Synthesis, radiolabeling and initial in vivo evaluation of [¹¹C]KSM-01 for imaging PPAR-α receptors

Abstract

Peroxisome proliferator-activated receptor alpha (PPAR-α) is a ligand-activated nuclear receptor transcription factor that regulates the fatty acid β-oxidation. An in vitro assay identified the p-methoxy phenyl ureido thiobutyric acid derivative KSM-01 (IC₅₀=0.28±0.09 nM) having a higher affinity to activate PPAR-α than the PPAR-α agonist GW7647 ((IC₅₀=0.46±0.19 nM). In this study, we report the synthesis and initial in vivo evaluation of [¹¹C]KSM-01. The radiosynthesis was carried out by first alkylating the corresponding p-phenol precursor with [¹¹C]MeI in DMF using NaOH, followed by deprotection of the t-butyl ester group by TFA, yielding [¹¹C]KSM-01. SUV analysis of dynamic micro PET/CT imaging data showed that [¹¹C]KSM-01 accumulation was ~2.0-fold greater in cardiac-specific PPAR-α overexpressing transgenic mice compared to wild-type littermates. The post-PET biodistribution studies were consistent with these results and demonstrated 2.5-fold greater radiotracer uptake in the heart of transgenic mice compared to the wild-type littermates. These results demonstrate the potential utility of PPAR-α agonists as PET radiopharmaceuticals.

Peroxisome proliferator activated receptors (PPAR) are ligand-activated transcription factors which belong to the nuclear receptor gene family.^1–3 PPARs bind to endogenous ligands including eicosanoids, free fatty acids, leukotrienes, and prostaglandins and are classified into three subtypes: PPAR-α, PPAR-β/δ and PPAR-γ.^{1, 3} Ligand-activated PPARs bind to the retinoid X receptor (RXR) to form heterodimer complexes that trigger PPAR-response elements (PPRE) which modulate lipid, glucose, or cholesterol associated metabolic pathways, depending on the nature of ligand.^{3, 4} PPAR-α plays a vital role in regulating cellular fatty acid β-oxidation and ketogenesis and is activated by a wide range of fibrate drugs; this activation induces proliferation of peroxisomes. PPAR-α is highly expressed in tissues with significant breakdown of fatty acids including liver, heart, brown adipose tissue, kidney and intestine.^5–8 Due to its critical role in lipid metabolism, drugs which can modulate PPAR-α are being evaluated as targeted therapeutic strategies against cardiovascular diseases including type 2 diabetes mellitus, lipodistropy, and atherosclerosis.⁹

PPAR-α agonists have been studied extensively as therapeutic candidates for atherosclerosis because they down-regulate production of the adhesion molecule VCAM-1, which is responsible for endothelial cell activation in the arterial wall during atherosclerotic disease progression.^{10, 11} PPAR-α agonists may be involved in prevention of HDL cholesterol accumulation especially in cardiocytes and could affect the body weight by regulating fatty acid oxidation, thus also playing a potentially important role in diabetic cardiomyopathy.^{11, 12} Hypolipidemic fibrate drugs are an important class of PPAR-α ligands; however, fibrates which are considered highly selective in vivo activators of hepatic PPAR-α in rodents, often do not express the same level of selectivity in humans; many are also only moderately selective for PPAR-α over the PPAR-γ and PPAR-δ subtypes.³ Attempts to identify more potent PPAR-α ligands have led to synthesis of ureidofibrates that are active at lower concentrations in rodent models of hyperlipidemia. The ureido thioisobutyric acid (TiBA) derivative GW9578 was observed to be a more potent and selective PPAR-α agonist with lipid-lowering activity when compared to traditional fenofibrate derivatives.¹³ However, difficulties in handling GW9578, which is a viscous oil, and its poor selectivity for human PPAR-α led to development of GW7647, which demonstrated ~200 fold selectivity towards human PPAR-α over PPAR-γ and PPAR-δ.¹⁴ Considerable literature evidence suggests that sufficient PPAR-α density exists in cardiac cells for the evaluation of agonists in imaging studies.^{10, 12} Transgenic mice with cardiac-specific overexpression of PPAR-α display a phenotype similar to that of human diabetic cardiomyopathy. Furthermore, several diabetic metabolic abnormalities, including higher fatty acid and lower glucose uptake were observed in a transgenic mouse model with cardiac-specific overexpression of PPAR-α.^{10, 12, 15}

Previous attempts have been made to measure PPAR-α activity using PPRE luciferase transgenic mice through in vivo and ex vivo bioluminescence imaging.¹⁶ However, to date, no PET tracer has been reported for the in vivo imaging of PPAR-α in the heart. A PPAR-α agonist as a PET radiotracer would thus become a pivotal tool to fill critical gaps in understanding the pathogenesis of diabetic cardiomyopathy triggered by PPAR-α. The ureido-TiBA derivative GW7647 (Figure 1) has been investigated as potent PPAR-α agonist in connection with metabolic syndromes like dyslipidemia and atherosclerosis.¹⁴

Figure 1.

Structures of the ureido-TiBA PPAR-α agonists

GW7647 was used as a lead in designing a PET tracer by replacing the cyclohexyl group with a methoxyphenyl group; we synthesized the -2, -3, and -4 methoxyphenyl isomers of tert-butyl-2-(4-(2-(1-(4-cyclohexylbutyl)-3- methoxyphenyl)ureido)ethyl)phenylthio)-2-methylpropanoate as potential PPAR-α radiotracers. The structure activity relationships derived through in vitro binding assay results indicated the p-methoxyphenyl ureidothiobutyric derivative KSM-01 (Figure 1) IC₅₀=0.28±0.09 nM had the highest affinity to activate PPAR-α among the -o,-m,-p methoxyphenyl isomers. This report describes the synthesis, radiolabeling and initial in vivo microPET evaluation of [¹¹C]KSM-01 to image PPAR-α in cardiac-specific PPAR-α overexpressing mice.

The syntheses of KSM-01, 02, 03 are shown in Scheme 1 and described in detail in the online Supplemental Methods. Reduction of 4-mercaptobenzoic acid 1 with LiAlH₄ in THF resulted in the primary alcohol 2.¹⁷ The sulfur group on 2 was alkylated with α-bromoisobutyrate in KOH-EtOH¹³ to give compound 3 after which the hydroxyl group in 3 was converted to the corresponding chloro analog 4 using PPh₃ in CCl₄.¹⁸ Cyanation and reductive amination of compound 4 using KCN and BH₃.THF respectively gave the amine analog 5.¹⁹ Further reaction with 4-cyclohexyl-1-bromobutane²⁰ in DIPEA and THF gave the secondary amine 6,²¹ which was coupled with 2-methoxy, 3-methoxy, or 4-methoxy phenylisocyanate followed by TFA-assisted t-butyl ester deprotection to give analogs KSM-03, KSM-02 and KSM-01 respectively.¹³ The precursor KSM-01A for ¹¹C-radiolabeling was obtained from the secondary amine 6 coupled with 1,1¹-dicarbonyl imidazole followed by 4-amino phenol in THF.¹³

Scheme 1.

Reagents: (a) LiAlH₄, THF, 67%; (b) a-bromoisobutyrate, KOH, EtOH, 94%; (c) PPh₃, C₂Cl₆, 95%; (d) KCN, BH₃.THF, 78%; (e) 4-cyclohexyl-1-bromobutane, DIPEA, 68%; (f) 2/3/4-methoxyphenylisocyanate, TFA. Yields: KSM-03, 65%; KSM-02, 85%; KSM-01, 90%.

PPAR-α binding affinity of KSM-01, KSM-02 and KSM-03 was assessed to identify the most potent PPAR-α agonist. Using a beta-lactamase reporter-gene under control of the PPAR-α response element, a cell-based assay developed by Invitrogen was used to determine the IC₅₀ values of the novel PPAR-α ligands. Novel PPAR-α compounds were measured for their ability to inhibit reporter gene activity. The assay utilizing GeneBLAzer PPAR-α UAS-bla HEK293T cells is described in detail in the online Supplemental Methods. The fluorescence intensity was measured using a Victor³ plate reader after addition of the LiveBLAzer™-FRET B/G (CCF4-AM) substrate. Concentration-response titration points for each compound were fitted to the Hill equation yielding concentrations of halfmaximal inhibition (IC₅₀) and maximal response (efficacy) values. The IC₅₀ values of the analogs are shown in Table 1.^{16, 21} Compound KSM-01 showed higher potency towards PPAR-α compared to the two isomers and the previously reported PPAR-α agonist GW7647, suggesting that KSM-01 could be a suitable PPAR-α PET imaging agent.

Table 1.

In vitro binding affinities of ureido-TiBAs for PPAR-α

Compound	IC50 (nM)
KSM-01	0.28±0.09
KSM-02	0.59±0.22
KSM-03	1.93±0.99
GW7647	0.46±0.19

The radiochemical synthesis of [¹¹C]KSM-01 is described in detail in the online Supplemental Methods. It was achieved by first alkylating the corresponding p-phenol precursor KSM-01A with [¹¹C]MeI in DMF using NaOH at 90°C for 5 min and then deprotecting the t-butyl ester group with trifluoroacetic acid at 90°C for 3 min as depicted in Scheme 2. The total time required for the synthesis of [¹¹C]KSM-01, including [¹¹C]MeI production, purification and formulation was approximately 50 min. The radiochemical purity of [¹¹C]KSM-01 was >98% and was confirmed by co-elution with non-radioactive KSM-01. The chemical purity of [¹¹C]KSM-01 determined by the HPLC UV mass was >97%. The calculated radiochemical yield was ~19% and the final product had a specific activity of 987 mCi/µmol (decay-corrected to end of synthesis).

Scheme 2.

Reagents: (a) [¹¹C]MeI, 5N NaOH, DMF, 90°C, 5 min; (b) TFA, 90°C, 3 min.

The uptake of [¹¹C]KSM-01 was compared in transgenic mice with cardiac-specific overexpression of PPAR-α (PPAR-α +/+) and wild type littermates (PPAR-α−/−) through microPET imaging and post-PET biodistribution studies as described in the online Supplemental Methods. Transgenic mice were produced as previously described.¹⁰ PPAR-α protein levels are approximately 15-fold more abundant in the cardiac ventricles of these transgenic animals when compared with wildtype littermates; this model has been used to evaluate the role of PPAR-α mediated lipid metabolism in the development of diabetic cardiomyopathy.¹⁰

Dynamic PET imaging was performed for 0–60 min post intravenous injection of [¹¹C]KSM-01. Standard uptake values (SUVs) analysis of the PET data revealed a 2- fold greater accumulation of radioactivity in PPAR-α overexpressing transgenic mice (0.68±0.007) over the control animals (0.37±0.09). Although liver uptake is significant, microPET images show higher [¹¹C]KSM-01 accumulation in the heart of PPAR-α overexpressing transgenic mice compared to wild-type mice (Figure 2).

Figure 2.

Representative Focus 220 microPET image of [¹¹C]KSM-01 in male MHC-PPAR-α transgenic and wild type (WT) mice. The images were summed from 0–60 min after iv injection of 170 µCi [¹¹C]KSM-01.

A post-PET biodistribution study was also conducted and the results are presented in Table 2. Radiotracer accumulation was observed to be high in liver tissue, which is a primary organ for PPAR-α expression^5–8 in both transgenic and control animals: %ID/g = 55.031 ± 4.926 and 60.699 ± 0.774 respectively. Additionally, the radiotracer demonstrated ~2.5 fold greater distribution in the cardiac tissue of PPAR-α overexpressing transgenic mice (1.09 ± 0.246) when compared to the wild type littermates (0.479 ± 0.007) (Figure 3). The biodistribution results were consistent with the microPET imaging data.

Table 2.

Post-PET biodistribution of [¹¹C]KSM-01 in non-target tissues of male MHC PPAR-α overexpressing transgenic (TG) and wild type mice (WT). Results are expressed as %ID/g ± standard deviation.

Organ	TG	WT
Blood	0.42±0.14	0.42±0.10
Lung	2.65±0.34	2.11±0.31
Liver	55.30±4.92	60.69±0.77
Spleen	1.33±0.29	1.55±0.14
Kidney	2.58±0.56	2.01±0.21
Muscle	0.36±0.15	0.35±0.18

Figure 3.

Post-PET uptake of [¹¹C]KSM-01 in cardiac tissue (%ID/g) of male MHC-PPAR-α transgenic (TG) and wild type mice (WT).

In summary, the ureido TiBAC analog KSM-01, a PPAR-α agonist with high potency (IC₅₀ =0.28 nM) was synthesized and [¹¹C]-radiolabeled for the first time. MicroPET imaging data comparing transgenic mice which selectively overexpress PPAR-α in the heart with wild-type littermates indicate the potential utility of PPAR-α agonists as PET radiopharmaceuticals. Although liver uptake is high due to the normal expression of PPAR-α, elevated cardiac uptake can be clearly visualized in PPAR-α over-expressing transgenic mice. Further experiments evaluating this strategy with ¹⁸F-radiolabled PPAR-α agonists are currently underway.