Automated quantification of photoreceptor outer segments in developing and degenerating retinas on microscopy images across scales

Suse Seidemann, Florian Salomon, Karl B. Hoffmann, Thomas Kurth, Ivo F. Sbalzarini, Robert Haase, Marius AderFrontiers in Molecular Neuroscience 2024 · DOI

Abstract

The functionality of photoreceptors depends on their outer segments (POS), which contain organized membranous structures for light signal transduction. While POS health is often assessed qualitatively, quantitative methods are needed for objective analysis. This paper introduces QuaPOS, a suite of automated tools for quantifying POS maturation and quality using light microscopy (LM) and transmission electron microscopy (TEM). By applying machine learning and gradient-based analysis, the researchers successfully tracked POS development in wild-type mice and degeneration in various disease models.

Introduction: The Need for Objective Metrics

Photoreceptors (rods and cones) are essential for vision, relying on their outer segments (POS) to convert light into electrical signals. Structural defects or degeneration in these segments lead to vision loss in conditions like Retinitis Pigmentosa. Historically, assessing POS health on microscopy slides has relied on subjective "eye-balling" or labor-intensive manual counting. This lack of standardization makes it difficult to compare results across studies or handle large-scale datasets.

Methodology: Bridging Scales with QuaPOS

The researchers developed two distinct workflows to provide a comprehensive view of POS health:

  1. QuaPOS-LM (Light Microscopy): Utilizes a Random Forest classifier to segment cone POS in fluorescently stained retinal sections (S-opsin marker). It automatically extracts features such as volume, height, and shape.
  2. QuaPOS-TEM (Transmission Electron Microscopy): Analyzes the nanometer-scale alignment of membrane discs using image intensity gradients. It calculates "coherency," a mathematical metric for structural order, distinguishing between "chaotic" and "perfectly stacked" membranes.
Overview of the QuaPOS methodology and developmental imaging data across light and electron microscopy scales.

The tools were validated using wild-type (WT) mice to establish developmental benchmarks and three mutant lines (Cpfl1, RhoKO, and rd19) to test sensitivity to retinal degeneration.

Key Results and Insights

The automated analysis revealed that POS maturation involves both physical growth and increasing internal organization.

Developmental Growth

In wild-type mice, the tools recorded an 8-fold increase in summed POS volume between postnatal days P8 and P24. The internal membrane alignment (local coherency) also increased significantly, marking the transition from unorganized precursors to functional light-sensing stacks.

Quantitative analysis showing the increase in POS bounding box height during postnatal development (P8 to P24).

Quantifying Degeneration

In disease models, QuaPOS detected significant drops in all health metrics. For example, in the RhoKO model, global coherency (a measure of overall membrane order) was nearly zero (0.07) compared to 0.52 in healthy controls. In the Cpfl1 model, while the total number of cones decreased, the remaining segments often appeared structurally normal, suggesting an "all-or-nothing" failure mode at the individual segment level.

Conclusion and Future Directions

QuaPOS provides a robust, open-source framework for the objective quantification of retinal health. While the current LM tool faces challenges with densely packed "clusters" and the TEM tool requires manual region-of-interest selection, these methods represent a significant leap over qualitative assessment. Future work may incorporate advanced instance segmentation (like StarDist) and super-resolution microscopy to further refine these metrics, particularly for the more numerous and thinner rod photoreceptors.

Summary updated

Abstract

The functionality of photoreceptors, rods, and cones is highly dependent on their outer segments (POS), a cellular compartment containing highly organized membranous structures that generate biochemical signals from incident light. While POS formation and degeneration are qualitatively assessed on microscopy images, reliable methodology for quantitative analyses is still limited. Here, we developed methods to quantify POS (QuaPOS) maturation and quality on retinal sections using automated image analyses. POS formation was examined during the development and in adulthood of wild-type mice via light microscopy (LM) and transmission electron microscopy (TEM). To quantify the number, size, shape, and fluorescence intensity of POS, retinal cryosections were immunostained for the cone POS marker S-opsin. Fluorescence images were used to train the robust classifier QuaPOS-LM based on supervised machine learning for automated image segmentation. Characteristic features of segmentation results were extracted to quantify the maturation of cone POS. Subsequently, this quantification method was applied to characterize POS degeneration in "cone photoreceptor function loss 1" mice. TEM images were used to establish the ultrastructural quantification method QuaPOS-TEM for the alignment of POS membranes. Images were analyzed using a custom-written MATLAB code to extract the orientation of membranes from the image gradient and their alignment (coherency). This analysis was used to quantify the POS morphology of wild-type and two inherited retinal degeneration ("retinal degeneration 19" and "rhodopsin knock-out") mouse lines. Both automated analysis technologies provided robust characterization and quantification of POS based on LM or TEM images. Automated image segmentation by the classifier QuaPOS-LM and analysis of the orientation of membrane stacks by QuaPOS-TEM using fluorescent or TEM images allowed quantitative evaluation of POS formation and quality. The assessments showed an increase in POS number, volume, and membrane coherency during wild-type postnatal development, while a decrease in all three observables was detected in different retinal degeneration mouse models. All the code used for the presented analysis is open source, including example datasets to reproduce the findings. Hence, the QuaPOS quantification methods are useful for in-depth characterization of POS on retinal sections in developmental studies, for disease modeling, or after therapeutic interventions affecting photoreceptors.
BibTeX
@inproceedings{seidemann2024automated,
  author = {Suse Seidemann and Florian Salomon and Karl B. Hoffmann and Thomas Kurth and Ivo F. Sbalzarini and Robert Haase and Marius Ader},
  title = {{Automated quantification of photoreceptor outer segments in developing and degenerating retinas on microscopy images across scales}},
  booktitle = {Frontiers in Molecular Neuroscience},
  year = {2024},
  doi = {10.3389/fnmol.2024.1398447},
}
Figure