The Retinal Basis of Light Aversion in Neonatal Mice

Franklin Caval-Holme, Marcos L. Aranda, Andy Quaen Chen, Alexandre Tiriac, Yizhen Zhang, Benjamin E. Smith, Lutz Birnbaumer, Tiffany M. Schmidt, Marla B. FellerbioRxiv 2022 · DOI

TL;DR

Neonatal mice exhibit an innate aversion to bright light, a behavior crucial for survival before vision develops. This study identifies a specific type of retinal neuron, M1 ipRGCs, as the primary driver of this photoaversion, acting as a dedicated light-sensing channel that avoids interference from other developing neural activity.

What Problem Does This Paper Address?

Developing mammals, including mice, show an aversion to bright light (photoaversion) even before their visual system is fully mature. This behavior is thought to be mediated by intrinsically photosensitive retinal ganglion cells (ipRGCs), which contain the photopigment melanopsin. However, ipRGCs are not a single cell type; they comprise multiple subtypes (M1-M6 in mice) with distinct properties. The key question is: which of these ipRGC subtypes are responsible for signaling this aversive light response in the developing brain, and how do they achieve this reliably amidst the noisy activity of a developing retina?

Methods

The researchers employed a multi-pronged approach:

  • Behavioral Assays: They measured photoaversion in neonatal mice (postnatal day 8) by observing their head turns away from a light source in a controlled chamber. This was done using automated tracking of a reflective sticker on the mouse's head.
  • Genetic Manipulations: They used genetically modified mice:
  • TRPC6/7 Knockout (KO) Mice: These mice lack specific ion channels (TRPC6 and TRPC7) that are known to be involved in M1 ipRGC phototransduction.
  • Opn4Cre Brn3bDTA Mice: These mice have had all ipRGC types ablated except for a specific subset of M1 ipRGCs that lack the Brn3b transcription factor.
  • Two-Photon Calcium Imaging: This technique allowed them to visualize the light-evoked activity (calcium transients) of individual neurons in ex vivo retinal slices. This helped them assess the photosensitivity of different ipRGC subtypes.
  • Pharmacological Interventions: They used drugs to block gap junction coupling (Meclofenamic Acid - MFA) or modulate dopamine signaling (SCH23390), which are known to affect the light sensitivity of non-M1 ipRGCs.
  • Retinal Wave Analysis: They used calcium imaging to observe how ipRGCs responded to spontaneous retinal waves, a major source of neural activity in the developing retina.

Datasets

Not explicitly discussed as training/evaluation datasets in the context of machine learning. The study generated and analyzed novel experimental data from mouse models.

Key Insights and Ideas

  • M1 ipRGCs are the primary drivers of neonatal photoaversion: This is a major finding, pinpointing a specific cell type responsible for a critical developmental behavior.
  • TRPC6/7 channels are crucial for M1 ipRGC-mediated photoaversion: The absence of these channels severely impairs the photoaversion response, directly linking molecular machinery to behavior.
  • M1 ipRGCs are largely insulated from retinal waves: This is a critical insight for understanding how the developing brain can reliably detect light signals. Retinal waves are a source of "noise" that could otherwise mask light-evoked signals. M1s' weak response to these waves allows them to act as a dedicated "clean" channel for light information.
  • Photoaversion is largely independent of non-M1 ipRGCs: Manipulating the function of other ipRGC subtypes (M2-M6) had minimal impact on photoaversion, further emphasizing the role of M1s.

Key Results

  • TRPC6/7 KO mice showed a dramatically reduced photoaversion response to threshold light intensities, failing to turn away from the light.
  • Two-photon imaging revealed reduced photosensitivity in M1 ipRGCs of TRPC6/7 KO mice, while other ipRGC types were largely unaffected.
  • Mice with only M1 ipRGCs (Brn3b-negative) exhibited normal photoaversion, demonstrating the sufficiency of M1s for this behavior.
  • Pharmacological blockade or genetic deletion of gap junction channels, which primarily affect M2-M6 ipRGCs, had only minor effects on photoaversion, and only at the brightest light intensities.
  • M1 ipRGCs were only weakly depolarized by spontaneous retinal waves, unlike other ipRGC types which were strongly affected. This suggests a specialized role for M1s in distinguishing light signals from background neural activity.

Limitations and Future Work

  • Contribution of M2-M6 at saturating intensities: While M1s are the primary drivers, the paper acknowledges that M2-M6 might contribute to photoaversion at very high light intensities, though this is not definitively proven.
  • Connexin expression in development: The study characterized connexin expression in developing ipRGCs, finding Cx45 to be prominent. While Cx45 KO mice showed some effects on M4 ipRGCs and a slight reduction in photoaversion at saturating intensities, the full implications for neonatal photoaversion require further investigation.
  • Future Directions: The authors suggest that M1 ipRGCs form a distinct information channel that is crucial for mediating photoaversion during early development. Future work could explore how this M1-driven signal is processed in the brain to elicit the aversive behavior and how this pathway contributes to the refinement of visual circuits. The role of M1s in other non-image-forming visual functions also warrants further investigation.

Summary updated

Abstract

SUMMARY Aversive responses to bright light (photoaversion) require signaling from the eye to the brain. Melanopsin-expressing intrinsically photosensitive retinal ganglion cells (ipRGCs) encode absolute light intensity and are thought to provide the light signals for photoaversion. Consistent with this, neonatal mice exhibit photoaversion prior to the developmental onset of image vision, and melanopsin deletion abolishes photoaversion in neonates. It is not well understood how the population of ipRGCs, which constitutes multiple physiologically distinct types (denoted M1-M6 in mouse) encodes light stimuli to produce an aversive response. Here, we provide several lines of evidence that M1 ipRGCs that lack the Brn3b transcription factor drive photoaversion in neonatal mice. First, neonatal mice lacking TRPC6 and TRPC7 ion channels failed to turn away from bright light, while two photon Ca 2+ imaging of their acutely isolated retinas revealed reduced photosensitivity in M1 ipRGCs, but not other ipRGC types. Second, mice in which all ipRGC types except for Brn3b-negative M1 ipRGCs are ablated, exhibited normal photoaversion. Third, pharmacological blockade or genetic knockout of gap junction channels expressed by ipRGCs, which reduces the light sensitivity of M2-M6 ipRGCs in the neonatal retina, had small effects on photoaversion only at the brightest light intensities. Finally, M1s were not strongly depolarized by spontaneous retinal waves, a robust source of activity in the developing retina that depolarizes all other ipRGC types. M1s therefore constitute a separate information channel between the neonatal retina and brain that could ensure behavioral responses to light but not spontaneous retinal waves.
BibTeX
@misc{cavalholme2022retinal,
  author = {Franklin Caval-Holme and Marcos L. Aranda and Andy Quaen Chen and Alexandre Tiriac and Yizhen Zhang and Benjamin E. Smith and Lutz Birnbaumer and Tiffany M. Schmidt and Marla B. Feller},
  title = {{The Retinal Basis of Light Aversion in Neonatal Mice}},
  year = {2022},
  doi = {10.1101/2022.01.20.477130},
  howpublished = {bioRxiv},
}
Figure