PET Ligands for Imaging Mutant Huntingtin Aggregates: A Case Study in Non‐For‐Profit Scientific Management

Positron emission tomography imaging of misfolded proteins with high‐affinity and selective radioligands has played a vital role in expanding our knowledge of neurodegenerative diseases such as Parkinson's and Alzheimer's disease. The pathogenesis of Huntington's disease, a CAG trinucleotide repeat disorder, is similarly linked to the presence of protein fibrils formed from mutant huntingtin (mHTT) protein. Development of mHTT fibril–specific radioligands has been limited by the lack of structural knowledge around mHTT and a dearth of available hit compounds for medicinal chemistry refinement. Over the past decade, the CHDI Foundation, a non‐for‐profit scientific management organisation has orchestrated a large‐scale screen of small molecules to identify high affinity ligands of mHTT, with lead compounds now reaching clinical maturity. Here we describe the mHTT radioligands developed to date and opportunities for further improvement of this radiotracer class.


Introduction
Huntington's Disease (HD) is an autosomal dominant, neurodegenerative disorder which leads to progressive motor, cognitive and psychiatric problems. [1,2]Affecting around 5 in 100,000 people globally, [3] HD runs a progressive course over 15 to 20 years, ultimately leading to death.The mean age of onset of Huntington's Disease is 40 years of age, however neurobiological changes are evident before the presentation of clinical signs. [1]HD is caused by a single genetic mutation: an expanded CAG trinucleotide repeat in the huntingtin (HTT) gene.Discovered in 1993, [4] the mutant HTT (mHTT) gene, located on chromosome 4p16.3,encodes the production of a mutant huntingtin (mHTT) protein which is characterised by an abnormally long N-terminal polyglutamine (polyQ) tract.Individuals with greater than 39 CAG repeats develop Huntington's disease with the age of onset decreasing with increasing CAG repeat length. [1]It is generally considered that Huntington's disease pathology is linked to the toxic gain-of-function of its hallmark protein, mHTT.Currently, there is no cure for HD.Pharmacological intervention is limited to symptom management, however, there are new mHTT lowering therapeutics on the horizon, [5,6] with several under clinical evaluation in human trials. [7]Table 1 details terms relevant to the topics of HD and mHTT-PET radiotracer development, discussed herein.

Cryo-EM
Structural biology technique which allows 3D resolution of non-crystalline macromolecules.As resolution afforded by cryo-EM continues to improve, so will our understanding of the structures of misfolded proteins.

Exon-1 mHTT protein
The translated produce of spliced exon-1 mRNA.It is thought to be more pathogenic than full-length mHTT.
CHDI Cure Huntington's Disease Initiative SUV Standard Uptake Value; a measure of PET radiotracer uptake, corrected for administered dose and subject body weight.Allows for comparison of radiotracer uptake between subjects.

New Therapies on the Horizon for HD
The advent of novel disease-modifying treatments promises to shift the therapeutic paradigm in HD by addressing its root cause, mHTT.At the time of writing there are more than 50 active trials exploring the efficacy of various pharmacological interventions in HD. [8] Broadly, the most promising HD mHTT-targeted therapies fall into the following mechanistic classes: [9]

The Case for mHTT-PET
With the translation of disease-modifying therapies to first-inhuman clinical trials, there is an increasing unmet need for accurate quantification of mHTT burden in HD brains.This is was recently highlighted by the case study of Tominersen, the flagship ASO developed by Ionis Pharmaceuticals (later acquired by Roche) for reducing mHTT expression (Figure 1). [11]ominersen is a non-selective ASO, which targets exon 36 of human HTT mRNA (both mutant and wild-type alleles).Tominersen is delivered intrathecally to enable delivery of the ASO to brain regions adjacent to the CSF.In Phase 1/2a trials, Tominersen was shown to reduce concentrations of mHTT and neurofilament light chain protein in the CSF of patients.In subsequent Phase 3 trials however, Tominersen failed to meet its primary endpoint: the trial cohort performed worse on clinical rating scales compared to the control group.Despite this, CSF levels of mHTT were reduced in the trial group and post-hoc analysis indicated that younger patients at an earlier disease stage could benefit from Tominersen treatment. [8,11]he failure of Tominersen in Phase 3 trials raised many pertinent questions around successful trial design of ASOs for HD: the importance of conducting Phase 2 dose-finding studies and selecting the best disease-stage cohort, namely the one which would benefit most from ASO treatments.Additionally, questions concerning the relevant therapeutic dose of Tominersen arose over the course of the trial, and the risk of offtarget side effects (e. g. hydrocephalus). [12]ritically, the Tominersen trial highlighted the importance of selecting clinical trial endpoints that are sensitive enough to detect changes in disease progression within short clinical trial timeframes, as well as therapeutic target engagement at the relevant CNS site.This is particularly salient for studies performed in pre-manifest cohorts, where clinical changes in HD would be absent over the course of a clinical trial.
Indeed, in 2012 a large observational trial of over 200 participants with HD, the TRACK-HD clinical trial, pooled a battery of neuroimaging, cognitive, UHDRS, quantitative motor and neuropsychiatric measures of disease progression to identify the most sensitive marker of decline over 24-months.Of the measures tested, brain atrophy identified by MRI, not changes in clinical symptoms, was found to be the most sensitive measure of disease progression over the 24-month period in pre-manifest and early HD patients, compared to typical clinical rating scales used to stage disease (e. g.UHDRS).MRI-measured atrophy was capable of detecting changes 16 years before symptom onset. [13,14]Despite the utility of MRImeasured atrophy as a biomarker of HD progression, it is still only sensitive to physiological changes after the commencement of neuronal loss.Moreover, brain atrophy does not furnish any information on mHTT burden: there is a need for a sensitive, quantitative measure of mHTT in the brain to better determine treatment efficacy of emerging mHTT-lowering therapies.
Positron emission tomography (PET) imaging, using a radiotracer which binds specifically mHTT aggregates, is a promising methodology for achieving this.A mHTT-binding PET radiotracer could provide an accurate clinical measure for mHTT lowering therapies by providing a sensitive measure of mHTT levels in the brains of HD patients.

PET in Neurodegenerative Disease
PET is a sensitive non-invasive molecular imaging technique that is used to study (patho-) physiological processes by detecting positron-emitting radiotracers in vivo.PET imaging can provide physiological information by targeting biochemical pathways or receptors which are specific to a certain neural circuit or disease state. [15]PET is a mainstay of neuroscience research as it lends vision and quantification to the molecular processes underlying neurodegenerative disease.
PET radiotracers consist of a targeting moiety; either a small-molecule drug or antibody, tethered to a positronemitting radionuclide.Carbon-11 and fluorine-18 are the most widely used radionuclides for CNS-PET applications, with respective radioactive half-lives of 20.4 and 109.7 minutes. [15]pon injection, the radiotracers emit positrons that travels a short distance (~2 mm) before being captured by an electron.The collision of an electron and positron (matter and antimatter) results in an annihilation event producing two γ-rays.These gamma rays are detected by photon detectors and subsequent image reconstruction produces an image showing the distribution of the PET probe throughout the body.PET imaging is typically used in conjunction with anatomical imaging modalities such as CT and MRI.
A salient example of the utility of CNS-PET in neurodegenerative disease research is in the development of amyloid-β clearing antibodies for the treatment of Alzheimer's disease.The recent TRAILBLAZER-ALZ 2 Phase III clinical trial exploring the efficacy of amyloid-beta-clearing antibody Donanemab in early symptomatic AD, used both Aβ and tau-PET imaging profiles as secondary outcome measures. [16]Other examples of CNS targets under investigation as PET targets, include isoform-specific tau for various tauopathies and alphasynuclein for Parkinson's disease and related disorders. [17]

CNS-PET Radiotracer Development
PET radiotracer development for CNS disorders is a uniquely challenging venture in medicinal chemistry.The foremost hurdle in designing a CNS-PET radiotracer is the requirement for the radiotracer to permeate the blood-brain-barrier.There are several physiochemical frameworks (e. g.Lipinksi, Veber's rules, in silico methods such as CNS-Multiple Parameter Optimisation) [18,19] for screening small molecules for blood-brain penetrance.Broadly, a candidate radioligand should fulfill the following criteria: [20][21][22][23] 6. Challenges Specific to mHTT-PET Radiotracer Development

Low Abundance of the Target Protein and Selectivity Requirements
mHTT aggregates are present in much lower abundance compared with analogous misfolded proteins in other proteinopathies such as Alzheimer's and Parkinson's disease. [24]The significance of this from a PET radiotracer development perspective is the requirement for extremely high affinity ligands for mHTT.This is because the lower the abundance of aggregated protein in the brain (i.e. the Bmax value), the higher the affinity of ligand that is needed to provide adequate resolving power in vivo (recall that the B max /K d quotient should exceed 10). [22]or example, in the case of alpha-synucleinopathies, it is estimated that the total concentration of alpha-synuclein is 10 to 50-fold lower than that of amyloid beta in Alzheimer's disease. [23,25]This would indicate sub-nanomolar affinities are required for candidate α-synuclein PET ligands.Biochemical staging suggests that alpha-synuclein is present in concentrations of 50 to 200 nM in the brainstem and subcortical regions in late-stage α-synucleinopathies. [26] By comparison, the total concentration of mHTT averaged across the whole brain in HD has been estimated at 150 nM, [27,28] placing mHTT in a similar concentration regime to that of α-synuclein in α-synucleinopathies.
In addition to having high affinity, the selectivity profile of the radiotracer for the desired target over other misfolded proteins, which are present, must be considered.Indeed, the co-occurrence of other protein aggregates has been observed in HD. [29,30] One quantitative study on post-mortem HD brain tissue from 56 HD patients found that there was increased aggregation of TDP-43, α-synuclein and phosphorylated tau in HD, which correlated with mHTT accumulation. [31]Interestingly there was a decrease in amyloidopathy in HD in this study.

Chemical Space Containing Small-Molecule Binders of mHTT is Relatively Underexplored
Large-scale commercial and public interest in CNS targets such as α-synuclein, tau and amyloid-β has resulted in decades of structural refinement of small-molecule ligands for these targets. [32]With a prevalence of 1 in 10,000 in European populations, [33] HD is not a commercially attractive therapeutic target to industry, let alone one for PET radiotracer development.As such, therapeutic and PET tracer development for these misfolded proteins is further advanced compared that for mHTT.
Chemical space around high affinity ligands for mHTT is relatively underexplored compared to that of amyloid-β, tau and α-synuclein.In the case of mHTT, a handful of highthroughput screens have been conducted towards inhibiting mHTT aggregation and developing mHTT protein degraders; [34][35][36] however, none have successfully yielded lead compounds with nanomolar potency. [37]

Lack of Structural Biological Information
As is the case with other misfolded proteins, there is relatively little known about the binding site of small molecule PET ligands.Owing to their inherent amorphous state, fibrillar proteins cannot be resolved by protein crystallography however the burgeoning field of cryo-electron microscopy has begun to tease out structural detail of protein fibrils such as alphasynuclein and amyloid fibrils, as well as their interactions with small-molecule PET ligands. [38]ryo-EM could in future drive rational PET tracer design for these neurodegenerative targets.Indeed ssNMR structures have already been leveraged for in silico screening of ligands for alpha-synculein. [39][42] The structure of recombinant mHTT fibrils has been studied with cryo-electron microscopy, but the reported structures are not sufficiently resolved to elucidate potential binding modes of small-molecule PET ligands. [43]

CHDI and mHTT PET Ligand Development
The CHDI foundation is a non-for-profit biomedical foundation that aims to "rapidly discover and develop drugs that delay or slow the progression of Huntington's Disease".CHDI are unique in that they lie somewhere between an academic and industrial organisation in scientific approach; CHDI is a privately funded, non-for-profit organisation which delivers against set scientific goals by enabling partnership between biotech, pharmaceutical companies and academic research groups.Project management is centrally managed by CHDI and directed towards rapid delivery of disease modifying treatments in HD.
Over the past decade, CHDI has spearheaded a comprehensive programme of work to address the challenge of developing an mHTT PET radiotracer.A survey of over 15 related patents suggests that hundreds of compounds have been synthesised and evaluated for their binding to mHTT aggregates.Using the tau ligand FDDNP as a hit compound, this large screen has culminated in the identification of three ligands with nanomolar affinity for mHTT which are amenable to radiolabelling with either carbon-11 or fluorine-18: CHDI-180R, [24] CHDI-626, [44] and CHDI-650, [45] the results of which have been published in 2020, 2021, and 2023 respectively.The timeline of the CHDI mHTT PET radiotracer development is summarised in Figure 3.
In their first-in-class study, Liu and co-workers reported the development of an mHTT ligand, CHDI-180 with low nanomolar affinity against mHTT fibrils.This medicinal chemistry approach relied on a central screening radioligand binding assay (RBA) based on the aggregation of recombinantly expressed mHTT monomer with 46 glutamine repeats as a surrogate for the physiologically relevant mHTT species.Aggregation of the mHTT monomer to the fibrillar mHTT form created a platform for high throughput screening of ligands for their ability to bind to these mHTT fibrils in vitro.CHDI-180R was subsequently N-[À 11 C] methylated to yield [ 11 C]CHDI-180R in greater than 99 % radiochemical purity, with molar activities in the range of 170 to 575 GBq/μmol.Uptake of [ 11 C]CHDI-180R was then studied in 9-month-old zQ175 homozygous and heterozygous mice; increased binding was observed in both heterozygous and homozygous mice compared to controls.In follow up PET imaging studies in four healthy Rhesus macaques, [ 11 C]CHDI-180R showed excellent brain uptake and elimination kinetics, with the standard uptake value (SUV) value in the brain reaching a maximum value of 2.5 to 2.8, one minute after administration.No accumulation of the radiotracer was observed in the brain over time as would be expected in healthy NHP study subjects.On the basis of this favourable pharmacokinetic profile, [ 11 C]CHDI-180R was progressed to Phase 1 human trials. [46]Extensive preclinical evaluation and kinetic modelling of this radiotracer in the zQ175DN and R6/2 HD mouse models was carried out by Bertoglio and coworkers. [47]n their development of [ 11 C]CHDI-180R, Liu and coworkers noticed extensive nonspecific binding of CHDI-180R to in Alzheimer's brain homogenate (containing Aβ/tau), highlighting the low mHTT-selectivity of this ligand.To address this selectivity issue, a second counter screening assay was established to identify ligands with low binding to Aβ/tau in Alzheimer's disease brain tissue.In this assay, a known tritiated amyloid-β ligand was incubated with AD brain homogenate and subsequently challenged with a large excess of the mHTT ligand under investigation.The extent of radioligand displacement by each ligand was used as a surrogate measure for offtarget binding to non-mHTT aggregates.
Using this screening protocol, a second family of mHTT ligands were developed which yielded the lead compound CHDI-626.CHDI-626 was found to have a higher affinity to mHTT than CHDI-180R, as well as a lower Aβ displacement; 16 % compared to 49 % for CHDI-180R.[ 11 C]CHDI-626 was radiosynthesised via O-[ 11 C]methylation with greater than 99 % radiochemical purity and molar activities ranging from 69 to 615 GBq/μmol.[ 11 C]CHDI-626 showed highly specific binding on R6/2 mice brain slices in autoradiography experiments compared to WT mice tissue.Similarly, [ 11 C]CHDI-626 showed rapid brain uptake and washout in Rhesus macaques (maximal brain SUV of 2.5 to 4.0 after 1 minute) and no radiotracer accumulation was observed.Longitudinal preclinical evaluation of [ 11 C]CHDI-626 in the zQ175DN mouse model is described by Bertoglio and co-workers. [48]n 2022, the results from a first-in-human study of [ 11 C]CHDI-180R and [ 11 C]CHDI-626 in three healthy volunteers were published. [49]Both radiotracers were deemed safe for in vivo PET imaging in humans, with brain uptakes of up to 3.5 %ID/g.]CHDI-180R to non-mHTT proteins or low affinity for mHTT. [50]Additionally, it is also possible that significant differences would be detected in cohorts at more advanced stages of disease.Nevertheless, owing to the lack of efficacy observed in these Phase 2a trials, development of [ 11 C]CHDI-180R has been discontinued. [50]o enable simpler imaging protocols for subjects and wider availability, it is desirable to design longer-lived fluorine-18 radiotracers.CHDI's achieved this by altering the CHDI-180R scaffold to include a fluorine-18 radiolabelling site, specifically adding tetra-deutero 2-fluoroethoxy group to reduce defluorination in their fluorine-18 candidate [ 18 F]CHDI-650. [45]The 2fluoroethoxy is a common functional group employed to introduce fluorine-18 into ligands owing to its modularity, however this chemical group is prone to defluorination in vivo.This is undesirable owing to the high oxaphilicity of fluoride anions, which leads to skull uptake.Addition of deuterium atoms can suppress oxidative difluorination via the kinetic isotope effect.
[  CHDI-650 was radiolabelled with fluorine-18 to yield [ 18 F]CHDI-650 in radiochemical purities greater than 99 % and molar activities more than 200 GBq/μmol.[ 18 F]CHDI-650 PET imaging in Rhesus monkeys showed rapid uptake and elimination (maximal SUV values of 2.4-2.7 after 0.6 min) of the radiotracer, as well as exhibiting a favourable metabolism profile, including low bone retention.Phase 1 trials of [ 18 F]CHDI-650 are expected to take place between 2024 and 2026.
In the meantime, CHDI continues to diversify their pipeline, adding three new fluorine-18 radiotracers ([ 18 F]CHDI-747, 385, 386) as well as a new carbon-11 radiotracer, [ 11 C]CHDI-009 into their preclinical pipeline (structures have not yet been disclosed). [51]dditionally, it should be noted that Kaur and co-workers reported the synthesis of a -CH 2 CF 3 derivative of CHDI-180R in 2021 (compound 1 depicted in Fig. 3). [52][ 18 F]1 was radiosynthesised from a gem-difluoroalkene precursor in 4.1 % decay-corrected radiochemical yield and a molar activity of 16.5 � 12.5 GBq/umol.The authors determined the binding of their radiotracer to mHTT via autoradiography on a small cohort of human post-mortem HD tissue samples, yielding a K d value of 2.3 nM and a binding potential (B max /K d ) of 13.8.Kinetic analysis of [ 18 F]1 PET imaging in two rhesus monkeys showed a peak brain uptake (SUV max ) of 2.0 after 90 seconds, followed by 50-60 % clearance after 40 minutes.The authors concluded [ 18 F]1 is a promising radiotracer candidate with a similar uptake and washout properties as [ 11 C]CHDI-180R.

Future Outlook
CHDI has spearheaded the design of mHTT PET radiotracers, however further work is required to yield a clinically viable radiotracer.Below we outline future opportunities for the improvement of mHTT radiotracers.
* Greater structural diversity is required to increase chances for success.The current CHDI family of ligands lie within two distinct scaffold families (the benzoxazole family of CHDI-180 and the benzo [4,5]imidazo[1,2-a]pyrimidine core of CHDI-626 molecule).
* Machine-learning approaches could be leveraged to explore new chemical space outside the CHDI ligand family.
* All potential ligands should be filtered through in silico scoring systems, which predict likelihood of in vivo CNS penetrance and metabolic stability.
* Future resolution of mHTT fibrils via cryo-EM will enable more rational radiotracer design via in silico docking methods.Structural information should drive ligand design where possible.
* Increased availability of in vitro models (i.e. recombinant mHTT fibrils) is required to develop high-throughput binding assays to better screen potential mHTT ligands for mHTT fibrils.Currently mHTT protein fragments are not commercially available for this purpose, although full length mHTT protein can be purchased.
* Protein selectivity should be embedded in the in vitro screening of compound libraries.Compounds should be counter-screened for binding with co-occurring protein aggregates (namely amyloid-β, tau, α-synuclein and TDP-43).
* Moreover, as the main high throughput screening method for mHTT ligands is based on recombinant mHTT fibrils, differences in the in vitro morphology of recombinantly expressed mHTT constructs and in vivo morphology of human-derived mHTT fibrils should be considered.
* To this end, use of seeding approaches to reproduce pathological forms of mHTT protein fibrils should be explored in the screening of ligands in assays using recombinantly expressed mHTT exon-1 [53] Indeed, it has been found that mHTT exon-1 fibrils exhibit seeding activity. [47]hese challenges facing the design of a viable mHTT radiotracer will be best met by increased academic participation in mHTT radiotracer design to introduce a variety of perspectives into the in vitro validation of these PET ligands.

Conclusions
The recent development of mHTT targeted small-molecule radiotracers, represents an exciting development in HD research and CNS-PET radiotracer development in general.Such a radiotracer has the potential to revolutionise how emerging treatments for HD could be evaluated in clinical trials and lend new insight into how mHTT pathology drives disease in vivo.However, further work is required to identify a clinically viable PET radiotracer for mHTT quantitation in vivo.Increased participation by third-party academic and industrial groups will increase chances for success by diversifying the scientific approaches taken towards the development of an mHTT PET ligand.
This work has been supported by the Alzheimer's Research UK East Network Centre grant ARUK-NC2021-EAST.CGFD is supported by the Cambridge Australia Bragg Scholarship given by the Cambridge Commonwealth, European and International Trust.

Figure 1 .
Figure 1.Underlying pathology of Huntington's disease and targeted species for PET ligand development in HD.

Figure 2 .
Figure 2. Major hurdles in the design of an mHTT-specific radiotracer.
[ 11 C]CHDI-626 formed an unidentified brain-penetrant radiometabolite, which precluded it from further clinical evaluation.In 2022, preliminary results of a Phase 1b study (the iMageHTT trial) which quantified [ 11 C]CHDI-180R in six HD patients and three healthy controls were published.Updated Phase 2a results reported in 2023 showed no difference in [ 11 C]CHDI-180R brain uptake between six early-intermediate stage HD subjects and healthy controls.Maximal SUV uptake of [ 11 C]CHDI-180R ranged between 2 and 2.5 in both HD individuals and healthy controls.Metabolic sampling showed that after 70 minutes approximately 50 % of the radiotracer remained intact.The lack of difference between HD and non-HD individuals may be due to high off-target binding of [ 11 C 18 F]CHDI-650 improved on [ 11 C]CHDI-180R by replacing the pyridazinone ring of [ 11 C]CHDI-180R with a pyrimidine ring which reduced off-target binding.The introduction of a ring nitrogen into the central bicycle of [ 11 C]CHDI-180R improved brain penetration.

Table 1 .
Glossary of terms relevant to mHTT PET radiotracer development.
mHTTMutant huntingtin protein encoded by the mHTT gene with an extended CAG tract; the pathogenic species in HD containing an expanded polyQ tract).mHTTliesat the heart ofcellular pathology in HD and forms neuronal aggregates.K dBinding affinity of a ligand for the target (i.e. mHTT fibrils).Given the low concentration of mHTT in HD (B max ), a picomolar to nanomolar affinity of candidate ligands is ideal.