Dynamic changes in cell size and corresponding cell fate after optic nerve injury

Abstract

Identifying disease-specific patterns of retinal cell loss in pathological conditions has been highlighted by the emergence of techniques such as Detection of Apoptotic Retinal Cells and Adaptive Optics confocal Scanning Laser Ophthalmoscopy which have enabled single-cell visualisation in vivo. Cell size has previously been used to stratify Retinal Ganglion Cell (RGC) populations in histological samples of optic neuropathies, and early work in this field suggested that larger RGCs are more susceptible to early loss than smaller RGCs. More recently, however, it has been proposed that RGC soma and axon size may be dynamic and change in response to injury. To address this unresolved controversy, we applied recent advances in maximising information extraction from RGC populations in retinal whole mounts to evaluate the changes in RGC size distribution over time, using three well-established rodent models of optic nerve injury. In contrast to previous studies based on sampling approaches, we examined the whole Brn3a-positive RGC population at multiple time points over the natural history of these models. The morphology of over 4 million RGCs was thus assessed to glean novel insights from this dataset. RGC subpopulations were found to both increase and decrease in size over time, supporting the notion that RGC cell size is dynamic in response to injury. However, this study presents compelling evidence that smaller RGCs are lost more rapidly than larger RGCs despite the dynamism. Finally, using a bootstrap approach, the data strongly suggests that disease-associated changes in RGC spatial distribution and morphology could have potential as novel diagnostic indicators.

Figure 1

RBPMS and Brn3a colabelling of murine retinal ganglion cells in naive retina and after induction of optic nerve injury, using an established optic nerve crush model. (A) Illustration of the protocol for automatically measuring RGC cytoplasmic (RBPMS) and nuclear (Brn3a) area as described in the text. (B) Plots of Brn3a labelled area versus RBPMS labelled area in each retina with linear regression fit shown in red. Matching between Brn3a and RBPMS areas was achieved by calculating the cross-nearest neighbour distance between populations as described in the text.

Figure 2

The distribution of RGC nucleus area changes over the course of the OHT (A, C) and pONT (B, D) models. (A) Average RGC nucleus area does not significantly change over the course of the OHT model. (B) However, a significant increase in nuclear area is observed 3 days after pONT induction, followed by a reduction in area from 7 days post pONT (One-way ANOVA, Tukey post-hoc test, *p < 0.05, ***p < 0.001). Data points represent individual animals, to illustrate inter-animal variability. (C, D) Overall, the RGC density declines and the distribution of RGC density by nucleus area shifts to the right over the course of the (C) OHT model and (D) pONT models, although this trend is more subtle in the OHT model.

Figure 3

(A–D) Assessment of RGC nucleus size probability distribution functions (i), and difference versus control condition (ii), over the natural history of the OHT model. (E–H) Assessment of RGC nucleus size probability distribution functions in the pONT model (i), and difference versus control condition (ii), over the natural history of the pONT model. Mean ± SE.

Figure 4

The size dependence of RGC loss in the pONT model is related to susceptibility to primary and secondary neurodegeneration. (Ai–Ei) PDF of RGC soma area from paired superior vs inferior retinal quadrants at each timepoint. (Aii–Eii) Difference between superior and interior PDFs from (Ai–Ei). At first, retinal neurodegeneration is comparable in both superior and inferior retinal quadrant regions. As the model progresses, however, a greater proportion of smaller RGCs are observed in the superior than inferior retina, suggestive that RGCs becoming smaller precedes primary degeneration or that larger RGCs are more susceptible to secondary degeneration. No such trend was observed in the OHT model (data not shown), most likely due to the milder, more chronic nature of this model. (F) Schematic representation of retinal regions more prone to primary vs secondary neurodegeneration in the pONT model.

Figure 5

Relationship between nucleus area and rate of loss (half-life) over the course of model. RGC densities binned by nucleus area were plotted over the natural history of the OHT and pONT models before fitting to exponential decay equation (A). In both (B) OHT and (C) pONT models, smaller RGCs typically have shorter half-lives than larger cells, suggesting preferential loss of smaller RGCs from the retina.

Figure 6

Determination of the diagnostic utility of subsampling the RGC population in the OHT model using a Spatial Bootstrapping versus naive controls. (A–D) Difference between OHT and Naive retina randomly subsampled 100 000 times as described in the text and corresponding ROC curves (E–H). Results compared include (A, E) Nearest Neighbour Distance (NND), (B, F) Regularity Index (RI), (C, G) Mean RGC size and Mean absolute deviation in RGC size (D, H). A table with the median and 95% confidence intervals for the bootstrapping results has been added in supplement (Suppl. Table 2).

Figure 7

Determination of the diagnostic utility of subsampling the RGC population in the pONT model using a Spatial Bootstrapping versus naive controls. (A–D) Difference between pONT and Naive retina randomly subsampled 100 000 times as described in the text and corresponding ROC curves (E–H). Results compared include (A, E) Nearest Neighbour Distance (NND), (B, F) Regularity Index (RI), (C, G) Mean RGC size and Mean absolute deviation in RGC size (D, H). A table with the median and 95% confidence intervals for the bootstrapping results has been added in supplement (Suppl. Table 2).