Purpose: The pupillary light reflex (PLR) is critical for proper visual function, regulating the amount of light entering the eye. It has been thought that the melanopsin-expressing, intrinsically photosensitive retinal ganglion cells (ipRGCs) drive the PLR via activation of brain circuits involving the olivary pretectal nucleus (OPN) and ultimate release of acetylcholine from parasympathetic fibers in the iris muscle. However, it was recently reported that melanopsin is capable of mediating the PLR in isolation from the brain. We therefore investigated the relative contributions of these two mechanisms in driving the PLR.

Methods: We utilized neuronal injury models, anatomical tracing of neuronal projections, and genetic and pharmacological silencing of cholinergic signaling to examine the relative contribution of retinal, central, and direct inputs to the PLR.

Results: We identified a new pathway by which ipRGCs drive the PLR via a direct projection to the iris. Using genetic tracing methods, we identified ipRGC axons innervating both the ciliary body and iris muscle. To determine whether this projection is functional, we then performed optic nerve injury to completely isolate the eye from brain circuitry. In the initial days following injury, we observed constriction of the pupil that was independent of brain circuitry. Interestingly, this constriction was completely absent 5 weeks following optic nerve injury, coincident with the death of ganglion cells in this injury paradigm, and indicating the RGCs are required for this brain-independent PLR. We further show that cholinergic signaling is required for this reflex, because silencing cholinergic signaling pharmacologically or genetically abolishes the PLR. We next examined the ipRGC subtypes involved in this reflex. Ablation of the M1 subtype of ipRGC results in loss of both consensual and ipsilateral PLR, indicating that ipRGCs are in fact required for this reflex. Interestingly, mice lacking ipRGCs that project to the OPN, but retaining a subset of M1 ipRGCs, lack a consensual PLR, but retain normal ipsilateral PLR. This indicates that the centrally-mediated and brain-independent PLR are driven by distinct populations of ipRGC.

Conclusions: We have identified a novel mechanism by which ipRGCs drive the PLR via a direct projection to the iris muscle. This pathway combines with the centrally-mediated PLR to drive the ipsilateral PLR through modulation of cholinergic signaling.