Adding color perception: A wide-field color LED multifocal electroretinogram and a neural circuit model for coding color

James A Kuchenbecker, Marquette University

Abstract

Recently, gene therapy has shown potential in treating color blindness through the insertion of new long- (L-) wavelength spectral sensitivity into protanopic monkey and gerbil retinas. The opsin was inserted into an existing subset of the middle- (M-) wavelength sensitive cone photoreceptors via an adeno-associated virus. Knowing the retinal area transfected by the virus in vivo is an important aspect to the project because an important goal is to cover a portion of the fovea. To this end, a new wide-field color LED-based multifocal-electroretinogram (mf-ERG) was developed and validated for specific aim I. The new mf-ERG can isolate responses from L-cones using a 650 nm wavelength light, improves existing mf-ERG technology by measuring over 70 degrees of visual angle (clinical mf-ERGs stimulate over 30 degrees), and maintains viable SNR out in the far periphery. Evidence from messenger RNA data from human cadavers has shown the ratio of L:M cone photopigment increases from fovea to far periphery. This result has been difficult to corroborate in vivo because current imaging techniques cannot gauge cone-photoreceptor type beyond ∼eight degrees. The mf-ERG developed in specific aim I was used in specific aim II to gauge sensitivity of green light versus red light as a function of eccentricity among color-normal individuals. The goal was to corroborate the in vitro result. In the vision community there is a growing body of literature showing high variation in cone photoreceptor mosaics. Despite these differences individuals show no difference in the ability to discriminate colors when tested on a battery of color vision tests. Additionally, the expression of a new spectral photopigment in dichromatic retinas has shown, behaviorally, the animal is immediately able to take advantage of the new photopigment to discriminate colors. Specific aim III encompasses developing a mathematical model that can, using the spectral sensitivities of the S-, M-, and L-cones, extract the six fundamental percepts of whiteness, blackness, greenness, redness, yellowness, and blueness while being constrained by known anatomy and experimental findings listed above. Developing such a model may elucidate underlying circuitry of the retina.

This paper has been withdrawn.