Back to search

University of Edinburgh

Measuring neurodegeneration in mouse models of multiple sclerosis using SV2A PET and microstructural MRI

Abstract

dc:description.abstract

Multiple sclerosis (MS) is a neurodegenerative disease characterised by the formation of focal demyelinated lesions in the central nervous system associated with axonal damage, synaptic alterations, and neuronal loss leading to progressive and permanent disability. Conventional structural magnetic resonance imaging (MRI) used to diagnose and monitor MS has limited specificity and sensitivity for microstructural changes that may be related to disease severity and progression. More recently, microstructural MRI techniques such as magnetisation transfer imaging (MTI) and diffusion-weighted imaging (DWI), and positron emission tomography (PET) imaging of the presynaptic vesicle protein SV2A using [18F]SynVesT-1 have been described as promising markers of myelin, neurite, and synaptic integrity, respectively. I thus hypothesise that SV2A PET/MRI may be able to detect and characterise microstructural changes indicative of neurodegeneration in relevant mouse models of MS. To test this, global and focal white and grey matter demyelination was induced in wildtype mice with the administration of either cuprizone orally or of lysophosphatidylcholine (LPC) directly into the corpus callosum or cortex. I then performed complementary histological analyses using immunofluorescence and autoradiography with [3H]SynVesT-1 to validate and supplement in vivo results. My key PET findings include the first report of ex vivo autoradiography using [3H]SynVesT-1 that demonstrates high ligand affinity, binding potential, and specificity for its target not only under control conditions, but also in the presence of significant demyelination and remyelination in the cuprizone model, overall highlighting its sensitivity to SV2A and its potential as a marker of synaptic changes. Notably, I also found that non-displaceable binding was highly variable between timepoints. As this is typically assumed to be uniform across brain regions and pathological conditions, I showed that these changes can significantly bias commonly reported in vivo PET outcome measures for this tracer and should thus be corrected for to ensure accurate quantification and interpretation of imaging results. To further investigate the biological relevance of these changes, I validated these results with immunofluorescence staining of SV2A. Since SV2A is a presynaptic vesicle protein, changes in tracer binding may be due not only to changes in synaptic density, but also to changes in vesicle number or copy number per vesicle. Moreover, SynVesT-1 cannot differentiate between synapse subtypes due to the broad expression of SV2A. I thus characterised changes in SV2A puncta density, intensity, size, and colocalisation with common markers of inhibitory and excitatory synapses in various brain subregions in the cuprizone model. These results confirmed the global increase in SV2A with remyelination that I previously found in vivo with [18F]SynVesT-1, and further showed novel increases in excitatory synaptic density in the cortex during early remyelination, and general SV2A+ inhibitory synapse increases with late remyelination. Collectively, this study demonstrates the excellent imaging and binding characteristics of SynVesT-1 as well as its sensitivity to underlying changes in synaptic density and function with demyelination and remyelination, which is an essential step towards future translation into human MS studies. A central aim of the MRI study was to perform direct, quantitative comparisons of commonly derived but confounded metrics of myelin and neurite integrity (MTR and DTI) with more novel measures (MTsat and NODDI) that may be better able to resolve subtle microstructural changes in the tissue. To do this, I utilised a novel analysis pipeline that involves stacking multiple tissue slices to more accurately represent the larger volume of MRI slices, and performed direct voxel-by-voxel comparisons between the two modalities in the cuprizone model. Magnetisation transfer imaging demonstrated that both MTR and MTsat were sensitive to in vivo changes with demyelination and remyelination. Subsequent voxelwise and linear regression analyses with immunofluorescence markers of myelin, axonal, glial, and neuronal density, however, revealed that changes in the two measures may be driven by distinct biological drivers; while MTR showed greater sensitivity to inflammatory and glial responses during demyelination, MTsat had greater sensitivity to later reparative processes. These findings support the complementary use of both metrics for more comprehensive assessment of tissue pathology. Diffusion MRI and associated voxelwise histological and linear regression analyses further demonstrated that both Watson and Bingham-NODDI derived metrics were sensitive to microstructural changes associated with demyelination and remyelination, and that even subtle changes in model assumptions and parameter choices can substantially influence in vivo sensitivity to pathology. These measures were moreover differentially influenced by underlying tissue features, with the Watson model showing reduced sensitivity to glial-related confounds in white matter, and the Bingham model showing increased sensitivity for changes in neurite orientation and density in grey matter. Additionally, both Watson and Bingham-NODDI models overall had higher specificity for microstructural changes during demyelination and remyelination than traditional DTI-derived metrics. Taken together, these findings further support the use of diffusion MRI metrics as sensitive and biologically informed markers of tissue pathology. Exploratory receiver operating characteristic analyses in the cuprizone model demonstrated that combinations of MTI and DWI-derived metrics provided improved classification for demyelination, early remyelination, late remyelination, and axonal degeneration than any individual metric alone, supporting the use of multimodal imaging biomarkers. Collectively, these findings show that microstructural MRI metrics have differential sensitivities for biological processes at varying stages of disease progression, and that their combined use enables more robust and detailed characterisation of myelin integrity and neurodegeneration relevant to MS. Lastly, I performed SV2A PET/MRI on LPC focal white and grey matter demyelination models to complement results in the cuprizone model. I found significant changes in both in vivo MTI and DWI measures in the corpus callosum lesion, and significant changes in in vivo PET and DWI, but not MTI, measures for the focal cortical model during demyelination and remyelination. These results further demonstrate the sensitivity of SV2A PET/MRI for subtle tissue changes, highlighting their translational relevance. In conclusion, the results of this thesis demonstrate that SV2A PET with [18F]SynVesT-1 and microstructural MRI approaches are able to detect dynamic and spatially heterogeneous changes associated with demyelination, remyelination, and neurodegeneration. The preclinical validation work performed overall offered quantitative insights into how pathological processes relevant to MS contribute to changes in imaging signals, providing a complementary reference for the analysis, interpretation, and selection of these imaging biomarkers in clinical studies of pwMS.

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Panichnantakul, Pantila (Kelly)
Advisors dc:contributor.advisor
  • Waldman, Adam
  • Tavares, Adriana
  • Williams, Anna

Subjects

dc:subject × 5

Rights

Language dc:language.iso
en

Identifiers

dc:identifier.*

Chain of custody

source
Harvested from
University of Edinburgh
Base URL
era.ed.ac.uk/server/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Panichnantakul, Pantila (Kelly). Measuring neurodegeneration in mouse models of multiple sclerosis using SV2A PET and microstructural MRI. 2026. https://era.ed.ac.uk/handle/1842/44974