Synthesis of [18F]F-γ-T-3, a Redox-Silent γ-Tocotrienol (γ-T-3) Vitamin E Analogue for Image-Based In Vivo Studies of Vitamin E Biodistribution and Dynamics

Vitamin E, a natural antioxidant, is of interest to scientists, health care pundits and faddists; its nutritional and biomedical attributes may be validated, anecdotal or fantasy. Vitamin E is a mixture of tocopherols (TPs) and tocotrienols (T-3s), each class having four substitutional isomers (α-, β-, γ-, δ-). Vitamin E analogues attain only low concentrations in most tissues, necessitating exacting invasive techniques for analytical research. Quantitative positron emission tomography (PET) with an F-18-labeled molecular probe would expedite access to Vitamin E’s biodistributions and pharmacokinetics via non-invasive temporal imaging. (R)-6-(3-[18F]Fluoropropoxy)-2,7,8-trimethyl-2-(4,8,12-trimethyltrideca-3,7,11-trien-1-yl)-chromane ([18F]F-γ-T-3) was prepared for this purpose. [18F]F-γ-T-3 was synthesized from γ-T-3 in two steps: (i) 1,3-di-O-tosylpropane was introduced at C6-O to form TsO-γ-T-3, and (ii) reaction of this tosylate with [18F]fluoride in DMF/K222. Non-radioactive F-γ-T-3 was synthesized by reaction of γ-T-3 with 3-fluoropropyl methanesulfonate. [18F]F-γ-T-3 biodistribution in a murine tumor model was imaged using a small-animal PET scanner. F-γ-T-3 was prepared in 61% chemical yield. [18F]F-γ-T-3 was synthesized in acceptable radiochemical yield (RCY 12%) with high radiochemical purity (>99% RCP) in 45 min. Preliminary F-18 PET images in mice showed upper abdominal accumulation with evidence of renal clearance, only low concentrations in the thorax (lung/heart) and head, and rapid clearance from blood. [18F]F-γ-T-3 shows promise as an F-18 PET tracer for detailed in vivo studies of Vitamin E. The labeling procedure provides acceptable RCY, high RCP and pertinence to all eight Vitamin E analogues.


Introduction
The Vitamin E family is comprised of eight chromanol analogues that fall into two subfamilies, tocopherols (TPs) and tocotrienols (T-3s). These families are characterized by the hydrocarbon chain at chromanol-C2. Whereas T-3s have a C2 isoprenoid side chain with three non-conjugated double bonds (C3′-C4′, C7′-C8′, C11′-C12′), the TPs have an identical but fully saturated chain. The four naturally occurring TP and T-3 analogues (α-, β-, γ-, δ-) differ from each other only through variations in methyl substitutions at C5, C7 and C8 on the aromatic chromanol ring (Figures 1 and 2) [1]. Vitamin E analogues are important natural antioxidants that protect cells by interacting with free radicals. Importantly, there is growing evidence that other medicinal properties including anticancer, anti-inflammatory, and neuroprotective activity may be as important as their antioxidant action. . Specific biochemical interactions may lower blood cholesterol and blood pressure, reverse atherosclerosis, minimize stroke-related brain damage, stimulate hair regrowth, and prevent sun damage to skin [2][3][4][5]. The Vitamin E analogue γ-T-3, present in many plant oils but especially palm, annatto and rice bran oils, is of particular interest to cancer researchers [6], and has been selected as the representative T-3 analogue in the current manuscript.
The pharmacology and molecular biology of Vitamin E are not fully resolved [5]. In general, the bicyclic portion (functional domain; Figure 2) is responsible for the antioxidant properties, whereas the hydrocarbon side chain at C2 has two functions: the proximal portion imparts signaling activity, and the distal (terminal) hydrocarbon tail imparts additional hydrophobicity which facilitates interaction with lipophilic cell components [2]. The side-chain terminus is subject to oxidative attack by cytochrome P450 4F2 (CYP4F2) at C-13′, leading to ω-hydroxylation followed by a cascade of five beta oxidations starting with C-13′. This truncation of the side chain ends with the water-soluble carboxyethylhydroxychroman or carboxymethylbutylhydroxychroman cores [7][8][9]. Metabolism of the aromatic ring of the chroman moiety has been reported to give rise to minor metabolites [10].
Gamma tocotrienol (γ-T-3), the analogue of interest because of its anti-cancer effects, has methyl substituents at C7 and C8 ( Figure 1) and acts at several molecular foci [11]. The scientific literature addressing Vitamin E is too extensive to review in this paper; interested readers are referred to the many contemporary reviews of its dietary and medicinal attributes [5,[12][13][14].  Vitamin E analogues are important natural antioxidants that protect cells by interacting with free radicals. Importantly, there is growing evidence that other medicinal properties including anti-cancer, anti-inflammatory, and neuroprotective activity may be as important as their antioxidant action. Specific biochemical interactions may lower blood cholesterol and blood pressure, reverse atherosclerosis, minimize stroke-related brain damage, stimulate hair regrowth, and prevent sun damage to skin [2][3][4][5]. The Vitamin E analogue γ-T-3, present in many plant oils but especially palm, annatto and rice bran oils, is of particular interest to cancer researchers [6], and has been selected as the representative T-3 analogue in the current manuscript.
The pharmacology and molecular biology of Vitamin E are not fully resolved [5]. In general, the bicyclic portion (functional domain; Figure 2) is responsible for the antioxidant properties, whereas the hydrocarbon side chain at C2 has two functions: the proximal portion imparts signaling activity, and the distal (terminal) hydrocarbon tail imparts additional hydrophobicity which facilitates interaction with lipophilic cell components [2]. The side-chain terminus is subject to oxidative attack by cytochrome P450 4F2 (CYP4F2) at C-13 , leading to ω-hydroxylation followed by a cascade of five beta oxidations starting with C-13 . This truncation of the side chain ends with the water-soluble carboxyethylhydroxychroman or carboxymethylbutylhydroxychroman cores [7][8][9]. Metabolism of the aromatic ring of the chroman moiety has been reported to give rise to minor metabolites [10].
Gamma tocotrienol (γ-T-3), the analogue of interest because of its anti-cancer effects, has methyl substituents at C7 and C8 ( Figure 1) and acts at several molecular foci [11]. The scientific literature addressing Vitamin E is too extensive to review in this paper; interested readers are referred to the many contemporary reviews of its dietary and medicinal attributes [5,[12][13][14].

Introduction
The Vitamin E family is comprised of eight chromanol analogues that fall into two subfamilies, tocopherols (TPs) and tocotrienols (T-3s). These families are characterized by the hydrocarbon chain at chromanol-C2. Whereas T-3s have a C2 isoprenoid side chain with three non-conjugated double bonds (C3′-C4′, C7′-C8′, C11′-C12′), the TPs have an identical but fully saturated chain. The four naturally occurring TP and T-3 analogues (α-, β-, γ-, δ-) differ from each other only through variations in methyl substitutions at C5, C7 and C8 on the aromatic chromanol ring (Figures 1 and 2) [1]. Vitamin E analogues are important natural antioxidants that protect cells by interacting with free radicals. Importantly, there is growing evidence that other medicinal properties including anticancer, anti-inflammatory, and neuroprotective activity may be as important as their antioxidant action. . Specific biochemical interactions may lower blood cholesterol and blood pressure, reverse atherosclerosis, minimize stroke-related brain damage, stimulate hair regrowth, and prevent sun damage to skin [2][3][4][5]. The Vitamin E analogue γ-T-3, present in many plant oils but especially palm, annatto and rice bran oils, is of particular interest to cancer researchers [6], and has been selected as the representative T-3 analogue in the current manuscript.
The pharmacology and molecular biology of Vitamin E are not fully resolved [5]. In general, the bicyclic portion (functional domain; Figure 2) is responsible for the antioxidant properties, whereas the hydrocarbon side chain at C2 has two functions: the proximal portion imparts signaling activity, and the distal (terminal) hydrocarbon tail imparts additional hydrophobicity which facilitates interaction with lipophilic cell components [2]. The side-chain terminus is subject to oxidative attack by cytochrome P450 4F2 (CYP4F2) at C-13′, leading to ω-hydroxylation followed by a cascade of five beta oxidations starting with C-13′. This truncation of the side chain ends with the water-soluble carboxyethylhydroxychroman or carboxymethylbutylhydroxychroman cores [7][8][9]. Metabolism of the aromatic ring of the chroman moiety has been reported to give rise to minor metabolites [10].
Gamma tocotrienol (γ-T-3), the analogue of interest because of its anti-cancer effects, has methyl substituents at C7 and C8 ( Figure 1) and acts at several molecular foci [11]. The scientific literature addressing Vitamin E is too extensive to review in this paper; interested readers are referred to the many contemporary reviews of its dietary and medicinal attributes [5,[12][13][14].  Vitamin E analogues remains a major concern for harnessing their medicinal properties. For example, the bioavailability of γ-T-3 has been reported to be~10% following a single oral bolus dose of 10 mg/kg to rats [15].
The syntheses (Scheme 1) of F-γ-T-3 and [ 18 F]F-γ-T-3 are now reported, together with a PET image of the biodistribution of [ 18 F]F-γ-T-3 following i.v. injection into a murine tumor model. The objective of this program is to develop reagents that will facilitate administration, bioavailability, clearance, and distribution studies of Vitamin E dynamics in animal models and humans through the application of imaging techniques [16,17]. The disposition of T-3′s within body tissues and association with various membranes are reported to be superior to those of other members of the natural Vitamin E group. The low bioavailability of all Vitamin E analogues remains a major concern for harnessing their medicinal properties. For example, the bioavailability of γ-T-3 has been reported to be ~10% following a single oral bolus dose of 10 mg/kg to rats [15].
The syntheses (Scheme 1) of F-γ-T-3 and [ 18 F]F-γ-T-3 are now reported, together with a PET image of the biodistribution of [ 18 F]F-γ-T-3 following i.v. injection into a murine tumor model. The objective of this program is to develop reagents that will facilitate administration, bioavailability, clearance, and distribution studies of Vitamin E dynamics in animal models and humans through the application of imaging techniques [16,17].

Chemistry
F-γ-T-3 and TsO-γ-T-3 were synthesized from γ-T-3 in acceptable chemical yields of 61% and 48%, respectively (Scheme 1). The isocratic HPLC system developed for this work provided good separation between TsO-γ-T-3 and F-γ-T-3. A chromatogram of a mixture of TsO-γ-T-3 and F-γ-T-3 showed the presence of several minor impurities which were not identified ( Figure 3). Although no quantitative stability studies were undertaken, standard solutions of TsO-γ-T-3 and F-γ-T-3 in ethanol showed minimal chemical change over a period of four months when stored at room temperature unprotected from light. Solutions of TsO-γ-T-3 in propylene glycol were less stable, appearing to break down within a few hours. More work would be required to establish the stability of TsO-γ-T-3 and F-γ-T-3 under other storage conditions.

Chemistry
F-γ-T-3 and TsO-γ-T-3 were synthesized from γ-T-3 in acceptable chemical yields of 61% and 48%, respectively (Scheme 1). The isocratic HPLC system developed for this work provided good separation between TsO-γ-T-3 and F-γ-T-3. A chromatogram of a mixture of TsO-γ-T-3 and F-γ-T-3 showed the presence of several minor impurities which were not identified ( Figure 3). Although no quantitative stability studies were undertaken, standard solutions of TsO-γ-T-3 and F-γ-T-3 in ethanol showed minimal chemical change over a period of four months when stored at room temperature unprotected from light. Solutions of TsO-γ-T-3 in propylene glycol were less stable, appearing to break down within a few hours. More work would be required to establish the stability of TsO-γ-T-3 and F-γ-T-3 under other storage conditions.

Radiofluorination
Radiofluorination of TsO-γ-T-3 provided adequate radiochemical yield (RCY 12%) without further optimization (Scheme 1). HPLC purification afforded 0.5-1 GBq high-purity (RCP > 99%) [ 18 F]F-γ-T-3. The gradient HPLC method produced high radiochromatographic purity product with no detectable radiochemical impurities. As expected, the high-specific-activity product obtained via nucleophilic radiofluorination was not detectable by UV absorption, but addition of internal standard F-γ-T-3 produced a peak at 14.4 min, directly underneath the radioactive peak of [ 18 F]F-γ-T-3 upon co-chromatography ( Figure 3). The other UV-active peaks were not identified, but a larger peak at approximately 7.2 min showed UV absorbance at both 254 and 291 nm, and could therefore be the

Radiofluorination
Radiofluorination of TsO-γ-T-3 provided adequate radiochemical yield (RCY 12%) without further optimization (Scheme 1). HPLC purification afforded 0.5-1 GBq high-purity (RCP >99%) [ 18 F]F-γ-T-3. The gradient HPLC method produced high radiochromatographic purity product with no detectable radiochemical impurities. As expected, the high-specific-activity product obtained via nucleophilic radiofluorination was not detectable by UV absorption, but addition of internal standard F-γ-T-3 produced a peak at 14.4 min, directly underneath the radioactive peak of [ 18 F]F-γ-T-3 upon co-chromatography ( Figure 3). The other UV-active peaks were not identified, but a larger peak at approximately 7.2 min showed UV absorbance at both 254 and 291 nm, and could therefore be the Tos-γ-T-3 hydrolysis product formed during work up of the radiofluorination reaction ( Figure 3). The calculated partition coefficients (logP) of F-γ-T-3 and γ-T-3 were 9.07 and 8.80, respectively.

Radiolabeling
There are very few options for the introduction of elemental radiolabels into Vitamin E analogues based on their chemical structure and ascribed regional structure-activity relationships (Figures 1 and 2). The radiolabeling options fall into two main categories: those that impact the redox/bio-oxidation properties (e.g., alteration/substitution at C6-OH; redox silencing), and those

Radiolabeling
There are very few options for the introduction of elemental radiolabels into Vitamin E analogues based on their chemical structure and ascribed regional structure-activity relationships (Figures 1 and 2). The radiolabeling options fall into two main categories: those that impact the redox/bio-oxidation properties (e.g., alteration/substitution at C6-OH; redox silencing), and those that alter signaling, docking and metabolic degradation ascribed to the C2 chain (e.g., substitution on isoprenoid double bonds of T-3s). The latter option requires creation of a reactive centre somewhere along the C2 chain, a challenging chemical synthesis, and creates substantial alterations in the biophysical properties of these molecules. Incorporation of tracer elements at aromatic sites other than C6-O would similarly necessitate extensive chemical modification. In short, chemical alteration at either the side chain or the aromatic moiety will temper interaction with membranes and macromolecular targets, alter redox characteristics and/or modulate absorption and biodistribution patterns. The search for more active synthetic derivatives of natural TPs and T-3s focus primarily on substitution at C6-O [18][19][20][21]. Introduction of F-18 at C6-O via a prosthetic linking arm seems to be the best option for radiofluorination of all Vitamin E analogues.
Two options are available for radiofluorination at C6-O. The approach based on electrophilic radiofluorination with [ 18 F]F 2 has several drawbacks including technical challenges of working with electrophilic fluorine, lower specific activity of the final labeled products, and low regioselectivity of substitution in aromatic systems [22]. The latter constraint could be addressed using organometallic precursors, but the synthesis of these precursors presents its own special challenges [23]. The second radiolabeling option is nucleophilic radiofluorination of an appropriate alkyl linking group introduced as a substituent on C6-O. The relative ease of nucleophilic substitution, the high specific activities available, and controlled regiospecificity all mitigate in its favor [24]. Electrophilic and nucleophilic radiofluorination at C6-O both give rise to new compounds that are redox silent, and although this affects the antioxidant properties, this has not been considered to be a major impediment to the development of more active derivatives [20].
The catabolism of Vitamin E analogues occurs via sequential oxidations, first by ω-oxidation of a terminal methyl group of the isoprenoid side chain, followed by sequential β-oxidations to the corresponding carboxyethyl hydroxychromans (CEHCs) and eventual urinary excretion in urine as glucuronated or sulfated forms [8,25]. Although qualitative metabolic effects have not been demonstrated as a result of aromatic ring substitution patterns and substituents, quantitative effects (rates) have been reported and ascribed to differences in binding to tocopherol-ω-hydroxylase (CYP4F2), the enzyme responsible for ω-hydroxylation [26].
The current demonstration is based on the nucleophilic radiofluorination approach using bifunctional fluoroalkanes, e.g., 18 F(CH 2 ) n X (n = 1-3, X = Br, OMes, OTos) [27]. This approach (Scheme 1) required the syntheses of reference F-γ-T-3 and the radiofluorination precursor TsO-γ-T-3, both of which proceeded in a straight-forward manner. As depicted in Scheme 1, mesylation of 3-fluoro-1-propanol afforded intermediate 2 which was subjected to nucleophilic substitution of its mesyl group with γ-T-3 to furnish the target F-γ-T-3 in 61% yield. The tosyl-radiofluorination precursor, TsO-γ-T-3, was prepared in a 47% yield via nucleophilic displacement of one tosylate group of the di-tosylate 3 with γ-T-3. Standard work up of these compounds required attention to their instability under silica gel chromatography and sensitivity to UV light.
NMR analyses of the non-radioactive compounds focused on resonance shifts arising from substitution of the C6 hydroxyl group of the aromatic ring by ether-linked propyl groups (3-tosylpropyland 3-fluoropropyl-). Assignment of proton and carbon resonances of all natural Vitamin E analogues remains incomplete despite the application of multidimensional NMR (HSQC, COSY, ADEQUATE(1,1), C-H HMBC, and NOESY) techniques [28]. Resonances introduced by the 3-(F/TsO)-propyloxy substituent at C6, replacing hydrogen on the C6 hydroxy of γ-T-3, are reported in the experimental section as CH 2 -1, CH 2 -2 and CH 2 -3 ( Figure 5).
3 with γ-T-3. Standard work up of these compounds required attention to their instability under silica gel chromatography and sensitivity to UV light.
NMR analyses of the non-radioactive compounds focused on resonance shifts arising from substitution of the C6 hydroxyl group of the aromatic ring by ether-linked propyl groups (3tosylpropyl-and 3-fluoropropyl-). Assignment of proton and carbon resonances of all natural Vitamin E analogues remains incomplete despite the application of multidimensional NMR (HSQC, COSY, ADEQUATE(1,1), C-H HMBC, and NOESY) techniques [28]. Resonances introduced by the 3-(F/TsO)-propyloxy substituent at C6, replacing hydrogen on the C6 hydroxy of γ-T-3, are reported in the experimental section as CH2-1, CH2-2 and CH2-3 ( Figure 5).     3 with γ-T-3. Standard work up of these compounds required attention to their instability under silica gel chromatography and sensitivity to UV light.
NMR analyses of the non-radioactive compounds focused on resonance shifts arising from substitution of the C6 hydroxyl group of the aromatic ring by ether-linked propyl groups (3tosylpropyl-and 3-fluoropropyl-). Assignment of proton and carbon resonances of all natural Vitamin E analogues remains incomplete despite the application of multidimensional NMR (HSQC, COSY, ADEQUATE(1,1), C-H HMBC, and NOESY) techniques [28]. Resonances introduced by the 3-(F/TsO)-propyloxy substituent at C6, replacing hydrogen on the C6 hydroxy of γ-T-3, are reported in the experimental section as CH2-1, CH2-2 and CH2-3 ( Figure 5).    The preparation of radiofluorination precursors in which alkyl tosylates are appended to the core molecule is a technique frequently used to radiofluorinate aryl cores. In such reactions, propyl tosylates may be preferred for reasons of stability and radiochemical yield. It has been reported that in the radiofluorination of the propyl tosylate or ethyl tosylate carbocyclic nucleoside precursors, the respective radiochemical yields were 10% and 4-5%, respectively [29]. Another approach to radiofluorinate Vitamin E analogues at C6-O would be to first prepare [ 18 [31,32]. At 30-150 GBq/µmol, an injected radioactivity dose of 14.8 MBq would represent approximately 0.48 to 0.1 nmol of F-γ-T-3 (end of synthesis), an amount not likely to modulate most Vitamin E-related processes and in line with the low Vitamin E concentrations found in tissues. Pharmacokinetic parameters of Vitamin E analogues, in humans at least, do not appear to be dose dependent over a large dose range [33], keeping in mind that studies using non-labeled analogues are normally undertaken at comparatively high doses.

PET Imaging
PET imaging was limited to a demonstration study using two tumor-bearing mice. One hour after injection, there was no thoracic retention of the [ 18 F]F-γ-T-3 tracer, whereas retention in the upper abdominal region including liver, gall bladder and upper intestine was prominent at 1 h and persisted through to 3 h post-tracer injection. Although formulation of highly lipophilic drugs for i.v. administration can be difficult, this challenge is markedly reduced for high SA radiopharmaceuticals. None the less, radiopharmaceutical preparation and formulation problems, and radiopharmaceutical administration techniques, are among the several important factors affecting the biodistribution of radiopharmaceuticals [34]. PPG is an inexpensive, non-toxic, and well-tolerated cosolvent [34] that has been used in a range of formulations as a co-solvent [35].
Both radiotracers depict tumor and blood pool, but [ 18 F]FDG concentrations visually appear to be higher in the tumor and in blood. The elevated uptake of radioactivity in the thorax and head following [ 18 F]FDG injection reflects high glucose utilization by heart and brain. In contrast, high uptake of radioactivity in the abdominal region following [ 18 F]F-γ-T-3 injection (Figure 4) is likely to be indicative of extensive first-pass accumulation of this large (MW 624) lipophilic molecule.
The cancer cell line ATCC ® HTB-26™, aka MDA-MB-231, a human triple-negative epithelial mammary gland/breast adenocarcinoma initially derived from a metastatic site pleural effusion, was selected for this work because it serves as a versatile tool for many cancer-related research studies. Examples include comparative biodistributions of diverse radiotracers and studies of cell response to various agents [39,40]. Uptake of [ 18 F]F-γ-T-3 by the MDA-MB-231 tumor was sufficient to visualize the tumor, but quantification of the image radioactivity disposition was not attempted. Rapid clearance of radioactivity from the blood stream was visually apparent, and with concurrent hepatic uptake, would be consistent with first-pass clearance and in line with previously reported extensive reticuloendothelial (RES) clearance and short plasma half life (~2.5-3 h) following i.v. intravenous delivery to rats [15]. However, uptake by tumor is visible in the PET image, indicating that an appreciable fraction of the injected radioactivity was not subject to first-pass clearance from blood even given the very small i.v. dose of [ 18 F]F-γ-T-3.
Same-animal imaging with [ 18 F]FDG 24 h after [ 18 F]F-γ-T-3 injection produced expected FDG-biodistribution images. Differences in biodistributions of the two tracers included the high head, thorax and tumor uptake of [ 18 F]FDG with lingering blood pool radioactivity. This biodistribution was in stark contrast to the virtual absence of uptake in head and thorax and the low levels of radioactivity in tumor and blood after [ 18 F]F-γ-T-3 injection and apparent rapid clearance from blood. Emulsification of the highly lipophilic [ 18 F]F-γ-T-3 in EtOH:PPG:saline (1:1:2) upon injection into the blood stream and consequent skewing of the biodistribution of [ 18 F]F-γ-T-3 was considered unlikely because the chemical load of [ 18 F]F-γ-T-3 was estimated to be less than 1 nmol. Importantly, the PET image provides no evidence of extravasation or a depot effect at or downstream from the injection site, and the animals displayed no acute toxicity/side effects over the 24 h period between the [ 18

F]F-γ-T-3 and [ 18 F]FDG injections. These observations confirm that dose formulation and injection technique were not contributing factors to the observed biodistribution of either [ 18 F]F-γ-T-3 or [ 18 F]FDG.
The PET image (Figure 4) represents the first 'big picture' of the whole-body distribution of any member of the Vitamin E family and specifically of the T-3 subfamily. This image confirms rapid blood clearance and strong hepato-biliary-intestinal accumulation of [ 18 F]F-γ-T-3 after i.v. injection, which although not quantitative, augments biodistribution data for δ-T-3 from an oral dosing study [41]. The traditional use of Vitamin E analogues (as mixtures or as pure substitutional isomers) in nutraceuticals and dietary supplements underlines the importance of natural source constituents as 'co-factors' in their bioavailability and adds another level of complexity to the interpretation of dynamic data such as metabolism and bioavailability [26,42].

General
Solvents for reactions were purified by successive passage through columns of alumina and copper under an argon atmosphere. Reagents were purchased from commercial sources and used without further purification unless noted otherwise. All reactions were carried out under a positive-pressure argon atmosphere and monitored by thin-layer chromatography (TLC) on Whatman MK6F silica gel micro TLC plates (25 µm thickness) or Silica Gel G-25 UV254 (0.25 mm) microplates using hexanes:EtOAc Powdered Cs 2 CO 3 (3.18 g, 9.77 mmol) was added to a mixture of 2 (1.52 g, 9.77 mmol) and γ-T-3 (1.50 g, 3.66 mmol) in DMF (15 mL). The resulting mixture was stirred at r.t. overnight, and then diluted with Et 2 O (100 mL) and washed with H 2 O (50 mL). The aqueous solution was extracted with Et 2 O (2 × 50 mL) and the resulting organic solution was washed with brine (50 mL), dried over anhydrous Na 2 SO 4 and filtered. The filtrate was concentrated and purified by silica gel column chromatography eluted with 0-5% EtOAc in hexane to afford F-γ-T-3 as a yellowish oil (1.05 g, 61%):  Powdered Cs 2 CO 3 (4.24 g, 13.0 mmol) was added to a mixture of γ-T-3 (2 g, 4.87 mmol) and 3 (5 g, 13.0 mmol) in DMF (20 mL). The resulting mixture was stirred at r.t. overnight. The mixture was then diluted with EtOAc (20 mL) and water (50 mL) and finally extracted with EtOAc (2 × 20 mL). The combined organic solution was washed with H 2 O (50 mL). The organic solution was dried over Na 2 SO 4 . After filtration, the filtrate was concentrated and purified by silica gel column chromatography eluted with DCM to afford TsO-γ-T-3 as a yellow oil in 48% yield (1.45 (3)(4)(5)(6)(7) was eluted to the reactor with a solution consisting of K 2 C 2 O 4 (2.5 mg), K 222 (10 mg) and K 2 CO 3 (10 mL of 5 mg/mL solution) in CH 3 CN and H 2 O (1 mL, 80:20). This solution was evaporated to dryness at 65 • C under helium flow and vacuum for 7 min followed by heating at 120 • C under vacuum for a further 7 min. Tosylate precursor (TsO-γ-T-3; 10 mg) in CH 3 CN was added to the anhydrous K[ 18 F]F/K 222 residue, followed by heating at 100 • C for 10 min. The radioactive reaction mixture was then diluted with mobile phase (EtOH-H 2 O, 1.5 mL) and transferred to the loop injection vial. The reaction vial was washed further with mobile phase (1.5 mL) and transferred to the loop injection vial. Preparative HPLC [(Nucleosil C18, 300 mm × 16 mm), mobile phase H 2 O:EtOH (90:10), flow rate 3 mL/min] afforded [ 18 F]F-γ-T-3 (0.5-1.0 GBq; 12% RCY; >99% RCP), in a total preparation time of 45 min. [ 18 F]F-γ-T-3 was formulated in EtOH:propylene glycol (PPG):saline(0.9%) (25:25:50) and used directly for small-animal imaging studies.

Animal Imaging by Positron Emission Tomography (PET)
Female Balb/c nude mice (age 9 weeks) (Animal Resources Centre, Western Australia) were inoculated subcutaneously on the flank with 4 × 10 6 MDA-MB-231 (Human Caucasian breast adenocarcinoma; ATCC ® HTB-26™) cells in phosphate-buffered saline (PBS)-Matrigel (1:1). Mice were weighed and tumors measured three times weekly using electronic calipers. Tumor volume (mm 3 ) was calculated as length (mm)/2 × width (mm) 2 . Nineteen days later, mice with tumors (~70-380 mm 3 ) were injected with [ 18 F]F-γ-T-3 (14.8 MBq) in formulation solvent (140 µL). At 1 h post-dosing, the mice were anesthetised using isoflurane and imaged on a Philips Mosaic small-animal PET scanner over 10 min. The animals were returned to their housing box to recover before being re-anesthetised and imaged again at 2 h post-dose. The recovery-re-anesthetization was repeated for 3 h post-dose imaging. The following day the mice were fasted for 3 h before being injected with 14.8 MBq [ 18 F]FDG and then imaged under isoflurane anaesthesia 90 min post-injection.
All procedures using animals were approved by the Peter MacCallum Cancer Centre Animal Experimentation Ethics Committee (Permit E475) and were performed in accordance with the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes (ISBN 1864965975; https://www. nhmrc.gov.au/about-us/publications/australian-code-care-and-use-animals-scientific-purposes).

Conclusions
Syntheses of the reference (F-γ-T-3) and labeling precursor (TsO-γ-T-3) compounds were facile, and radiofluorination of TsO-γ-T-3 provided [ 18 F]F-γ-T-3 in acceptable radiochemical yield and high radiochemical purity. Formulation of the highly lipophilic [ 18 F]F-γ-T-3 in EtOH:PPG:saline (1:1:2) was well tolerated by mice following i.v. injection. Initial PET images after administration of [ 18 F]F-γ-T-3 showed evidence of first-pass hepatobiliary uptake, with rapid blood clearance and no uptake in brain, heart or lungs. [ 18 F]F-γ-T-3 is considered to be a useful tracer in pharmacokinetic studies of F-γ-T-3, and a highly sensitive marker for dynamic studies of Vitamin E-containing products and formulations.