Computational Model of Myelin Degeneration in Optic Neuritis

Computational Model of Myelin Degeneration in Optic Neuritis INTRODUCTION Multiple sclerosis (MS) is a potentially debilitating autoimmune disorder that targets the nervous system and affects nearly 2 million adults worldwide. Among others, optic neuritis (ON) occurs as a common precursor and symptom of MS. ON causes unilateral blindness, color vision loss, vision field loss, and/or pain with eye movement. ON is rarely linked with diseases besides MS. RESULTS Figure 1 A Stimulus-Dependent Correlations B Matched Constant Correlations Both ON and MS result from deterioration of axonal myelin sheaths, affecting cellular communication. Myelin sheaths allow for efficient saltatory conduction. ON can self-resolve with unspecific underlying mechanisms in 95% of cases, but most MS relapses and progresses without cure (1). Fig. 1 Stimulus dependent information proportional to the geometric means of the cells’ tuning curves with different maximum possible correlation coefficients. Obtained for cell pairs where both cells fire at their maximum possible rates, pmax. Stimulus-dependent correlations (A) depict difference in slopes of pairs of cells with different firing rates. Shows direct correlation between the number of cells and activity between the pair. Matched constant correlations (B) depict replaced correlation coefficients for each cell pair with the average over stimuli of that pair’s stimulus-dependent correlation. Shows an inverse relationship. Figure 2 B Myelinated Axon Unmyelinated Axon Action Potential (mV) Focus is typically placed on myelin thickness and internode length when assessing communication between neurons. Studies show that longer nodes of Ranvier decrease conduction speed, possibly affecting cognitive processes (2). Assessing the impact of factors including and besides myelin thickness may reveal other processes of symptom generation in MS and ON. A Action Potential (mV) https://www.firstvieweyecare.com/wpcontent/uploads/2011/05/?ND METHODS This project utilizes two computational models. Both models were slightly modified with restrictions to imitate conditions of ON with varying degrees of decreased axon myelin thickness. The first original computational model is derived from jZ Lab (4). It uses MATLAB to model mouse retina direction selective ganglion cells. The code computes the amount of information conveyed about the stimulus between control and manipulated cell models. The variable pmax, which represents the maximum possible firing rate of the neurons, is restricted in intervals of 0.2. The second original computational model is derived from ModelDB (3). This model utilizes NEURON to simulate myelinated CNS axons in mouse optic nerves using a double cable model. By representing small fibers, the simulation accurately represents saltatory conduction in CNS regions of interest. Time (ms) Time (ms) Fig. 2 Action potential (mV) depicted in axon. 0.6 µm diameter axon represented in both graphs through a double cable model (DCM). Control group axon (A) is compared to a partially-unmyelinated axon (B). Action potential generated by unmyelinated axon was significantly lower than myelinated axon as predicted. The concentrations of sodium and potassium inside and outside of the cell were manipulated to simulate a loss in myelin. DISCUSSION • Overactivity of Na+ channels occurs in ON, modeled by increasing the [Na+] in the cell • The codes compute the information values for both the “normal” responses of the model, closely matching our experimental observations from the mouse retina • 1A makes sense because as the number of cells increase, larger firing rates should show more stimulus information and communication between the cells • 1B makes sense because with stimulus-independent correlations, in stark contrast to stimulus-dependent ones, larger correlations correspond to weaker population codes