Synthesis and preclinical evaluation of a novel fluorine-18 labeled small-molecule PET radiotracer for imaging of CXCR3 receptor in mouse models of atherosclerosis

Abstract

Background: CXCR3 is a chemokine receptor and is expressed on innate and adaptive immune cells. It promotes the recruitment of T-lymphocytes and other immune cells to the inflammatory site in response to the binding of cognate chemokines. Upregulation of CXCR3 and its chemokines has been found during atherosclerotic lesion formation. Therefore, the detection of CXCR3 by positron emission tomography (PET) radiotracer may be a useful tool to detect atherosclerosis development noninvasively. Herein, we report the synthesis, radiosynthesis, and characterization of a novel fluorine-18 (F-18, ¹⁸ F) labeled small-molecule radiotracer for the imaging of the CXCR3 receptor in mouse models of atherosclerosis. Methods: The reference standard ( S )-2-(5-chloro-6-(4-(1-(4-chloro-2-fluorobenzyl)piperidin-4-yl)-3-ethylpiperazin-1-yl)pyridin-3-yl)-1,3,4-oxadiazole ( 1 ) and its corresponding precursor 9 were synthesized using organic syntheses. The radiotracer [ ¹⁸ F] 1 was prepared in one-pot, two-step synthesis via aromatic ¹⁸ F-substitution followed by reductive amination. Cell binding assays were conducted using 1 , [ ¹²⁵ I]CXCL10, and CXCR3A- and CXCR3B-transfected human embryonic kidney (HEK) 293 cells. Dynamic PET imaging studies over 90 min were performed on C57BL/6 and apolipoprotein E (ApoE) knockout (KO) mice that were subjected to a normal and high-fat diet for 12 weeks, respectively. Blocking studies were conducted with preadministration of the hydrochloride salt of 1 (5 mg/kg) to assess the binding specificity. Time-activity curves (TACs) for [ ¹⁸ F] 1 in both mice were used to extract standard uptake values (SUVs). Biodistribution studies were performed on C57BL/6 mice, and the distribution of CXCR3 in the abdominal aorta of ApoE KO mice was assessed by immunohistochemistry (IHC). Results: The reference standard 1 and its precursor 9 were synthesized over 5 steps from starting materials in good to moderate yields. The measured K _i values of CXCR3A and CXCR3B were 0.81 ± 0.02 nM and 0.31 ± 0.02 nM, respectively. [ ¹⁸ F] 1 was prepared with decay-corrected radiochemical yield (RCY) of 13 ± 2%, radiochemical purity (RCP) >99%, and specific activity of 44.4 ± 3.7 GBq/µmol at the end of synthesis (EOS) ( n =6). The baseline studies showed that [ ¹⁸ F] 1 displayed high uptake in the atherosclerotic aorta and brown adipose tissue (BAT) in ApoE KO mice. The uptake of [ ¹⁸ F] 1 in these regions was reduced significantly in self-blocking studies, demonstrating CXCR3 binding specificity. Contrary to this, no significant differences in uptake of [ ¹⁸ F] 1 in the abdominal aorta of C57BL/6 mice were observed in both baseline and blocking studies, indicating increased CXCR3 expression in atherosclerotic lesions. IHC studies demonstrated that [ ¹⁸ F] 1 -positive regions were correlated with CXCR3 expression, but some atherosclerotic plaques with significant size were not detected by [ ¹⁸ F] 1 , and their CXCR3 expressions were minimal. Conclusion: The novel radiotracer, [ ¹⁸ F] 1 was synthesized with good RCY and high RCP. In PET imaging studies, [ ¹⁸ F] 1 displayed CXCR3-specific uptake in the atherosclerotic aorta in ApoE KO mice. [ ¹⁸ F] 1 visualized CXCR3 expression in different regions in mice is in line with the tissue histology studies. Taken together, [ ¹⁸ F] 1 is a potential PET radiotracer for the imaging of CXCR3 in atherosclerosis.

Figure 1

The cartoon of the docking model for compound 1 that has 14 interactions with the amino acids of CXCR3.

Figure 2

(a) The semi-preparative HPLC of [¹⁸F]1 and (b) The analytical HPLC of [¹⁸F]1. Co-injection of nonradioactive standard 1 confirmed that the collected product was an authentic product.

Figure 3

Representative CXCR3 PET images of [¹⁸F]1 in two ApoE KO mice and one C57BL/6 control mouse. The specific uptake of [¹⁸F]1 was observed in the atherosclerotic aorta (red arrow) and BAT (blue arrow) of ApoE KO mice and the BAT of C57BL/6 control mice.

Figure 4

The average time-activity curves (TACs) of ApoE KO mice (N=3) and C57BL/6 control mice (N=5). The TACs indicate the blocking effects of [¹⁸F]1, demonstrating the specificity of [¹⁸F]1.

Figure 5

The areas under the curve (AUCs) of the BAT and atherosclerotic aorta of ApoE KO mice and the BAT and corresponding aorta of C57BL/6 mice.

Figure 6

The results of biodistribution using male and female control C57BL/6 mice. The uptake of [¹⁸F]1 in each organ was measured at 10, 30, and 60 min after its intravenous injection.

Figure 7

CXCR3 PET imaging using [¹⁸F]1 and their corresponding cardiovascular tissues obtained from the ApoE KO mice. (a) Region I (red arrow) is a [¹⁸F]1-positive region, and the corresponding aorta tissue was highly expressing CXCR3 (Green color); meanwhile, region II (blue arrow) has a significant atherosclerotic plaque. Nevertheless, it is a [¹⁸F]1-negative region of PET imaging and indicated low CXCR3 expression levels. These results suggest CXCR3 expression-dependent PET detection by [¹⁸F]1. (b) Region III also indicated [¹⁸F]1-positivity, and subsequent histologic analysis revealed a significant atherosclerotic plaque with high CXCR3 expression levels (green detection in the fluorescence image), consistent with CXCR3 expression-dependent detection by the radiotracer.