Structure And Function In Retinal Neurons

National Institute Of Neurological Disorders And Stroke

Investigators

Linked publications & trials

Abstract

In the eye, neural circuits process images. These circuits maintain image constancy over illuminations ranging from starlight to noontime sun. Rods evolved for nocturnal vision, and cones for diurnal vision. Separate interneurons process these rod or cone signals. Mammals such as cats, rabbits, and rodents, are reasonable models for human rod circuitry. In primates further cone types and cone circuits evolved for color vision, but in common laboratory mammals cone density is low, and color sense is weak. Zebrafish, like old-world primates, evolved color vision through a proliferation of spectrally tuned cone types. The difference is that old-world primates developed three cone types, but zebrafish employ eight. Nonetheless the circuitry strategies for processing chromatic information and controlling cone development may be similar. The ease of genetic manipulation in zebrafish is advantageous, and there are extensive libraries of visual-system mutants and transgenics. For these reasons, this lab and others have worked to develop zebrafish as a model for electrophysiological and neuroanatomical studies of photoreceptor and visual system development, circuitry, and function. The thyroxin beta 2 nuclear receptor (trb2) has been a recent focus. This gene is required for differentiation and development of red cones. Transgenic lines from the Rachel Wong lab (crx:mYFP-2A-trb2 and gnat2:mYFP-2A-trb2) are gain-of-function lines, either expressing trb2 in the wrong cone types (gnat2:trb2), or in retinal progenitors, prematurely in retinal development. For crx:trb2, the resultant density of differentiated larval cones immunoreactive for red opsin doubles, while immunoreactivity for cones with green, blue or UV opsin is greatly reduced. Concomitantly larval red cone signals greatly increase in amplitude, with the gene-duplicated LWS2 red opsin generating the signals, as also occurs in control larval eyes. Green, blue and UV opsin cones generate greatly diminished signal amplitudes in crx:trb2 larvae. For gnat:trb2 the density of red-opsin immunoreactive cones also doubles, but early larval signal amplitudes of red cones increase less than 10%, and the signal amplitudes of other cone types remains unchanged. The difference is that trb2 is inducing red opsin production in cones already synthesizing the native green, blue or UV opsins so that most cones become immunoreactive for two opsins, an original and a newcomer. With the introduction of red opsin this late in development, the process of transformation into a red cone takes longer, as original opsins are slowly removed by disk shedding. In UV cones, this turnover must occur, as activity reporters for the UV opsin gene are silenced in the gnat2:trb2 retinas. By adulthood in both transgenics only, or nearly only, red cone amplitudes are measured, unlike controls, which continue to express 5 to 6 different cone opsin signals. A further transformation occurs. Whereas control red cone amplitudes in adults are evenly split between amplitudes for LWS2 and LWS1 red opsins, LWS1 opsins amplitudes are larger in the transgenics. In larvae we have also noted that cone morphology is distorted by the trb2 gain-of function transgenes, bringing about wider inner segments and shorter axons than control cones. It appears that trb2 has a very wide range of actions in retinal development, in addition to being a requirement for red-cone specification. A trb2 -/- mutant line with no larval or adult red-cone signals, and enhanced UV-cone signals was established by Crispr corruption of the first-exon in the trb2 reading frame. We previously found this line lacked larval optomotor or optokinetic responses for both red/black and green/black color contrasts, and lacked ERG responses to long-wavelength stimuli. The arrestin 3a antigen, (zpr-1 antibody), normally expressed in red-green double cones, was lost.