The afferent visual pathway as a model for assessing neurodegeneration in experimental autoimmune encephalomyelitis – a translational optical coherence tomography (OCT) approach

Abstract

Multiple sclerosis is an inflammatory and neurodegenerative autoimmune disease of the central nervous system that presents with perivenous lymphocytic infiltration, focal demyelination and neuro-axonal degeneration. Neuro-axonal injury is a key contributor to non-reversible long-term disability in multiple sclerosis, however, the exact mechanisms underlying neurodegeneration as well as the sequence of events leading to structural and functional impairment are not yet fully understood. Visual impairment, including optic neuritis, is a common and early clinical feature of multiple sclerosis, with up to 75% of patients experiencing some form of visual disability during their disease course. Frequently, episodes of multiple sclerosis related acute optic neuritis are followed by structural retinal damage that can be quantified in vivo using optical coherence tomography (OCT). Reduced retinal peripapillary nerve fibre layer and macular ganglion cell layer thickness assessed by OCT have been shown to correlate strongly with functional visual impairment and- on a broader level - brain atrophy among multiple sclerosis patients. Similar to findings in human studies, alterations in the retina including the retinal ganglion cells and optic nerve have been reported in a mouse model of multiple sclerosis called experimental autoimmune encephalomyelitis (EAE). However, the exact biological basis and temporal dynamics behind the retinal abnormalities observed clinically and in OCT studies, including the complex interplay between inflammatory and degenerative mechanisms have not yet been fully elucidated. EAE can be used to investigate structural damage in the afferent visual pathway, providing a unique model for assessing neurodegenerative changes and their functional consequences. In addition, it has the potential to generate deeper insight into basic mechanisms underlying structural tissue damage in multiple sclerosis. To assess visual pathway damage in EAE, an OCT and magnetic resonance imaging platform was developed and established. OCT detected inner retinal layer thickness changes in EAE mice, resembles what is observed in human multiple sclerosis related acute papillitis. Where an initial phase of swelling was associated with inner retinal layer thickening, followed by a significant decrease in thickness over time, once the swelling had subsided, likely representative of neuro-axonal pathology. Diffusion tensor imaging measurements of the optic nerve and tract provided complementary results suggestive of demyelination and axonal damage in the visual pathway. OCT retinal thickness also correlated significantly with diffusion tensor imaging measurements providing support for retrograde and anterograde degeneration in the visual pathway following optic neuritis. Immunofluorescence analysis contributed further evidence for a strong inflammatory response in the retina and optic nerve in EAE mice at the final observational time point. A combination of both OCT and magnetic resonance imaging provided a reliable and sensitive quantitative tool to assess structural damage in the visual pathway and the temporal sequence of neurodegeneration in EAE. To further characterize the biological mechanisms of visual pathway impairment in EAE and to unravel the morphological correlates of retinal thickness changes, a longitudinal OCT study including a systematic assessment of retinal and optic nerve immunohistochemical analysis was performed. Signs of inflammatory edema 11 days post immunisation coincided with inner retinal layer thickening, while neuro-axonal degeneration throughout the disease course contributed to inner retinal layer thinning observed at later time points. Early retinal pathology, including axonal transport impairment, was observed prior to cellular infiltration (i.e. T-cells) in the optic nerve 11 days post immunisation. However, the consequences of early retinal damage on OCT-derived measurements were offset by the initial inflammatory edema. Microgliosis and astrocytosis was detected in the retina prior to optic neuritis and persisted until the final observational time point. Early glial activity likely contributed to initial signs of retinal pathology that appeared in the absence of cellular infiltration, suggesting a need for early intervention of optic neuritis. Subsequent to inflammation, Müller cells responded to retinal pathology with possible neuroprotective behaviour. Müller cell reactivity (i.e. aquaporin-4 and glutamine synthetase decrease) occurred after 11 days post immunisation in the inner retinal layer. Future studies should explore Müller cell reactivity and its potentially neuroprotective role. Severe neuro-axonal degeneration was observed in the optic nerve and retina until 33 days post immunisation. Although, retrograde degeneration likely promoted the majority of observed inner retinal layer damage following optic neuritis, primary pathology – possibly due to gliosis – also contributed to inner retinal layer thinning. These results added morphological substrate to the OCT findings, further solidifying this tool as a valuable method to assess visual pathway damage in EAE and multiple sclerosis. Although, retinal ganglion cell pathology had been observed using immunohistochemical analysis and corroborated inner retinal layer thickness findings, it is still unclear to what extent retinal ganglion cell loss plays a role in OCT observed thinning of the retina. In vivo measurements of retinal ganglion cells can provide additional information on the time-course of neuro-axonal pathology within the same mouse, which histology is not capable of. A transgenic mouse line with yellow fluorescent proteins expressed in the Thy1 sequence (a marker for neurons including retinal ganglion cells) was used in conjunction with green autofluorescence imaging to examine neuro-axonal damage in the retina longitudinally. Feasibility and longitudinal reliability of green autofluorescence imaging to assess yellow fluorescent protein expressing cells was confirmed in healthy transgenic mice. To measure the number of yellow fluorescent protein expressing cells in the retina both a manual and automated cell counting method was employed. Although the automated cell counting method had high variance compared to the manual cell count, it was significantly faster and allowed for a greater area on the fundus image to be examined, giving a more unbiased representative read-out of neurodegeneration and allowing for spatial statistics. In EAE mice, hyperintensity of the yellow fluorescent signal was observed around the optic nerve head at onset of clinical symptoms, likely coinciding with the occurrence of optic neuritis in this model. Interestingly, this phenomenon dissipated at peak of symptoms in almost all the EAE mice. Green autofluorescence imaging may be a useful way to identify the occurrence of optic neuritis in EAE through the hyperintensity observed at clinical onset of symptoms, especially compared to standard magnetic resonance imaging or ex-vivo methods of confirming optic neuritis. Future studies should evaluate the value of this model for assessing retinal ganglion cell death or neuropathology at later stages of experimental optic neuritis. The EAE induced Thy1 transgenic mouse model would be valuable in future pre-clinical trials that want to target retinal ganglion cells. Overall, investigating structural anterior visual pathway damage may constitute a unique model for assessing mechanism and temporal sequence of neurodegeneration in multiple sclerosis. The extent and rapid onset of axonal and neuronal damage in this model appears relevant for pre-clinical trials and for testing interventions scaled to multiple sclerosis.