Biological Correlations and Confounders for Quantification of Retinal Ganglion Cells by Optical Coherence Tomography Based on Studies of Outbred Mice.

Purpose: Despite popularity of optical coherence tomography (OCT) in glaucoma studies, it's unclear how well OCT-derived metrics compare to traditional measures of retinal ganglion cell (RGC) abundance. Here, Diversity Outbred (J:DO) mice are used to directly compare ganglion cell complex (GCC) thickness measured by OCT to metrics of retinal anatomy measured ex vivo with retinal wholemounts and optic nerve histology. Methods: J:DO mice (n = 48) underwent fundoscopic and OCT examinations, with automated segmentation of GCC thickness. RGC axons were quantified from para-phenylenediamine-stained optic nerve cross-sections and somas from BRN3A-immunolabeled retinal wholemounts, with total inner retinal cellularity assessed by TO-PRO and subsequent hematoxylin staining. Results: J:DO tissues lacked overt disease. GCC thickness, RGC abundance, and total cell abundance varied broadly across individuals. GCC thickness correlated significantly to RGC somal density (r = 0.58) and axon number (r = 0.44), but not total cell density. Retinal area and nerve cross-sectional area varied widely. No metrics were significantly influenced by sex. In bilateral comparisons, GCC thickness (r = 0.95), axon (r = 0.72), and total cell density (r = 0.47) correlated significantly within individuals. Conclusions: Amongst outbred mice, OCT-derived measurements of GCC thickness correlate significantly to RGC somal and axon abundance. Factors limiting correlation are likely both biological and methodological, including differences in retinal area that distort sampling-based estimates of RGC abundance. Translational Relevance: There are significant-but imperfect-correlations between GCC thickness and RGC abundance across genetic contexts in mice, highlighting valid uses and ongoing challenges for meaningful use of OCT-derived metrics.

AxonDeep: Automated Optic Nerve Axon Segmentation in Mice With Deep Learning.

Purpose: Optic nerve damage is the principal feature of glaucoma and contributes to vision loss in many diseases. In animal models, nerve health has traditionally been assessed by human experts that grade damage qualitatively or manually quantify axons from sampling limited areas from histologic cross sections of nerve. Both approaches are prone to variability and are time consuming. First-generation automated approaches have begun to emerge, but all have significant shortcomings. Here, we seek improvements through use of deep-learning approaches for segmenting and quantifying axons from cross-sections of mouse optic nerve. Methods: Two deep-learning approaches were developed and evaluated: (1) a traditional supervised approach using a fully convolutional network trained with only labeled data and (2) a semisupervised approach trained with both labeled and unlabeled data using a generative-adversarial-network framework. Results: From comparisons with an independent test set of images with manually marked axon centers and boundaries, both deep-learning approaches outperformed an existing baseline automated approach and similarly to two independent experts. Performance of the semisupervised approach was superior and implemented into AxonDeep. Conclusions: AxonDeep performs automated quantification and segmentation of axons from healthy-appearing nerves and those with mild to moderate degrees of damage, similar to that of experts without the variability and constraints associated with manual performance. Translational Relevance: Use of deep learning for axon quantification provides rapid, objective, and higher throughput analysis of optic nerve that would otherwise not be possible.

Blast Preconditioning Protects Retinal Ganglion Cells and Reveals Targets for Prevention of Neurodegeneration Following Blast-Mediated Traumatic Brian Injury.

Purpose: The purpose of this study was to examine the effect of multiple blast exposures and blast preconditioning on the structure and function of retinal ganglion cells (RGCs), to identify molecular pathways that contribute to RGC loss, and to evaluate the role of kynurenine-3-monooxygenase (KMO) inhibition on RGC structure and function. Methods: Mice were subjected to sham blast injury, one single blast injury, or three blast injuries separated by either 1 hour or 1 week, using a blast intensity of 20 PSI. To examine the effect of blast preconditioning, mice were subjected to sham blast injury, one single 20-PSI injury, or three blast injuries separated by 1 week (5 PSI, 5 PSI, 20 PSI and 5 PSI, 5 PSI, 5 PSI). RGC structure was analyzed by optical coherence tomography (OCT) and function was analyzed by the pattern electroretinogram (PERG). BRN3A-positive cells were quantified to determine RGC density. RNA-seq analysis was used to identify transcriptional changes between groups. Results: Analysis of mice with multiple blast exposures of 20 PSI revealed no significant differences compared to one 20-pounds per square inch (PSI) exposure using OCT, PERG, or BRN3A cell counts. Analysis of mice exposed to two preconditioning 5-PSI blasts prior to one 20-PSI blast showed preservation of RGC structure and function. RNA-seq analysis of the retina identified multiple transcriptomic changes between conditions. Pharmacologic inhibition of KMO preserved RGC responses compared to vehicle-treated mice. Conclusions: Preconditioning protects RGC from blast injury. Protective effects appear to involve changes in KMO activity, whose inhibition is also protective.

Mouse models and strain-dependency of Chédiak-Higashi syndrome-associated neurologic dysfunction.

Chédiak-Higashi syndrome (CHS) is a lethal disorder caused by mutations in the LYST gene that involves progressive neurologic dysfunction. Lyst-mutant mice exhibit neurologic phenotypes that are sensitive to genetic background. On the DBA/2J-, but not on the C57BL/6J-background, Lyst-mutant mice exhibit overt tremor phenotypes associated with loss of cerebellar Purkinje cells. Here, we tested whether assays for ataxia could measure this observed strain-dependency, and if so, establish parameters for empowering phenotype- and candidate-driven approaches to identify genetic modifier(s). A composite phenotypic scoring system distinguished phenotypes in Lyst-mutants and uncovered a previously unrecognized background difference between wild-type C57BL/6J and DBA/2J mice. Accelerating rotarod performance also distinguished phenotypes in Lyst-mutants, but at more advanced ages. These results establish that genetic background, Lyst genotype, and age significantly influence the severity of CHS-associated neurologic deficits. Purkinje cell quantifications likewise distinguished phenotypes of Lyst-mutant mice, as well as background differences between wild-type C57BL/6J and DBA/2J mice. To aid identification of potential genetic modifier genes causing these effects, we searched public datasets for cerebellar-expressed genes that are differentially expressed and/or contain potentially detrimental genetic variants. From these approaches, Nos1, Prdx2, Cbln3, Gnb1, Pttg1 were confirmed to be differentially expressed and leading candidates.