Oxidation of p-[125I]Iodobenzoic Acid and p-[211At]Astatobenzoic Acid Derivatives and Evaluation In Vivo

The alpha particle-emitting radionuclide astatine-211 (211At) is of interest for targeted radiotherapy; however, low in vivo stability of many 211At-labeled cancer-targeting molecules has limited its potential. As an alternative labeling method, we evaluated whether a specific type of astatinated aryl compound that has the At atom in a higher oxidation state might be stable to in vivo deastatination. In the research effort, para-iodobenzoic acid methyl ester and dPEG4-amino acid methyl ester derivatives were prepared as HPLC standards. The corresponding para-stannylbenzoic acid derivatives were also prepared and labeled with 125I and 211At. Oxidization of the [125I]iodo- and [211At]astato-benzamidyl-dPEG4-acid methyl ester derivatives provided materials for in vivo evaluation. A biodistribution was conducted in mice with coinjected oxidized 125I- and 211At-labeled compounds. The oxidized radioiodinated derivative was stable to in vivo deiodination, but unfortunately the oxidized [211At]astatinated benzamide derivative was found to be unstable under the conditions of isolation by radio-HPLC (post animal injection). Another biodistribution study in mice evaluated the tissue concentrations of coinjected [211At]NaAtO3 and [125I]NaIO3. Comparison of the tissue concentrations of the isolated material from the oxidized [211At]benzamide derivative with those of [211At]astatate indicated the species obtained after isolation was likely [211At]astatate.


Introduction
There are only a few alpha-emitting radionuclides that have decay properties suitable for use in Targeted Alpha Therapy [1][2][3]. One radionuclide of high interest is the alpha particle-emitting radionuclide astatine-211 ( 211 At). 211 At is particularly attractive as it is a halogen, and in principle can be directly incorporated into biological targeting agents. Importantly, 211 At is very efficient in killing cancer cells as it decays by a branched-chain process which effectively provides an alpha particle emission in 100% of the decays; furthermore, important is the fact that it does not have any alpha-emitting daughter nuclides after decay, which could cause toxicity in non-target tissues. While 211 At is attractive for the development of Targeted Alpha Therapy agents, issues with in vivo stability of the label have been an impediment to the development of effective therapeutic agents [4,5].
Early studies demonstrated that 211 At bonded to non-activated or deactivated aromatic ring pendant groups provided monoclonal antibody conjugates that are relatively stable towards in vivo deastatination [6][7][8]; however, it was found that this was not the case for more rapidly metabolized molecules. Thus, a number of investigators have evaluated chemical methods of stabilizing 211 At on biomolecules to in vivo deastatination [4,9], but in most cases where there is a carbon-211 At bond some release of 211 At occurred in vivo. Interestingly, an example where the in vivo stability of the 211 At is relatively high is in meta-[ 211 At]astatobenzylguanidine ([ 211 At]MABG) [10][11][12]. The reason for the higher in vivo stability of [ 211 At]MABG is not known. Recently, a neopentyl glycol approach to labeling has shown promise for labeling with 211 At [9], but this method may not be applicable to all biomolecules to be studied. We have investigated alternate 211 At-bonding agents, such as anionic aryl-borates, which have been shown to provide an astatine-boron bond that is highly stable in vivo [13]; however, the conjugation of an anionic boron cage moiety to biomolecules can potentially affect their in vivo targeting properties, so for some molecules it might be best if the astatine could be incorporated directly into the organic molecule through a carbon-astatine bond. Thus, we have been interested in evaluating alternate 211 At-carbon bonding approaches as a means of obtaining higher in vivo stability of the astatine bond.
In this investigation, we evaluated the effect on stability brought about by oxidizing an astatine atom that was bonded to a deactivated aryl-carbon atom. Since there are no stable isotopes of astatine, we used information on the stability of non-radioactive aryl-iodine molecules that have a higher oxidation state of the iodine atom [14,15]; it was noted that many aryliodine molecules with the iodine atom in a higher oxidation state are very effective organic oxidizing agents and thus would not be useful for in vivo applications; however, an example of a highly oxidized aryliodine, iodoxybenzoic acid, that was found to be quite stable in vitro seemed to warrant investigation of the in vivo stability of the corresponding radioiodinated and astatinated compounds [16]. Thus, we synthesized some iodo-, radioiodo-and astatobenzoic acid derivatives, then oxidized the bonded (radio)halogen, isolated the products from that oxidation process, and evaluated biodistributions in mice of the isolated products in mice. The results of that investigation are described herein.

Syntheses of Compounds
The compounds synthesized for this investigation are shown in Figure 1. A literature report indicated that p-iodoxybenzoic acid, 2, was stable in vitro [16], so that was our first synthetic target. An early report of the preparation of p-iodoxybenzoic acid involved the preparation of p-iodosobenzoic acid (also named as p-iodosybenzoic acid), which disproportionates to p-iodobenzoic acid, 1, and p-iodoxybenzoic acid, 2 [16]; however, that approach was not viewed as efficient, so a more direct synthesis was sought; it had been reported that iodoxybenzene derivatives could be prepared using refluxing aqueous sodium iodate, but under the conditions used p-iodosylbenzoic acid was obtained rather than 2 [17]. Another approach to the synthesis of iodoxybenzene derivatives from iodobenzene derivatives was the use of oxone as the oxidant [18,19]. Oxidation of benzoic acid methyl ester 3 with mCPBA under microwave heating at 70 °C for 20 min provided para-iodoxybenzoic acid methyl ester 6 to use as an HPLC standard. Compound 6 had the correct exact mass by HRMS analysis and the NMR spectra showed consistent changes in the aryl proton shifts for converting 5 [δ 7.75 (d), 7  In this study, we investigated the use of K 2 S 2 O 8 , NaIO 4 and m-chloroperbenzoic acid (mCPBA) as oxidizing agents in solvent mixtures of MeOH/H 2 O, DMF/H 2 O or DMSO/H 2 O, heating from 60 • C to 150 • C in the microwave. The studies began with oxidation of (non-radioactive) para-iodobenzoic acid, 1 (Figure 1), to prepare the iodoxy-benzoic acid. Initial oxidation studies to prepare the 2 by reaction of 1 with NaIO 4 or K 2 S 2 O 8 resulted in mixtures of products, none of which could be identified as the desired para-iodoxybenzoic acid, 2. Interestingly, it was found that HCl could not be used in the reaction as that resulted in the formation of para-chlorobenzoic acid under the reaction conditions studied; it appeared that the carboxylate interferes with the oxidation reaction, so the methyl ester, 3, was prepared and used in subsequent studies. Further studies demonstrated that the use of mCPBA provided the highest yields under the conditions studied, so those reactions are described.
Oxidation of benzoic acid methyl ester 3 with mCPBA under microwave heating at 70 • C for 20 min provided para-iodoxybenzoic acid methyl ester 6 to use as an HPLC standard. Compound 6 had the correct exact mass by HRMS analysis and the NMR spectra showed consistent changes in the aryl proton shifts for converting 5 [δ 7.75 (d), 7 Figure 2. Oxidation of benzoic acid methyl ester 3 with mCPBA under microwave heating at 70 °C for 20 min provided para-iodoxybenzoic acid methyl ester 6 to use as an HPLC standard. Compound 6 had the correct exact mass by HRMS analysis and the NMR spectra showed consistent changes in the aryl proton shifts for converting 5 [δ 7.75 (d), 7 Figure 2. It was apparent that the oxidized (radio)halobenzoic acid derivatives needed to be more soluble in aqueous solutions to be of value in the development of new radiopharmaceuticals; this fact led to the synthesis of amino-dPEG4-carboxylic acid, 16, adducts of para-tri-n-butylstannylbenzoic acid and para-iodobenzoic acid, to form para-substituted benzamidyl-dPEG4-carboxylate methyl esters 9 and 10 respectively, as shown in Figure 3. It was apparent that the oxidized (radio)halobenzoic acid derivatives needed to be more soluble in aqueous solutions to be of value in the development of new radiopharmaceuticals; this fact led to the synthesis of amino-dPEG 4 -carboxylic acid, 16, adducts of para-tri-n-butylstannylbenzoic acid and para-iodobenzoic acid, to form para-substituted benzamidyl-dPEG 4 -carboxylate methyl esters 9 and 10 respectively, as shown in Figure 3.
The synthesis of the arylstannane derivative 9, which can be used for radiolabeling, began by reaction of para-iodobenzoic acid, 1, with trifluoroacetic acid tetrafluorophenyl ester (TFA-OTFP) to prepare the para-iodobenzoic acid tetrafluorophenyl (TFP) ester 14 [20]. In the next step, the para-tri-n-butylstannylbenzoic acid TFP ester 15 was prepared using a palladium catalyzed stannylation reaction [20]. Preparation of the amino-dPEG 4 -acid adduct, 17a, was accomplished by reaction of 15 with amino-dPEG 4 -acid, 16, and Et 3 N in DMF at room temperature. The para-tri-n-butylstannylbenzamidyl-dPEG 4 -carboxylic acid TFP ester, 18a, was prepared by reaction of TFA-OTFP with 17a and Et 3 N in CHCl 3 at room temperature. Conversion of 18a to the methyl ester 9 was accomplished by reaction in anhydrous methanol and dimethylaminopyridine (DMAP).
Non-radioactive para-iodobenzamidyl-dPEG 4 -carboxylic acid methyl ester, 10 was prepared as an HPLC standard. In that synthesis 17b was prepared by reaction of 14 with 16 in DMF containing Et 3 N at room temperature. Compound 17b was subsequently esterified using TFA-OTFP to form the TFP ester 18b. Following that 18b was converted to the methyl ester 10 by reaction with anhydrous methanol and triethylamine Et 3 N. The synthesis of the arylstannane derivative 9, which can be used for radiolabeling, began by reaction of para-iodobenzoic acid, 1, with trifluoroacetic acid tetrafluorophenyl ester (TFA-OTFP) to prepare the para-iodobenzoic acid tetrafluorophenyl (TFP) ester 14 [20]. In the next step, the para-tri-n-butylstannylbenzoic acid TFP ester 15 was prepared using a palladium catalyzed stannylation reaction [20]. Preparation of the amino-dPEG4acid adduct, 17a, was accomplished by reaction of 15 with amino-dPEG4-acid, 16, and Et3N in DMF at room temperature. The para-tri-n-butylstannylbenzamidyl-dPEG4-carboxylic acid TFP ester, 18a, was prepared by reaction of TFA-OTFP with 17a and Et3N in CHCl3 at room temperature. Conversion of 18a to the methyl ester 9 was accomplished by reaction in anhydrous methanol and dimethylaminopyridine (DMAP).
Non-radioactive para-iodobenzamidyl-dPEG4-carboxylic acid methyl ester, 10 was prepared as an HPLC standard. In that synthesis 17b was prepared by reaction of 14 with 16 in DMF containing Et3N at room temperature. Compound 17b was subsequently esterified using TFA-OTFP to form the TFP ester 18b. Following that 18b was converted to the methyl ester 10 by reaction with anhydrous methanol and triethylamine Et3N.

Radiohalogen Labeling and Oxidation of Radioiodine and Astatine Compounds
Reactions of tri-n-butylstannylbenzoic acid methyl ester, 4 (

Radiohalogen Labeling and Oxidation of Radioiodine and Astatine Compounds
Reactions of tri-n-butylstannylbenzoic acid methyl ester, 4 (

Biodistribution Studies
Two biodistribution studies were conducted. One biodistribution study evaluated the tissue distribution in athymic mice of isolated products from the reactions to produce [ 125 I]11 and [ 211 At]13. Isolation of [ 125 I]11 from the reaction mixture and reinjection after evaporation of solvents and dissolution in PBS showed that it did not change during that process ( Figure 4E). In contrast, the product obtained by isolation of the radiochromatogram peak at a retention time of 5.8 min ( Figure 4D), believed to be [ 211 At]13, had a major peak at 4.3 min and a minor peak at 3.4 min ( Figure 4F Table S1). The fact that the concentrations (%ID/g) of 125 I in stomach and neck tissues are low is an indicator that iodoxyaryl compounds are stable to in vivo deiodination. In contrast, the tissue concentrations obtained for the isolated (decomposed) [ 211 At]13 are similar to free 211 At [13], suggesting that an 211 At species was injected, most likely a mixture of [ 211 At]astatide and [ 211 At]astatate based on the oxidative conditions used prior to isolation and the radio-HPLC retention times observed ( Figure 4F). isolated solution containing [ 211 At]13 was conducted after coinjection of [ 125 I]11 and [ 211 At]13 into mice due to timing required to conduct the biodistribution studies. The tissue concentrations of 125 I and 211 At at 1, 4 and 24 h post injection (p.i.) are plotted in Figure  5 (tabulation of data in Supplemental Table S1). The fact that the concentrations (%ID/g) of 125 I in stomach and neck tissues are low is an indicator that iodoxyaryl compounds are stable to in vivo deiodination. In contrast, the tissue concentrations obtained for the isolated (decomposed) [ 211 At]13 are similar to free 211 At [13], suggesting that an 211 At species was injected, most likely a mixture of [ 211 At]astatide and [ 211 At]astatate based on the oxidative conditions used prior to isolation and the radio-HPLC retention times observed ( Figure 4F).   Figure 6A,B) the reaction solutions were isolated from the HPLC effluent to remove oxidant and solvents were evaporated with heating. The residues remaining were dissolved in PBS and re-examined by radio-HPLC. Radiochromatograms of the isolated products are shown in Figure Figure 6E,F), which allowed them to be used directly in   (Figure 6A,B) the reaction solutions were isolated from the HPLC effluent to remove oxidant and solvents were evaporated with heating. The residues remaining were dissolved in PBS and reexamined by radio-HPLC. Radiochromatograms of the isolated products are shown in Figure   in the lung (7.1 %ID/g @ 1 h p.i., 5.8 %ID/g @ 4 h p.i.) and spleen (6.8 %ID/g @ 1 h p.i., 6.4 %ID/g @ 4 h p.i.) is higher than that in the liver, intestine and muscle (<1.6 %ID/g @ 1 and 4 h p.i.). Essentially, all [ 125 I]IO 3 activity cleared out from the body by 24 h p.i., but a small percentage of injected 211 At activity remained in the neck (4.3 %ID/g), stomach (2.0 %ID/g) and lung (1.0 %ID/g) at 24 h p.i. tissues, whereas, the concentration of [ 211 At]AtO3 -in the lung (7.1 %ID/g @ 1 h p.i., 5.8 %ID/g @ 4 h p.i.) and spleen (6.8 %ID/g @ 1 h p.i., 6.4 %ID/g @ 4 h p.i.) is higher than that in the liver, intestine and muscle (<1.6 %ID/g @ 1 and 4 h p.i.). Essentially, all [ 125 I]IO3activity cleared out from the body by 24 h p.i., but a small percentage of injected 211 At activity remained in the neck (4.3 %ID/g), stomach (2.0 %ID/g) and lung (1.0 %ID/g) at 24 h p.i.    Table S2 in Supplementary Materials for a tabulation of %ID/g data.

Discussion
The primary goal of this study was to evaluate whether a specific type of astatinated aryl compound that has the At atom in a higher oxidation state might be stable to in vivo deastatination. We were aware that iodoaromatic compounds which have the iodine atom in a higher oxidation state were generally unstable due to their propensity to oxidize other organic molecules [15,21], but a specific example of a para-iodoxybenzoic acid was found to be quite stable in vitro [16]. Thus, we chose to use para-[ 125 I]iodoxylbenzoate and para-[ 211 At]astatoxybenzoate derivatives as model compounds to evaluate their in vivo stability; these model compounds were attractive as this approach could potentially be readily adapted to many astatinated and radioiodinated small molecules.
Initial studies were conducted to prepare p-(radio)haloxybenzoic acids and their methyl esters as simple model compounds. While the benzoic acid methyl esters could be prepared, radiolabeled and oxidized as desired, those compounds were found to be quite insoluble in all solvents tried; this led to conjugation of a linker molecule, amino-dPEG4carboxylic acid, 16, with para-iodobenzoic acid and para-tri-n-butylstannylbenzoic acid esters to increase the aqueous solubility of the compounds to be tested. Preparation of the stannylbenzoate derivatives was conducted as intermediates for the radiohalogenation reactions [22][23][24]. The para-(radio)halobenzoic acid derivatives containing the amino-dPEG4 acid linker had improved solubility and the corresponding radioiodinated and astatinated derivatives [ 125 I]10 and [ 211 At]12, could be readily prepared from the stannyl intermediate 9. Subsequent oxidation with mCPBA provided the para-[ 125 I]iodoxybenzamide derivative [ 125 I]11 based on its radio-HPLC retention time, which corresponded with the NMR and MS characterized non-radioactive standard 10. Contrary to that, the radio-HPLC retention time of the astatinated molecule [ 211 At]13 was significantly different from that of the radioiodinated product, making it more difficult to be certain of the identity of the product; however, the change in retention time for the astatinated compounds was consistent with those of the radioiodinated compounds and consistent with other compounds studied in our laboratory. Thus, it seems very likely that the astatinated compounds in Figure 3 have the structures shown.
The issue addressed in this study was stability of astatinated and radioiodinated aromatic compounds when injected into living animals. Nature provides deiodinase en-  Table S2 in Supplementary Materials for a tabulation of %ID/g data.

Discussion
The primary goal of this study was to evaluate whether a specific type of astatinated aryl compound that has the At atom in a higher oxidation state might be stable to in vivo deastatination. We were aware that iodoaromatic compounds which have the iodine atom in a higher oxidation state were generally unstable due to their propensity to oxidize other organic molecules [15,21], but a specific example of a para-iodoxybenzoic acid was found to be quite stable in vitro [16]. Thus, we chose to use para-[ 125 I]iodoxylbenzoate and para-[ 211 At]astatoxybenzoate derivatives as model compounds to evaluate their in vivo stability; these model compounds were attractive as this approach could potentially be readily adapted to many astatinated and radioiodinated small molecules.
Initial studies were conducted to prepare p-(radio)haloxybenzoic acids and their methyl esters as simple model compounds. While the benzoic acid methyl esters could be prepared, radiolabeled and oxidized as desired, those compounds were found to be quite insoluble in all solvents tried; this led to conjugation of a linker molecule, amino-dPEG 4 -carboxylic acid, 16, with para-iodobenzoic acid and para-tri-n-butylstannylbenzoic acid esters to increase the aqueous solubility of the compounds to be tested. Preparation of the stannylbenzoate derivatives was conducted as intermediates for the radiohalogenation reactions [22][23][24]. The para-(radio)halobenzoic acid derivatives containing the amino-dPEG 4 acid linker had improved solubility and the corresponding radioiodinated and astatinated derivatives [ 125 I]10 and [ 211 At]12, could be readily prepared from the stannyl intermediate 9. Subsequent oxidation with mCPBA provided the para-[ 125 I]iodoxybenzamide derivative [ 125 I]11 based on its radio-HPLC retention time, which corresponded with the NMR and MS characterized non-radioactive standard 10. Contrary to that, the radio-HPLC retention time of the astatinated molecule [ 211 At]13 was significantly different from that of the radioiodinated product, making it more difficult to be certain of the identity of the product; however, the change in retention time for the astatinated compounds was consistent with those of the radioiodinated compounds and consistent with other compounds studied in our laboratory. Thus, it seems very likely that the astatinated compounds in Figure 3 have the structures shown.
The issue addressed in this study was stability of astatinated and radioiodinated aromatic compounds when injected into living animals. Nature provides deiodinase enzymes to recycle iodine for thyroid hormones [25] and also provides enzymes that dehalogenate compounds to detoxify them [26][27][28]. There are several different mechanisms that are used by enzymes to dehalogenate compounds, including hydrolytic substitution reactions and reductive or oxidative elimination reactions [26]. The family of deiodinase enzymes removes specific iodine atoms from thyroid hormones through a reductive elimination of iodine from the aromatic ring [29]. While the deiodinase enzymes may play a role in deiodination of some radioiodinated phenolic compounds [30], radioiodination of "deactivated" aromatic compounds (e.g., benzoates) has proven to stabilize them from in vivo deiodination; these non-activated aromatic ring compounds have been used to help stabilize carbon-astatine bonds [31,32], but many rapidly metabolized 211 At-labeled compounds have been found to be susceptible to deastatination [4]; it has been shown that oxidative dehalogenation can induce loss of the 211 At atom from astatobenzoates [33]. In the same study, quantum mechanical calculations indicated that astatobenzoates could undergo oxidative deastatination some 6 million times faster at 37 • C than the corresponding iodobenzoates.
The concept that oxidative deastatination as a mechanism for releasing 211 At in vivo led to a hypothesis that greater stability might be obtained if the astatine atom in arylastatine is stable in a higher oxidation state and cannot be further oxidized by the enzymatic oxidants in cellular compartments (e.g., lysosomes). In this study we found that oxidation of the At atom provided a new compound (presumably) the arylastatoxy compound, but that compound was unstable to the conditions used for isolation. While this result might seem to warrant further studies to determine if that compound could be isolated without decomposition, it was thought that the low stability of the compound would ultimately release the At in vivo, so no further studies were conducted.

General
Caution must be taken when working with aryliodine compounds where the iodine is oxidized to the + 5 oxidation state as some derivatives have been reported as explosive under excessive heating or impact [34]; however, it is thought that the explosive properties may have been from bromate impurities when potassium bromate in sulfuric acid was used to prepare the aryliodine (V) compound.
All reagents used were obtained from commercial sources as analytical grade or better and were used without further purification. meta-Chloroperbenzoic acid, chloramine-T, bis-tributyltin, and tetrakis (triphenylphosphine)palladium [Pd(Ph 3 P) 4 ] were obtained from Sigma-Aldrich (St. Louis, MO, USA). p-Iodobenzoic acid, 1, and its methyl ester 3 were obtained from commercial sources (i.e., Sigma-Aldrich). Amino-dPEG 4 -acid, 16, was provided as a gift from Quanta BioDesign (Plain City, OH, USA). 2,3,5,6-Tetrafluorophenyl trifluoroacetate (TFA-OTFP) was prepared as previously described [35]. Microwave reactions were conducted in a Biotage Initiator 2.0 using 0.5-2 mL vials. Samples were taken from the vial for HPLC analyses by puncturing the vial septum with a needle and withdrawing 10 µL. Reaction progress could be evaluated at multiple time points using this method.

Radioactive Materials
All radioactive materials were used under a UW Radiation Use Authorization and handled according to approved Radiation Safety protocols at the University of Washington.  [36] and was isolated from the irradiated Bi targets using a "wet chemistry" method as described previously [37]. Quantification of 125 I was accomplished on a Capintec CRC-15R Radioisotope Calibrator using the manufacturer's recommended calibration number. Quantification of 211 At was accomplished on a Capintec CRC-15R and a Capintec 55tR using cross-calibration against a High Purify Germanium detector as described previously [37].

Chromatography Equipment and Conditions
Non-radioactive compounds were purified using a Biotage Flash Purification System (Charlottesville, VA, USA). The purifications were conducted on a reversed-phase C18 FLASH 25 + M column. The column was eluted by a gradient mixture composed of MeOH and 0.1% aqueous acetic acid, pH 3.25. Starting with 25% MeOH, the initial solvent mixture was held for 2 min, increased to 100% MeOH over the next 12 min, and then held at 100% MeOH for 8 min. Fractions were collected based on UV detection at 254 nm. Identification of fractions containing the desired product was achieved using analytical HPLC. The fractions containing pure compound were combined, the solvent was evaporated, and the product was isolated to provide the yields listed.
HPLC analyses of non-radioactive compounds were performed using a system composed of a Hewlett-Packard quaternary 1050 gradient pump, a variable wavelength UW detector (254 nm) and an evaporative light scattering detector (ELSD 2000, Alltech, Deerfield, IL, USA). Radio-HPLC analysis of product mixtures from radiohalogenation reactions were analyzed using a system composed of two Beckman model 110 B pumps, a Beckman 420 controller, a Beckman model 153 UV detector (254 nm), and a Beckman model 170 radiation detector. Analysis of the HPLC data was conducted on Hewlett-Packard HPLC ChemStation software. Reversed-phase HPLC analyses were carried out on an Alltech Altima C18 column (5 µm, 250 × 4.6 mm) using a gradient solvent system at a flow rate of 1 mL/min. The gradient mixture was composed of MeOH and 0.1% aqueous HOAc (pH 3.25). The gradient started with 40% MeOH for 2 min and increased linearly to 100% MeOH over 10 min, followed by elution with 100% MeOH for 8 min. The retention time (t R ) of each compound is provided with the experimental procedure. Radio-HPLC analyses of 125 I, 211 At, and stable iodine ions were conducted by using a JMC J'Sphere ODS M80 column (5 µm, 250 × 4.6 mm) eluting isocratically at 1 mL/min with a 1:1 mixture of acetonitrile (ACN):50 mM aqueous tetrabutylammonium dihydrogen phosphate (pH 4.5).

Spectral Analyses
All new non-radioactive compounds were characterized using 1 H NMR and high resolution mass spectral (HRMS) analyses. The spectral data obtained is provided with the synthesis procedures. 1 H NMR spectra were collected on a Bruker AV-500 (500 MHz for 1 H) spectrometer, referencing to tetramethylsilane as an internal standard (δ = 0.0 ppm). HRMS data were obtained on a Bruker Apex-Qe FT/ICR instrument. The elemental composition calculations were conducted using the Bruker software (DataAnalysis 3.4).

Para-Tri-n-Butylstannylbenzoic Acid TFP Ester, 15
Para-Tri-n-butylstannylbenzoic acid TFP ester, 15, was prepared in 83% yield by reaction of 14 with bis-tributyltin, (Bu 3 Sn) 2 , and tetrakis(triphenylphosphine)palladium [Pd(Ph 3 P) 4 ] as previously described [13]. 4.5.9. Para-Tri-n-Butylstannylbenzamidyl-dPEG 4 -Carboxylic Acid, 17a Intermediate 17a was synthesized by adding H 2 N-dPEG 4 -CO 2 H, 16 (0.522 g, 1.967 mmol) to a solution of 15 (1 g, 1.788 mmol), Et 3 N (0.498 mL, 3.58 mmol), CH 2 Cl 2 (20 mL), and anhydrous DMF (5 mL). The mixture was stirred at room temperature for 4 h. After that period volatile materials were removed by rotary evaporator under vacuum, then the crude product was purified by silica gel column (2.5 cm × 25 cm) eluting with a gradient solution from 100% EtOAc to 50:50 MeOH:EtOAc to yield 61% of 17a as a colorless tacky solid. HRMS (ES + ) calcd for C 30  To a solution containing 1 g (0.98 mmol) of 17a and 0.206 mL (1.48 mmol) Et 3 N in 15 mL of anhydrous CHCl 3 was added 0.204 mL (1.185 mmol) of TFA-OTFP. The solution was stirred at room temperature for 30 min to prepare the TFP ester 18a. The solution containing 18a was used in the next step without purification. 4.5.12. Para-Tri-n-Butylstannylbenzamidyl-dPEG 4 -Carboxylic Acid Methyl Ester, 9 To the crude solution of 18a was added 1 mL of anhydrous methanol and 0.121 g (0.987 mmol) of DMAP. The resultant solution was stirred at room temperature for another 2 h; then, the volatile materials were removed using a rotary evaporator under vacuum. The crude 9 was purified by silica gel column (2.5 cm × 25 cm) eluting with a gradient beginning with 1:1 ethyl EtOAc:hexanes to 100% EtOAc, then to 10% MeOH/EtOAc and finally to 20% MeOH:EtOAc. Compound 9 was isolated in 86% yield as a colorless oil. HRMS (ES + ) calcd for C 31  To a solution containing 1 g (2.08 mmol) of 17b and 0.434 mL (3.12 mmol) Et 3 N in 15 mL anhydrous CHCl 3 was added 0.43 mL (2.49 mmol) TFA-OTFP. That solution was stirred at room temperature for 5 min then analyzed by HPLC. The HPLC confirmed that the TFP ester, 18b, was formed quantitatively. The solution containing 18b was used in the next step without purification. 4.5.14. Para-Iodobenzamidyl-dPEG 4 -Carboxyic Acid Methyl Ester, 10 To the crude solution of 18b was added 2 mL of anhydrous MeOH, and the resultant solution was stirred at room temperature for another hour. Volatile materials were removed by rotary evaporation under vacuum. The crude 10 was purified by silica gel column (1.5 cm × 25 cm) eluting with a gradient solution from 40% EtOAc/hexanes to 10% MeOH/EtOAc. The methyl ester 10 was isolated in 92% yield as a colorless tacky solid. 1 3 mCi), was added 20 µL of a 1 mg/mL aqueous solution of ChT. The mixture sat at room temperature for 5 min before conducting radio-HPLC analysis for the preparation of [ 211 At]12. Radio-HPLC indicated that a 100% conversion was obtained. Radio-HPLC t R = 12.5 min.

Para-Iodoxybenzamidyl-dPEG 4 -Carboxylic Acid Methyl Ester, 11
To a 10 mL MeOH solution containing 100 mg (0.196 mmol) of 10 was added 149 mg (0.432 mmol; 50% purity) mCPBA. The resultant solution was heated at 70 • C in a microwave reactor for 20 min. The conversion of 10 to 11 was followed by HPLC. The reaction mixture was evaporated to dryness using rotary evaporator under vacuum. The crude 11 was dissolved in 1: To the reaction mixture containing [ 211 At]12 (Section 4.5.14. above) was added 100 µL of a 0.05 M solution of mCPBA in NaOH. The solution was stirred and heated at 50 • C for 20 min under microwave heating. After that time the reaction solution was analyzed by radio-HPLC to give a 74% radiochemical yield of [ 211 At]13 by radio-HPLC analysis (t R = 5.8 min).

Preparation of [ 125 I]NaIO 3
To a 600 µL solution containing 0.01 M NaIO 4 in 0.1 N H 2 SO 4 was added a 2 µL solution containing 41 kBq (1.27 µCi) of [ 125 I]NaI. The reaction mixture sat at room temperature for 3 days to achieve complete conversion of [ 125 I]NaI to [ 125 I]NaIO 3 . The conversion was followed by radio-HPLC where non-radioactive NaIO 3 was used as a retention time reference standard [38].
(Animal study) To a 500 µL solution containing 0.025 M NaIO 4 in 0.1 N H 2 SO 4 was added a 1 µL solution containing (1.22 mCi) of [ 125 I]NaI. The solution was stirred and heated to 120 • C for 10 min under microwave heating. The conversion was followed by radio-HPLC where non-radioactive NaIO 3 was used as a retention time reference standard. The [ 125 I]NaIO 3 was purified by collecting the radio-HPLC peak. A 62% radiochemical yield was obtained. A portion of the isolated [ 125 I]NaIO 3 was used in the biodistribution study after the removal of HPLC solvents and dissolution in PBS.

Biodistribution Studies
All animal studies were approved by the Institutional Animal Care and Use Committee (IACUC) protocol 2485-05 at the University of Washington and were conducted in accordance with the NIH guidelines. Athymic mice (Crl:NU-Foxn1nu) were obtained from Charles River Laboratories (Hollister, CA, USA). In each experiment, the radioactive product was isolated from the radio-HPLC effluent, the organic solvent was removed by rotary evaporation, then the residue was dissolved in phosphate buffered saline (PBS) to prepare the doses for injection. The isolated 125 I-labeled product was coinjected with the corresponding isolated 211 At-labeled product to minimize the number of mice required.
In one biodistribution study, products from the reactions to make [ 125 I]11 and [ 211 At]13 were coinjected. A 100-µL volume of injectate containing~3 µCi (~81 kBq) of each radionuclide was administered to each of 15 mice via tail vein injection. In a second biodistribution study~5 µCi (~135 kBq) of [ 125 I]NaIO 3 and [ 211 At]NaAtO 3 were coinjected in a 100-µL dose to each mouse. Groups of 5 mice were euthanized at 1, 4 and 24 h post injection (p.i.) under ketamine/xylazine (130 mg/8.8 mg/kg) anesthesia by cervical dislocation and selected tissues were excised. Cardiac blood samples were collected under anesthesia immediately before euthanasia. Total blood weight was estimated to be 6% of the body weight. The tissues excised includee muscle, lung, kidney, spleen, liver, intestine, neck (thyroid) and stomach. Excised tissues were blotted free of blood and weighed prior to measuring the radioactivity in them.
Quantification of percent injected dose per gram (%ID/g) in the tissues was calculated by averaging the count of 5 × 1 µL samples of the injectate as standards for a total of 125 I and 211 At counts administered. The total quantity injected (µL) was obtained by weighing the syringe before and after injection. Radioactivity counting was conducted in a gamma counter and each sample was counted twice. The first count obtained total activity ( 211 At + 125 I). A second count of the samples was done at 4-5 days post the first count to allow all of the 211 At to decay. Thus, the second count represented only 125 I counts. The quantity of 211 At in each sample was obtained by subtracting the second counts from the first count. The 211 At counts were corrected for decay from the beginning of the gamma counting process.

Conclusions
This study did not provide a new approach to 211 At-labeling that would result in compounds stable to in vivo deastatination. The results showed that the oxidized arylastatine compound [ 211 At]13 is too unstable to be useful in the development of new radiopharmaceuticals; however, this study provided information that will be useful for future studies. The study was based on a hypothesis that arylastatine in higher oxidation states might be stable to in vivo deastatination and might provide a new method for labeling 211 At and radioiodine. Some [ 125 I]iodoxylaryl and [ 211 At]astatoxylaryl compounds were synthesized in relatively high radiochemical yields (>70%) using mCPBA oxidation at 50-70 • C. While the [ 125 I]iodoxylaryl compound was found to be quite stable to in vivo deiodination, the [ 211 At]astatoxyaryl was found to be unstable under the conditions used for isolation. The biodistribution studies revealed drastically different in vivo tissue distributions of 125 I-labeled [ 125 I]11 and the material isolated from the reaction to produced p-[ 211 At]astatoxylbenzoyl-dPEG-methyl ester, [ 211 At]13. Comparison of the radio-HPLC retention times and the tissue distributions of the isolated material from the oxidation reaction to produce [ 211 At]13 with those of [ 211 At]NaAtO 3 indicated that [ 211 At]13 likely decomposed to [ 211 At]AtO 3 during the isolation process. We continue to believe that pre-oxidation of astatination compounds may provide an alternate labeling approach for stabilizing 211 At-labeled aromatic compounds to in vivo deastatination; it is known that in vivo oxidation of arylastatine compounds can destabilize them towards deastatination, so pre-oxidized 211 At-labeled compounds, where the At atom is already in a higher oxidation state, might be resistant to further oxidation. For example, molecules containing At in a higher oxidation state can be formed by At-bonding with two aryl rings, forming a positive charged astatonium ion or an Ar 2 -At=O compound, which potentially can add stability to the oxidized At-complex. As in vivo deastatination is a major impediment to the development of 211 At-labeled radiopharmaceuticals, it seems that such pre-oxidized 211 At-labeled compounds warrant investigation.

Data Availability Statement:
The data presented in this study are available in the Supplementary Materials as Tables S1 and S2 (listed above).