Comparison of 64Cu and 68Ga for Molecular Imaging of Atherosclerosis using the Apolipoprotein A-I Mimetic Peptide FAMP

Background: Molecular imaging for detection of the atherosclerotic plaque burden has been highlighted as a modality for the diagnosis of atherosclerosis. We recently developed a novel and noninvasive positron emission tomography (PET) that was functionalized with an apolipoprotein (Apo) A-I mimetic peptide [known as Fukuoka University Apo A-I mimetic peptide (FAMP)] radiolabeled with gallium-68 (68Ga) 1, 4, 7, 10-tetraazacyclododecane-1, 4, 7, 10-tetraacetic acid (DOTA) to specifically image the status of atherosclerotic plaque in myocardial infarctionprone Watanabe heritable hyperlipidemic rabbits (WHHL-MI). Methods and Results: To achieve more sensitive molecular imaging, FAMP was modified with 4, 11 bis (carboxymethyl) 1, 4, 8, 11 tetraazabicyclo (6.6.2) hexadecane (CB-TE2A) and radiolabeled with copper-64 (64Cu) for PET, and the ability of 64Cu-TE2A-FAMP to image plaque was compared with that of 68Ga-DOTA-FAMP. Japanese white normal (JW) and WHHL-MI rabbits were intravenously injected with 64Cu-CB-TE2A-FAMP or 68GaDOTA-FAMP, and subjected to continuous PET (25-30 MBq). Interestingly, 64Cu-CB-TE2A-FAMP was not taken up by atherosclerotic lesions in the aorta of WHHL-MI, whereas 68Ga-DOTA-FAMP was dramatically illuminated in the aorta of WHHL-MI. Moreover, 64Cu-CB-TE2A-FAMP was rapidly decomposed and 64Cu was excreted to the intestine, liver or urinary bladder in both JW and WHHL-MI rabbits. Conclusions: These results demonstrated that FAMP may be a target molecule for atherosclerotic molecular imaging with 68Ga-DOTA, but not with 64Cu-CB-TE2A. The selection of a suitable radio-nuclide and chelator might be important for HDL functioning imaging. Journal of Cardiovascular Diseases & Diagnosis J o u r n a l o f C ar dio vas cular Dases & ia g n o s i s


Introduction
Cardiovascular disease is the leading cause of death in the world [1]. Atherosclerosis affects arterial blood vessels due to chronic inflammation and induces cardiovascular disease. Molecular imaging aims to identify and guide the treatment of vulnerable atherosclerotic plaque. A novel and noninvasive positron emission tomography (PET)-based method for imaging inflammatory plaque with 18 F-fluorodeoxyglucose (FDG) is currently of interest in the clinical setting [2]. Although atherosclerotic disease may produce no or only mild diffuse uptake along the vessel wall in FDG-PET, both the sensitivity and molecular specificity for targeting atherosclerosis by FDG-PET are relatively low [3][4][5].
Since the physical interactions between apolipoprotein A-I (Apo A-I) and ATP-binding cassette transporter 1 (ABCA1) modulate not only binding to Apo A-I but also internalization and transcytosis in macrophages and aortic endothelial cells, Apo A-I mimetic peptide may be a candidate for functional high-density lipoprotein (HDL) imaging in atherosclerosis. We recently synthesized a novel 24-amino acid Apo A-I mimetic peptide without phospholipids [known as Fukuoka University Apo A-I mimetic peptide (FAMP)], which potently removes cholesterol in vitro via specific ABCA1 [6]. In that study using a cholesterol-fed mouse model, we found that analyzed by matrix-assisted laser desorption/ionization time-offlight mass spectrometry (MALDI-TOF MS; found: m/z 3,137.2 (M+H)+, calculated for (M+H)+: 3,137.8) ( Figure 1B). DOTA-Nhydroxysuccinimidyl (NHS) ester was prepared by the preferential activation of 1 carboxyl group of the DOTA chelator in dimethyl sulfoxide (DMSO), as described previously [7,8].

PET imaging protocol
Imaging was carried out using a microPETR Focus220 (Siemens, Knoxville, TN, USA) as described previously [7]. Briefly, emission data acquisition was started at the same time as the intravenous administration of 64 Cu-CB-TE2A-FAMP and 68 Ga-DOTA-FAMP for 6 h with an energy window of 400-650 keV and a coincidence timing window of 6 ns. The emission images were reconstructed by using filtered back projection with a Ramp filter and a cutoff at the Nyquist frequency. The time course of 64 Cu activity in the region of the abdominal aortic bifurcation was obtained using ASIPro software included with the microPET system, and the results were adjusted for the background.

Time-dependent attenuation of 64 Cu-CB-TE2A-FAMP at the abdominal aortic bifurcation
We performed a PET analysis 6 h after the injection of 64 Cu-CB-TE2A-FAMP at the abdominal aortic bifurcation, when radioactivity should be completely cleared from the circulation. PET images of the abdominal aorta revealed little accumulation of 64 Cu-CB-TE2A-FAMP in either JW or WHHL-MI rabbits ( Figure 2). The time-dependent attenuation of 64 Cu-CB-TE2A-FAMP in the abdominal aortic bifurcation in WHHL-MI rabbits was mild compared with that in JW rabbits.

PET imaging after injection of the tracer
PET images clearly showed a high uptake of the 68 Ga-DOTA-FAMP tracer in the aorta of WHHL-MI rabbits ( Figure 3A), and little uptake in the aorta of JW rabbits ( Figure 3B). On the other hand, PET images showed little uptake of 64 Cu-CB-TE2A-FAMP in the aorta of either WHHL-MI or JW rabbits ( Figure 3C, 3D). Furthermore, the clear uptake of 64 Cu-CB-TE2A-FAMP in the intestine, liver and bladder was seen in both WHHL-MI and JW rabbits ( Figures 3C, 3D and 4A). Finally, there were no uptake in the aorta and the clear uptake of 64 Cu-CB-TE2A-FAMP in the intestine, liver and bladder at 24 h after the injection in both WHHL-MI and JW rabbits ( Figure 4B, 4C).

Discussion
We previously developed a novel PET tracer, Apo A-I mimetic peptide FAMP radiolabeled with 68 Ga-DOTA, to specifically image the status of atherosclerotic plaque in WHHL-MI [7]. To achieve more sensitive molecular imaging, FAMP was modified with CB-TE2A and radiolabeled with 64 Cu for PET, and the ability of 64 Cu-TE2A-FAMP to image plaque was compared with that of 68 Ga-DOTA-FAMP. 68 Ga-DOTA-FAMP was interestingly superior to 64 Cu-CB-TE2A-FAMP in atherosclerotic molecular imaging. The selection of a suitable radionuclide and chelator should be important for HDL functioning imaging.
In this study, we selected 64 Cu as a tracer because it has a longer half-life than 68 Ga (12.7 h and 68 min, respectively) for more feasibility clinical use. The half-life of fluorescence-labeled FAMP in blood was about 4 h, according to data on blood clearance of FAMP in mice [6]. Although we thought that a combination of 64 Cu and FAMP could provide superior molecular imaging, PET images showed no uptake of 64 Cu-CB-TE2A-FAMP in the aorta at 6 h after injection ( Figure 3C). Since we did not observe any uptake of 64 Cu-CB-TE2A-FAMP in the aorta at 1-2 h after injection (data not shown), we consider that 64 Cu-  Cu-CB-TE2A-FAMP may be rapidly decomposed, since 64 Cu was excreted to the intestine, liver or urinary bladder in both JW and WHHL-MI rabbits (Figures 3CD and 4A). Previous reports have shown that 64 Cu-DOTA and 64 Cu-TETA complex are moderately unstable in vivo due to the release of uncoordinated 64 Cu by decomposition in the blood or transchelation in the liver, which leads to high uptake in non-target tissues [11,12]. 64 Cu-labeled DOTA/TETA conjugates often show poor blood and liver clearance, high uptake in non-target organs, and increased background radioactivity levels, resulting in reduced imaging sensitivity [13,14]. As for 4Cu-CB-TE2A complexes, Sprague et al. indicated that the 64 Cu-CB-TE2A-Tyr3-octreotate shows improved blood, liver, and kidney clearance compared with the analogous 64 Cu-TETA agent [15]. 64 Cu-CB-TE2A-c is a potential candidate for imaging tumor angiogenesis [16]. In addition, 64 Cu-CB-TE2A-ReCCMSH(Arg11) also showed greatly improved liver and blood clearance as well as higher tumor-to-non-target tissue ratios compared with 64 Cu-DOTA-ReCCMSH (Arg11) [17]. Thus, many researchers have successfully performed molecular imaging using 64 Cu-CB-TE2A. We do not know why 64 Cu-CB-TE2A-FAMP may be rapidly decomposed in vivo at this stage, although we successfully synthesized and purified 64 Cu-CB-TE2A-FAMP as shown in Figure 1. Further studies are needed to solve this issue.
Although 68 Ga-DOTA-FAMP was interestingly superior to 64 Cu-CB-TE2A-FAMP for atherosclerotic molecular imaging, the PET images obtained with 68 Ga-DOTA-FAMP tracer show atherosclerotic plaques more clearly than those obtained using previously reported tracers ( 18 F-FDG and 111 In-low-density lipoprotein, etc.) [18][19][20]. FAMP has been shown to have a high capacity for cholesterol efflux from A172 cells and mouse and human macrophages in vitro, and this efflux was mainly dependent on ABCA1 transporter [6]. 68 Ga-DOTA-FAMP may accumulate in atherosclerotic plaque, which requires FAMP. Mizoguchi et al. reported that a pioglitazone-treated group demonstrated significantly greater suppression of FDG imaging of carotid and aortic plaque inflammation compared with a glimepiride group in patients with diabetes mellitus [21]. Their study only found a decrease in inflammation after treatment, and could not assess changes in plaque volume or vulnerability. In addition, the precise mechanisms that underlie FDG uptake in atheroma are not clear. Recently, Santulli et al. proposed an innovative strategy based on a selective microRNA in the treatment of atherosclerosis and restenosis [22]. Since the final aim of molecular imaging is to identify and guide the treatment of vulnerable atherosclerotic plaques that are at high risk of rupture and subsequent thrombosis, we need to identify a better strategy for achieving more sensitive and specific molecular imaging using FAMP in the future.
There are several study limitations in this study. First, it is important to match PET-positive lesions with the results of immunohistochemical analysis to determine the cell make-up, although we did not perform the analysis because of the difficulties associated with radioactive tissue. Second, the target molecule for 68 Ga-DOTA-FAMP has not yet been identified. Third, PET imaging cannot include the heart, we targeted the descending aorta (bifurcation of aorta) and could not perform coronary imaging.
In conclusion, these results demonstrated that FAMP may be a target molecule for atherosclerotic molecular imaging with 68 Ga-DOTA, but not with 64 Cu-CB-TE2A. The selection of a suitable radionuclide and chelator might be important for HDL functioning imaging.