Abstract
Purpose :
Automatic or semi-automatic estimation of photoreceptor density is a huge challenge for retinal mapping. However it is a necessary method to understand neurodegenerative diseases. Respective density of both intact cones and rods is invaluable information to track evolution of pathologies and to optimize ex-vivo and in-vivo pathological models. Here we propose to use a periciliary membrane's modelization to realize a non ambigous individual counting of rods and cones in the retina.
Methods :
Eyes, obtained from perfused and non-perfused macaques and pigs were dissected and fixed in paraformaldehyde (PFA). Retina then underwent either tissue-freezing medium included cryosection (30µm-thick sections) or were directly processed for immunofluorescence (whole mounted retinas.
Immunostaining was performed with a primary antibody mix directed against the following proteins: rhodopsin, opsin or PNA, protocadherin 15 (pcd15) and the very large G protein 1 (Vlgr1). Retinas were imaged using classic confocal microscope. 3D reconstruction was performed using Imaris software.
Results :
Examination of Vlgr1 showed a punctuate staining of the base of the outer segment. This punctate staining was clearly associated to pcd15 (a calyceal processes’ protein) and to either rhodopsin or cone opsin . This association prevents any mistake due to artefactual staining. Counting all the Vlgr1 puncta enabled us to define the number of photoreceptors while their association to either cone opsin or rhodopsin provided an adequate means to differentiate rod and cone photoreceptors. We were able to modelize periciliary membrane localization of all photoreceptors on 100µmx100µm confocal imagery fields. Depending on the tissue piece, rods were individually discriminated up to 170 000 cells/mm2, a density approximately corresponding to their maximum density in primates.
Conclusions :
The proposed technique enabled us precise and semi-automatized counting of photoreceptors even in their thinnest morphologies. Moreover, the outer/inner segment is highly susceptible to be lost during photoreceptor degeneration, its counting then becoming a physiological assessment of structurally intact photoreceptors in the tested retina. This method will provide an adequate retinal mapping for characterizing photoreceptor degeneration in animal models for the development of therapies.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.