Synthesis and Evaluation of a Fluorine-18 Radioligand for Imaging Huntingtin Aggregates by Positron Emission Tomographic Imaging

Mutations in the huntingtin gene (HTT) triggers aggregation of huntingtin protein (mHTT), which is the hallmark pathology of neurodegenerative Huntington’s disease (HD). Development of a high affinity 18F radiotracer would enable the study of Huntington’s disease pathology using a non-invasive imaging modality, positron emission tomography (PET) imaging. Herein, we report the first synthesis of fluorine-18 imaging agent, 6-(5-((5-(2,2-difluoro-2-(fluoro-18F)ethoxy)pyridin-2-yl)methoxy)benzo[d]oxazol-2-yl)-2-methylpyridazin-3(2H)-one ([18F]1), a radioligand for HD and its preclinical evaluation in vitro (autoradiography of post-mortem HD brains) and in vivo (rodent and non-human primate brain PET). [18F]1 was synthesized in a 4.1% RCY (decay corrected) and in an average molar activity of 16.5 ± 12.5 GBq/μmol (445 ± 339 Ci/mmol). [18F]1 penetrated the blood-brain barrier of both rodents and primates, and specific saturable binding in post-mortem brain slices was observed that correlated to mHTT aggregates identified by immunohistochemistry.


INTRODUCTION
Huntington's disease (HD) (Jenkins and Conneally, 1989;Bhattacharyya, 2016) is a neurodegenerative disease that progressively damages the motor, cognitive and psychiatric functions of patients (Dominguez and Munoz-Sanjuan, 2014;Cybulska et al., 2020). There is currently no approved therapy capable of delaying or slowing down HD onset or its progression (Estevez-Fraga et al., 2020). HD is primarily caused by the mutation in a single autosomal dominant gene leading to the formation of the mutant huntingtin gene (mHTT). mHTT with expanded CAG trinucleotide repeats (>36 CAG), encodes elongated polyQ repeats in the N-terminus, which in turn triggers aggregation of the huntingtin protein (Moldovean and Chiş, 2020). The pathology of HD is characterized by huntingtin protein aggregates. Although normal HTT is expressed throughout the body, the mHTT selectively targets brain cells and results in deteriorating medium spiny neurons of the striatum and cortex regions (DiFiglia, 2020). Design of suitable imaging agents for quantification of mHTT using positron emission tomography (PET) imaging will fill a critical gap in HD research by enabling non-invasive identification and tracking of huntingtin protein aggregates. PET imaging is a highly sensitive and noninvasive technique for quantifying biological targets within a living human and enabling their use as biomarkers of a disease. Thus a PET imaging agent for mHTT could be expected to have similar benefits to the amyloid, tau and α-synuclein PET imaging agents currently used for dementia imaging (Mathis et al., 2017), allowing diagnosis of HD, monitoring of disease progression, and evaluating patient response to HD modifying therapies targeting mHTT. The only example to date for imaging mHTT is with the 11 C-labeled agents [ 11 C]CHDI-180R and [ 11 C]CHDI-626 which exhibit high affinity (low nanomolar IC 50 ) toward mHTT and have selectivity over other protein aggregates like amyloid and tau (Dominguez et al., 2016;Liu et al., 2020Liu et al., , 2021Bertoglio et al., 2021). To the best of our knowledge, there are currently no 18 F-labeled PET imaging agents for imaging huntingtin protein aggregates described in the literature. Driven by the remarkable results achieved by [ 11 C]CHDI series in HD imaging, herein we describe the first synthesis of an 18 F analog ([ 18 F]1) to image mHTT in HD patients (Figure 1).
An 18 F-labeled analog is highly desirable because of the longer half-life of fluorine-18 (109.8 min) (Ross and Wester, 2011) compared to carbon-11 (20.4 min), which will enable its use in longer imaging studies to improve signal to background in the image, facilitate more sophisticated imaging studies with blocking agents, and eventually distribution to off-site PET imaging facilities. Looking at the structure of [ 11 C]CHDI-180R, we chose not to incorporate 18 F either as the 2-fluoroethyl group (Figure 2A) due to potential metabolic instability and generation of toxic side products (Pan, 2019), nor on the pyridine ring due to undesirable stereoelectric effects that could affect the imaging agents binding to huntingtin aggregates ( Figure 2B). So, we decided to incorporate 18 F on the CHDI scaffold as a trifluoromethyl group ( Figure 2C) to limit the potential for metabolic instability and to avoid negatively impacting target engagement.
The trifluoromethyl (CF 3 ) group is a common motif in small molecule-based drug scaffolds (Yale, 1959;Lien and Riss, 2014). Moreover, 2-[ 18 F]trifluoromethyl groups not only acts as a prosthetic group but also have improved metabolic stability when compared to 2-fluoromethyl groups, while still providing a similar straightforward means to incorporate fluorine-18 into drug scaffolds for PET imaging (Riss et al., 2012). Installation of trifluoromethyl groups on drug scaffolds serves as a common way of lead optimization in drug development to improve metabolic stability and overall pharmacokinetics. The strong electron withdrawing nature and higher stability of CF 3 groups mean they have also received widespread interest from PET radiochemists, resulting in the development of a variety of novel labeling methods (Chen et al., 2015;Taddei et al., 2021). The synthesis of CF 3 groups can be achieved by electrophilic fluorination (Chirakal et al., 1995;Teare et al., 2007), isotopic exchange reactions (Suehiro et al., 2011), nucleophilic fluorination (Kramer et al., 2020), transition metal catalyzed reactions (Huiban et al., 2013) and other methods (Chen et al., 2015). By far, the most common and preferred route is to carry out nucleophilic fluorination by [ 18 F]fluoride ion (Ross and Wester, 2011). The required difluorovinylfunctionalized labeling precursors (Rafique et al., 2018) can be predominantly assessed by numerous synthetic routes e.g., CH activation and elimination (Yang et al., 2020), Wittig reaction (Fuqua et al., 1965;Li et al., 2017), Julia-Kocienski-type reactions (Burton et al., 1996;Zhao et al., 2010;Zheng et al., 2013) and other methods (Ichikawa et al., 1991). Herein, we describe synthesis of desired [ 18 F]trifluoroethyl product [ 18 F]1 through a nucleophilic addition of H[ 18 F]F to a difluorovinylfunctionalized precursor. We also report the preclinical evaluation of [ 18 F]1 in vitro (binding affinity experiments, and saturation binding autoradiography experiments with post-mortem HD brain tissue samples), and in vivo (rodent and nonhuman primate (NHP) PET imaging).

General Considerations
Unless otherwise stated all the chemicals were purchased from commercial suppliers and used without purification. Automated flash chromatography was performed with a Biotage Isolera Prime system. High-performance liquid chromatography (HPLC) was performed using a Shimadzu LC-2010A HT. 1 H and 13 C NMR spectra were collected on a Varian 500 NMR (500 MHz for 1 H NMR and 125 MHz for 13 C NMR), in DMSO-d 6 or CDCl 3 unless otherwise indicated, δ in ppm rel. to tetramethylsilane (δ = 0), J in Hz. Mass spectra were measured on an Agilent Q-TOF HPLC-MS.

General Considerations
Unless otherwise stated, reagents and solvents were commercially available and used without further purification: sodium chloride, 0.9% USP, and sterile water for injection, USP, were purchased from Hospira; ethanol was purchased from American Regent; HPLC grade acetonitrile was purchased from Fisher Scientific. Other synthesis components were obtained as follows: sterile filters were obtained from Millipore; sterile product vials were purchased from Hollister-Stier; C18 Sep-Paks were purchased from Waters Corporation. C18 Sep-Paks were flushed with 10 mL of ethanol followed by 10 mL of water prior to use. Radio-HPLC was performed using a Shimadzu LC-2010A HT system equipped with a Bioscan B-FC-1000 radiation detector.
Fluorine-18 was produced via the 18 O(p,n) 18 F nuclear reaction using a GE PET Trace cyclotron equipped with a high yield fluorine-18 target at 55 µA to produce 74 GBq (2 Ci) of fluorine-18. F]Fluoride was then eluted into the reaction vessel using aqueous potassium carbonate (3.5 mg in 0.5 mL of water). A solution of kryptofix-2.2.2 (15 mg in 1.0 mL of acetonitrile) was then added to the reaction vessel and the [ 18 F]fluoride was dried by azeotropic evaporation of the water-acetonitrile mixture. Evaporation was achieved by heating the reaction vessel to 100 • C. The reactor was then cooled to 90 • C, and precursor was added with stirring for 3 min. Subsequently, the reaction mixture was cooled to 50 • C, followed by the addition of HPLC buffer (3.0 mL). The reaction was loaded onto a semipreparative column (Luna PFP, 250 × 10 mm, 35% Acetonitrile, 20 mM NH 4 OAc, 0.2% AcOH, flow rate = 4 mL/min). The product peak (∼82-86 min retention time, see Supplementary Figure 2 for a typical HPLC trace) was collected and diluted into a round-bottom flask containing 50 mL of water. The solution was then passed through a C-18 Sep-Pak to trap the product on the C-18 cartridge. The C18 cartridge was washed with 10 mL of sterile water. The product was eluted with 0.82 mL of ethanol/Tween 80 solution (0.66 mL of ethanol in 0.16 mL in Tween 80), followed by 9.5 mL of normal saline. The final formulation was passed through a 0.2 µm sterile filter into a sterile dose vial. The final product was obtained in 2446 ± 17.8 MBq (66.1 ± 17.7 mCi), 4.1% decay corrected yield,>98% RCP (see Supplementary Figure 3), pH = 5-5.5, n = 3 in 120 min from the end of bombardment. Identity was confirmed via co-injection with unlabeled reference standard (Supplementary Figure 4) and the product was stable for at least 150 min post-end-of-synthesis (Supplementary Figure 5).

General Considerations
All animal studies were performed in accordance with the standards set by the University of Michigan Institutional Animal Care and Use Committee (IACUC).

Rodent Small Animal Positron Emission Tomography Imaging Protocol
PET imaging studies were performed for [ 18 F]1 in Sprague-Dawley female rats (n = 3, animal weights = 283-411 g) using a Concorde MicroPET P4 gantry (Knoxville, TN) scanner. The animals were anesthetized (isoflurane), placed on a nose cone and positioned in the scanner for imaging. Anesthesia was maintained throughout the entire study. Following a transmission scan, the animals were injected intravenously (i.v.) via tail vain catheter as a bolus over 1 min with [ 18 F]1 (447-476 µCi in 140-150 µL of saline) and the head was imaged for 120 min. In each case, emission data were corrected for attenuation and scatter and reconstructed using the 3D maximum a priori (3D MAP) method. By using a summed image, regions of interest were defined for the whole brain on multiple planes. The volumetric regions of interest were then applied to a full dynamic data set to generate timeradioactivity curves (TACs).

Non-human Primate Positron Emission Tomography Imaging Protocol
Imaging studies were performed on Microsystem (Knoxville, TN) R4 microPET in two intact, mature female rhesus monkeys (n = 2, animal weights 9.6-10.2 kg). The animals were anesthetized in the home cage with telazol and transported to the PET facility. Subjects were intubated for mechanical ventilation, and anesthesia was continued with isoflurane. Anesthesia was maintained throughout the duration of the PET scan. A venous catheter was inserted into one hind limb and the monkey was placed on the PET gantry with its head secured to prevent motion artifacts. Ten minutes later, 3.5-5.3 mCi of [ 18 F]1 was administered in a bolus dose over 1 min, and the brain imaged for 120 min (5 × 2 min frames -4 × 5 min frames -9 × 10 min frames). Emission data were collected beginning with the injection, and continued for 120 min. Collection of vitals (HR, SPO 2 , EtCO 2 , and respiratory rate) was carried out during the whole scan. Data were corrected for attenuation and scatter and reconstructed using the three-dimensional-maximum a priori method (3D MAP algorithm). By using a summed image, regions of interest were defined for the whole brain and different brain regions on multiple planes. The volumetric regions of interest were then applied to a full dynamic data set to generate TACs.

Autoradiography
Frozen blocks (1×1 inch) of postmortem brain tissue samples from HD patients and a normal control (age range from 49 to 85) were used for the autoradiography binding studies ( Table 1). Tissue was obtained from the University of Michigan Alzheimer's Disease Center Brain Bank. Frozen blocks were sliced into 20 µm sections using a Hacker Instruments cryostat set to -15 • C. Tissue was thaw-mounted on the 1×3 inch polylysine-subbed glass slides. Sections used for autoradiography experiments were preconditioned for 5 min with phosphate buffer saline (PBS) pH 7.4 at 25 • C, and incubated in varying concentrations of [ 18 F]1 (0.05-5 nM) for 30 min at room temperature. Nonspecific binding (NSB) was determined by coincubation in the presence of 10 µM unlabeled 1F (dissolved in 1 mL methanol). Sections were washed twice (1 min each) in PBS followed by a distilled water rinse, all at 4 • C. Sections were dried under a stream of air and opposed to a phosphoimager screen for 10 min. Aliquots of the stock solutions were placed on a TLC plate and co-exposed with tissue sections as a standard curve. After development (GE/Fuji Typhoon FLA 7,000), image densitometry was analyzed with ImageQuant software (Fuji). Phosphoimager units were converted to femtomoles on the basis of image densities overlying the standards and the specific activity of the radioligand. Data was analyzed with Excel and graphs made with SigmaPlot.

Tissue Fixing
Brain tissue sections were removed from storage at -80 • C and thawed for 5 min before incubating in Davidson's fixative (8.1% formaldehyde, 33.3% ethanol, 11.1% acetic acid, Eosin Y stain) for 24 h at room temp. Sections were then quickly rinsed in 70% ethanol to remove residual formaldehyde. All incubations were carried out at room temperature.

Secondary Antibody Staining
Tissue sections were washed 3 × 5 min in PBS-T and incubated in a 1:200 dilution of secondary antibody (anti-goat-IgG, Vector Laboratories BA-5,000, anti-rabbit-IgG, Vector Laboratories BA-1000) in PBS-TBA for 2 h and washed 3 × 5 min with PBS-T. All incubations were carried out at room temperature.

Visualization
Slides were developed as instructed using the VECTASTAIN Elite ABC Kit (Standard) (Vector Laboratories PK-6100). Tissue sections were then washed 3 × 5 min in PBS-T before incubating for 4 min in a 0.5% w/v solution of diaminobenzidine in PBS-T (filtered) with 0.001% hydrogen peroxide and finally counterstaining with Giemsa stain. All incubations were carried out at room temperature.

Development of a Radiosynthesis of [ 18 F]1
The synthesis of required gem-difluoroalkene precursor 11 was envisioned as shown in Scheme 1, as an additional step of the route to the required standard. The selective protection of phenolic group of 6-(hydroxymethyl)pyridin-3-ol 2 was carried out with 2,2,2-trifluoroethylmethanesulfonate yielding trifluoroethyl protected intermediate 3 in 41% yield. Analog 3 was reacted with 2-methyl-1,3-benzoxazol-5-ol and Tsunoda reagent (cyanomethylenetributylphosphorane) to construct the benzoxazole intermediate 5 in 73% yield. Next, ring opening of benzoxazole intermediate 5 was achieved by treatment with 2M HCl in ethanol which yielded amino alcohol 7 in quantitative yield. The amino alcohol 7 was subjected to peptide coupling with 1-methyl-6-oxo-1,6-dihydropyridazine-3-carboxylic acid using EDCI/pyridine to give amide product 9 in 54% yield. The cyclization of amide 6 to synthesize reference standard 1 was carried out using at reflux in toluene using a Dean-Stark trap and p-toluene sulfonic acid as an acid catalyst in an 80% yield. After successful synthesis of standard 1, the next step was to construct gem-difluoro enol ether precursor 11 for its application SCHEME 1 | Synthesis of cold standard 1 and precursor 11 for radiosynthesis of [ 18 F]1.
to try radiochemistry. We attempted generation of the precursor by subjecting 1 with n-BuLi at -78 • C, which has been reported for other PET imaging agent precursors (Fawaz et al., 2014), but the conditions did not result in the desired product and opened pyridazine ring of 1 was instead observed.
To accomplish the radiolabeling to generate [ 18 F]1, previously reported radiochemical conditions for this chemistry developed by the Riss group and used for [ 18 F]lansoprazole from our group were evaluated (Fawaz et al., 2014). Gem-difluoroalkene precursor 11 (2.3 mg) was dissolved in a mixture of anhydrous dimethyl sulfoxide (500 µL) and isopropanol (36 µL) and reacted with azeotropically dried [ 18 F]fluoride at 90 • C for 3 min (Entry 1). When subjected to these conditions, we obtained the desired product [ 18 F]1 in only 0.01% radiochemical yield (RCY) and in 1:10 ratio of trifluoromethyl ([ 18 F]1)/difluoroalkene (12) ( Table 2). While laying the foundation of this chemistry, Pike and co-workers (Aigbirhio et al., 1993) explained that anhydrous conditions favored the generation of radiolabeled alkene via an elimination-addition mechanism. Further Riss et al. (2012) explained that protic additives enhanced the ratio of trifluoromethyl derivatives over the alkene (Riss and Aigbirhio, 2011;Riss et al., 2012), owing to the quenching of the intermediate anion before eliminating a fluoride anion (Landini et al., 1989). Next, we tried other protic additives (e.g., ammonium chloride, triflate or carbonate) that have been previously utilized for production of [ 18 F]N-methyl-lansoprazole (NML) for clinical use (Fawaz et al., 2014;Kramer et al., 2020). The optimal reaction solvent consisted of a mixture of anhydrous DMSO (950 µL) and saturated ammonium chloride (5.0 µL), as optimized for NML (Fawaz et al., 2014). Gratifyingly, the reaction resulted in product formation in a 4-5% (n = 3) radiochemical yield, and a satisfactory ratio of Upon establishing the radiochemistry conditions, we turned our attention to the development of a suitable HPLC method for purification of [ 18 F]1, which we expected to be challenging given the structural similarities of [ 18 F]1 and 12. Reflecting this, traditional reverse-HPLC stationary phases like C18 failed to achieve reasonable separation between the two compounds. Fortunately, perfluorophenyl-capped matrix , Phenomenex] as the stationary phase worked well in both the semipreparative and analytical separations (see representative HPLC traces in the Supporting Information). Purification of [ 18 F]1 was achieved using semipreparative HPLC conditions (see Supplementary Figure 2 for a typical HPLC trace), which enabled separation of the desired product [ 18 F]1 (t R = 82-86 min) and [ 18 F]gem-difluoroalkene 12 (t R = 73-80 min). While retention times are somewhat long, and there is likely further scope for optimization, this method provided adequate separation of [ 18 F]1 and 12 for this preliminary study. For reformulation, the purified [ 18 F]1 was trapped on a C18 (Waters, 1cc vac) cartridge, the cartridge was rinsed with water to remove the residual HPLC solvent/buffers and eluted with ethanol (0.5 mL) and saline (9.5 mL) for injection. During the sterile filtration step, we noticed that the dose was retained on the filter membrane, losing 40% of the imaging agent on the filter. In order to avoid this loss of dose, we screened different filters (see Supplementary Material) and conditions. Gratifyingly, we were able to successfully reformulate the dose in a mixture of ethanol (660 µL), Tween-80 (160 µL), and saline (9.5 mL), which facilitated sterile filtration without loss of dose. The formulation had the added advantage that the dose was also no longer retained on the syringes utilized for i.v. injection during preclinical imaging studies. Full automated synthesis provided [ 18 F]1 in 4-5% yield (2,446 ± 17.8 MBq, 66.1 ± 17.7 mCi, n = 3, decay corrected radiochemical yield, based upon 2.0 Ci of [ 18 F]fluoride),>98% RCP and molar activities = 16.5 ± 12.5 GBq/µmol (445 ± 339 Ci/mmol). The radiochemical purity of formulated [ 18 F]1 was analyzed with radio-HPLC to determine stability of a dose kept at room temperature for 2.5 h post-endof-synthesis; [ 18 F]1 did not show any evidence of decomposition and RCP remained>95% (see Supplementary Figure 6).

In vitro Autoradiography and Immunohistochemistry
To determine the suitability of [ 18 F]1 for in vivo experiments, we first undertook an in vitro autoradiography using post-mortem brain tissue samples from HD patients as well as a control subject ( Table 1 and Supplementary Figures 7-11). In vitro binding experiments were used to measure [ 18 F]1 affinity to mHTT aggregates in the HD brain sections, and the K d of [ 18 F]1 for mHTT was 2.30 nM. A series of saturation binding experiments using postmortem HD brain slices was next performed ( Table 3).    HD brain slices suggested that the binding in caudate and putamen is specific and saturable ( Table 3, entries 1 and 2), consistent with HD being a disease known to predominantly impact neurons in the basal ganglia, particularly in the earliest stages (Reiner et al., 2011;Tang and Feigin, 2012). Scatchard analysis suggested that the binding fitted to a single binding site (see Supplementary Material). As HD progresses there is also involvement of the cerebral cortex and other subcortical structures (Burgold et al., 2019). We observed saturable binding of [ 18 F]1 in HD cortical tissue samples ( Table 3, Entry 3), but it was substantially lower that the caudate and putamen sections. Lastly, as expected, we also observed no evidence of saturable binding of [ 18 F]1 in a control putamen sample ( Table 3, Entry 4). Non-specific binding (identified by co-incubation with 10 µM 1F) was observed in the white matter. To validate the autoradiography data, we conducted immunohistochemistry with anti-huntingtin antibody (ABN903) to identify mHTT aggregates in adjacent brain sections. Considering individual brain samples revealed a weak correlation (R = 0.433) between mHTT positive cells per µm 2 and disintegrations per µg of tissue/decay corrected dose ( Figure 3A). This trend might only be weakly discernible given heterogenous distribution of mHTT aggregates and the small sample size available from our brain brank. Indeed, a stronger correlation (R = 0.714) was apparent when considering the trend between mHTT positive cells per µm 2 identified by IHC and averaged across samples of a given brain region, and the calculated B max from the binding studies for the same brain region ( Figure 3B).

In vivo Positron Emission Tomography Imaging
The in vivo behavior of [ 18 F]1 was initially investigated in rodents. PET imaging studies were performed with [ 18 F]1 in female Sprague-Dawley rats (n = 3). A region-of-interest (ROI) was defined for the whole brain in the reconstructed PET data, and the summed data was used to generate a whole brain TAC (Figure 4). The data was converted to standardized uptake values (SUVs) and plotted for the 120 min dynamic imaging window. PET scans of the 3 rats revealed rapid uptake of [ 18 F]1 in the brain, with peak uptake occurring in 90 s (∼1,750 nCi/cc, corresponding to SUV max of ∼1.0) and subsequent washout throughout the duration of the scan. We observed about 30% clearance in 30 min, and about 45% clearance in 40 min. These encouraging results prompted us to next examine the in vivo imaging properties of [ 18 F]1 in non-human primates (NHPs). PET imaging studies were performed for [ 18 F]1 in mature female rhesus monkeys (n = 2). ROIs were drawn for the whole brain, as well as numerous brain regions (cortex, cerebellum, thalamus, striatum), and the summed data was analyzed to generate regional TACs. The data was converted to SUV and plotted for the 120 min imaging window (Figure 5). Results were analogous to rodent scans, with high brain uptake of [ 18 F]1 apparent in both monkey scans and peak uptake occurring in ∼90 s (∼1,000 nCi/cc, corresponding to SUV max of ∼2.0). Following peak uptake, [ 18 F]1 washed out from the brain. The time-radioactivity curves revealed about 30-40% clearance in 20 min, and about 50-60% clearance in 40 min. The results confirmed that the scaffold is blood-brain barrier (BBB) permeable and demonstrated quick wash-out from the brain, and low background/nonspecific signal.
After an intravenous injection, [ 18 F]1 penetrated the intact blood-brain barrier (BBB) of both rodents and NHPs efficiently, and peak uptake occurred in ∼90 s. Because there were no mutant HTT aggregates in the healthy monkey or rodent brain, as expected, [ 18 F]1 did not display any specific binding or prolonged retention in the brain. The performance is comparable with literature data for [ 11 C]CHDI-180R [e.g., SUV 2.7-3.0 in NHP imaging studies (Liu et al., 2020)].

CONCLUSION
In summary, an automated radiosynthesis of an 18 F PET imaging agent for mHTT has been developed for imaging patients with Huntington's disease. Highlights of the current method are its straightforward chemistry, simplicity, good radiochemical yields, and adaption to a commercial radiochemistry synthesis module for automated production of the radioligand in high purity. Imaging studies exhibited good brain uptake in rats and non-human primates, and autoradiography studies with post-mortem human HD brain tissue studies showed specific binding and evidence of correlation with mHTT protein aggregates identified by immunohistochemistry. Overall [ 18 F]1 is a promising candidate for imaging mHTT with PET to support disease management, track disease progression and evaluate experimental HD therapies. Future studies aimed at clinical translation of [ 18 F]1 will determine the safety profile of the radiotracer (pharmacology/toxicology and dosimetry studies), establish the metabolism, validate a synthesis to provide tracer suitable for clinical use and further validate the specificity of the signal for mHTT aggregates.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

ETHICS STATEMENT
The animal study was reviewed and approved by the University of Michigan IACUC.

AUTHOR CONTRIBUTIONS
PS designed the research. TK, AB, AL, TD, JS, JA, and WW performed the research. TK and AB contributed new reagents and analytical tools. TK, AB, AL, WW, TD, and PS analyzed the data. TK, AB, and PS wrote the manuscript. All authors reviewed the manuscript.

FUNDING
Financial support of this work from the University of Michigan, Department of Radiology, and the NIH (Award No. R01EB021155 to PS) is gratefully acknowledged.