[¹⁸F]Fluoro-hydroxyphenethylguanidines: Efficient Synthesis and Comparison of Two Structural Isomers as Radiotracers of Cardiac Sympathetic Innervation

Abstract

Fluorine-18 labeled phenethylguanidines are currently under development in our laboratory as radiotracers for quantifying regional cardiac sympathetic nerve density using PET imaging techniques. In this study, we report an efficient synthesis of ¹⁸F-hydroxyphenethylguanidines consisting of nucleophilic aromatic [¹⁸F]fluorination of a protected diaryliodonium salt precursor followed by a single deprotection step to afford the desired radiolabeled compound. This approach has been shown to reliably produce 4-[¹⁸F]fluoro-m-hydroxyphenethylguanidine ([¹⁸F]4F-MHPG, [¹⁸F]1) and its structural isomer 3-[¹⁸F]fluoro-p-hydroxyphenethylguanidine ([¹⁸F]3F-PHPG, [¹⁸F]2) with good radiochemical yields. Preclinical evaluations of [¹⁸F]2 in non-human primates were performed to compare its imaging properties, metabolism, and myocardial kinetics with those obtained previously with [¹⁸F]1. The results of these studies have demonstrated that [¹⁸F]2 exhibits imaging properties comparable to those of [¹⁸F]1. Myocardial tracer kinetic analysis of each tracer provides quantitative metrics of cardiac sympathetic nerve density. Based on these findings, first-in-human PET studies with [¹⁸F]1 and [¹⁸F]2 are currently in progress to assess their ability to accurately measure regional cardiac sympathetic denervation in patients with heart disease, with the ultimate goal of selecting a lead compound for further clinical development.

TABLE OF CONTENTS GRAPHIC

INTRODUCTION

Cardiac autonomic dysfunction contributes to morbidity and mortality in many diseases that damage the heart, including congestive heart failure, myocardial ischemia, myocardial infarction, and diabetic autonomic neuropathy.^1–3 Disease-induced alterations in the nervous control of the heart can be caused by changes in the outflow of nervous impulses to the parasympathetic and sympathetic branches of the autonomic nervous system arising from central sites in the brain, and by regional degeneration of postganglionic parasympathetic and/or sympathetic nerve fibers in the heart. The complex interaction between aberrant parasympathetic and sympathetic influences on the heart often evolves with the progression of disease. For example, early heart failure is characterized by hyperactivity of sympathetic nerve pathways, with a concomitant reduction in parasympathetic activation, a process that referred to as ‘parasympathetic withdrawal’.⁴ The chronically elevated levels of the neurotransmitter norepinephrine that result from this imbalance of parasympathetic and sympathetic influences are cardiotoxic and contribute to the progression of heart failure, including development of left ventricular dilation.^{5, 6} Throughout the time course of heart failure, autonomic dysfunction can promote malignant arrhythmias, leading to sudden cardiac death.^{7, 8}

Clinical imaging studies of cardiac sympathetic denervation using nuclear scintigraphy techniques have provided important insights into how different diseases affect this nerve population.^{9, 10} Our laboratory has previously developed several radiotracers for noninvasive imaging studies of cardiac sympathetic nerves, including [¹²³I]metaiodobenzylguanidine ([¹²³I]MIBG) for planar scintigraphy and SPECT imaging, and [¹¹C]-(–)-m-hydroxyephedrine ([¹¹C]HED) for PET imaging (Figure 1).^{11, 12} These radiolabeled analogs of the endogenous neurotransmitter norepinephrine are transported into presynaptic sympathetic nerve terminals by the norepinephrine transporter (NET). Since NET expression in the heart is only associated with sympathetic nerve varicosities, the cardiac retention levels of these tracers can be used as metrics of regional sympathetic nerve density. Clinical trials with [¹²³I]MIBG and [¹¹C]HED tracers in heart failure patients have demonstrated that cardiac sympathetic denervation is associated with a significantly increased risk of sudden cardiac death. For example, the multi-center ADMIRE-HF study of 961 heart failure patients showed that a low heart-to-mediastinum ratio (H/M) of [¹²³I]MIBG obtained from planar gamma camera images, a global measure of cardiac sympathetic denervation, was a predictor of mortality and fatal arrhythmic events in patients eligible for implantable cardioverter defibrillator (ICD) therapy.¹³ More recently, the prospective PAREPET trial used [¹¹C]HED to assess regional sympathetic denervation, [¹³N]ammonia to assess myocardial perfusion, and [¹⁸F]FDG to assess myocardial viability in 204 heart failure patients staged for an ICD.¹⁴ The PAREPET results showed that the regional extent of sympathetic denervation in the left ventricle (as defined by [¹¹C]HED retention deficits) was the strongest predictor of sudden cardiac arrest among all imaging parameters measured. Patients with regional sympathetic denervation encompassing more than 38% of their left ventricle defined the upper tertile of patients with the highest risk of sudden cardiac arrest (p < 0.001). The results of the ADMIRE-HF and PAREPET trials suggest that noninvasive imaging studies of cardiac sympathetic denervation may find a clinical role in improved risk stratification of heart failure patients for ICD placement.¹⁵

Figure 1.

Structures of norepinephrine and some radiotracers for imaging cardiac sympathetic nerve terminals.

We have recently been investigating radiolabeled phenethylguanidines as next generation sympathetic nerve tracers with improved myocardial kinetics for accurate and sensitive quantification of regional sympathetic nerve density using PET and tracer kinetic analysis methods.¹⁶ Our initial work focused on the carbon-11 labeled hydroxyphenethylguanidine N-[¹¹C]guanyl-(–)-m-octopamine ([¹¹C]GMO, Figure 1). PET studies with [¹¹C]GMO in non-human primates showed that the myocardial kinetics of this tracer could be successfully analyzed using compartmental modeling methods or Patlak graphical analysis to obtain an estimates of the ‘net uptake rate constant’ K_i (mL/min/g) as a measure of regional sympathetic nerve density.¹⁷ In control studies and pharmacological blocking studies with the NET inhibitor desipramine (DMI) in rhesus macaque monkeys, the K_i values from compartmental modeling analysis were found to decrease with increasing DMI doses along a sigmoidal dose-response curve, with an IC₅₀ of 0.087 mg/kg and a Hill slope n_H = –0.70. Similar results were obtained using the Patlak slopes K_p (mL/min/g) from Patlak graphical analysis of [¹¹C]GMO kinetics, as an alternative estimate of the K_i values (IC₅₀ = 0.068 mg/kg, n_H = –0.54). These results strongly suggest that regional estimates of K_i from tracer kinetic analysis of [¹¹C]GMO kinetics in human hearts would serve as a reproducible metric of regional cardiac sympathetic nerve density.

While the preclinical PET studies with [¹¹C]GMO were highly encouraging, the short half-life of carbon-11 (20.4 min) limits its use to PET centers with an on-site cyclotron. Conversely, fluorine-18 has a sufficiently long half-life (1.83 h) that the production and distribution of ¹⁸F-labeled radiopharmaceuticals from a central production facility to stand-alone PET imaging centers is feasible. To develop a fluorine-18 labeled phenethylguanidine, we initially chose to prepare 4-[¹⁸F]fluoro-m-hydroxyphenethylguanidine ([¹⁸F]4F-MHPG, [¹⁸F]1, Figure 1). In previous studies in an isolated rat heart model, the kinetics of [¹¹C]4F-MHPG were comparable to those of [¹¹C]GMO.¹⁶

A previous report described several preclinical evaluations of [¹⁸F]1, including PET imaging studies in non-human primates.¹⁸ The results of the PET studies of [¹⁸F]1 closely paralleled those obtained with [¹¹C]GMO,¹⁷ supporting further development of this agent for clinical translation. For these pilot studies, our initial approach to the preparation of [¹⁸F]1 was a multi-step automated radiosynthesis involving production of a ¹⁸F-labeled intermediate using a diaryliodonium salt precursor followed by a deprotection step, an N-guanylation step, a second deprotection step and finally HPLC purification to provide [¹⁸F]1.¹⁸ Because of the multiple steps required, this initial method relied on the use of two automated radiosynthesis modules in adjacent hot-cells, linked in series, to provide the final product. While this approach was adequate for preclinical evaluations of [¹⁸F]1, it was poorly suited to routine production of the radiotracer for clinical studies in human subjects. Thus, one goal of this work was to develop an efficient automated radiosynthesis of [¹⁸F]1 capable of providing sufficiently high radiochemical yields for clinical PET studies.

We report here a new two-step approach to the radiosynthesis of [¹⁸F]1 that utilizes a novel N,N’,N’’,N’’-tetrakis-Boc protected guanidinyliodoium salt precursor for nucleophilic aromatic [¹⁸F]fluorination, followed by a single deprotection step, to provide [¹⁸F]1 in good yields and high specific activities for clinical PET studies. This new approach is performed using a single automated radiosynthesis module and has proven to be very reliable.

With the establishment of this new synthetic approach to ¹⁸F-hydroxyphenethylguanidines, the second objective of this study was to prepare 3-[¹⁸F]fluoro-p-hydroxyphenethylguanidine ([¹⁸F]3F-PHPG, [¹⁸F]2, Figure 1), the structural isomer of [¹⁸F]1. Preclinical evaluations of this compound were performed to compare its characteristics as a cardiac innervation tracer with those of [¹⁸F]1. Studies in non-human primates showed that [¹⁸F]2 is as good or better than [¹⁸F]1 as a cardiac PET imaging agent. Based on the positive preclinical testing results with [¹⁸F]1 and [¹⁸F]2, these new cardiac sympathetic innervation radiotracers are currently being evaluated in first-in-human PET studies under an exploratory investigational new drug (IND) clearance from the FDA.

RESULTS AND DISCUSSION

Chemistry and Radiochemistry.

Our initial strategy for improving the radiosynthesis of [¹⁸F]1 was to utilize a N,N’-bis-Boc-protected diaryliodonium salt precursor for direct no-carrier-added nucleophilic aromatic [¹⁸F]fluorination via a two-step automated radiosynthesis (Scheme 1). However, many attempts at ¹⁸F-radiofluorination of the N,N′-bis-Boc-guanidinyliodonium salt precursor 3 with different [¹⁸F]fluoride sources under a variety of reaction conditions were unsuccessful in providing the desired ¹⁸F-labeled intermediate [¹⁸F]4. We hypothesized that this disappointing outcome was related to the two unprotected protons of the guanidinyl group being acidic, causing a disturbance in the ¹⁸F-radiofluorination reaction. To confirm this hypothesis, we explored the ¹⁸F-radiofluorination of a test compound with a fully protected guanidinyl moiety to eliminate the unprotected protons (Scheme 2). The N,N′,N′′,N′′-tetrakis-Boc-guanidinyliodonium salt precursor 5 was synthesized, and radiolabeling tests were carried out under various conditions, including different [¹⁸F]fluoride sources, reaction solvents (e.g., DMF and MeCN), reaction times and reaction temperatures to produce the ¹⁸F-labeled compound [¹⁸F]6 (Table 1).¹⁹ The reaction of K[¹⁸F]F-K₂₂₂ or [¹⁸F]TBAF with precursor 5 afforded [¹⁸F]6 in low yields (0–3%) at 150 °C for 10 min. When precursor 5 was reacted with Cs[¹⁸F]F under general anhydrous radiofluorination conditions, the yield of [¹⁸F]6 was improved (~ 9% at 150 °C for 10 min). However, when 5 was reacted with Cs[¹⁸F]F for a longer reaction time (150 °C for 25 min), the yield of the desired product decreased to ~ 4%. We hypothesized that the reduced radiochemical yields at longer reaction times might be due to the decomposition of the iodonium salt precursor 5 and the radiofluorinated product [¹⁸F]6 at the 150 °C reaction temperature. In tests, 5 started slow thermal decomposition at 150 °C in open capillary tubes in a melting point apparatus, supporting this hypothesis. The addition of TEMPO (1.0 mg) as a radical scavenger and a small amount of water (10 μL) in the reaction solvent (DMF, 0.5 mL) to increase the solubility of Cs[¹⁸F]F greatly increased the radiochemical yields, up to ∼ 51% for manual syntheses at 150 °C for 5 min. Additional tests using a reduced temperature of 130 °C for 5 min, decreased the yield of [¹⁸F]6 to ~ 31%. Tests at temperatures greater than 150 °C did not produce any improvements in the yields.

Scheme 1.

Ineffective Initial Approach to an Improved Radiosynthesis of [¹⁸F]1.

Scheme 2.

¹⁸F-Fluorination of a Test Compound with a Fully Protected Guanidinyl Moiety.

Table 1.

[¹⁸F]Fluorination of diaryliodonium salts (5, 11 and 21) under various conditions.

Entry	I+ Salta	18F-fluoride [18F]F−X+	Method	T (°C)	t (min)	Solventb	TEMPO & H2Ob	Product	Yield (%)c	n
1	5	[18F]F−K+	manual	150	10	MeCN	–	[18F]6	NR	2
2	5	[18F]F−K+	manual	150	10	DMF	–	[18F]6	3 ± 2	2
3	5	[18F]F−TBA+	manual	150	10	DMF	–	[18F]6	Trace	2
4	5	[18F]F−Cs+	manual	150	10	DMF	–	[18F]6	9 ± 2	2
5	5	[18F]F−Cs+	manual	150	5	DMF	+	[18F]6	51 ± 5	3
6	5	[18F]F−Cs+	manual	130	5	DMF	+	[18F]6	31 ± 5	3
7	5	[18F]F−Cs+	manual	120	5	DMF	+	[18F]6	28 ± 3	3
8	5	[18F]F−Cs+	manual	150	15	DMF	+	[18F]6	43 ± 5	2
9	5	[18F]F−Cs+	manual	150	25	DMF	+	[18F]6	30 ± 5	2
10	11	[18F]F−Cs+	manual	150	5	DMF	+	[18F]22	16 ± 5	5
11	11	[18F]F−Cs+	manual	150	15	DMF	+	[18F]22	11 ± 4	3
12	11	[18F]F−Cs+	automated	150	5	DMF	+	[18F]1	7.0 ± 3.5 d	13
13	21	[18F]F−Cs+	automated	150	5	DMF	+	[18F]2	8.0 ± 3.5 e	12

^aAmount of diaryliodonium salt precursor used was 4-6 mg.

^bReaction solvent (MeCN or DMF; 500 μL) prepared without (−) or with (+) 1 mg TEMPO and 10 μL of H₂O.

^cProgress of the reaction and yields for manual reactions were analyzed by radio-TLC (developing solvent: ethylacetate/hexane = 30:70, vol/vol).

^dYield of the isolated pure product [¹⁸F]1 by semipreparative column (Phenomenex Synergi 10μ Hydro-RP 80A, 250×10 mm) using HPLC (5% EtOH in 40 mM NH₄OAc buffer, λ=254 nm, 4.0 ml/min).

^eYield of the isolated pure product [¹⁸F]2 by semipreparative column (Phenomenex Synergi 10μ Hydro-RP 80A, 250×10 mm) using HPLC (3.5% EtOH in 40 mM NH₄OAc buffer, λ=254 nm, 4.0 ml/min).

This encouraging result led us to prepare the required N,N′,N′′,N′′-tetrakis-Boc-protected guanidinyliodonium salt precursor for a simplified automated production of [¹⁸F]1 for clinical studies (Scheme 3). The N′,N′′-bis-Boc-protected iodophenethylguanidine 7 was previously synthesized in our laboratory.¹⁸ This compound was converted to the N,N′,N′′,N′′-tetrakis-Boc-protected guanidine 8 by treatment with di-t-butyl dicarbonate in the presence of dimethylaminopyridine and triethylamine in THF. Reaction of 8 with bis(trimethyl)tin in the presence of tetrakis(triphenylphosphine)palladium provided the trimethylstannane compound 9. [2-hydroxy(tosyloxy)iodo]thiophene,²⁰ which was generated in situ by a mixture of 2-(diacetoxyiodo)thiophene²¹ with p-toluenesulfonic acid under nitrogen atmosphere, was reacted with trimethylstannane 9 to provide (2-thienyl)iodonium tosylate 10 in 87% yield as a yellow solid. Among various counteranions, bromide was found to be especially reactive, increasing the radiochemical yield.²² Therefore, (2-thienyl)iodonium tosylate 10 was converted into the corresponding bromide 11 as a grey powder in 86% yield using KBr in a solution of CH₃CN and H₂O.

Scheme 3.

Synthetic Route to the Diaryliodonium Salt Precursor 11^a

^a(a) (Boc)₂CO, DMAP, Et₃N, THF, rt, 48h, 92%; (b) Sn₂Me₆, Pd(PPh₃)₄, toluene, reflux, N₂, 30 min, 86%; (c) (i) 2-(Diacetoxyiodo)thiophene, p-TsOH·H₂O, MeCN, CH₂Cl₂, N₂, rt, 1 h; (ii) 9, MeCN, CH₂Cl₂, N₂, rt, 20 h, 87%; (d) KBr, MeCN, H₂O, 65 °C-rt, 1 h, 86%.

A similar approach was used to prepare the guanidinyl-iodonium salt precursor 21 for production of [¹⁸F]2 (Scheme 4). The N-Boc-protected tyramine 13 was iodinated in the presence of sodium iodide and sodium hypochlorite to afford 14 in low yields (43%). The iodophenol 14 was protected by benzylation with benzyl bromide followed by subsequent deprotection of the Boc group to afford the benzylated 3-iodotyramine intermediate 16. Condensation of 16 with 1,3-bis(t-butoxycarbonyl)-2-methyl-2-thiopseudourea afforded the N′,N′′-bis-Boc-protected iodophenethylguanidine 17. This compound was used to convert the N,N′,N′′,N′′-tetrakis-Boc-protected guanidininyliodonium salt 21 using the same sequence of steps used to prepare 11 as shown in Scheme 3. Briefly, reaction of the trimethylstannane 19 with [2-hydroxy(tosyloxy)iodo]-thiophene provided the (2-thienyl)iodonium tosylate 20 in 86% yield as a yellow solid. Conversion of 20 to the corresponding bromide 21 yielded a grey powder in 96% yield. The structures of 11 and 21 were confirmed by ¹H NMR, ¹³C-NMR, high-resolution mass spectrometry (HRMS) spectra and C, H and N elemental analysis.

Scheme 4.

Synthetic Route to the Diaryliodonium Salt Precursor 21^a

^a(a) (Boc)₂CO, Et₃N, THF, rt, 24 h, 98%; (b) NaI, NaOCl, KOH, MeOH, 0 °C, 43%; (c) BnBr, K₂CO₃, Acetone, 4 h, 78%; (d) HCl, 65 °C, 1 h, 97%; (e) 1,3-N,N’-bis(tert-butoxycarbonyl)-2-methyl-2-thiopseudourea, Et₃N, DMF, 0 °C-rt, 24 h, 58%; (f) (Boc)₂CO, DMAP, Et₃N, THF, rt, 20 h, 63%; (g) Sn₂Me₆, Pd(PPh₃)₄, toluene, reflux, N₂, 30 min, 98%; (h) (i) 2-(Diacetoxyiodo)thiophene, p-TsOH·H₂O, MeCN, CH₂Cl₂, N₂, rt, 1 h; (ii) 19, MeCN, CH₂Cl₂, N₂, rt, 44 h, 86%; (i) KBr, MeCN, H₂O, 65 °C-rt, 1 h, 96%.

A TRACERlab FX_FN radiosynthesis module (GE Healthcare) was used to achieve a fully automated radiosynthesis of [¹⁸F]1 and [¹⁸F]2 using a two-step reaction as shown in Scheme 5. The first step incorporates [¹⁸F]F^– into the aromatic ring of the protected diaryliodonium salt precursor to yield an ¹⁸F-labeled intermediate ([¹⁸F]22 or [¹⁸F]23) using Cs[¹⁸F]F in DMF (150 °C, 5 min, 1 mg TEMPO, 10 μL H₂O). The second step is the simultaneous cleavage of the benzyl ether and the N,N′,N′′,N′′-tetrakis-Boc protecting groups by treatment with 3.0 N HBr at 120 °C for 15 min followed by HPLC purification to provide the desired radiotracer, [¹⁸F]1 or [¹⁸F]2. The structure of the ¹⁸F-labeled product was confirmed by comparing its retention time against the corresponding non-radioactive fluorine-19 standard using reverse-phase HPLC. The total synthesis time of the fully automated process is 90 min. For [¹⁸F]1, the purified product was obtained in 7.0 ± 3.5% yield at end of synthesis (EOS), (n = 13, decay-corrected based on starting activity) with >98% radiochemical purities. [¹⁸F]2 was obtained in 8.0 ± 3.5% yield (n = 12) with >99% radiochemical purities (Table 1). Activity levels in the final product ranged from 0.6–2.0 GBq for [¹⁸F]1 and 1.0–2.4 GBq for [¹⁸F]2. Specific activities were 56 ± 19 GBq/μmol for [¹⁸F]1 and 104 ± 26 GBq/μmol for [¹⁸F]2 at EOS. Other than the radiolabeled product, no other non-radioactive compounds were observed in the HPLC/UV trace during quality control of the final product. Mass concentrations averaged 0.48 ± 0.29 μg/mL for [¹⁸F]1 and 0.27 ± 0.14 μg/mL for [¹⁸F]2. Stability tests of 3 batches of each compound showed that both compounds are completely stable in the ammonium acetate buffer at room temperature for at least 6 hours.

Scheme 5.

Efficient Radiosynthetic Route to [¹⁸F]1 and [¹⁸F]2^a

^a(a) Cs[¹⁸F]F, TEMPO, H₂O, DMF, 150 °C, 5 min; (b) 3.0 N HBr, 120 °C, 15 min.

Isolated Rat Heart Studies.

The kinetics of [¹⁸F]2 were measured in the isolated working rat heart model, which has been used previously to measure the neuronal uptake and retention kinetics of [¹⁸F]1,¹⁸ the carbon-11 analog of MIBG ([¹¹C]MIBG) and [¹¹C]HED.¹⁶ Extraneuronal uptake (‘uptake-2’), which transports some norepinephrine analogs into myocytes, was blocked pharmacologically by adding 54 μM corticosterone to all heart perfusates.²³ At the beginning of the study, a 10 min constant infusion of tracer in the heart perfusate is performed to measure a neuronal uptake rate, K_up (mL perfusate/min/g wet). After this, the heart is switched to normal heart perfusate to measure clearance rates from neuronal spaces, expressed as a major clearance half-time, T_1/2 (h). As shown in Figure 2, [¹⁸F]2 had almost identical kinetics as [¹⁸F]1 in this model, with neuronal uptake rates K_up = 0.79 mL/min/g wet for [¹⁸F]2 and 0.77 mL/min/g wet for [¹⁸F]1. Major clearance halftimes were T_1/2 > 50 h for [¹⁸F]2 and T_1/2 > 24 h for [¹⁸F]1. The long neuronal retention times of these compounds are due to their efficient uptake and storage in norepinephrine storage vesicles. This shows that [¹⁸F]1 and [¹⁸F]2 are good substrates of the vesicular monoamine transporter, isoform 2 (VMAT2), which is localized in cell membranes of storage vesicles in peripheral sympathetic nerve terminals.²⁴ In comparison, [¹¹C]MIBG and [¹¹C]HED have much faster neuronal uptake rates, with K_up = 3.65 mL/min/g wet for [¹¹C]MIBG and K_up = 2.35 mL perfusate/min/g wet for [¹¹C]HED. They also have much faster clearance rates in this experimental model. [¹¹C]MIBG has a major clearance half-time T_1/2 = 2.1 h and [¹¹C]HED has T_1/2 = 1.6 h. These clearance rates are faster than are seen in human imaging studies with [¹²³I]MIBG and [¹¹C]HED because the isolated rat heart has coronary flow rates that are about 10 times higher than resting blood flow in a human heart.²⁵ If tracer molecules diffuse out of sympathetic neurons in the isolated rat heart they are more likely to be cleared from interstitial spaces into the capillary bed rather than re-entering nerve terminals by NET transport. In contrast, several studies have shown that [¹¹C]HED continuously diffuses out of nerve terminals only to be taken back up into the nerves during PET studies in normal human hearts.^25–27 The slower neuronal uptake rates and long neuronal retention times of [¹⁸F]1 and [¹⁸F]2 are favorable for tracer kinetic analysis,¹⁷ including Patlak graphical analysis, which requires irreversible trapping of a tracer in a tissue compartment.²⁸

Figure 2.

Kinetics of [¹⁸F]2 in isolated rat heart. Data for [¹⁸F]1, [¹¹C]MIBG and [¹¹C]HED are shown for comparison. The tracer is infused for 10 min to measure its neuronal uptake rate (K_up, mL/min/g wet), then the heart is switched to normal perfusate to measure tracer clearance rates (T_1/2, h). [¹⁸F]1 and [¹⁸F]2 have very long neuronal retention times due to efficient storage inside norepinephrine storage vesicles. ADV: apparent distribution volume.

PET Studies in Non-human Primates.

Dynamic PET studies with [¹⁸F]2 were performed in rhesus macaque monkeys to characterize the imaging properties, metabolism and myocardial kinetics of the tracer (n = 4). Representative cardiac PET images of [¹⁸F]2 show that the tracer provides high quality images of the distribution of sympathetic nerve terminals in the heart (Figure 3). Corresponding PET images for [¹⁸F]1 are also provided. There is little retention of either tracer in the lungs, providing very high heart-to-lung contrast. [¹⁸F]2 exhibited good heart-to-blood (H/B) contrast in the final PET images (t = 85 min), with H/B = 4.0 ± 0.3. This was better heart-to-blood contrast than the ratio measured for [¹⁸F]1, which was H/B = 3.0 ± 0.5.^{17, 18} Final heart-to-liver (H/L) ratios for [¹⁸F]2 were 2.2 ± 0.8. Due to faster clearance of activity from the liver, contrast between heart and liver was a little better for [¹⁸F]1, with final H/L = 2.5 ± 0.3. No significant uptake of free fluorine-18 was observed in the vertebral bones of the spine for either compound, indicating that these tracers are not susceptible to defluorination in non-human primates.

Figure 3.

Representative PET images of [¹⁸F]1 and [¹⁸F]2 in rhesus macaque monkeys.

Metabolism of [¹⁸F]2.

To assess the metabolism of [¹⁸F]2, six venous blood samples were drawn during the dynamic PET studies. Plasma was separated from blood and processed for analysis on a reverse-phase HPLC system with an in-line radiation detector optimized for detection of 511 keV positron annihilation photons to determine the fraction of plasma activity in the form of radiolabeled metabolites.¹⁸ Representative data showing the metabolic breakdown of [¹⁸F]2 in plasma are shown in Figure 4. Data on the metabolic breakdown of [¹⁸F]1 in rhesus macaques are shown for comparison. For both compounds, the metabolism time course was bi-phasic, with an initially rapid phase of metabolism followed by a slower phase. [¹⁸F]2 was metabolized more slowly than [¹⁸F]1. The mean time at which 50% of parent radiotracer was still intact was 6.7 ± 2.4 min for [¹⁸F]2 vs. 2.3 ± 0.9 min for [¹⁸F]1 (n = 4 each).

Figure 4.

Time course of the metabolic breakdown of [¹⁸F]1 and [¹⁸F]2 in the plasma of rhesus macaque monkeys.

Using in vitro incubations of [¹⁸F]1 with a monkey liver cytosol fraction and the cofactor for sulfur conjugation, 3’-phospho-adenosine-5’-phosphosulfate (PAPS), we previously showed that the main metabolic pathway for this compound in rhesus macaques was sulfate conjugation at its m-hydroxyl group.¹⁸ Incubations of [¹⁸F]2 under identical conditions were unable to produce any radiometabolite, suggesting that this compound is not a substrate for sulfate conjugation in rhesus monkeys. Additional in vitro tests with [¹⁸F]2 using a monkey liver microsomal fraction and the glucuronidation cofactor uridine 5’-diphospho-glucuronic acid (UDPGA) were also negative. Further tests are needed to identify the unknown polar metabolite of [¹⁸F]2 produced in rhesus macaque monkeys.

Blood Activity Partitioning.

Using aliquots of plasma and whole blood, the ratio of the activity concentration in plasma (C_p) over that in whole blood (C_wb) was determined for each venous blood sample. For [¹⁸F]2, the mean ratio C_p/C_wb measured was 1.23 ± 0.05 (n = 4). The C_p/C_wb ratio tended to be constant throughout the PET study. This is similar to previous results for [¹⁸F]1 which found C_p/C_wb = 1.25 ± 0.06 (n = 4).¹⁸

Tracer Kinetic Analysis.

The myocardial tissue kinetics C_t(t) of [¹⁸F]2 and its estimated kinetics in plasma C_p(t) were analyzed using compartmental modeling and Patlak graphical analysis as previously described.^{17, 18} For compartmental modeling analysis, the model used has two tissue compartments, one for extraction into and clearance from extracellular spaces, with rate constants K₁ (mL/min/g) and k₂ (min^–1), respectively, and the second for irreversible uptake into sympathetic neurons, with rate constant k₃ (min^–1). A blood volume fraction term BV (dimensionless) was also included to account for activity in myocardial blood. The estimates of K₁, k₂, and k₃ from compartmental modeling were used to calculate a ‘net uptake rate constant’ K_i (mL/min/g) = (K₁k₃)/(k₂ + k₃), which reflects the net rate of tracer influx into tissue compartments. K₁ estimates were fairly consistent for [¹⁸F]2, while k₂ and k₃ were more variable, similar to previous findings with [¹⁸F]1. The mean values of K₁ for [¹⁸F]2 were K₁ = 0.46 ± 0.05 mL/min/g, compared with K₁ = 0.56 ± 0.07 mL/min/g for [¹⁸F]1. Estimates of k₃ (the rate constant related to neuronal uptake of the tracer) were too variable to be useful as a quantitative measure of nerve density, but the values of the net uptake rate constant K_i calculated from the rate constant estimates were very consistent. For [¹⁸F]2, the mean net uptake rate constant was K_i = 0.221 ± 0.029 mL/min/g, which can be compared with K_i = 0.341 ± 0.041 mL/min/g for [¹⁸F]1. Patlak graphical analysis also gave consistent results. This analysis approach uses a mathematical transformation of C_p(t) and C_t(t) to construct a ‘Patlak plot’ which has a distinct linear phase. The slope of the linear phase, the Patlak slope K_p (mL/min/g), is theoretically equal to the calculated net uptake rate constant K_i derived from the compartmental modeling rate constants. For [¹⁸F]2, the measured Patlak slopes averaged K_p = 0.169 ± 0.021 mL/min/g, comparied with K_p = 0.302 ± 0.031 mL/min/g for [¹⁸F]1. For any given study, relative to their corresponding K_i values, the Patlak slopes are biased a little lower in magnitude because Patlak analysis does not account for myocardial blood activity in the tissue kinetics C_t(t), while the blood volume term BV used in compartmental modeling does account for the blood activity. Representative examples of the results of compartmental modeling and Patlak analysis of [¹⁸F]2 are shown in Figure 5. In previous PET studies in non-human primates with [¹¹C]GMO and [¹⁸F]1, using the NET inhibitor desipramine (DMI) to pharmacologically induce different degrees of NET transporter occlusion showed that net uptake rate constants K_i and Patlak slopes K_p each tracked sensitively and reproducibly with DMI-induced declines in available NET transporters.^{17, 18} Taken together, the results of these studies suggest that values of K_i or K_p obtained from tracer kinetic analysis of the kinetics of [¹⁸F]1 or [¹⁸F]2 can be used as quantitative metrics of regional sympathetic nerve density in the heart.

Figure 5.

Myocardial tracer kinetic analyses for [¹⁸F]2 kinetics in the same monkey. Parameter estimates from compartment modeling (A) were used to calculate a net uptake rate constant K_i (mL/min/g). Patlak graphical analysis (B) provided a Patlak slopes K_p (mL/min/g).

Advantages of ¹⁸F-Hydroxyphenethylguanidines.

[¹⁸F]1 and [¹⁸F]2 offer some advantages over existing cardiac sympathetic innervation radiotracers. The longer half-life of a fluorine-18 (1.83 h) compared with the 20.4 min half-life of carbon-11 in compounds such as [¹¹C]HED and [¹¹C]GMO was one advantage discussed previously. In terms of structure-activity relationships, there are four major factors that govern the neuronal uptake and retention kinetics of sympathetic nerve radiotracers: (a) the rate of neuronal uptake by NET (equal to V_max/K_m for NET transport); (b) the rate of vesicular uptake (equal to V_max/K_m for VMAT2 transport); (c) vulnerability to intraneuronal metabolism by enzymes such as monoamine oxidase (MAO); and (d) membrane diffusion rates, which influence vesicular storage and neuronal retention, and are partly determined by a tracer’s lipophilicity (log P). [¹⁸F]1 and [¹⁸F]2 were developed in an effort to design a tracer with optimal properties in each of these four categories. We hypothesized that such a tracer would possess myocardial kinetics that could be successfully analyzed using established tracer kinetic analysis techniques to provide quantitative measures of regional sympathetic nerve density. Specifically, the targeted properties were: (a) slower NET transport rates than [¹¹C]HED and [¹²³I]MIBG; (b) rapid vesicular uptake due to efficient VMAT2 transport; (c) resistance to intraneuronal metabolism; and (d) low lipophilicity to promote long neuronal retention times. The slower neuronal uptake rate makes the rate constant associated with this process more identifiable from the cardiac PET kinetics, and the rapid uptake and long retention in storage vesicles eliminates one rate constant from the kinetic model, leading to more reliable and consistent parameter estimates.¹⁷

To achieve this goal, radiolabeled phenethylguanidines were investigated because several phenethylguanidines are potent neuron blocking agents due to their prolonged retention inside norepinephrine storage vesicles.^29–31 The isolated rat heart studies with [¹⁸F]1 and [¹⁸F]2 presented above (Figure 2) demonstrate that these compounds satisfy the targeted properties (a), (b) and (d). PET studies showing very long neuronal retention times for [¹⁸F]1 and [¹⁸F]2 in non-human primate myocardium are consistent with (b) rapid vesicular uptake and (d) very little diffusion of the tracers from storage vesicles. For (c), the guanidine group of the side chain of [¹⁸F]1 and [¹⁸F]2 confers stability against neuronal enzymes such as tyrosine hydroxylase, MAO, dopamine-β-hydroxylase, or DOPA decarboxylase.³² Previous studies with the carbon-11 analog of [¹⁸F]1 in rats showed that 100% of the activity in the myocardium at t = 30 min after tracer administration was in the form of the parent tracer, consistent with no intraneuronal metabolism of the tracer.³³ Thus, [¹⁸F]1 and [¹⁸F]2 have been found to satisfy all of our targeted neuronal tracer properties.

Structure-activity studies of a series of ¹¹C-labeled phenethylguanidines in the isolated rat heart showed that several hydroxyphenethylguanidines had extremely long retention times in this model. This was particularly true for ring hydroxylation at the m- and p- positions, the beta- position of the side chain, or combinations of these hydroxyl substitutions.¹⁶ In contrast, none of the ¹¹C-labeled benzylguanidine compounds tested, including [¹¹C]-p-hydroxybenzylguanidine, were found to have long neuronal retention times in the isolated rat heart. These results suggest that hydroxyphenethylguanidines are better VMAT2 substrates than benzylguanidines, or are retained in storage vesicles much more efficiently, leading to high uptake and prolonged retention in storage vesicles. The highly acidic interior of vesicles has been described as an ‘amine trap’, since the low pH inside vesicles further encourages high pK_a molecules to exist in their ionized form.³⁴ The very high pK_a of the guanidine group of hydroxyphenethylguanidines (pK_a = 10 to 13) likely causes them to be highly protonated inside vesicles, further enhancing its vesicular retention.

Two ¹⁸F-labeled analogs of [¹²³I]MIBG, m-[¹⁸F]fluorobenzylguanidine ([¹⁸F]MFBG) and p-[¹⁸F]fluorobenzylguanidine ([¹⁸F]PFBG) were synthesized previously for PET studies of cardiac sympathetic innervation and neuroendocrine tumors.^{35, 36} Studies of [¹⁸F]PFBG in the isolated rat heart model showed it had a neuronal uptake rate K_up = 2.80 mL/min/g wet, ~ 77% of the rate of 3.65 mL/min/g wet measured for [¹²³I]MIBG.^{16, 37} The major neuronal clearance halftime for [¹⁸F]PFBG was T_1/2 = 0.59 h, compared with T_1/2 = 2.1 h for [¹²³I]MIBG. Thus, [¹⁸F]PFBG has rapid NET transport rate into sympathetic neurons and clears from the neurons faster than [¹²³I]MIBG. We are not aware of any human cardiac imaging studies with [¹⁸F]PFBG or [¹⁸F]MFBG, although one report used [¹⁸F]PFBG to assess cardiac nerve damage in a canine model of infarction.³⁸ While these tracers are certainly capable of imaging cardiac sympathetic nerves, their kinetics would not satisfy the four tracer properties we had targeted in the development of [¹⁸F]1 and [¹⁸F]2. Thus, another advantage of the ¹⁸F-hydroxyphenethylguanidines [¹⁸F]1 and [¹⁸F]2 over radiolabeled benzylguanidines is the ability to analyze their irreversible myocardial kinetics using standard tracer kinetic methods, such as Patlak analysis, to obtain quantitative regional estimates of sympathetic nerve density.

CONCLUSION

Our previous multi-step radiosynthesis of [¹⁸F]1 was not a practical approach for routine production of [¹⁸F]fluoro-hydroxyphenethylguanidines in a clinical PET radiochemistry facility. A more efficient method involving ¹⁸F-labeling of a N,N’,N’’,N’’-tetrakis-Boc protected guanidinyliodoium salt precursor followed by a single deprotection step and HPLC purification has been developed. This new radiosynthetic approach has proven to be very reliable and provides sufficient radiochemical yields and specific activities for clinical studies in human subjects. This new method was used to synthesize [¹⁸F]2, a structural isomer of [¹⁸F]1. Preclinical tests of [¹⁸F]2 demonstrated that it has imaging properties and kinetics that are very similar to those observed with [¹⁸F]1, and that quantitative metrics of regional cardiac sympathetic nerve density can be obtained from tracer kinetic analyses of its myocardial kinetics.

Due to the different metabolic pathways of [¹⁸F]1 and [¹⁸F]2, both agents are currently being evaluated in first-in-human studies under an exploratory IND clearance from the FDA (ClinicalTrials.gov NCT02385877). The results of these studies will be used to select a lead radiotracer for further clinical development. If PET studies with one of these radiotracers can provide accurate and sensitive regional measures of cardiac sympathetic nerve density in human hearts, it could be used to investigate the contribution of cardiac sympathetic denervation to mechanisms that lead to sudden cardiac death from heart diseases. Based on previous clinical studies with [¹²³I]MIBG and [¹¹C]HED, a potential clinical role for [¹⁸F]1 or [¹⁸F]2 would be in improved risk stratification of heart failure patients being staged for implantable cardioverter defibrillator (ICD) therapy.

METHODS

Chemicals and General Instrumentation.

NMR spectra were obtained on a Varian Inova 500 (499.90 MHz for ¹H; 125.70 MHz for ¹³C) spectrometer. ¹H and ¹³C NMR chemical shifts (δ) are reported in parts per million (ppm) relative to internal standard TMS and coupling constants (J) are in Hz. High-resolution mass spectra were obtained on a VG (Micromass) 70–250S spectrometer using electrospray ionization (ESI) in positive ion mode, direct chemical ionization (DCI) or electron impact (EI) at 70 eV. Melting points were determined on a Mel-Temp capillary melting point apparatus in open capillary tubes. Flash column chromatography was performed with E. Merck 230–400 mesh silica gel. Analytical TLC was performed with Analtech 0.25 mm glass-backed plates with fluorescent background. Visualization of TLC plates was achieved by UV illumination or treatment with phosphomolybdic acid (PMA). High pressure liquid chromatography (HPLC) was performed on a Hitachi pump L-7100 instrument equipped with Hitachi D-7500 integrator and Hitachi L-4000 UV detector. Radioactivity detection was done with Bioscan coincidence (model B-FC-4000) detector. Reverse-phase HPLC analysis of the formation of radiometabolites in plasma samples was performed on a Perkin-Elmer Series 410 LC instrument equipped with an Ortec Model 905–4 NaI(Tl) radiodetector (Oak Ridge, TN).

Reagents and solvents were purchased from commercial sources and used without further purification unless otherwise noted. [¹⁹F]4F-MHPG and [¹⁹F]3F-PHPG (standards for HPLC analysis) and N,N′-bis(tert-butoxycarbonyl)-N-3-benzyloxy-4-iodophenethylguanidine (7) were prepared in our laboratory using previously reported methods.^{18, 39}

N,N′,N′′,N′′-Tetrakis(tert-butoxycarbonyl)-N-3-benzyloxy-4-iodophenethylguanidine (8).

A solution of di-tert-butyl dicarbonate (6.1 mmol, 6.1 mL of 1.0 M solution in THF) was added to a solution of N′,N′′-bis(tert-butoxycarbonyl)-N-3-benzyloxy-4-iodophenethylguanidine 7 (600 mg, 1.0 mmol), dimethylaminopyridine (74 mg, 0.61 mmol) and triethylamine (0.85 mL, 6.1 mmol) in anhydrous THF (12 mL) at room temperature. The mixture was stirred for 48 h and then poured over water (50 mL). The mixture was diluted with ethyl acetate (50 mL) and extracted with ethyl acetate (2 × 50 mL). The combined extracts were washed with brine, dried over Na₂SO₄ and concentrated under reduced pressure. The residue was purified by flash column chromatography (silica gel, 15% ethyl acetate in hexane) to afford the product 8 (730 mg, 92%) as a white oil; ¹H NMR (500 MHz, CDCl₃) δ 7.69 (d, J = 7.9 Hz, 1H), 7.53 (d, J = 7.4 Hz, 2H), 7.40 (t, J = 7.4 Hz, 2H), 7.32 (t, J = 7.4 Hz, 1H), 6.83 (d, J = 1.7 Hz, 1H), 6.66 (dd, J = 7.9, 1.7 Hz, 1H), 5.15 (s, 2H), 3.94 (td, J = 8.0, 4.9 Hz, 2H), 2.90 (td, J = 8.0, 4.9 Hz, 2H), 1.52–1.46 (m, 36H); ¹³C NMR (125 MHz, CDCl₃) δ 157.95, 157.48, 151.39, 147.63, 143.88, 141.05,139.51,136.74, 131.09, 128.72, 128.05, 127.40, 123.71, 113.89, 84.30, 83.92, 83.90, 82.29, 71.05, 48.78, 33.36, 28.25, 28.17, 28.13; MS (ESI) m/z 796 (M+H)⁺, HRMS (ESI) calcd for C₃₆H₅₀IN₃O₉ 818.2484 (M+Na)⁺, found 818.2491; Anal. Calcd. For C₃₆H₅₀IN₃O₉: C, 54.34; H, 6.34; N, 5.28. Found: C, 54.54; H, 6.41; N, 5.14.

N,N′,N′′,N′′-Tetrakis(tert-butoxycarbonyl)-N-3-benzyloxy-4-trimethylstannylphenethyl guanidine (9).

Hexamethylditin (0.36 mL, 1.73 mmol) was added to a solution of compound 8 (688 mg, 0.86 mmol) and tetrakis(triphenylphosphine)palladium (50 mg, 0.04 mmol) in anhydrous toluene (8.0 mL) at room temperature under nitrogen atmosphere. The resulting mixture was heated to 130 ^oC for 30 min, cooled down to room temperature and filtered through a Celite pad. Celite pad was washed with ethyl acetate and the solvent was removed under reduced pressure. The residue was purified by flash column chromatography (silica gel, 100% hexane to 10% ethyl acetate in hexane) to afford the product 9 (621 mg, 86%) as a yellow oil; ¹H NMR (500 MHz, CDCl₃) δ 7.40–7.34 (m, 4H), 7.31–7.29 (m, 2H), 6.88 (d, J = 7.2 Hz, 1H), 6.82 (s, 1H), 5.02 (s, 2H), 3.96 (t, J = 8.1 Hz, 2H), 2.92 (t, J = 8.1 Hz, 2H), 1.53–1.31 (m, 36H), 0.17 (s, 9H); ¹³C NMR (126 MHz, CDCl₃) δ 158.0, 157.5, 151.4, 147.6, 143.9, 141.1 139.5, 136.7, 128.7, 128.5, 127.4, 123.7, 113.9, 84.3, 83.9, 83.9, 82.3, 71.1, 48.8, 33.4, 28.3, 28.2, 28.1, 0.2; MS (ESI) m/z 834 (M+H)⁺, HRMS (ESI) calcd for C₃₉H₅₉N₃O₉Sn 856.3166 (M+Na)⁺, found 856.3183; Anal. Calcd. For C₃₉H₅₉N₃O₉Sn: C, 56.26; H, 7.14; N, 5.05. Found: C, 56.00; H, 7.14; N, 4.88.

2-Benzyloxy-4-{2′-(N,N′,N′′,N′′-tetrakis(tert-butoxycarbonyl)guanidinyl)ethyl}phenyl (2-thienyl)iodonium tosylate (10).

A solution of 2-(diacetoxy)iodothiophene (92 mg, 0.278mmol) in CH₂Cl₂ (1.0 mL) was added to a solution of p-toluenesulfonic acid hydrate (53 mg, 0.278 mmol) in MeCN (1.0 mL) at room temperature under nitrogen atmosphere. The white precipitate was immediately generated and the mixture was stirred for 1 h. A solution of compound 9 (232 mg, 0.278 mmol) in CH₂Cl₂ (1.0 mL) and MeCN (1.0 mL) was added slowly to the reaction mixture. After the white precipitate was disappeared, the mixture was stirred at room temperature for 20 h under nitrogen atmosphere. The solvent was removed under reduced pressure and the residue was purified by flash column chromatography (silica gel, 100% CH₂Cl₂ to 20:1=CH₂Cl₂:MeOH) to afford the tosilate salt 10 (255 mg, 87%) as a yellow solid. mp 79–84 °C; ¹H NMR (500 MHz, DMSO-d₆) δ 8.26 (d, J = 8.1 Hz, 1H), 7.89 (dd, J = 5.3, 1.3 Hz, 1H), 7.76 (dd, J = 3.8, 1.3 Hz, 1H), 7.52–7.37 (m, 7H), 7.24 (d, J = 1.6, 1H), 7.11–7.08 (m, 3H), 6.95 (dd, J = 8.2, 1.6 Hz, 1H), 5.34 (s, 2H), 3.91 (t, J = 7.4 Hz, 2H), 2.91 (t, J = 7.4 Hz, 2H), 2.28 (s, 3H), 1.43–1.39 (m, 27H), 1.29 (s, 9H); ¹³C NMR (126 MHz, DMSO-d₆) δ 157.05, 155.10, 150.54, 146.676, 146.23, 145.82, 143.64, 139.44, 137.51, 136.87, 136.33, 135.67, 129.23, 128.63, 128.40, 128.01, 127.96, 127.28, 125.48, 123.77, 114.20, 108.08, 101.56, 83.43, 83.21, 81.51, 70.90, 48.86, 47.40, 46.17, 40.01, 32.64, 27.49, 27.41, 27.19, 20.77; HRMS (EI) calcd for C₄₀H₅₃IN₃O₉S 878.2542 (M-OTs)⁺, found 878.2546.

2-Benzyloxy-4-{2′-(N,N′,N′′,N′′-tetrakis(tert-butoxycarbonyl)guanidinyl)ethyl}phenyl (2-thienyl)iodonium bromide (11).

A solution of KBr (98 mg, 0.82 mmol) in H₂O (1.0 mL) was added to a solution of compound 10 (200 mg, 0.19 mmol) in MeCN (1.0 mL) at 60 ^oC for 5 min. The reaction mixture was stirred at room temperature for 1 h. The precipitate was washed with ice H₂O (10 mL), filtered, washed further with hexane several times and dried in vacuo to afford the bromide salt 11 (156 mg, 86%) as a light yellow solid. mp 116–119°C; ¹H NMR (500 MHz, DMSO-d₆) δ 8.25 (d, J = 8.1 Hz, 1H), 7.86 (d, J = 5.3 Hz, 1H), 7.73 (d, J = 3.7 Hz, 1H), 7.52–7.38 (m, 5H), 7.22 (s, 1H), 7.08 (dd, J = 5.3, 3.7 Hz, 1H), 6.94 (d, J = 8.1 Hz, 1H), 5.33 (s, 2H), 3.91 (t, J = 7.4 Hz, 2H), 2.90 (t, J = 7.4 Hz, 2H), 1.43–1.39 (m, 27H), 1.30 (s, 9H); ¹³C NMR (126 MHz, DMSO-d₆) δ 157.05, 155.06, 150.53, 146.73, 146.00, 143.63, 139.03, 136.90, 135.96, 135.70, 129.10, 128.61, 128.35, 127.92, 123.72, 114.17, 109.53, 108.87, 102.82, 83.43, 83.19, 81.49, 70.87, 47.38, 32.64, 27.48, 27.40, 27.20; HRMS (ESI) calcd for C₄₀H₅₃IN₃O₉S 878.2542 (M-Br)⁺, found 878.2555; Anal. Calcd. For C₄₀H₅₃BrIN₃O₉S: C, 50.11; H, 5.57; N, 4.38. Found: C, 50.04; H, 5.69; N, 4.37.

2-(4-hydroxyphenyl)ethylamine tert-butylcarbamate (13).

To a solution of 2-(4-hydroxy phenyl)ethylamine 12 (4.00 g, 29.16 mmol) in THF (48 mL) was added triethylamine (4.3 mL, 30.82 mmol). Di-tert-butyl-dicarbonate (6.70 g, 30.70 mmol) was added to the resulting solution and the reaction mixture was stirred at room temperature overnight. The solvent was evaporated under reduced pressure and the residue was purified by flash column chromatography (silica gel, 20% ethyl acetate in hexane) to afford the desired compound 13 (6.78 g, 98%) as a colourless oil; ¹H NMR (500 MHz, CDCl₃) δ 7.03 (d, J = 8.1 Hz, 2H), 6.77 (d, J = 8.1 Hz, 2H), 4.57 (br. s, NH), 3.33 (t, J = 6.5 Hz, 2H), 2.71 (t, J = 6.5 Hz, 2H), 1.47 (s, 9H); ¹³C NMR (126 MHz, CDCl₃) δ 154.43, 129.88, 115.43, 98.46, 67.47, 42.01, 33.25, 28.46. CAS Registry Number: 64318–28-1.

2-(3-iodo-4-hydoxyphenyl)ethylamine tert-butylcarbamate (14).

Sodium iodide (7.41 g, 49.45 mmol) and NaOH (1.54 g, 38.50 mmol) were added to a solution of compound 13 (7.82 g, 32.98 mmol) in MeOH (60 mL). The resulting solution was cooled to 0 °C. Sodium hypochlorite (4.0–4.9% in water, 81.8 mL, 49.45 mmol) was added slowly to the solution by dropping funnel. The reaction temperature was kept at 0–3 °C. After adding NaOCl, the reaction mixture was stirred for one more hour at 0–5 °C. A solution of sodium thiosulfate (10% in H₂O, 70 mL) was added, and the pH was adjusted to 6.5 by addition of HCl (2.0 N solution). The product was extracted by ethyl acetate (3 × 100 mL), and the combined extracts were washed with brine, dried over Na₂SO₄ and concentrated under reduced pressure. The residue was purified by flash column chromatography (silica gel, 20% ethyl acetate in hexane) to afford the compound 14 (5.15 g, 43%) as a white powder; mp 112–115 °C; ¹H NMR (500 MHz, CDCl₃) δ 7.49 (s, 1H), 7.06 (d, J = 8.0 Hz, 1H), 6.92 (d, J = 8.0 Hz, 1H), 4.53 (br. s, NH), 3.31 (t, J = 6.5 Hz, 2H), 2.70 (t, J = 6.5 Hz, 2H), 1.47 (s, 9H); ¹³C NMR (126 MHz, CDCl₃) δ 155.86, 153.51, 138.23, 133.11, 130.64, 115.02, 85.67, 79.40, 44.56, 34.78, 28.54. CAS Registry Number: 788824–50-0.

2-(4-benzyloxy-3-iodophenyl)ethylamine tert-butylcarbamate (15).

Benzyl bromide (0.52 g, 3.03 mmol) and K₂CO₃ (0.63 g, 4.55 mmol) were added to a solution of compound 14 (1.10 g, 3.03 mmol) in acetone (17 mL). The resulting solution was stirred at 70 °C for 4 hours. The reaction mixture was filtered and the solvent was evaporated under reduced pressure. Water was added to the residue and the product was extracted by ethyl acetate (3 × 100 mL), and the combined extracts were washed with brine, dried over Na₂SO₄ and concentrated under reduced pressure. The residue was purified by flash column chromatography (silica gel, 10% ethyl acetate in hexane) to afford the compound 15 (1.07 g, 78%) as a white powder; mp 70–72 °C; ¹H NMR (500 MHz, CDCl₃) δ 7.63 (s, 1H), 7.49 (d, J = 7.4 Hz, 2H), 7.39 (t, J = 7.4 Hz, 2H), 7.32 (t, J = 7.4 Hz, 1H), 7.09 (d, J = 8.3 Hz, 1H), 6.79 (d, J = 8.3 Hz, 1H), 5.13 (s, 2H), 4.52 (br. s, NH), 3.32 (t, J = 6.5 Hz, 2H), 2.70 (t, J = 6.5 Hz, 2H), 1.44 (S, 9H); ¹³C NMR (126 MHz, CDCl₃) δ 155.90, 155.80, 139.67, 136.56, 133.56, 130.87, 128.79, 128.54, 127.86, 126.98, 112.72, 86.92, 79.33, 70.97, 41.78, 34.80, 28.45, 28.41. CAS Registry Number: 788824–73-7.

2-(4-Benzyloxy-3-iodophenyl)ethylamine hydrochloride (16).

Hydrochloric acid (45 mL of 4.0M solution in 1,4-dioxane, 180 mmol) was added to a solution of compound 15 (0.80 g, 17.6 mmol) in ethyl acetate (5 mL). The resulting solution was stirred at 65 °C for 1 hour. The solvent was evaporated under reduced pressure. The crude product was dissolved again in ethyl acetate and concentrated under reduced pressure. The residue was purified by flash column chromatography (silica gel, 5% methanol and 0.3% NH₄OH in methylene chloride) to afford the compound 16 (0.61 g, 97%) as a yellow powder; mp 162–167 °C; ¹H NMR (500 MHz, CDCl₃) δ 7.70 (s, 1H), 7.46 (d, J = 7.4 Hz, 2H), 7.37 (t, J = 7.4 Hz, 2H), 7.30 (t, J = 7.4 Hz, 1H), 7.15 (d, J = 8.4 Hz, 1H), 6.76 (d, J = 8.4 Hz, 1H), 5.07 (s, 2H), 3.20 (t, J = 7.8 Hz, 2H), 3.00 (t, J = 7.8 Hz, 2H); ¹³C NMR (126 MHz, CDCl₃) δ 156.50, 139.75, 136.35, 130.52, 129.83, 128.55, 127.89, 126.97, 112.91, 110.01, 87.48, 87.22, 70.92, 45.82, 41.12. CAS Registry Number: 794507–50-9.

N′, N′′-Bis(tert-butoxycarbonyl)-N-4-benzyloxy-3-iodophenethylguanidine (17).

To a cooled (0 °C) solution of compound 16 (0.37 g, 1.06 mmol) and triethylamine (0.75 mL, 5.38 mmol) in anhydrous DMF (3.5 mL) was added in portion 1,3-bis(tert-butoxycarbonyl)-2-methyl-2-thiopseudourea (0.34 g, 1.16 mmol). The resulting mixture was stirred at 0 °C for 1 h, warmed to room temperature and stirred overnight. The mixture was diluted with ethyl acetate (50 mL), washed with saturated NH₄Cl solution (200 mL), and extracted with ethyl acetate (2 × 150 mL). The combined extracts were washed with brine, dried over Na₂SO₄ and concentrated under reduced pressure. The residue was purified by flash column chromatography (silica gel, 10% ethyl acetate in hexane) to afford the product 17 (0.37 g, 58%) as a white solid; mp 115–117 °C. ¹H NMR (500 MHz, CDCl₃) δ 11.46 (br. s, 1NH), 8.38 (br. s, 1NH), 7.67 (s, 1H), 7.49 (d, J = 7.6 Hz, 2H), 7.39 (t, J = 7.6 Hz, 2H), 7.32 (t, J = 7.6 Hz, 1H), 7.11 (d, J = 9.0 Hz, 1H), 6.78 (d, J = 9.0 Hz, 1H), 5.13 (s, 2H), 3.62 (q, J = 6.7 Hz, 2H), 2.77 (t, J = 6.7 Hz, 2H), 1.50 (s, 18H); ¹³C NMR (126 MHz, CDCl₃) δ 156.11, 155.99, 139.76, 136.57, 133.15, 129.69, 128.53, 127.84, 126.98, 112.74, 86.91, 83.12, 79.26, 70.96, 42.15, 33.93, 28.31, 28.09; HRMS calcd for C₂₆H₃₄IN₃O₅ 596.1616, found 596.1615.

N, N′,N′′,N′′-Tetrakis(tert-butoxycarbonyl)-N-4-benzyloxy-3-iodophenethylguanidine (18).

A solution of di-tert-butyl dicarbonate (40.7 mmol, 40.7 mL of 1.0 M solution in THF) was added to a solution of compound 17 (4.04 g, 6.78 mmol), N,N-dimethylaminopyridine (497 mg, 4.07 mmol) and triethylamine (5.67mL, 40.1 mmol) in anhydrous THF (82 mL) at room temperature. The mixture was stirred for 48 h and then poured over water (200 mL). The mixture was diluted with ethyl acetate (200 mL). After decantation, the aqueous layer was extracted with ethyl acetate (2 × 200 mL). The combined extracts were washed with brine, dried over Na₂SO₄ and concentrated under reduced pressure. The residue was purified by flash column chromatography (silica gel, 10% ethyl acetate in hexane) to afford the product 18 (3.41 g, 63%) as a white oil; ¹H NMR (500 MHz, CDCl₃) δ 7.71 (s, 1H), 7.49 (d, J = 7.6 Hz, 2H), 7.39 (t, J = 7.6 Hz, 2H), 7.32 (t, J = 7.6 Hz, 1H), 7.17 (d, J = 8.4 Hz, 1H), 6.78 (d, J = 8.4 Hz, 1H), 5.13 (s, 2H), 3.93 (t, J = 8.0 Hz, 2H), 2.85 (t, J = 8.0 Hz, 2H), 1.50 (s, 36H); ¹³C NMR (126 MHz, CDCl₃) δ 157.68, 155.85, 147.34, 139.77, 136.58, 133.51, 129.92, 128.54, 127.85, 126.98, 112.71, 86.84, 83.66, 82.03, 70.96, 48.73, 31.87, 28.04, 27.98, 27.91; HRMS calcd for C₃₆H₅₀IN₃O₉ 818.2484, found 818.2479.

N,N′,N′′,N′′-Tetrakis(tert-butoxycarbonyl)-N-4-benzyloxy-3-trimethylstannylphenethylguanidine (19).

Hexamethylditin (2.0 mL, 9.60 mmol) was added to a solution of compound 18 (2.79 g, 3.50 mmol) and tetrakis(triphenylphosphine)palladium (200 mg, 0.16 mmol) in anhydrous toluene (30 mL) at room temperature under nitrogen atmosphere. The resulting mixture was heated to 130 °C for 30 min, cooled down to room temperature and filtered through a Celite pad. Celite pad was washed with ethyl acetate and the solvent was removed under reduced pressure. The residue was purified by flash column chromatography (silica gel, 100% hexane to 10% ethyl acetate in hexane) to afford the product 19 (2.87, 98%) as a yellow oil; ¹H NMR (500 MHz, CDCl₃) δ 7.42–7.37 (m, 4H), 7.32 (t, J = 6.5 Hz, 2H), 6.91 (d, J = 7.1 Hz, 1H), 6.84 (s, 1H), 5.04 (s, 2H), 3.98 (t, J = 8.1 Hz, 2H), 2.94 (t, J = 8.1 Hz, 2H), 1.51– 1.49 (m, 36H), 0.19 (s, 9H); ¹³C NMR (126 MHz, CDCl₃) δ 163.2, 157.8, 151.2, 147.4, 143.6, 141.2, 137.1, 136.5, 128.4, 128.1, 127.8, 127.7, 121.8, 110.9, 83.6, 83.4, 82.0, 70.0, 48.9, 33.6, 28.1, 28.0, 27.9, 14.1; HRMS (ESI) calcd for C₃₉H₅₉N₃O₉Sn 856.3166 (M+Na)⁺, found 856.3172; Anal. Calcd. For C₃₉H₅₉N₃O₉Sn: C, 56.26; H, 7.14; N, 5.05. Found: C, 56.00; H, 7.23; N, 4.90.

2-Benzyloxy-5-{2′-(N,N′,N′′,N′′-tetrakis(tert-butoxycarbonyl)guanidinyl)ethyl}phenyl(2-thienyl)iodonium tosylate (20).

A solution of 2-(diacetoxy)iodothiophene (209 mg, 0.637 mmol) in CH₂Cl₂ (5.0 mL) was added to a solution of p-toluenesulfonic acid hydrate (121 mg, 0.637 mmol) in MeCN (5.0 mL) at room temperature under nitrogen atmosphere. The white precipitate was immediately generated and the mixture was stirred for 1 h. A solution of compound 19 (530 mg, 0.637 mmol) in CH₂Cl₂ (3.0 mL) and MeCN (3.0 mL) was added slowly to the reaction mixture. After the white precipitate disappeared, the mixture was stirred at room temperature for 20 h under nitrogen atmosphere. The solvent was removed under reduced pressure and the residue was purified by flash column chromatography (silica gel, 100% CH₂Cl₂ to 20:1=CH₂Cl₂:MeOH) to afford the tosylate salt 20 (482 mg, 86%) as a yellow solid; mp 140–141°C; ¹H NMR (500 MHz, DMSO-d₆) δ 8.21 (s, 1H), 7.90 (t, J = 5.3 Hz, 1H), 7.77 (t, J = 5.3 Hz, 1H), 7.49–7.37 (m, 7H), 7.32 (d, J = 8.5 Hz, 2H), 7.12–7.08 (m,3H), 5.33 (s, 2H), 3.88 (t, J = 7.8 Hz, 2H), 2.85 (t, J = 7.8 Hz, 2H), 2.28 (s, 3H), 1.43–1.36 (m, 36H); ¹³C NMR (126 MHz, DMSO-d₆) δ 157.1, 153.7, 150.6, 146.7, 145.8, 143.6, 139.5, 137.5, 136.5, 136.3, 135.8, 134.7, 133.7, 129.2, 128.6, 128.3, 128.0, 127.7, 125.5, 113.9, 110.3, 101.3, 83.6, 83.2, 81.6, 70.9, 47.5, 31.1, 27.9, 27.52, 27.4, 27.3, 27.2, 20.9.

2-Benzyloxy-5-{2′-(N,N′,N′′,N′′-tetrakis(tert-butoxycarbonyl)guanidinyl)ethyl}phenyl(2- thienyl)iodonium bromide (21).

A solution of KBr (583 mg, 4.90 mmol) in H₂O (5.0 mL) was added to a solution of compound 20 (430 mg, 0.49 mmol) in MeCN (7.0 mL) at 60 °C for 5 min. The reaction mixture was stirred at room temperature for 1 h. The precipitate was washed with ice H₂O (10 mL), filtered, washed further with hexane several times and dried in vacuo to afford the bromide salt 21 (411 mg, 96%) as a light yellow solid; mp 140–141 °C; ¹H NMR (500 MHz, CDCl₃) δ 7.97 (s, 1H), 7.53 (d, J = 5.2 Hz, 1H), 7.46 (d, J = 5.2 Hz, 1H), 7.41–7.35 (m, 6H), 6.96–6.91 (m, 2H), 5.21 (s, 2H), 3.92 (t, J = 8.1 Hz, 2H), 2.90 (t, J = 8.1 Hz, 2H), 1.52–1.48 (m, 36H); ¹³C NMR (126 MHz, CDCl₃) δ 157.6, 153.6, 151.1, 147.3, 143.5, 137.9, 136.5, 135.2, 134.7, 134.6, 133.8, 128.8, 128.7, 128.5, 127.5, 114.4, 113.8, 83.89, 83.8, 82.2, 71.9, 48.4, 33.8, 32.0, 29.7, 28.1, 28.0, 27.9; HRMS (ESI) calcd for C₄₀H₅₃IN₃O₉S 878.2542 (M-Br)⁺, found 878.2540; Anal. Calcd. For C₄₀H₅₃BrIN₃O₉S: C, 50.11; H, 5.57; N, 4.38. Found: C, 51.86; H, 5.95; N, 4.24.

Radiosynthesis of [¹⁸F]1.

A TRACERlab FX_FN computer-controlled radiosynthesis module (GE Healthcare) was used to achieve fully automated radiosyntheses of [¹⁸F]1 and [¹⁸F]2. [¹⁸F]F^– was prepared by the ¹⁸O(p,n)¹⁸F reaction using H₂¹⁸O as the target material in a GE PETrace cyclotron. [¹⁸F]F^– was isolated from the enriched water by trapping on a Waters Sep-Pak^® Light QMA cartridge (pre-activated with 10 mL of ethanol and 10 mL of H₂O) and eluted from the cartridge into the glassy-carbon reactor vial of the TRACERlab FX_FN system with a solution of 0.5 mL Cs₂CO₃ (0.05 M in H₂O). MeCN (1.0 mL) was added to the reactor vessel and then water/acetonitrile is evaporated at 80 °C under vacuum with a nitrogen stream to yield dried Cs[¹⁸F]F. After cooling to 60 °C, a mixed solution of 0.5 mL of DMF and 20 μL of MilliQ H₂O, containing 5.5–6.0 mg of the diaryliodonium salt precursor 11 and 1.0 mg of TEMPO (2,2,6,6-tetramethylpiperidine-N-oxyl) were added to the reactor vessel containing Cs[¹⁸F]F. The sealed reaction mixture was heated at 150 °C for 5 min to produce 3-benzyloxy-4-[¹⁸F]fluorophenethyl-N,N′,N′′,N′′-tetrakis-BOC-guanidine [¹⁸F]22 as intermediate. After cooling to 70 °C, a solution of 48% HBr (0.5 mL) and MeCN (0.5 mL) was added to the reaction mixture. The reaction solution was heated at 120 °C for 15 min. and then cooled to 50 °C. Next, a mixture solution of NaOH solution (1.0 mL, 4.0 M in H₂O) and buffer solution (1.8 mL, 5% EtOH in 40 mM NH₄OAc) was added into the reactor vessel. This mixture was injected onto a reverse-phase HPLC column (Phenomenex Synergi 10μ Hydro-RP 80A, 250×10 mm, 5% EtOH in 40 mM NH₄OAc buffer_, flow rate 4.0 mL/min, λ = 254 nm) and [¹⁸F]1 was collected at R_t = 30–32 min. The collected [¹⁸F]1 fraction was passed through a 0.22 μm sterilizing filter directly into a 10 mL septum-sealed sterile pyrogen-free glass vial.

Specific activity (SA) was determined by injecting a sample of [¹⁸F]1 with known activity (kBq) onto an HPLC system used for quality control. The area under the UV absorbance peak associated with the [¹⁸F]1 radioactivity peak was compared against a predetermined standard curve to estimate the total mass ([¹⁸F]4F-MHPG + [¹⁹F]4F-MHPG) in μg. The ratio of ¹⁸F-activity to total mass (converted from μg to μmol using the molecular weight of [¹⁹F]4F-MHPG) gave the specific activity.

Radiosynthesis of [¹⁸xsF]2.

The same methods described for the synthesis of [¹⁸F]1 were used to prepare [¹⁸F]2, except that the appropriate precursor 21 was used instead of precursor 11. HPLC purification conditions were slightly different: Phenomenex Synergi 10μ Hydro-RP 80A, 250×10 mm, 3.5% EtOH in 40 mM NH₄OAc buffer_, flow rate 4.0 mL/min, λ = 254 nm. In this system, [¹⁸F]2 was collected at R_t = 31–33 min.

Isolated Rat Heart Studies.

Hearts from male Sprague-Dawley rats (225 – 500 g) were perfused under moderate workload conditions (7.3 mmHg preload, 73 mmHg afterload) using a working heart preparation.⁴⁰ Two parallel perfusion circuits were connected to the left atrial cannula with a 3-way connector to allow for rapid switching from one circuit to the other. The heart perfusate was Krebs-Henseleit (KH) bicarbonate buffer (118 nM NaCl, 4.7 mM KCl, 2.55 mM CaCl₂, 1.2 mM MgSO₄, 1.2 mM KH₂PO₄, and 25 mM NaHCO₃) containing 5 mM glucose, oxygenated with a 95% O₂/5% CO₂ gas mixture and held at 37 °C. Corticosterone (54 μM) was added to the perfusate to block extraneuronal uptake (uptake-2) of the radiotracer into the rat myocardium.²³

Fluorine-18 activity in the heart was measured externally using a pair of cesium fluoride (CsF) scintillation detectors with crystal size 51 cm diameter and 51 cm thick (Crismatec 51Y51; Saint-Gobain, Nemours, France). The front faces of the two CsF detectors were positioned directly opposite each other, ~ 4 cm apart, with the heart centered between them. Each detector was enclosed in a large cylindrical lead collimator (2 cm wall thickness, 25 cm long) to shield against background counts from radioactive sources outside the heart. Two coincidence detection circuits were established between the detectors using standard Nuclear Instrumentation Module (NIM) electronic modules. One circuit measured total coincident events between the two detectors (true + random coincident events), and the second measured only random coincident events. A computerized data acquisition system interfaced to the NIM-module coincidence circuits was used to acquire and record the whole-heart radioactivity data throughout the study.⁴¹

Hearts were initially perfused for a 30 min stabilization period using KH buffer in the first perfusion circuit. During this time, [¹⁸F]2 was added to 1.0 L of KH buffer circulating in the second perfusion circuit and allowed to equilibrate over several minutes. Three 1.0 mL aliquots were drawn from the second perfusion circuit for counting in a gamma counter (Cobra II Auto-Gamma, Perkin-Elmer, Waltham, MA) to determine the radioactivity concentration in the perfusate (C_p), which was ~ 74 kBq/mL perfusate. After initiating data acquisition from the CsF detectors, the heart was rapidly switched to the second perfusion circuit to begin a constant infusion of [¹⁸F]2 for 10 min. Then the heart was switched back to the first perfusion circuit for 120 min to measure clearance rates of the tracer from the heart.

The acquired whole-heart radioactivity data (counts per second; cps) at each time point were converted to an ‘apparent distribution volume’ (ADV; mL perfusate/g wet), by dividing by the perfusate radioactivity concentration C_p (kBq/mL perfusate), the detector system calibration factor Z_calib (cps/kBq), and the measured wet mass of the heart M_w (g wet). Neuronal uptake rates of the radiotracers (K_up; mL perfusate/min/g wet) were calculated as the slope of a linear regression of the ADV data between t = 1 min and t = 4 min of the 10 min infusion study. Clearance rates were estimated by fitting the ADV data during the clearance phase of the study to multiple exponential decay processes. The exponential clearance rate constants (λ_i) were used to calculate corresponding clearance half-times: T_1/2 = ln(2)/λ_i. The slowest rate, associated with clearance from sympathetic neurons, is reported for each compound.

PET imaging Studies with [¹⁸F]2.

Cardiac PET studies (n = 4) were performed in rhesus macaque monkeys using a microPET P4 primate scanner (Siemens/CTI Concorde Microsystems, Knoxville, TN). After anesthetizing the animal, a percutaneous angiocather was placed in the saphenous vein of each leg, one for tracer injection, the other for blood sampling. Vital signs, including heart rate (bpm), blood oxygen saturation levels (SpO₂) and body temperature were monitored continuously (model V3404P, SurgiVet, Norwell, MA). Dynamic PET data were acquired in list-mode for 90 min after intravenous injection of 155 – 230 MBq of [¹⁸F]2. List-mode emission data were rebinned into a 27-frame dynamic sequence (12×10 s, 2×30 s, 2×60 s, 2×150 s, 2×300 s, 7×600 s). Rebinned emission data were corrected for attenuation and scatter, and transaxial images reconstructed using maximum a posteriori (MAP) reconstruction⁴², an iterative method that accounts for the detector point spread function in the model of the system.

Blood Partitioning and Radiometabolite Analysis.

Venous blood samples (1.5–2.0 mL) were centrifuged for 1 min at 12000 × g to separate plasma and red blood cells. Plasma was deproteinized by adding perchloric acid (HClO₄; final concentration 0.4N) and centrifuging for 5 min at 12000 × g. The supernatant was neutralized with KOH (pH 7.0–7.5) and filtered twice (Millex GS 0.22 μm, Millipore, Billerica, MA). Aliquots (0.1 mL) of whole blood, plasma and the final supernatant were counted in a gamma counter. Count data for plasma and whole blood aliquots were decay corrected and used to calculate the relative concentrations of [¹⁸F]2 in plasma and whole blood (C_p/C_wb). The final supernatant was analyzed by HPLC (Synergi 10 μm Hydro-RP column, 4.6 × 250 mm, 60 mM sodium phosphate buffer, pH 5.4 with 8% ethanol, flow rate 1.0 mL/min) with an in-line radiation detector. Under the HPLC conditions used [¹⁸F]2 had a retention time R_t = 12.7 min, while the main polar radiometabolite formed had R_t = 9.3 min. Peak area analysis of the radiation detection curve was used to estimate the percentage intact parent tracer fraction (f_intact) for each plasma sample.

In Vitro Metabolism Tests.

To test for sulfate conjugation of [¹⁸F]2, a 20 μL aliquot of monkey liver cytosol (#452461, BD Biosciences, San Jose, CA) was added to a 10 μL aliquot of 10 mM sulfotransferase cofactor PAPS (adenosine-3′-phosphate-5′-phosphosulfate lithium salt hydrate; #A1651, Sigma-Aldrich, Milwaukee, WI) dissolved in 50 μL of 1.0 mM Tris-HCl buffer (pH 7.4) and 170 μL of ultrapure water (18 MΩ·cm MilliQ, Millipore, Billerica, MA) and incubated at 37 °C for 5 min.⁴³ Next, 740 kBq of [¹⁸F]2 in 250 μL of ultrapure water was added to the reaction mixture (final volume 500 μL) and incubated at 37 °C for 20 min. The reaction was terminated by centrifugation at 16,000 × g for 5 min at 4 °C. The supernatant was filtered (Millex GS 0.22 μm, Millipore, Billerica, MA) and analyzed using HPLC with radiation detection as described above for the rhesus macaque plasma samples. To test for glucuronidation of [¹⁸F]2, a 25 μL aliquot containing 0.5 mg of monkey liver microsomes (#452413, BD Biosciences, San Jose, CA) was added to a reaction mixture of 50 μL of 20 mM glucuronidation cofactor UDPGA (uridine 5′-diphosphoglucuronic acid; #U5625 Sigma-Adrich, Milwaukee, WI) and 50 μL of 30 mM DTT (dithiothreitol, #D0632, Sigma-Aldrich, Milwaukee, WI) dissolved in 1.0 M glycine/NaOH buffer containing 50 mM MgCl₂ (pH 9.2). An additional 125 μL of the glycine/NaOH buffer was added and the mixture incubated at 37 °C for 5 min. Next, 740 kBq of [¹⁸F]2 in 250 μL of ultrapure water was added to the reaction mixture (final volume 500 μL) and incubated at 37 °C for 20 min. Using the same procedures for the sulfate conjugation test, the reaction mixture was centrifuged and the supernatant filtered for analysis using radio-HPLC.

Plasma Time-Activity Curve.

Region-of-interest analysis of the dynamic PET images was used to determine a time-activity curve of the activity concentration in whole blood, C_wb(t). The whole blood time-activity curve C_wb(t) was then multiplied by the measured ratio of activity in plasma over whole blood (C_p/C_wb) and by the percentage of intact parent tracer fraction in plasma data, f_intact(t), to estimate the kinetics of the plasma concentration of intact tracer, C_p(t). The estimated plasma curve C_p(t) was used as the input function for tracer kinetic analyses.

Tracer Kinetic Analysis.

For each PET study, the final four dynamic image frames were summed and used to draw a region-of-interest on the left ventricular wall, encompassing 3–4 transaxial slices, to extract a time-activity curve for myocardial tissue C_t(t). For compartmental modeling analysis, C_t(t) and the plasma kinetics C_p(t) for [¹⁸F]2 were analyzed using a two-tissue compartment model with irreversible trapping to estimate the rate constants K₁ (mL/min/g), k₂ (min^–1), k₃ (min^–1) and a blood volume fraction BV (dimensionless). The C_t(t) and C_p(t) data were also analyzed using Patlak analysis to estimate a Patlak slope, K_p (mL/min/g).

Animal Care.

The care of all animals used in this study was done in accordance with the Animal Welfare Act and the National Institute of Health’s Guide for the Care and use of Laboratory Animals.⁴⁴ Animal protocols were approved by the Institutional Animal Care and Use Committee (IACUC) at the University of Michigan.

ABBREVIATIONS USED

DMI

desipramine

[¹¹C]HED

[¹¹C]-(–)-m-hydroxyephedrine

[¹¹C]GMO

N-[¹¹C]guanyl-(–)-m-octopamine

[¹²³I]MIBG

[¹²³I]m-iodobenzylguanidine

NET

norepinephrine transporter

TEMPO

2,2,6,6-tetramethylpiperidine-N-oxyl

VMAT2

vesicular monoamine transporter, isoform 2.