Reduced D₂/D₃ receptor binding and glucose metabolism in a macaque model of Huntington’s disease

Abstract

Background:

Dopamine system dysfunction and altered glucose metabolism are implicated in pre- and early symptomatic stages of Huntington’s disease (HD), a neurological disease caused by mutant huntingtin (mHTT) expression.

Objectives:

To characterize alterations in cerebral dopamine D₂/D₃ receptor density and glucose utilization in a newly developed AAV-mediated NHP model of HD that expresses mHTT throughout the striatum and cortex.

Methods:

PET imaging was performed with [¹⁸F]Fallypride to quantify D₂/D₃ receptor density and [¹⁸F]FDG to measure cerebral glucose utilization in these HD macaques.

Results:

Compared to controls, HD macaques showed significantly reduced dopamine D₂/D₃ receptor densities in basal ganglia (p<0.05). Additionally, HD macaques displayed significant glucose hypometabolism throughout the cortico-basal ganglia network (p<0.05).

Conclusions:

[¹⁸F]Fallypride and [¹⁸F]FDG are PET imaging biomarkers of mHTT-mediated disease progression that can be used as noninvasive outcome measures in future therapeutic studies with this AAV-mediated HD macaque model.

INTRODUCTION

Huntington’s disease (HD) is a genetic polyglutamine repeat disorder that results in progressive cognitive and psychiatric disturbances, along with a complex movement disorder^1–3. HD is caused by a mutation in exon 1 of the HTT gene on chromosome 4, where a CAG repeat >40 leads to the production of a mutant HTT protein (mHTT) with an expanded glutamine stretch at the N-terminus⁴. mHTT misfolds and forms intracellular aggregates in the brain and periphery—a hallmark of HD disease pathogenesis^5–7. Therapies aimed at preventing or delaying the neuropathology, and ensuing symptoms, of HD are under investigation, including a variety of (m)HTT-lowering approaches, some of which have recently progressed to evaluation in clinical trials^8,9. While genetically modified rodent models of HD have advanced our understanding of many of the pathophysiological mechanisms of mHTT-induced cellular dysfunction, they do not fully recapitulate the behavioral symptomology of humans with HD including cognitive deficits, chorea and bradykinesia. Additionally, species differences in genetics, brain size, neuroanatomy and connectivity limit the ability to translate findings from rodents to predict clinical responses in human patients. To address this issue, we recently created a rhesus macaque model of HD characterized by mild motor and cognitive phenotypes, striatal and cortical atrophy, white matter degeneration, and reduced cortico-striatal functional connectivity, as well as neuronal mHTT+ aggregates in many of the brain regions in which they are observed in human HD patients (e.g. caudate, putamen, prefrontal cortex, anterior cingulate cortex, motor cortex, globus pallidus, thalamus, amygdala, among others)^10,11.

Given that multiple aberrant cellular mechanisms, including dopamine system dysfunction and altered cerebral metabolism, have been implicated in premanifest/early symptomatic stages of HD, this study sought to test the hypothesis that mHTT expression altered regional D₂/D₃ receptor density and glucose metabolism in this NHP model by using [¹⁸F]Fallypride and 2-[¹⁸F]Fluoro-2-deoxy-D-glucose ([¹⁸F]FDG), respectively. The overall aim was to further validate this large-animal model of HD for use in pre-clinical studies, and to define appropriate imaging biomarkers that can be employed to assess the efficacy of novel experimental therapies in this model.

METHODS

Study design

The objective of this study was to characterize alterations in cerebral dopamine D₂/D₃ receptor density and glucose utilization in a NHP model of HD^10–12. Adult macaques were injected into the caudate and putamen with a 1:1 mixture of AAV2.retro and AAV2 expressing the N171 fragment of human mHTT bearing either a pathological number of CAG repeats (HTT85Q, n=6) or a control number of repeats (HTT10Q, n=6). A third group was injected with vector diluent only (Buffer, n=5). PET imaging of the brain was performed between 16–19 months post-AAV delivery with [¹⁸F]Fallypride to measure D₂/D₃ receptor density and [¹⁸F]FDG to measure cerebral glucose utilization. [¹⁸F]FDG was quantified by glucose standardized uptake values (SUV_glc), whereas [¹⁸F]Fallypride using the Logan reference model with the cerebellum as reference region¹³. To delineate brain regions and generate regional parcellation maps, PET data was aligned with the ONPRC18 species-specific MRI atlas¹². The same spatial normalization approach was used for both the [¹⁸F]Fallypride and [¹⁸F]FDG PET images (summarized in Figure S1). Detailed information for each methodological procedure is available in the Supplementary materials.

Data analysis

After applying parcellations from the ONPRC18 atlas, average BP_ND and SUV_glc values were calculated for each volume of interest (VOI). Due to the lack of statistically significant hemispheric differences, BP_ND and SUV_glc data were averaged across hemispheres for subsequent analyses. Despite the sample size of the groups (n=5–6), the linear mixed-model analysis was selected for the analysis of the quantitative data as it allowed sex and age of the animals to be included as covariates. Post-hoc group comparisons with Tukey corrections for multiple comparisons were performed for each region separately. All processing and analyses were performed by the researchers while blind to experimental group allocation. Statistical threshold for all analysis was set at p<0.05. Additional details are available in Supplementary materials.

RESULTS

Reduced D₂/D₃ dopamine receptor binding in the basal ganglia of HD macaques

To assess the extent of D₂/D₃ receptor loss in the novel AAV2:2retro-mediated NHP model of HD, we employed [¹⁸F]Fallypride PET imaging and compared BP_ND between HTT85Q-treated macaques and HTT10Q- or Buffer-treated controls throughout the brain. HTT85Q-treated macaques show several brain areas with significantly reduced D₂/D₃ receptor binding. Averaged BP_ND parametric maps illustrating coronal, axial, and sagittal receptor density for each group are shown in Figure 1A. Statistical analysis of the average D₂/D₃ receptor density (BP_ND) in each brain region revealed a significant group effect (F_(2,378)=3.307, p=0.0377), a significant brain region effect (F_(26,378)=332.8, p<0.0001), and a significant group*brain region interaction (F_(52,378)=1.540, p=0.0129). Post-hoc analyses indicated that HTT85Q macaques had significantly lower BP_ND in the caudate, putamen, and external segment of the globus pallidus (GPe) compared to both Buffer (CD: t=3.38; p=0.0341, −10.2%; PUT: t=4.15; p=0.0142, −21.6%; GPe: t=3.09; p=0.0488, −22.4%) and HTT10Q control groups (CD: t=4.03; p=0.0161, −10.0%; PUT: t=3.32; p=0.0368, −14.9%; GPe: t=3.20; p=0.0425, −21.8%) (Figure 1B and Supplementary Table 3). There were no significant group differences in any of the other brain regions examined.

Figure 1. Reduced D₂/D₃ receptor binding in the basal ganglia of NHPs expressing HTT85Q.

(A) Mean [¹⁸F]Fallypride parametric maps of D₂/D₃ receptor density for each group. PET is overlaid onto the ONPRC18 T2w MRI template for anatomical reference. (B) Regional [¹⁸F]Fallypride BP_ND group comparisons indicate a significant reduction in D₂/D₃ receptor density in the caudate, putamen and external globus pallidus of HTT85Q macaques compared to controls. (Buffer, n=5; HTT10Q, n=6; HTT85Q, n=6; induvial animals within each group are plotted using unique shapes). *p<0.05. Abbreviations: CD = caudate, PUT = putamen, GPe = external globus pallidus, GPi = internal globus pallidus.

Widespread reduction of glucose metabolism in HD macaques

To assess regional changes in brain glucose metabolism resulting from AAV2:2retro-mediated expression of mHTT throughout the cortico-basal ganglia circuit, we employed [¹⁸F]FDG PET imaging and compared SUV_glc between HTT85Q treated animals and controls. Averaged glucose uptake images illustrating coronal, axial, and sagittal views of SUV_glc for each group are depicted in Figure 2A. Statistical analysis revealed a significant main effect of group (F_(2,350)=90.03, p<0.0001), and brain region (F_(24,350)=3.423, p<0.0001), but no significant interaction between these factors (F_(48,350)=0.08065, p>0.9999). Post-hoc tests comparing groups for each region separately indicated that HTT85Q macaques had significantly lower glucose uptake (SUV_glc) compared to animals in both control groups in several cortical and subcortical brain regions, (14.5–20.7% reduction compared to buffer group and 24.1–30.5% reduction compared to HTT10Q; p<0.05 each, see Supplementary Table 4). In contrast, the HTT10Q and Buffer groups did not differ significantly in any of the 28 investigated brain regions. Comparisons are illustrated in Figure 2B and reported in Supplementary Table 4.

Figure 2. Reduced glucose uptake in the brain of NHPs expressing HTT85Q.

(A) Mean [¹⁸F]FDG SUV_glc images of glucose uptake for each group. PET is overlaid onto the ONPRC18 T2w MRI template for anatomical reference. (B) Regional [¹⁸F]FDG SUV_glc group comparisons indicate significant wide-spread reductions in glucose uptake in HTT85Q macaques compared to controls. (Buffer, n=5; HTT10Q, n=6; HTT85Q, n=6; induvial animals within each group are plotted using unique shapes). *p<0.05, **p<0.01. Abbreviations: DLPFC: dorsolateral prefrontal cortex; VLPFC: ventrolateral PFC; OPFC: orbitofrontal cortex; VMPFC: ventromedial PFC; DMPFC: dorsomedial PFC; ACC: anterior cingulate cortex; DPMC: dorsal premotor cortex; VPMC: ventral PMC; SMC: supplemental motor cortex; MC: primary MC; STC: superior temporal cortex; ITC: inferior TC; RC: rhinal cortex; IC: insular cortex; SSC: somatosensory cortex; PC: parietal cortex; PCC: posterior cingulate cortex; OCC: occipital cortex; CD: caudate; PUT: putamen; GPi: internal globus pallidus; GPe: external GP; LatTH: lateral thalamus; MdTH: medial thalamus; HC: hippocampus; AMY: amygdala; SN: substantia nigra; CB: cerebellum.

DISCUSSION

A central goal of this study was to establish in vivo neuroimaging biomarkers that can be used to evaluate promising therapeutics in this newly developed AAV-mediated macaque model of HD. We characterized cortico-basal ganglia circuits using PET imaging of D₂/D₃ receptor density and glucose metabolism, and identified region-specific changes at 16–19-months post-AAVmHTT85Q administration. HTT85Q-treated animals exhibited a 10–22% reduction in D₂/D₃ receptor binding in the basal ganglia and a 12–30% reduction in glucose utilization throughout several brain regions compared to controls. These findings recapitulate data from pre-manifest and early-stage HD gene carriers who show 8–22% (early premanifest) and 12–29% (late premanifest) reductions basal ganglia D₂/D₃ receptor binding, and 5–7% (pre-manifest) and 30–38% (early-stage manifest) reductions in cerebral glucose metabolism, compared to healthy controls^14–17. In addition to the PET findings described here, work with this model has also established that HTT85Q treated macaques exhibit working memory impairments, motor phenotypes, reductions in cortico-striatal functional connectivity, microstructural alterations in white matter tracts, and mild (3–5%) cortical and striatal atrophy¹¹. In terms of the new HD Integrated Staging System (HD ISS), these animals, at 16–19 months post-surgery, present similarly to patients at Stage 2, ie. imaging biomarkers are present which demonstrate disease pathology and mild motor and cognitive behavioral symptoms are evident. However, a caveat is that HTT85Q macaques have a longer Q length compared to the Q-length range seen in adult-onset HD (85Q vs 40–55Q).

Moving forward, due to the cross-sectional nature of this study, additional work will need to be conducted to elucidate the longitudinal time course of these effects in this model, to determine the earliest point at which changes in D₂/D₃ receptor density and glucose utilization are detectable, and to query D₁ receptor density to characterize changes in both the direct and indirect basal ganglia circuits. Additionally, the relatively mild degree of striatal and cortical atrophy observed in these animals¹¹, taken in conjunction with larger magnitude of the PET findings reported here, suggests the possibility of synaptic degeneration or dysfunction in this model—as has been recently reported in mice¹⁸ and people with HD¹⁹. Therefore, SV2A PET imaging for quantification of synaptic changes represents another potential biomarker for this model, although additional validation will be required. Ongoing studies are investigating the biodistribution of mHTT protein throughout the brain in these animals using a novel radiotracer designed to bind to the aggregated form of mutant HTT^20,21. Once complete, animals will undergo necropsy, and histological and molecular readouts for disease processes will be collected to corroborate the PET findings reported here.

Together, these findings parallel the imaging and behavioral observations in HD patients and establish a macaque model of HD that is reproducible, able to be generated in large numbers, expresses mHTT in functional brain circuits, and appropriate for screening therapies at stages of disease similar to patients participating in current HD Phase I/II clinical trials. Whereas mouse models have already contributed significantly in the understanding of the disease, including mechanisms beyond polyQ fragment toxicity such as RNA toxicity, NHP models such as this feature important benefits including the presence of anatomial regions more closely related to the human brain (e.g. division between caudate and putamen or cortical folds), and behavioral phenotypes (e.g. chorea, working memory deficits) which are not present in mouse and other large animal models²². In the context of testing therapeutics, the higher homology of NHP brain to the human one offers a more profound and reliable understanding of the therapeutic biodistribution and efficacy following, for instance, intrathecal or intrastriatal delivery of new mHTT-lowering agents. The current results provide important validation for the utility of this new macaque model to recapitulate specific features of early-stage HD, including D₂/D₃ dopamine receptor dysregulation and altered cerebral metabolism, that can be quantified with PET imaging, setting the stage for evaluating the efficacy of promising therapeutic strategies in this model.

FINANCIAL DISCLOSURES OF ALL AUTHORS (FOR PRECEDING 12 MONTHS):

ARW:

Stock Ownership in medically-related fields- none

Intellectual Property Rights-none

Consultancies- none

Expert Testimony-none

Advisory Boards-none

Employment- Full time employment as a postdoctoral fellow at the Oregon National Primate Research Center

Partnerships- none

Contracts- none

Honoraria- none

Royalties- none

Grants- NIH-NIA T32AG055378

Other- none

DB:

Stock Ownership in medically-related fields- none

Intellectual Property Rights- none

Consultancies- Anonymous consultancy for GLG

Expert Testimony- none

Advisory Boards- none

Employment- Full time employment as a postdoctoral fellow at the University of Antwerp, Belgium

Partnerships- none

Contracts- none

Honoraria- Honoraria for seminar at Roche

Royalties- none

Grants- FWO (K201222N, 1229721N, I007522N), University of Antwerp (KP BOF FFB210050 & FFB210314), and GSKE (FFP220132)

Other- none

WAL:

Stock Ownership in medically-related fields- none

Intellectual Property Rights- none

Consultancies- none

Expert Testimony- none

Advisory Boards- none

Employment- Full time employment as a Senior Research Assistant at the Oregon National Primate Research Center

Partnerships- none

Contracts- none

Honoraria- none

Royalties- none

Grants- Supported by NIH grants NS099136, U01 NS116752, Bev Hartig HD Foundation, CHDI Foundation and Noah’s Hope Foundation

Other-none

KB:

Stock Ownership in medically-related fields- none

Intellectual Property Rights- none

Consultancies- none

Expert Testimony- none

Advisory Boards- none

Employment- Full time employment at Rice University as a graduate student (KB performed all work on this grant while a research assistant in Dr. McBride’s lab at the ONPRC prior to enrolling in graduate school)

Partnerships- none

Contracts- none

Honoraria- none

Royalties- none

Grants- Dean of Natural Sciences Worden Fellowship, Dean of Graduate Studies First Year Fellowship

Other- none

JT:

Stock Ownership in medically-related fields-none

Intellectual Property Rights- none

Consultancies- none

Expert Testimony- none

Advisory Boards- none

Employment- Full time manager and technologist at the Primate Multimodal Imaging Center, Oregon National Primate Research Center

Partnerships- none

Contracts- none

Honoraria- none

Royalties- none

Grants- none

Other- none

JL:

Stock Ownership in medically-related fields- none

Intellectual Property Rights-none

Consultancies- none

Expert Testimony- none

Advisory Boards- none

Employment- Professor, Oregon Health & Sciences University (0.82 FTE)

Partnerships- none

Contracts- none

Honoraria- none

Royalties- none

Grants- NIH/NHLVI HL140577 no cost extension, 5 P51 OD011092 NIH/OD, NIH/NICHD1 P01 HD106485-01 subcontract through UNC Chapel Hill Medical Research Foundation nonfederal unrelated to this project

Other - none

JLM:

Stock Ownership in medically-related fields- none

Intellectual Property Rights- none

Consultancies- Sanofi, Voyager Therapeutics and Capsigen (regarding gene therapy for neurodegenerative disorders)

Expert Testimony- none

Advisory Boards- none

Employment- Full time employment as an Associate Professor at the Oregon National Primate Research Center

Partnerships- none

Contracts- none

Honoraria- Honoraria for delivering seminars to Voyager Therapeutics and Sanford Research

Royalties- none

Grants- Supported by NIH grants NS099136, U01 NS116752, Bev Hartig HD Foundation, CHDI Foundation and Noah’s Hope Foundation

Other- none