1H), 7.94 (dd, J = 9.9, 2.7 Hz, 1H), 7.49–7.53 (m, 2H), 7.34 (s, 1H), 4.98–4.94 (m, 1H), 4.42 (dd, J = 17.5, 5.6 Hz, 1H), 4.23–3.88 (m, 4H), 2.97–2.67 (m, 2H), 2.28 (d, J = 15.5 Hz, 1H), 2.03 (s, 2H), 1.32–1.22 (m, 3H).13C NMR (75 MHz, CDCl3) δ: 171.18, 167.43, 167.08, 162.84, 159.53, 148.85, 148.81, 145.61, 139.94, 139.86, 132.29, 132.17, 125.30, 125.16, 120.79, 120.44, 119.45, 116.09, 109.19, 108.87, 60.39, 52.49, 52.06, 51.63, 44.36, 42.24, 37.64, 37.30, 36.96, 28.54, 21.03, 14.18. HRMS calcd for C18H15F3N3O2+ 362.1111 [M + H]+; found, 362.1113.

1b

Yield: 32.6%. Rf: 0.54 (100% ethyl acetate); FC conditions: (100% ethyl acetate); purity > 95% (detected by HPLC). 1H NMR (300 MHz, CDCl3) δ: 8.94 (s, 1H), 8.31 (dd, J = 9.0, 6.0 Hz, 1H), 7.76 (d, J = 7.5 Hz, 1H), 7.49 (d, J = 3.8 Hz, 1H), 7.40 (t, J = 7.4 Hz, 1H), 7.18 (s, 1H), 5.00 (t, J = 6.3 Hz, 1H), 4.51–3.91 (m, 4H), 2.86–2.76 (m, 2H). 13C NMR (75 MHz, CDCl3) δ: 167.29, 167.11, 159.70, 156.29, 149.93, 140.44, 139.07, 138.91, 128.46, 127.80, 127.69, 125.84, 125.13, 121.80, 120.92, 119.74, 116.00, 114.34, 114.09, 53.40, 44.34, 42.26, 29.67. HRMS calcd for C18H15F3N3O2+ 362.1111[M + H]+; found, 362.1112.

1c

Yield: 31.6%. Rf: 0.51 (100% ethyl acetate); FC conditions: (100% ethyl acetate); purity > 95% (detected by HPLC). 1H NMR (300 MHz, CDCl3) δ 8.78 (d, J = 3.8 Hz, 1H), 7.95 (d, J = 8.3 Hz, 1H), 7.71 (s, 1H), 7.53–7.32 (m, 3H), 4.96 (dd, J = 8.0, 4.6 Hz, 1H), 4.36 (dd, J = 17.2, 5.6 Hz, 1H), 4.01 (dd, J = 13.9, 7.7 Hz, 3H), 2.88–2.62 (m, 2H).13C NMR (75 MHz, CDCl3) δ: 166.34, 166.11, 158.63, 155.22, 148.90, 139.45, 139.31, 137.98, 137.82, 127.46, 126.78, 126.67, 124.81, 124.13, 120.79, 119.98, 119.92, 118.74, 115.04, 113.34, 113.09, 52.41, 51.48, 51.05, 50.62, 43.33, 41.24, 36.64, 36.31, 35.97, 28.67. HRMS calcd for C18H15F3N3O2+ 362.1111[M + H]+; found, 362.1116.

Molecular Docking

1a–c and UAMC1110 were generated as three-dimensional mol2 files using Chem3D Pro 14.0 software (Cambridge Soft), and the most stable configurations were obtained using MM2 calculations optimization. The output files were converted to .pdb files with OpenBabel 2.4.1. The proteins (PDB ID: 1Z68 and 6Y0F) was prepared by addition of hydrogen atoms by AutoDockTools-1.5.6 (Scripps Research Institute). All ligand bonds were identified as flexible. The FAP catalyst residues Ser624, Asp702, and His734 triad were incorporated into the grid box for reference of the binding site of Linagliptin, and the docking parameters were determined for molecular docking. The docking grid box was generated using AutoGrid 4.0 with 20 × 20× 24 Å3 dimensions and a grid spacing of 1.000 Å, which was large enough to cover the catalytic site of FAPα. The center of the grid box was allocated using x, y, and z coordinates of 43.528, −71.650, and 82.800, respectively. Autodock Vina software (Scripps Research Institute) was used for docking experiments to study the influence of fluorine atoms introduced into positions 6, 7, and 8 of UAMC1110 on the affinity of FAP.28 The lowest-energy/top-ranked docked pose for each compound was visualized using the PyMol software.

IC50 Determination of Ligand for FAP

Assay buffer (25 mM Tris, 250 mM NaCl, pH 7.4) was used for dilution and reaction. Recombinant human FAP protein was purchased from Bio-Techne Co. Ltd. (Cat. 3715-SE-010) with initial concentration 0.2 mg/mL and diluted to 0.4 μg/mL. The substrate Gly-Pro-AMC (Macklin) was diluted to 40 μM. The concentrations of 1a–c and UAMC1110 were diluted from 40 μM to 4 pM with eight gradients. The compound 1a, 1b, 1c, or UAMC1110 (25 μL) and substrate (25 μL) were equally added to 96-well plates, and then FAP protein (50 μL) was added and incubated at 37 °C for 1 h. Each concentration was performed in triplicate. The fluorescence intensity was measured (Ex/Em = 380/460 nm). IC50 value was defined as the inhibitor concentration that caused a 50% decrease in activity under assay condition. Data were processed by GraphPad Prism and dose–response was used to calculate IC50.

Radiolabeling

[18F]Fluoride solution in [18O]H2O and [68Ga]Gallium was purchased from HTA Co., Ltd. Solid-phase extraction (SPE) cartridges such as Sep-Pak QMA Light and Oasis HLB cartridges were purchased from Waters (Milford, MA). HPLC was performed on Agilent 1100 series system with different HPLC columns. The radiolabeling of [68Ga]FAPI04 followed the method of the published paper.11

The [18F]fluoride (30 mCi) was dried three times azeotropically with 1 mL of ACN at 140 °C under a flow of nitrogen. In a typical reaction, 40 μL (0.02 mmol) of copper(II) trifluoromethanesulfonate and 60 μL of pyridine were added to the vial containing dried [18F]fluoride. Next, 1 mg of precursor 2a, 2b, or 2c dissolved in 0.4 mL of DMA was added to the above solution. The reaction solution was held at 140 °C for 10 min. The mixture was then cooled, and 9 mL of water. The mixture was loaded onto an activated Oasis HLB 3 cm3 cartridge, and washed with 10 mL of water. The crude product [18F]1a, [18F]1b, or [18F]1c was eluted with 1 mL of ethanol, diluted with 1 mL of deionized water, loaded onto a semipreparative column (9.4 mm × 250 mm, 5 μm, ZORBAX Eclipse XDB-C18, Agilent), and eluted with gradient mobile phase (A = 0.1% formic acid, B = ACN, 90% A in 0–2 min, 90–20% A in 2–30 min, 20% A in 30–31 min, 20–90% A in 31–31.1 min, 90% A in 31.1–35 min) at a flow rate of 3 mL/min for separation. The fraction containing [18F]1a, [18F]1b, or [18F]1c at 21.53, 21.56, and 20.49 min, respectively, was collected, diluted with DI water (30 mL), and loaded on a C18 Sep-Pak. The SepPak was washed with DI water (10 mL) and dried. The product was eluted with ethanol (1 mL), concentrated, then diluted with 1 mL 0.9% W/V saline containing 5% ethanol, and passed through a sterile membrane filter (0.22 μm). Finally, a formulated solution ready for administration was obtained.

Partition Coefficient and In Vivo Stability

Partition Coefficient (Log P)

The partition coefficients were measured by mixing [18F]1a–c (37 kBq) with 3 mL each of 1-octanol and buffer (pH 7.4, 0.1 M phosphate) in a test tube. The test tube was then vortexed for 2 min and centrifuged for 10 min at room temperature. Samples (2 mL) from the 1-octanol and buffer layers were weighed and counted on a gamma counter, respectively. The partition coefficient was determined by calculating the ratio of counts per milligram in octanol to that of the buffer. Samples of the 1-octanol layer were repartitioned until consistent partition coefficient values were obtained. The measurement was repeated three times.

In Vivo Stability

ICR mice (male, weight, 30–34 g, 6–7 weeks old) were injected with 37MBq [18F]1a–c (5% ethanol). The mice were sacrificed under isoflurane anesthesia at 30 min after the injection. Blood was collected to heparin sodium treated-EP tube and centrifuged at 5000 rpm for 2 min at 4 °C to separate the plasma. Equal volume ACN was added to plasma and the mixture was vortexed for 10 s and deproteinized by centrifugation at 14000 rpm for 5 min. Supernatant was collected and resuspended in equal volume ACN and centrifuged for another 5 min to precipitate proteins. The supernatant from plasma was concentrated by nitrogen and injected into HPLC system with radioactivity detector and using a semipreparative column (9.4 mm × 250 mm, 5 μm, ZORBAX Eclipse XDB-C18, Agilent) eluting with gradient mobile phase (A = 0.1% formic acid, B = ACN, 90% A in 0–2 min, 90–20% A in 2–30 min, 20% A in 30–31 min, 20–90% A in 31–31.1 min, 90% A in 31.1–35 min) at a flow rate of 3 mL/min.

Cellular Uptake

Briefly, A549-FAP cells were seeded in 12-well plates and cultivated for 48 h to a final confluence of approximately 80–90% (3–8 × 105 cells/well). The medium was replaced by using 1 mL of fresh medium without fetal bovine serum. [18F]1a,b (0.2 MBq) was added to the cell culture and incubated at time intervals ranging from 5–120 min at 37 °C. For blocking studies, the A549-FAP cells were pretreated with UAMC1110 for 5 min as a competitor (2.5 μmol/mL) before the addition of [18F]1a,b (about 0.2 MBq). For internalization experiments, A549-FAP cells were incubated with the radiolabeled compound for 60 min at 37 °C and was terminated by removing the medium from the cells and washing twice with 1 mL of PBS. Subsequently, cells were incubated with 1 mL of glycine HCl (1 mol/L, pH 2.2) for 10 min at 37 °C to remove the surface-bound activity. Next, the cells were washed with 2 mL of ice-cold PBS and lysed with 1.4 mL of lysis buffer (1 mol/L NaOH, 0.2% SDS) to determine the internalized fraction. For efflux experiments, the radioactive medium was removed after incubation for 60 min and replaced with nonradioactive medium over time intervals ranging from 0 to 120 min. In all experiments, the cells were washed twice with 1 mL of PBS (pH 7.4) and subsequently lysed with 1.4 mL of lysis buffer. Radioactivity was determined using a γ-counter and the results are expressed as % ID/1 mio. cells. Each experiment was performed three times with three replicates for each independent experiment.

Biodistribution

Nude mice (female, weight 12–16 g) bearing U87MG xenografts were purchased from Guangdong GemPharmatech Co., Ltd. All animal experiments were approved by Animal Experiments and Experimental Animal Welfare Committee of Capital Medical University and carried out according to the guidelines of the Animal Welfare Act (AEEI-2022–150). Approximately 1.13 MBq [18F]1a in 0.1 mL of solution was administrated via tail vein injection in conscious animals. Five mice per group were euthanized at the 60 min time point, and organs of interest were collected and weighed in preweighed plastic bags. Activities in organs were measured by a WIZARD2 2480 automatic γ-counter (PerkinElmer, ∼70% efficiency). A 0.1 mL (same volume as injected) sample of a 100× dilution of the injected dose as 1% ID was counted under the same treatment. The results were expressed as the percent uptake of injected dose per gram of tissue (% ID/g) and presented as the mean ± SD.

MicroPET-CT Imaging

Nude mice (female, weight 12–16 g) bearing U87MG xenografts and intracranial tumors were purchased from Guangdong GemPharmatech Co., Ltd. Data were recorded on a Madiclab PSA146 PET/CT/FMT instrument. Dynamic microPET-CT imaging studies were conducted with [18F]1a, [18F]1b, or [68Ga]FAPI04, similar to that reported previously.29 A total of 8–11 MBq of activity was injected intravenously via the lateral tail vein. For nude mice bearing U87MG xenograft tumors, 1 h dynamic PET images were collected after administration of 8–11 MBq of [18F]1a, [18F]1b or [68Ga]FAPI04. And static PET imaging was collected at 2 or 4 h postinjection. For the blocking study, U87MG subcutaneous tumor mice were injected with UAMC1110 (2 mg/kg) 5 min in advance, followed by the injection of [18F]1a or [18F]1b through the tail vein.

For nude mice bearing U87MG intracranial tumors, dynamic PET images were collected for 1 h after the administration of 8–11 MBq [18F]1a or [18F]1b. Static PET imaging was collected at 2 and 4 h postinjection. All animals were sedated with isoflurane anesthesia (2–3%, 1 L/min oxygen) and then placed on a heating pad to maintain body temperature throughout the procedure. The animals were visually monitored for breathing and any other signs of distress throughout the entire imaging period. The data acquisition began after an intravenous injection of the tracer. MicroPET-CT images were analyzed by using Pmod software (version 4.0, PMOD Technologies Ltd., Zurich, Switzerland). Each microPET image was manually coregistered to the Mirrione T2 mouse brain template by using rigid body transformation.30 Then, the resulting transformation parameters were applied to the corresponding microPET images. Four volumes of interest (VOIs) were selected from the Mirrione atlas.30 TACs were extracted from all the VOIs and performed as the percentage injection dose per cubic centimeter (% ID/g).

Statistical Analysis

Quantitative data were described as mean ± SD, and statistical differences between groups were analyzed by a Student’s t test using GraphPad Prism 8.0 software. A p-value less than 0.05 was statistically significant.

Results

Design of 1a–c

The initial skeleton of the current FAP tracer design starts from UAMC1110, which has a molecular weight of 343 (<500) and a theoretical calculation of CLogP = 1.5. This molecule meets the Lipinski brain entry rule,31 so it is expected to penetrate the blood–brain barrier. Therefore, by introducing a fluorine atom into the UAMC1110 skeleton, the newly designed compound still meets Lipinski’s rule and is also expected to penetrate the blood-brain barrier. Consequently, 1a–c introducing fluorine atoms at different positions on the quinoline ring of UAMC1110 were designed.

Molecular Docking

Currently, only Linagliptin and human FAPα complex crystal structures (PDB ID:6Y0F) is available. According to the X-ray crystal structure of FAPα (6Y0F and 1Z68),32 each subunit of the protein contains two topologically distinct domains: the β-propeller (residues 54–492) and the α/β-hydrolase domain (residues 27–53 and 493–760). The FAP catalytic center consisted of Ser624, Asp702 and His734 residues and was located at the junction of β-propeller and α/β-hydrolase domains. Inhibitors can approach the active site in two ways: through a central hole formed by the eight blades of the β-propeller domain or through a cavity formed between the β-propeller and the α/β hydrolase domains.32

Although Linagliptin and UAMC1110 both fit in the catalytic active center of FAPα, their molecular skeletons are very different, and their modes of binding are not exactly similar. Several studies of ligands docking with FAPα have been reported.33−36 Trujillow-Benitez et al. investigated the effect of the distance between the boron residue of pyrrolidone 2-boric acid derivatives (iFAP) and Ser-624 of FAPα on their inhibitory activity.36 Among the ligands mentioned above, the structure of iFAP is more similar to that of the compounds in this paper.

As shown in Figure 2A, Tables S1–S4, and Figure S1, 1a–c was fitted to the catalytic active center of the protein, and their molecular orientations were like that of UAMC1110. Their docking scores were close due to their molecular structures being very similar (−10.2, −10.0, −9.7, and −9.9 kcal/mol, 1a–c and UAMC1110, respectively). The quinoline ring of 1a had π–π stacking with residue Phe350 and hydrophobic interaction with Phe351 of the protein. This π–π interaction was not observed in the interactions of 1b, 1c, and UAMC1110 with FAP. More importantly, the cyano group of pyrrolidines interacted with residues Ser624 and His734, the key amino acid residues that affected enzyme activity. In addition, the pyrrolidine-N-carbonyl group of 1a has a double hydrogen bond with Tyr541 and Tyr660, but this hydrogen bond was not formed in the interactions of 1b, 1c, and UAMC1110 with the protein. In summary, 1a–c and UAMC1110 had similar binding modes, and 1a has a slightly better affinity than 1b, 1c, and UAMC1110.

Figure 2.

Figure 2

Open in a new tab

Docking analysis of the interactions of the ligands with FAPα. (a) Overlay of UAMC1110 (green) and 1a–c (blue, purple and yellow) bound to the active site of FAP protein. (b) 1a was docked with FAPα in a three-dimensional view. (gray dotted line: hydrophobic interactions; green lines: π–π stacking interactions; blue lines: hydrogen bonds).

Synthesis of 1a–c

After the structure of the target compound was established, the synthesis route of 1a–c was designed by retrosynthesis (Scheme 1). (S)-4,4-Difluoro-1-glycylpyrrolidine-2-carbonitrile was prepared according to reported methods.9 It is then coupled with bromoquinolinic acid 4a–c to give 3a–c in the presence of EDCI and HOBt. 3a–c react with hexamethylditin to obtain precursor 2a–c under the catalysis of Pd(PPh3)2Cl2. 2a–c were fluorinated to obtain the target compounds 1a–c. The synthesis yield of the entire route was not optimized in detail.

Scheme 1. Synthesis of 1a–c.

Scheme 1

Open in a new tab

Reagents and conditions: (a) (S)-4,4-difluoro-1-glycylpyrrolidine-2-carbonitrile, EDCI, HOBt, Et3N, DCM, rt, overnight; (b) Pd(PPh3)2Cl2, LiCl, hexamethyldistannane, toluene, 110 °C, 1.5 h; (c) Cu(OTf)2, KF, Py, DMA, 140 °C, 4 h.

Enzyme Inhibition Assay

To determine the affinity for FAP, we measured the inhibitory ability of 1a–c and UAMC1110 to affect the FAP catalytic activity of hydrolysis of the fluorescent FAP substrate (Ala-Pro-AMC). As Figure 3 shows, the IC50s of 1a–c and UAMC1110 to FAP are 6.11, 15.1, 92.1, and 4.17 nM, respectively. These compounds all exhibited high affinity for FAP and are worthy of further biological evaluation.

Figure 3.

Figure 3

Open in a new tab

Dose–response curves of the inhibition of FAP by UAMC1110 and 1a–c.

Radiolabeling

Taking the synthesis conditions of [18F]1a as an example (Scheme 2), we optimized the radiolabeling temperature, reactant concentration, and ratio using the method reported.37 According to the method in Table 1, entry1, [18F]fluoride were first rinsed with KOTf/K2CO3, dried, and then reacted with 2a in the present of Cu(OTf)2 and pyridine in DMA (Figure S1). The reaction solution was first diluted with 9 mL of H2O, and then purified by Oasis HLB to obtain 1 mL of ethanol eluent. This eluent was further diluted with 1 mL of H2O, and then purified by HPLC to obtain [18F]1a with a yield of 1.86 ± 1.21%. After 2 mL of clear loading solution was injected into HPLC, the column pressure increased rapidly to the point that HPLC could not function properly, so the crude product cannot be purified further. The KOTf/K2CO3 eluent solution has poor efficiency in rinsing [18F]fluoride, which will result in about 10% [18F]fluoride remaining on the QMA column. On further increasing the temperature to 140 °C, although the yield was improved to 6.79%, column high pressure still occurred during purification (entry 2, Table 1). Therefore, we tried to directly evaporate the aqueous solution of [18F]fluoride without QMA leaching. The dried fluoride ions were added to precursor 2a and Cu(OTf)2 dissolved by DMA and pyridine, and the total volume of the reaction was 0.5 mL (entry 3, Table 1). The reaction solution was first simply purified by Oasis HLB, and then purified by HPLC to obtain [18F]1a with a yield of 12.39 ± 3.31%. Under this condition, the post reaction treatment was simple, and there was no high column pressure phenomenon, indicating that KOTf/K2CO3 may be one of the reasons for high column pressure. Further use of the optimal ratio of catalysts in the literature,37 the reaction yield was improved to 15.67 ± 3.72% (entry 4, Table 1). The optimized radiolabeling method improves the yield and is expected to broaden the application of trimethyltin labeled precursor. [18F]1b and [18F]1c were successfully prepared following the radiolabeling conditions of [18F]1a (Figures S2 and S3). The specific activity of [18F]1a, [18F]1b, or [18F]1c was >120 MBq/nmol (n = 3) at the end of synthesis. The labeling conditions of well-known FAPI’s PET tracer [68Ga]FAPI04 are based on the reported methods.11

Scheme 2. Synthesis of [18F]1a–c.

Scheme 2

Open in a new tab

Reagents and conditions: (a) [18F]fluoride with Cu(OTf)2, Py, DMA, 140 °C, 10 min.

Table 1. Optimal Conditions for Radiosynthesis [18F]1aa.

entry reaction conditions yield
1 (1) KOTf/K2CO3, 110 °C, (2) 2 mM 2a, 8 mM Cu(OTf)2, 32 mM Py, 110 °C,10 min 1.86 ± 1.21%
2 (1) KOTf/K2CO3, 140 °C, (2) 2 mM 2a, 4 mM Cu(OTf)2, 32 mM Py, 140 °C, 10 min 6.79 ± 2.15%
3 (1) 140 °C, (2) 4 mM 2a, 4 mM Cu(OTf)2, 48 mM Py, 140 °C, 10 min 12.39 ± 3.31%
4 (1) 140 °C, (2) 4 mM 2a, 8 mM Cu(OTf)2, 52 mM Py, 140 °C, 10 min 15.67 ± 3.72%
Open in a new tab
a

Starting activities: about 740 MBq, yield corrected, n = 3.

Partition Coefficient and In Vivo Stability

After the radioactive target compound was obtained, the lipid–water partition coefficient was investigated. The logP values of [18F]1a–c are 1.23, 1.21, and 1.19, respectively, which fall within the range considered to be compatible with good blood-brain barrier permeability (1 < logP < 4).31,38,39 Further in vivo stability experiments showed that no significant metabolites were found in [18F]1a and [18F]1b in the plasma at 30 min post injection, while [18F]1c exhibited obvious defluorination (Figures 4A and S4). Therefore, [18F]1c was not subject to further detailed biological evaluation and was only further validated in PET imaging.

Figure 4.

Figure 4

Open in a new tab

(a) Radio-HPLC chromatogram of [18F]1a–c and their primary radiometabolites in plasma at 30 min postinjection. (b) Cellular uptake of [18F]1a,b and [68Ga]FAPI04 in A549-FAP cells with or without UAMC1110. (c) Internalized and membrane-bound [18F]1a,b and [68Ga]FAPI04 in A549-FAP cells. (d) Incubation of A549-FAP cells with [18F]1a,b and [68Ga]FAPI04 for 60 min followed by incubation for 30, 60, and 120 min with a compound-free medium.

Cell Uptake

In vitro cell experiments are performed to evaluate the specificity of [18F]1a,b for FAP by using highly FAP-expressing A549-FAP cells. Shown in Figure 4b, [18F]1a, [18F]1b, and [68Ga]FAPI04 enter cells, with uptake of 4.85, 5.15%, and 3.41 ID/1 mio. cells, respectively, within 5 min. As the time extended to 60 min, the uptake of three compounds increased. The difference is that at 120 min, the uptake of [18F]1a,b has decreased, while [68Ga]FAPI04 continues to increase. With pretreated UAMC1110, the uptake of [18F]1a,b and [68Ga]FAPI04 decreased by 86.9%, 94.2%, and 86.2%, respectively(Figure 4b), which demonstrated the specificity of binding to FAP. The internalization test showed that the internalization rates of [18F]1a,b and [68Ga]FAPI04 were comparable, with 61%, 57%, and 64% respectively (Figure 4c). Further efflux experiments showed that the efflux rate of [18F]1a,b was faster than that of [68Ga]FAPI04 at 60 min (Figure 4d). From cell experiments, [18F]1a,b are comparable to [68Ga]FAPI04 in cell uptake and internalization, which are worthy of further study.

U87MG Subcutaneous Tumor PET-CT Imaging

Dynamic and static PET imaging studies were performed in U87MG tumor-bearing nude mice to investigate the pharmacokinetics of [18F]1a,b and [68Ga]FAPI04. Figure 5a,b showed that [18F]1a,b can quickly cross the blood–brain barrier (uptake at 2.5 min, 1.61 ± 0.43, 1.05 ± 0.32% ID/g, respectively) and remain for a long time (uptake at 62.5 min, 1.06 ± 0.23, 1.09 ± 0.25% ID/g, respectively) (Figure S5). [18F]1a,b can be retained in the brain because FAP is also lowly expressed in the cerebellum, cortex, and hippocampus (data from protein atlas database). However, [68Ga]FAPI04 maintains a low uptake (at 2.5 and 62.5 min, 0.65 ± 0.29, 0.21 ± 0.10% ID/g, respectively) on the brain because it cannot penetrate the blood-brain barrier (Figure S5). Therefore, [18F]1a,b are expected to be applied to brain imaging. As can be seen from Figure S6, [18F]1a can be rapidly taken up by the tumor and has good retention (uptake = 5.44 ± 0.98, 5.41 ± 1.20, 5.82 ± 1.28, 0.83 ± 0.11% ID/g at 30, 60, 120, and 240 min, respectively). [18F]1b had lower tumor uptake compared to [18F]1a, but still maintained moderate uptake and good retention (3.46 ± 0.59, 3.52 ± 0.75, 2.15 ± 0.45, 1.62 ± 0.23% ID/g at 30, 60, 120, and 240 min, respectively, Figure S6). Consistent with previous reports,11 [68Ga]FAPI04 had low initial tumor uptake and poor retention (Figures 5c and S6). Compared with [68Ga]FAPI04, the muscle uptake of [18F]1a,b was higher, and therefore, the tumor-to-muscle ratio of [68Ga]FAPI04 was better than those of [18F]1a,b at 30 min post injection (Figure S7). From the tumor-to-liver ratio, it can be seen that the value of [18F]1a is much closer to [68Ga]FAPI04 than that of [18F]1b (Figure 5d), which is expected to diagnose liver-related diseases. Since [68Ga]FAPI04 is metabolized by the kidney, the tumor-to-kidney ratio of [18F]1a,b is greater than that of [68Ga]FAPI04, which is expected to diagnose kidney-related diseases (Figure 5e). Consistent with the high expression of FAP in joints,40[18F]1a exhibited high uptake in joints (knees and shoulders), with uptake values of 1.42 ± 0.16, 1.48 ± 0.13% ID/g at 30 min post injection, respectively(Figure 5a). However, [18F]1b did not show a significant bone joint uptake. [18F]1a,b showed no significant defluorination, while [18F]1c exhibited significant in vivo defluorination consistent with in vivo stability results (Figure S8). The blocking study for the tested [18F]1a,b in mice bearing U87MG-FAP tumors indicated a remarkable decrease in uptake by tumors and joints (Figure 5a,b), demonstrating the specificity of radiotracers for FAP. As can be seen from Figures 5a,b and S9, [18F]1a,b are mainly metabolized by the liver and kidney. Compared with [18F]1b, [18F]1a visualized tumors with higher uptake, higher retention, higher tumor-to-muscle ratio, tumor-to-liver ratio, and tumor-to-kidney ratio. Therefore, [18F]1a is suitable for PET imaging targeting FAP.

Figure 5.

Figure 5

Open in a new tab

Representative U87MG subcutaneous tumor microPET-CT images of [18F]1a (a), [18F]1b (b), and [68Ga]FAPI04 (c) with or without UAMC1110 at 30, 60, 120, or 240 min post injection. The white arrows represent the location of tumors. The tumor-to-liver ratio (d) and tumor-to-kidney ratio (e) of [18F]1a, [18F]1b and [68Ga]FAPI04 in U87MG subcutaneous tumors.

U87MG Subcutaneous Tumor Biodistribution

Further biodistribution experiments of tumor-bearing mice were performed to verify the results of the PET imaging. Consistent with PET results, [18F]1a,b are highly uptaken in the gallbladder and small intestine, which are mainly metabolized by the liver and kidneys. Compared with [18F]1b, [18F]1a had lower uptake in muscle, bone, lung, liver, stomach, blood, and pancreas at 60 min post injection (Figure 6). Therefore, consistent with the U87MG subcutaneous tumor imaging results, [18F]1a had higher tumor-to-liver ratio, tumor-to-muscle ratio, and tumor-to-blood ratio than [18F]1b (2.25 vs 1.38, 9.72 vs 4.59, 10.12 vs 5.19, respectively). The gallbladder uptake (133.28 ± 26.77 vs 109.7 ± 47.45, respectively) and tumor-to-kidney ratio (4.67 vs 3.25, respectively) of [18F]1a and [18F]1b are comparable. Therefore, [18F]1a is suitable as a PET tracer targeting FAP.

Figure 6.

Figure 6

Open in a new tab

Biodistribution studies of [18F]1a and [18F]1b in mice bearing U87MG xenograft at 60 min time points after intravenous injection (n = 5). *, p < 0.05 and **, p < 0.01.

U87MG Intracranial Tumor PET-CT Imaging

Further U87MG intracranial tumor PET imaging found that [18F]1a-b could penetrate the blood–brain barrier. To confirm whether there was an orthotopic tumor in the brain, we sacrificed the orthotopic mouse model from Figures 7 and Figure S10. The uptake of [18F]1b by the tumor was higher than that of [18F]1a, but the uptake of [18F]1b in the contralateral hemisphere was also higher than that of [18F]1a (Figure 7c). Therefore, the contrast between tumor and normal brain tissue in [18F]1a within 1 h of administration was higher than [18F]1b. As can be seen from Figure 7a, [18F]1a exhibited high contrast PET imaging at 30 and 60 min post injection, [18F]1b did not. [18F]1b was consistent with cellular uptake and had high tumor retention, and [18F]1b also exhibited high-contrast PET imaging with time extension to 120 min post injection (uptake, tumor 0.87 ± 0.04% ID/g vs brain 0.41 ± 0.03% ID/g, Figure 7b). With further extension of time, [18F]1b could still visualize gliomas at high contrast at 4 h (uptake, tumor 0.64 ± 0.03% ID/g vs brain 0.35 ± 0.02% ID/g, Figure 7b), when [18F]1a has been cleared away (uptake, tumor 0.17 ± 0.02% ID/g vs brain 0.22 ± 0.02% ID/g, Figure 7a). Comparing the imaging characteristics of [18F]1a,b, we found that [18F]1a is suitable for imaging gliomas.

Figure 7.

Figure 7

Open in a new tab

Representative U87MG intracranial tumor microPET-CT images of [18F]1a (a) and [18F]1b (b) at 30, 60, 120, or 240 min post injection. The white arrows represent the location of tumors. (c) Tumor and contralateral hemisphere of [18F]1a and [18F]1b in the U87MG intracranial tumor.

Discussion

The development of radiological diagnostic and therapeutic drugs targeting FAP is a hot topic in current radiopharmaceutical research. Improving the target-to-non target ratio of tumor imaging, reducing the toxic side effects of normal tissues, and increasing the therapeutic efficiency of tumors are important goals for the development of FAP-targeted radiopharmaceuticals. By introducing albumin binder moieties or PEG side chains,33,41−45 increasing the number of target molecules,22,46−48 conjugating with other target molecules,49 and using peptide structures as alternative target molecules,18 the short tumor residence time and low target-to-nontarget ratio of [68Ga]FAPI04 was optimized, which greatly improved the application of FAP-targeted radiopharmaceuticals in tumor diagnosis and treatment. A series of related tumor diagnosis and treatment clinical trials was carried out. In the emerging FAP radiotracers, most of them are radiolabeled with 68Ga-,18F-AlF chelated with DOTA, NOTA, or DODAGA, etc. Direct 18F-labeled tracers have rarely been reported.23 The advantage of metal chelating agents is that the β+ decaying nuclide, such as 68Ga-,18F-AlF could be easily changed to β– decaying nuclide, such as 177Lu- or 90Y-, which are therapeutic nuclides with excellent properties. However, due to their large molecular weight and nonelectric neutrality, they cannot penetrate the blood-brain barrier resulting in the failure of the diagnosis of brain diseases.

Compared with metal radionuclides, 18F has huge practical application advantages in the diagnosis of brain diseases owing to the comparable atomic radius between F atom and H atom. UAMC1110 is a small molecule FAP inhibitor with high affinity to FAP. To retain the affinity to the greatest extent, we replaced the H atom with the F atom. Theoretical calculations and in vitro enzyme activity inhibition experiments have confirmed that 1a–c can bind to FAP. The introduction of [18F]fluorine on the aromatic ring is more difficult than that on the aliphatic side chain. The [18F]fluorine introduced in this project are directly induced by the one-pot method without QMA rinsing to improve the radiolabeling yield and reduce the radiolabeling steps, which is conducive to the application of aromatic ring labeled [18F]fluorine. Although the structures of 1a–c are similar, there are significant differences in in vivo stability. Since the [18F]fluorine atom of [18F]1c is attracted by the adjacent N atom, the [18F]fluorine atom of [18F]1c is more likely to leave and show in vivo instability.

In terms of in vitro cell uptake, internalization, and efflux rates, [68Ga]FAPI04 and [18F]1a,b are relatively similar in terms of tumor uptake and retention. However, in terms of tumor imaging, [18F]1a,b has high uptake and retention of tumors, which may be due to the inability of cells to fully reflect the microenvironment of the tumor. Compared to [18F]1b, [18F]1a has a lower uptake and faster clearance in normal tissues in vivo. Correspondingly, [18F]1a visualizes tumors with higher contrast in a shorter time. Therefore, from the microPET imaging of U87MG subcutaneous tumors and U87MG orthotopic gliomas, [18F]1a can clearly outline the tumor at 30 min post injection. [18F]1a can also be clearly imaged in joint sites where FAP is highly expressed. FAP has low expression in the brain; therefore, tracer that can penetrate the blood-brain barrier should have a certain uptake and retention time. [68Ga]FAPI04 was rapidly cleared in the brain, while [18F]1a has higher initial brain uptake and better retention. PET imaging of the U87MG subcutaneous tumor with high expression of FAP showed that [18F]1a had higher uptake of the tumor and longer retention time compared to [68Ga]FAPI04. From the imaging results of [18F]1a–c, the position of the fluorine atom has a great influence on the stability and pharmacokinetics of the drug, which has certain reference significance for the design of other types of tracers in the future. Furthermore, PET imaging of U87MG intracranial gliomas indicates that [18F]1a can quickly visualize gliomas with a high contrast. Importantly, [18F]1a exhibits high stability in vivo; therefore, its metabolites do not interfere with brain imaging. Based on the above analysis, [18F]1a is expected to be applied in the diagnosis of FAP related brain diseases.

Acknowledgments

We thank Prof. Tang Ganghua from the South Hospital of Southern Medical University for providing A549-FAP cell. We acknowledge Professor Ai Lin from department of Nuclear Medicine, Beijing Tiantan Hospital, Capital Medical University.

Glossary

Abbreviations

ACN

Acetonitrile

DMA

N,N-Dimethylacetamide

Et3N

Triethylamine

EDCI

N-(3-(Dimethylamino)propyl)-N′-ethylcarbodiimide hydrochloride

PBS

Phosphate-buffered saline

HOBt

1-Hydroxybenzotriazole

HPLC

High-performance liquid chromatography

HRMS

High-resolution mass spectrometry

SD

Standard deviation

TAC

Time–activity curve

PET

Positron emission tomography

PET-CT

Positron emission tomography-computed tomography

Py

Pyridine

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsptsci.3c00187.

  • Data on docking, radiolabeling, U87MG subcutaneous tumor microPET-CT imaging, NMR of compound 1a–c (PDF)

Author Contributions

⊥ Z.Y. and Y.H. contributed equally to this paper. The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

This research was funded by the Beijing Natural Science Foundation (7222299) and Medical Innovation Capability Improvement Plan of Capital Medical University (12300124). National Cancer Center, National Clinical Research Center for Cancer, Cancer Hospital, and Shenzhen Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Shenzhen (SZ2020MS008), Shenzhen High-level Hospital Construction Fund, the National Natural Science Foundation of China (82372018, 82102115 and 81701753). The Shenzhen Science and Technology Program of China (JCYJ20220818101804009), Shenzhen Clinical Research Center for Cancer (No. (2021) 287).

The authors declare no competing financial interest.

Supplementary Material

pt3c00187_si_001.pdf (1.9MB, pdf)

References

  1. Xiao Y.; Yu D. Tumor microenvironment as a therapeutic target in cancer. Pharmacol Ther. 2021, 221, 107753 10.1016/j.pharmthera.2020.107753. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Feng H.; Zhuo Y.; Zhang X.; Li Y.; Li Y.; Duan X.; Shi J.; Xu C.; Gao Y.; Yu Z. Tumor Microenvironment in Hepatocellular Carcinoma: Key Players for Immunotherapy. J. Hepatocell Carcinoma 2022, 9, 1109–1125. 10.2147/JHC.S381764. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Park J. E.; Lenter M. C.; Zimmermann R. N.; Garin-Chesa P.; Old L. J.; Rettig W. J. Fibroblast activation protein, a dual specificity serine protease expressed in reactive human tumor stromal fibroblasts. J. Biol. Chem. 1999, 274, 36505–36512. 10.1074/jbc.274.51.36505. [DOI] [PubMed] [Google Scholar]
  4. Tchou J.; Zhang P. J.; Bi Y.; Satija C.; Marjumdar R.; Stephen T. L.; Lo A.; Chen H.; Mies C.; June C. H.; Conejo-Garcia J.; Puré E. Fibroblast activation protein expression by stromal cells and tumor-associated macrophages in human breast cancer. Hum Pathol. 2013, 44, 2549–2557. 10.1016/j.humpath.2013.06.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Mayola M. F.; Thackeray J. T. The Potential of Fibroblast Activation Protein-Targeted Imaging as a Biomarker of Cardiac Remodeling and Injury. Curr. Cardiol Rep. 2023, 25, 515–523. 10.1007/s11886-023-01869-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Sieweke J. T.; Grosse G. M.; Weissenborn K.; Derda A. A.; Biber S.; Bauersachs J.; Bavendiek U.; Tillmanns J. Circulating fibroblast activation protein α is reduced in acute ischemic stroke. Front Cardiovasc Med. 2022, 9, 1064157 10.3389/fcvm.2022.1064157. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Welt S.; Divgi C. R.; Scott A. M.; Garin-Chesa P.; Finn R. D.; Graham M.; Carswell E. A.; Cohen A.; Larson S. M.; Old L. J. Antibody targeting in metastatic colon cancer: a phase I study of monoclonal antibody F19 against a cell-surface protein of reactive tumor stromal fibroblasts. J. Clin Oncol. 1994, 12, 1193–1203. 10.1200/JCO.1994.12.6.1193. [DOI] [PubMed] [Google Scholar]
  8. Jansen K.; Heirbaut L.; Cheng J. D.; Joossens J.; Ryabtsova O.; Cos P.; Maes L.; Lambeir A.-M.; De Meester I.; Augustyns K.; Van der Veken P. Selective Inhibitors of Fibroblast Activation Protein (FAP) with a (4-Quinolinoyl)-glycyl-2-cyanopyrrolidine Scaffold. ACS Med. Chem. Lett. 2013, 4, 491–496. 10.1021/ml300410d. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Jansen K.; Heirbaut L.; Verkerk R.; Cheng J. D.; Joossens J.; Cos P.; Maes L.; Lambeir A.-M.; De Meester I.; Augustyns K.; Van der Veken P. Extended Structure–Activity Relationship and Pharmacokinetic Investigation of (4-Quinolinoyl)glycyl-2-cyanopyrrolidine Inhibitors of Fibroblast Activation Protein (FAP). J. Med. Chem. 2014, 57, 3053–3074. 10.1021/jm500031w. [DOI] [PubMed] [Google Scholar]
  10. Loktev A.; Lindner T.; Mier W.; Debus J.; Altmann A.; Jäger D.; Giesel F.; Kratochwil C.; Barthe P.; Roumestand C.; Haberkorn U. A Tumor-Imaging Method Targeting Cancer-Associated Fibroblasts. J. Nucl. Med. 2018, 59, 1423–1429. 10.2967/jnumed.118.210435. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Lindner T.; Loktev A.; Altmann A.; Giesel F.; Kratochwil C.; Debus J.; Jäger D.; Mier W.; Haberkorn U. Development of Quinoline-Based Theranostic Ligands for the Targeting of Fibroblast Activation Protein. J. Nucl. Med. 2018, 59, 1415–1422. 10.2967/jnumed.118.210443. [DOI] [PubMed] [Google Scholar]
  12. Giesel F. L.; Kratochwil C.; Lindner T.; Marschalek M. M.; Loktev A.; Lehnert W.; Debus J.; Jäger D.; Flechsig P.; Altmann A.; Mier W.; Haberkorn U. 68Ga-FAPI PET/CT: Biodistribution and Preliminary Dosimetry Estimate of 2 DOTA-Containing FAP-Targeting Agents in Patients with Various Cancers. J. Nucl. Med. 2019, 60, 386–392. 10.2967/jnumed.118.215913. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Farolfi A.; Calderoni L.; Mattana F.; Mei R.; Telo S.; Fanti S.; Castellucci P. Current and Emerging Clinical Applications of PSMA PET Diagnostic Imaging for Prostate Cancer. J. Nucl. Med. 2021, 62, 596–604. 10.2967/jnumed.120.257238. [DOI] [PubMed] [Google Scholar]
  14. Hennrich U.; Eder M. [177Lu]Lu-PSMA-617 (PluvictoTM): The First FDA-Approved Radiotherapeutical for Treatment of Prostate Cancer. Pharmaceuticals 2022, 15, 1292. 10.3390/ph15101292. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Jia A. Y.; Kashani R.; Zaorsky N. G.; Spratt D. E.; Kiess A. P.; Michalski J. M.; Zoberi J. E.; Kim H.; Baumann B. C. Lutetium-177 DOTATATE: A Practical Review. Pract Radiat Oncol. 2022, 12, 305–311. 10.1016/j.prro.2022.02.002. [DOI] [PubMed] [Google Scholar]
  16. Millul J.; Bassi G.; Mock J.; Elsayed A.; Pellegrino C.; Zana A.; Dakhel Plaza S.; Nadal L.; Gloger A.; Schmidt E.; Biancofiore I.; Donckele E. J.; Samain F.; Neri D.; Cazzamalli S. An ultra-high-affinity small organic ligand of fibroblast activation protein for tumor-targeting applications. Proc. Natl. Acad. Sci. U. S. A. 2021, 118, e2101852118 10.1073/pnas.2101852118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Backhaus P.; Gierse F.; Burg M. C.; Büther F.; Asmus I.; Dorten P.; Cufe J.; Roll W.; Neri D.; Cazzamalli S.; Millul J.; Mock J.; Galbiati A.; Zana A.; Schäfers K. P.; Hermann S.; Weckesser M.; Tio J.; Wagner S.; Breyholz H. J.; Schäfers M. Translational imaging of the fibroblast activation protein (FAP) using the new ligand [68Ga]Ga-OncoFAP-DOTAGA. Eur. J. Nucl. Med. Mol. Imaging. 2022, 49, 1822–1832. 10.1007/s00259-021-05653-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Zboralski D.; Hoehne A.; Bredenbeck A.; Schumann A.; Nguyen M.; Schneider E.; Ungewiss J.; Paschke M.; Haase C.; von Hacht J. L.; Kwan T.; Lin K. K.; Lenore J.; Harding T. C.; Xiao J.; Simmons A. D.; Mohan A.-M.; Beindorff N.; Reineke U.; Smerling C.; Osterkamp F. Preclinical evaluation of FAP-2286 for fibroblast activation protein targeted radionuclide imaging and therapy. Eur. J. Nucl. Med. Mol. Imaging. 2022, 49, 3651–3667. 10.1007/s00259-022-05842-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Ruan Q.; Zhou C.; Wang Q.; Kang F.; Jiang Y.; Li G.; Feng J.; Zong S.; Zhang J.; Wang J. A Simple Kit Formulation for Preparation and Exploratory Human Studies of a Novel 99mTc-Labeled Fibroblast Activation Protein Inhibitor Tracer for Imaging of the Fibroblast Activation Protein in Cancers. Mol. Pharmaceutics 2023, 20, 2942–2950. 10.1021/acs.molpharmaceut.2c01094. [DOI] [PubMed] [Google Scholar]
  20. Ma H.; Li F.; Shen G.; Cai H.; Liu W.; Lan T.; Yang Y.; Yang J.; Liao J.; Liu N. Synthesis and Preliminary Evaluation of 131I-Labeled FAPI Tracers for Cancer Theranostics. Mol. Pharmaceutics 2021, 18, 4179–4187. 10.1021/acs.molpharmaceut.1c00566. [DOI] [PubMed] [Google Scholar]
  21. Lee I. K.; Noguera-Ortega E.; Xiao Z.; Todd L.; Scholler J.; Song D.; Liousia M.; Lohith K.; Xu K.; Edwards K. J.; et al. Monitoring Therapeutic Response to anti-FAP CAR T Cells Using [18F]AlF-FAPI-74. Clin. Cancer Res. 2022, 28, 5330–5342. 10.1158/1078-0432.CCR-22-1379. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Pang Y.; Zhao L.; Fang J.; Chen J.; Meng L.; Sun L.; Wu H.; Guo Z.; Lin Q.; Chen H. Development of FAPI Tetramers to Improve Tumor Uptake and Efficacy of FAPI Radioligand Therapy. J. Nucl. Med. 2023, 64, 1449. 10.2967/jnumed.123.265599. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Toms J.; Kogler J.; Maschauer S.; Daniel C.; Schmidkonz C.; Kuwert T.; Prante O. Targeting Fibroblast Activation Protein: Radiosynthesis and Preclinical Evaluation of an 18F-Labeled FAP Inhibitor. J. Nucl. Med. 2020, 61, 1806–1813. 10.2967/jnumed.120.242958. [DOI] [PubMed] [Google Scholar]
  24. Schmidkonz C.; Rauber S.; Atzinger A.; Agarwal R.; Gotz T. I.; Soare A.; Cordes M.; Prante O.; Bergmann C.; Kleyer A.; et al. Disentangling inflammatory from fibrotic disease activity by fibroblast activation protein imaging. Ann. Rheum Dis. 2020, 79, 1485. 10.1136/annrheumdis-2020-217408. [DOI] [PubMed] [Google Scholar]
  25. Hotta M.; Rieger A. C.; Jafarvand M. G.; Menon N.; Farolfi A.; Benz M. R.; Calais J. Non-oncologic incidental uptake on FAPI PET/CT imaging. Br J. Radiol. 2023, 96, 20220463 10.1259/bjr.20220463. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Mori Y.; Dendl K.; Cardinale J.; Kratochwil C.; Giesel F. L.; Haberkorn U. FAPI PET: Fibroblast Activation Protein Inhibitor Use in Oncologic and Nononcologic Disease. Radiology. 2023, 306, 3. 10.1148/radiol.220749. [DOI] [PubMed] [Google Scholar]
  27. Dorrier C. E.; Jones H. E.; Pintarić L.; Siegenthaler J. A.; Daneman R. Emerging roles for CNS fibroblasts in health, injury and disease. Nat. Rev. Neurosci. 2022, 23, 23–34. 10.1038/s41583-021-00525-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Trott O.; Olson A. J. AutoDock Vina: Improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. J. Comput. Chem. 2010, 31, 455–461. 10.1002/jcc.21334. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Zheng W.; Huang Y.; Chen H.; Jiang Z.; Yu Z.; Yang T.; Zhang L.; Cheng X.; Liu Y.; Liu Q.; Ji X.; Wu Z. Synthesis and In Vitro and In Vivo Evaluation of 18F-Labeled Positron Emission Tomography Tracers for Imaging Aβ Plaques. ACS Chem. Neurosci. 2023, 14, 988–1003. 10.1021/acschemneuro.3c00025. [DOI] [PubMed] [Google Scholar]
  30. Mirrione M. M.; Schiffer W. K.; Fowler J. S.; Alexoff D. L.; Dewey S. L.; Tsirka S. E. A novel approach for imaging brain–behavior relationships in mice reveals unexpected metabolic patterns during seizures in the absence of tissue plasminogen activator. Neuroimage. 2007, 38, 34–42. 10.1016/j.neuroimage.2007.06.032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Lipinski C. A.; Lombardo F.; Dominy B. W.; Feeney P. J. Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings. Adv. Drug Delivery Rev. 2012, 64, 4–17. 10.1016/j.addr.2012.09.019. [DOI] [PubMed] [Google Scholar]
  32. Aertgeerts K.; Levin I.; Shi L.; Snell G. P.; Jennings A.; Prasad G. S.; Zhang Y.; Kraus M. L.; Salakian S.; Sridhar V.; Wijnands R.; Tennant M. G. Structural and kinetic analysis of the substrate specificity of human fibroblast activation protein alpha. J. Biol. Chem. 2005, 280, 19441–19444. 10.1074/jbc.C500092200. [DOI] [PubMed] [Google Scholar]
  33. Meng L.; Fang J.; Zhao L.; Wang T.; Yuan P.; Zhao Z.; Zhuang R.; Lin Q.; Chen H.; Chen X.; Zhang X.; Guo Z. Rational Design and Pharmacomodulation of Protein-Binding Theranostic Radioligands for Targeting the Fibroblast Activation Protein. J. Med. Chem. 2022, 65, 8245–8257. 10.1021/acs.jmedchem.1c02162. [DOI] [PubMed] [Google Scholar]
  34. Plescia J.; Hédou D.; Pousse M. E.; Labarre A.; Dufresne C.; Mittermaier A.; Moitessier N. Modulating the selectivity of inhibitors for prolyl oligopeptidase inhibitors and fibroblast activation protein-α for different indications. Eur. J. Med. Chem. 2022, 240, 114543. 10.1016/j.ejmech.2022.114543. [DOI] [PubMed] [Google Scholar]
  35. Deng L. J.; Wang L. H.; Peng C. K.; Li Y. B.; Huang M. H.; Chen M. F.; Lei X. P.; Qi M.; Cen Y.; Ye W. C.; Zhang D. M.; Chen W. M. Fibroblast Activation Protein α Activated Tripeptide Bufadienolide Antitumor Prodrug with Reduced Cardiotoxicity. J. Med. Chem. 2017, 60, 5320–5333. 10.1021/acs.jmedchem.6b01755. [DOI] [PubMed] [Google Scholar]
  36. Trujillo-Benítez D.; Luna-Gutiérrez M.; Ferro-Flores G.; Ocampo-García B.; Santos-Cuevas C.; Bravo-Villegas G.; Morales-Ávila E.; Cruz-Nova P.; Díaz-Nieto L.; García-Quiroz J.; et al. Design, Synthesis and Preclinical Assessment of 99mTc-iFAP for In Vivo Fibroblast Activation Protein (FAP) Imaging. Molecules. 2022, 27, 264. 10.3390/molecules27010264. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Makaravage K. J.; Brooks A. F.; Mossine A. V.; Sanford M. S.; Scott P. J. H. Copper-Mediated Radiofluorination of Arylstannanes with [18F]KF. Org. Lett. 2016, 18, 5440–5443. 10.1021/acs.orglett.6b02911. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Vermeulen K.; Vandamme M.; Bormans G.; Cleeren F. Design and Challenges of Radiopharmaceuticals. Semin Nucl. Med. 2019, 49, 339–356. 10.1053/j.semnuclmed.2019.07.001. [DOI] [PubMed] [Google Scholar]
  39. Pike V. W. PET radiotracers: crossing the blood-brain barrier and surviving metabolism. Trends Pharmacol. Sci. 2009, 30, 431–440. 10.1016/j.tips.2009.05.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Ge L.; Fu Z.; Wei Y.; Shi D.; Geng Y.; Fan H.; Zhang R.; Zhang Y.; Li S.; Wang S.; Shi H.; Song G.; Pan J.; Cheng K.; Wang L. Preclinical evaluation and pilot clinical study of [18F]AlF-NOTA-FAPI-04 for PET imaging of rheumatoid arthritis. Eur. J. Nucl. Med. Mol. Imaging. 2022, 49, 4025–4036. 10.1007/s00259-022-05836-3. [DOI] [PubMed] [Google Scholar]
  41. Lin J.-J.; Chuang C.-P.; Lin J.-Y.; Huang F.-T.; Huang C.-W. Rational Design, Pharmacomodulation, and Synthesis of [68Ga]Ga-Alb-FAPtp-01, a Selective Tumor-Associated Fibroblast Activation Protein Tracer for PET Imaging of Glioma. ACS Sens. 2021, 6, 3424–3435. 10.1021/acssensors.1c01316. [DOI] [PubMed] [Google Scholar]
  42. Xu M.; Zhang P.; Ding J.; Chen J.; Huo L.; Liu Z. Albumin Binder–Conjugated Fibroblast Activation Protein Inhibitor Radiopharmaceuticals for Cancer Therapy. J. Nucl. Med. 2022, 63, 952. 10.2967/jnumed.121.262533. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Ding J.; Xu M.; Chen J.; Zhang P.; Huo L.; Kong Z.; Liu Z. 86Y-Labeled Albumin-Binding Fibroblast Activation Protein Inhibitor for Late-Time-Point Cancer Diagnosis. Mol. Pharmaceutics 2022, 19, 3429–3438. 10.1021/acs.molpharmaceut.2c00579. [DOI] [PubMed] [Google Scholar]
  44. Wen X.; Xu P.; Shi M.; Liu J.; Zeng X.; Zhang Y.; Shi C.; Li J.; Guo Z.; Zhang X.; Khong P. L.; Chen X. Evans blue-modified radiolabeled fibroblast activation protein inhibitor as long-acting cancer therapeutics. Theranostics. 2022, 12, 422–433. 10.7150/thno.68182. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Wang Y.; Yuan H.; Liu N.; Tang S.; Feng Y.; Liu Y.; Cai P.; Xia L.; Zheng W.; Chen Y.; Zhou Z. High Affinity and FAP-Targeted Radiotracers: A Potential Design Strategy to Improve the Pharmacokinetics and Tumor Uptake for FAP Inhibitors. J. Med. Chem. 2023, 66, 8614–8627. 10.1021/acs.jmedchem.3c00259. [DOI] [PubMed] [Google Scholar]
  46. Ruan Q.; Feng J.; Jiang Y.; Zhang X.; Duan X.; Wang Q.; Yin G.; Xiao D.; Zhang J. Preparation and Bioevaluation of 99mTc-Labeled FAP Inhibitors as Tumor Radiotracers to Target the Fibroblast Activation Protein. Mol. Pharmaceutics 2022, 19, 160–171. 10.1021/acs.molpharmaceut.1c00712. [DOI] [PubMed] [Google Scholar]
  47. Zhao L.; Chen J.; Pang Y.; Fang J.; Fu K.; Meng L.; Zhang X.; Guo Z.; Wu H.; Sun L.; Su G.; Lin Q.; Chen H. Development of Fibroblast Activation Protein Inhibitor-Based Dimeric Radiotracers with Improved Tumor Retention and Antitumor Efficacy. Mol. Pharmaceutics 2022, 19, 3640–3651. 10.1021/acs.molpharmaceut.2c00424. [DOI] [PubMed] [Google Scholar]
  48. Zhong X.; Guo J.; Han X.; Wu W.; Yang R.; Zhang J.; Shao G. Synthesis and Preclinical Evaluation of a Novel FAPI-04 Dimer for Cancer Theranostics. Mol. Pharmaceutics 2023, 20, 2402–2414. 10.1021/acs.molpharmaceut.2c00965. [DOI] [PubMed] [Google Scholar]
  49. Zhao L.; Wen X.; Xu W.; Pang Y.; Sun L.; Wu X.; Xu P.; Zhang J.; Guo Z.; Lin Q.; et al. Clinical Evaluation of 68Ga-FAPI-RGD for Imaging of Fibroblast Activation Protein and Integrin αvβ3 in Various Cancer Types. J. Nucl. Med. 2023, 64, 1210–1217. 10.2967/jnumed.122.265383. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

pt3c00187_si_001.pdf (1.9MB, pdf)

Articles from ACS Pharmacology & Translational Science are provided here courtesy of American Chemical Society

RESOURCES