Demyelination and shrinkage of axons in the retinal nerve fiber layer in chickens developing deprivation myopia

Placing diffusers in front of the eyes induces deprivation myopia in a variety of animal models. As a result of the low pass filtering of the retinal images, less spatial information is available to the retina which should reduce neural activity. Since it has been found that myelination of axons in the central nervous system is modulated by neuronal activity, we have studied whether ganglion cell axons may shrink in response to the restricted visual input. Young chickens were treated for 5 h or 7 days with frosted diffusers to induce deprivation myopia. Nerve fiber layer thickness was measured in vivo, using B-scan OCT. Refractive states were tracked by IR photoretinoscopy, and UV fundus reflectivity by a custom-built device which flashed an LED centered in the camera aperture and recorded pupil brightness after refractive errors were corrected by trial lenses. Moreover, structure and histology of the retinal nerve fibers layer (RNFL) were analyzed ex vivo using transmission electron microscopy and immunohistochemistry. Since chicks have both non-myelinated and myelinated fibers in their RNFL, the thickness of myelin sheaths (G ratio) was measured, as well as the percentage of myelinated axons and the diameters of unmyelinated axons. Short-term deprivation caused an increase in UV fundus reflectivity already after 5 h (measured as pixel grey levels in the pupil: 28 ± 5 vs. 36 ± 10, p < 0.05) and thinning of the myelin sheaths (higher G ratio), compared to untreated control eyes (0.74 ± 0.01 vs. 0.79 ± 0.03, p < 0.05). Neither axon diameters (0.81 ± 0.05 μm vs. 0.82 ± 0.15 μm) nor thickness of the RNFL had changed after only 5 h (42.9 ± 1.3 μm vs. 42.3 ± 2.5 μm). However, after 7 days of diffuser wear, axons had become thinner (0.56 ± 0.14 μm vs. 0.78 ± 0.09 μm vs, p < 0.05), which could explain the thinning of the RNFL (36.3 ± 2.7 μm vs. 42.1 ± 2.4 μm, p < 0.01). Furthermore, myopic eyes had 38% less myelinated axons than untreated eyes as determined by immunohistochemical labelling against myelin basic protein (immunopositive areas in the central retina 1406 ± 341 μm² vs. 2185 ± 290 μm² in controls, p < 0.001). Myelin sheaths in the remaining axons remained unchanged (G ratio 0.76 ± 0.02 vs. 0.76 ± 0.03). Our study shows that deprivation myopia is associated with a significant loss of myelinated axons and shrinkage of the axon diameters of certain fibers in the RNFL. Early changes were already detected after 5 h and were accompanied by an increased fundus reflectivity in UV light. These parameters could therefore serve as the biomarkers for myopia development, at least in the chicken.