How neuronal processes develop and establish proper wirings with their synaptic partners is one of the fundamental questions of neuroscience. The vertebrate retina is an outstanding model system for studying dendritic development and neuronal connections. One of the critical visual functions, color vision, requires precise wiring of retinal neurons. In the mouse retina, there are two types of cone photoreceptors, the short wavelength sensitive cones (S-cones), which only express S-opsin, and the long wavelength sensitive cones (M-cones), many of which co-express S-opsin. In order to generate color opponency, signals from these two types of cones have to be segregated before they are contrasted at the ganglion cell level. S-cones only account for 2-5% of the total cone population. Thus, the downstream S-cone bipolar cells (SCBCs) face the daunting task of seeking out very sparse S-cones from a majority of M-cones. The outcome is that SCBCs develop a very unique dendritic arbor with long, meager dendrites that contact a handful of S-cones. This distinctive connection between S-cones and SCBCs makes it an excellent model system to study how presynaptic neurons affect the dendritic development and synaptic targeting of postsynaptic neurons. SCBs are labeled in a transgenic mouse line expressing Clomeleon (Clm) driven by thy1 promoter. Two mouse models were used to change the density and type of cones. In Thrb2-/- mice which lack thyroid hormone receptor 2 (TR2), M-opsin expression is abolished and all M-cones are turned into S-cones32. In S-opsin-/- mice, the S-opsin gene is knocked out. We crossed two lines and generated a double knockout (DKO) with neither S- nor M-opsin expression. We found that the numbers of SCBs in Thrb2-/-, S-opsin-/- and DKO mice are similar to those in wildtype. Morphologically, SCBs in Thrb2-/- and S-opsin-/- mice were indistinguishable from those in wildtype in terms of number of dendritic branches and cone contacts. SCBs in these mice appear to target specific true S-cones, even though the type of opsin expression is identical across all cones. Our results suggest that 1) dendritic development of SCBs appears to be intrinsic;2) S-cone identity may be specified by factors other than S-opsin expression a hypothesis that we will investigate in the future. In order to compare the transcriptome of S- and M-cones, we need to isolate true S-cones. An existing transgenic mouse line that expresses EGFP driven by an S-opsin promoter is not reliable in that some S-opsin negative cones also express GFP. This is likely due to the short S-opsin promoter region used in generating the mouse line. Due to low percentage of true S-cones (5%), this contamination from M-cones will be ruinous for RNAseq. We thus are currently working on generating our own mouse line using the Bacterial Artificial Chromosome (BAC) recombination approach to include longer upstream and downstream regulatory sequences at the endogenous S-opsin gene locus. We are sceening mouse lines that express yellow fluorescent protein YFP (Venus) only in S-opsin expressing cones.

National Institute of Health (NIH)
National Eye Institute (NEI)
Investigator-Initiated Intramural Research Projects (ZIA)
Project #
Application #
Study Section
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
U.S. National Eye Institute
Zip Code
Kim, Jung-Woong; Yang, Hyun-Jin; Oel, Adam Phillip et al. (2016) Recruitment of Rod Photoreceptors from Short-Wavelength-Sensitive Cones during the Evolution of Nocturnal Vision in Mammals. Dev Cell 37:520-32
Miyagishima, Kiyoharu J; Grunert, Ulrike; Li, Wei (2014) Processing of S-cone signals in the inner plexiform layer of the mammalian retina. Vis Neurosci 31:153-63