Abstract
Purpose :
Accurate quantification of retinal ganglion cells (RGCs) is crucial in studying many ocular diseases. Widely used pan-markers such as THY1 and RBPMS all have limitations when using automated counting programs. The goal of this study is to develop a nuclear stain that labels all RGCs, optimizing RGC staining for automated cell counting.
Methods :
Retinas from 21 C57BL/6J mice were immunostained with primary antibodies: anti-Brn3a (n=16), anti-POU6F2 (n=8) and anti-RBPMS (n=12). Five of these retinas were specifically used for quantitative analysis, undergoing triple staining with all three antibodies across different channels. ImageJ was used to merge the Brn3a and POU6F2 channels, creating a pseudocolor pan-RGC nuclear stain. To validate the efficacy of this combined nuclear staining approach, it was compared to the pan-RGC marker RBPMS, using the automated cell counting program RGCode.
Results :
RBPMS labels all RGC subtypes in the mouse retina; however, the cytoplasmic staining makes feature recognition difficult for automated counting programs. Brn3a labels RGC nuclei making the feature easy to distinguish and count. Unfortunately, Brn3a does not label all RGC subtypes. POU6F2 labels a group of RGC in the mouse retina, all of which are RBPMS immunoreactive and including all of the Brn3a negative RGCs. When retinas are stained with a combination of Brn3a and POU6F2, all RBPMS-positive RGCs have labeled nuclei. We hypothesize that the combination of Brn3a and POU6F2 can be used as a pan-RGC nuclear marker. To test this hypothesis, we used automated counting to quantify Brn3a, RBPMS and pseudocolor Brn3a with POU6F2 all in the same 5 retinas. Brn3a alone labels 44852 ± 778 RGCs/retina. RBPMS labels 49023 ± 939 RGCs/retina, significantly more than Brn3a (p=0.016, Mann-Whitney U Test). The highest number of counted RGCs occurred when merging the Brn3a and POU6F2 channels. The combination of Brn3a and POU6F2 labeled 52757 ± 1068 RGCs/retina, significantly higher than RBPMS (p=0.047). This difference was due to an underestimate of RBPMS-positive RGCs with adjacent pairs of cells being counted as a single cell.
Conclusions :
Combining Brn3a and POU6F2 antibodies effectively label the nuclei of all RGC subtypes, enabling a thorough and accurate assessment of RGC populations through automated counting methodologies.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.