Disease-causing gene network for retinal degeneration
The disease-causing gene network is similar to the blue print for fixing a machine. Having a blueprint would substantially help finding the underlying defects and strategies to fix a machine; while knowing the disease-causing gene network would aid finding the molecular targets for therapy and regenerating retina. To detect network components that have been perturbed by the disease state, we have established approaches to microdissected eye tissues including retina (Leung & Dowling 2005; Zhang & Leung 2010), retinal pigment epithelium (Leung et al., 2007; Zhang & Leung 2010), and lens from normal and mutant embryos at different stages of development. Novel statistical and bioinformatic analyses have been built for identification of the relevant gene regulatory network and gene families through expression profiling.
Microdissection of zebrafish retina (.mpg). Extracted from Zhang & Leung 2010.
For example, we have recently mapped a retinal differentiation network from smarca4 -> irx7 -> retinal lamination and terminal differentiation of photoreceptors through the study of smarca4 mutant (Leung et al., 2008; Hensley et al., 2011; Zhang et al., 2012).
A gene regulatory network for retinal dystrophic defects as regulated by irx7 gene. Adopted from Zhang et al., 2012
We are currently investigating gene network that controls three interesting aspects:
1. The establishment of retinal lamination
Retinal lamination is the proper organization of retinal cells into cell layers and formation of connection between them. Our studies on smarca4 and irx7 have revealed key downstream targets that control retinal lamination (Leung et al., 2008; Hensley et al., 2011; Zhang et al., 2012). We are particularly interested in studying how these final molecular targets mediate the extension of neurites from multiple retinal cell types and establish a proper connection to form the first synaptic layer – inner plexiform layer.
The retinal ganglion cells (green) in the retinas that are deficient in irx7 (left figure) do not extend neurites into the inner plexiform layer (red), as opposed to the normal projections in the controls (right figure; green signal on the red region, indicated by the yellow arrows). Modified from Zhang et al., 2012.
2. The control of the development of different cone photoreceptor sub-types
Through the characterization of signal transduction defect of smarca4 retinal dystrophic mutant (Leung et al., 2008), we have found several signaling molecules that are involved in the formation of different subtypes of cone photoreceptors (cones). Dissecting the molecular networks through which these signaling molecules mediate cones development can potentially aid our control of the regeneration of specific sub-types of cones in the patients.
In the embryonic eye of a mutant of a signal molecule, red cones are overproduced (C) when compared to the normal siblings (A) and a mutant of an isoform of this signal molecule, which actually shows an overproduction of another cone subtype (not shown). Image courtesy: Sylvia Bonilla, a graduate student in the laboratory.
3. The interaction between retina and retinal pigment epithelium (RPE) during disease state
Many retinal degenerative diseases are caused by defects in retina or RPE, but both tissues will finally degenerate if left untreated. We have established an approach to estimate RPE gene expression (Leung et al., 2007; Zhang & Leung 2010) and are using the results to elucidate signal crosstalk between retina and RPE. In particular, we have identified perturbations of novel signal transduction pathways in the RPE that would ultimately result in developmental defects in the retina. Studying disease-causing gene network of both retina and RPE will help us to obtain a unified and complete picture of retinal degeneration.