Abstract
Purpose :
Retinal nerve fiber layer (RNFL) reflectance is often assessed in clinical diagnosis of glaucoma. This study investigated the relationships between RNFL reflectance and axonal cytoskeleton, including microtubules (MTs), F-actin and neurofilaments (NFs).
Methods :
Whole-mounted rat retinas were imaged at 500 nm by means of imaging reflectometry. Reflectance (R) of nerve fiber bundles was measured in normal retinas and retinas treated with colchicine. After the measurements, retinas were fixed for immunohistological staining. Confocal images were used to measure bundle thickness (T) and count the strings of MTs, F-actin and NFs within bundles (Fig. 1A-C). Reflectance per unit thickness (σ = R/T) and string density (d) were calculated.
Results :
In normal RNFL, σ was correlated with dMT (r = 0.85) and dF-actin (r = 0.53) but not dNF (r = 0.03) (Fig. 1D-F). By assuming incoherent scattering of cytostructure, linear regression was used to describe the relationship between d and σ. The σ-intercept for dMT vs. σ was 0.046 (%/μm) which was 66% of the mean σ of normal bundles. The σ-intercept for dF-actin vs. σ was 0.055 (%/μm) or 81% of the mean. In colchicine treated retinas σ, dMT and dF-actin decreased significantly (p < 0.001); however, dNF did not change (p = 0.08). The mean σ was reduced to 0.043 (%/μm) which was a 38% reduction from the normal σ. With the linear model for dMT (Fig. 1D), the reduction was estimated to contain a 22% reduction due to loss of MTs (solid black circle) and 16% due to change of scattering components other than MTs. For F-actin (Fig. 1E), the reduction was estimated to contain a 6% reduction due to loss of F-actin and 32% due to loss of MTs and changes of other unknown components.
Conclusions :
Colchicine, a MT depolymerizing agent, also damages F-actin and other unknown structures. In normal retinas the contributions of MTs and F-actin to RNFL reflectance are not more than 34% and 19%, respectively. There are other structures contributing significantly to RNFL reflectance.
This is a 2021 ARVO Annual Meeting abstract.