Gene network for disease and drug action

When a machine is not working, you want to fix it. To do this properly, you need to know how the machine works, or even better, have a blue print of how it is built. This would allow you to find the underlying defects and devise strategies to repair the machine. When we want to find new ways to treat a retinal-degenerative disease, we also need a blue print—the disease-causing gene network. This network tells us how the molecular components in our body work together, and how a disease causes problems that ultimately affect our vision. If we know these details well, we may find a component in the molecular machine that should be targeted by new drugs to treat retinal degeneration. To this end, we study the disease-causing gene network in zebrafish eye-disease models. These models are prolific and their eyes shared many similarities with humans. Studying the network these models may reveal novel targets for treating patients.

picA gene regulatory network for retinal dystrophy. Adapted from Zhang et al., 2012.

The first step to elucidate the network is to detect the molecular components that are perturbed by retinal-degenerative diseases. This detection can be specifically done on the microdissected eye tissues, including retina (Leung & Dowling 2005; Zhang & Leung 2010) and retinal pigment epithelium (Leung et al., 2007; Zhang & Leung 2010). These tissues are collected from critical stages of embryonic development during which retinal degeneration occurs. Then, the normal tissues and the disease tissues are compared to reveal the changes in global gene expression. The differences in this comparison indicates the components that are affected in the molecular machine by retinal degeneration. 


Microdissection of zebrafish retina. Adapted from Zhang & Leung 2010.

Using these techniques, we have mapped a retinal differentiation network from smarca4 -> irx7 -> retinal lamination and terminal differentiation of photoreceptors in a smarca4 mutant (Leung et al., 2008; Hensley et al., 2011; Zhang et al., 2012; Zhang et al., 2013). We have also revealed a smarca4 -> egr1 -> p35/cdk5 network for promoting the differentiation of specific subtypes of amacrine cells in the retina (Zhang et al., 2013; Zhang et al., 2014). Furthermore, we have obtained the gene components in the RPE that are affected by smarca4 (Zhang et al., 2014).

We are using this approach to study various retinal-degeneration models. They are built based on mutations identified from patients, using the latest genome-editing tools named CRISPR-Cas. With these mutants, we can not only learn what goes wrong in their disease-causing gene network, but can also use these models to "discover new drugs". Once we find a promising drug, we can analyze how it exerts its therapeutic effect through influencing the gene network.