BACKGROUND:

​

There is an unmet need to identify consistent, early, and accessible biomarkers to detect and track neurodegenerative diseases. Cell apoptosis, microglial activation and neurodegeneration are neuropathological features in e.g., Alzheimer’s disease (AD), multiple sclerosis (MS) and glaucoma. Such cellular properties, usually referred to in the brain, also exist in the eye. Additionally, most neurodegenerative diseases have ocular manifestations. The clear optical media allows for non-invasive and accessible methods to visualise the cells in the eye, making it an ideal candidate as a potential solution to the unmet needs. My general aims have been to identify retinocellular features in models of glaucoma, AD and MS with plans to correlate this with some clinical data.

​

​

METHODOLOGY:

​

My latest focus has been on investigating the characteristics of retinal microglia and retinal ganglion cells (RGCs, type of retinal neuron) in a cuprizone-induced (CPZ) model of multiple sclerosis.

​

Male C57 mice were fed CPZ (0.2%w/w) mixed chow for 4 months (m) until sacrificed at 6m (6mCPZ) or 28m (28mCPZ) old. Age-matched controls (6mCtrl; 28mCtrl) were fed normal chow. Enucleated eyes were fixed in 4% PFA. Whole retinas were dissected, stained (RBPMS+RGCs; iba-1+microglia), mounted and imaged.

​

Microglia are immune cells of the CNS that dynamically change their shape to respond to its microenvironment to exert neuro-protective or -toxic effects, leading to the rescue or loss of neurons. To characterise retinal microglia, I developed a semi-automated method that quantifies and characterises each cell body from iba-1+microglia stained flat-mounts retinal images through a supervised machine learning algorithm based on support vector machines (SVM). The SVM categorises each counted cell into 5 morphology types including: ramified, hyper-ramified, activate, amoeboid and rod (Fig. 1), each of which have been associated with functions further explained here.

​

​

​

​

​

​

​

​

​

​

​

​

​

​

Figure 1. Microglia morphology types taken from Choi et al. 2021

 

To characterise retinal neurodegeneration, I developed a macro on ImageJ that uses functions to remove noise, equalise the background, enhance the contrast, and threshold the resulting image taken on a fluorescent microscope of a flat-mounted retina stained for RBPMS-positive RGC. The script extracts the total cell count per image and the RGC density is calculated based on the whole area of the retina.

 

RESULTS:

​

We performed a 2-way ANOVA with paired analysis (PA). There was a main effect of age for activated (p=0.028) and rod (p<0.001), and disease state for rod (p=0.04) microglia (Fig. 2A). PA showed higher activated microglia in 28mCtrl vs. 6mCtrl (p=0.049), and lower rod microglia in 28mCtrl vs. 6mCtrl (p=0.037) and in 28mCPZ vs. 6mCPZ (p=0.0007) (Fig. 2A). There was a main effect of age (p<0.001) and disease state (p=0.04) on RBPMS+RGC density (Fig. 2B). PA showed lower RBPMS+RGC density in 28mCtrlvs.6mCtrl (p<0.003) and in 28mCPZvs.6mCPZ (p<0.003) (Fig. 2B).

​

There is age-related loss of rod microglia and RGCs in healthy and CPZ groups, supporting reports that rod microglia could be protecting less damaged neurons. Age-related increased activated microglia may cause RGC loss, as transcription factors like TNF-alpha, bind to RGC receptors, inducing apoptosis.

​

​

​

​

​

​

​

​

​

 

 

 

 

 

 

 

 

 

 

 

Figure 2. Retinal microglia and RGC characteristics in a cuprizone-induced model of MS and age-matched controls

​

​

FUTURE WORK:

​

I plan to investigate retinocellular pathologies in a triple transgenic model of AD.

​

FUNDED BY:

​

​

​

​

​

​

​

​

 

 

 

CONTACT: