As a follow up to this question regarding the evolution of the eye, it was suggested that primitive eyes only needed to evolve a light sensor and could perhaps use the existing biochemical cascade infrastructure.
What is the minimum of this biological light sensor? Please provide a detailed biochemical description of how many proteins are involved and their interactions which can be "hooked up" to the existing biochemical cascade infrastructure. (to the best of our knowledge)
The minimum biological light sensor is just an existing signaling molecule that, by chance, also has some change in its activity due to light. There are several ways this could happen.
One extremely simple example would be based on a G-protein coupled receptor (GPCR). GPCRs are a family of related signal transduction proteins (the largest family) that sit on the surface of a cell's membrane. Each one is sensitive to a molecule or small group of related molecules. When exposed to the molecule it is sensitive to, the GPCR sets of a series of chemical cascades that trigger some activity in the cell. For example, it can serve as a signal for the cell to move towards the place where the chemical was detected, or move away.
Now, imagine that, just by chance, the molecule the GPCR is sensitive to has some sensitivity to light, in that when exposed to light the molecule either has a larger or smaller chance (even slightly) of binding to the GPCR. This is not improbable, a ton of biological molecules have this characteristic, in fact many have substantial responses to light even though they are not involved in light-sensing pathways. A good example is called "photoisomerization", where the same molecule can have more than one shape, and switches between two shapes when exposed to light.
This would be a biological light sensor. Not necessarily a very sensitive one, but once you have such a basic light sensor, natural selection would take over and lead to stronger and stronger effects.
In fact, this is pretty much how our eyes work today. It uses a GPCR whose target is a molecule that changes its shape when exposed to light, which makes it no longer able to bind to the receptor, shutting of the signaling cascade. What is more, this molecule is not only used in sensing light, it is a precursor to a molecule needed for gene transcription and signaling.
In fact, you can splice light-sensitive GPCRs from bacteria (which are different than those in animals) into non-light-sensitive animal cells, and those molecules will automatically integrate with existing GPCR signaling cascades, rendering the cell light sensitive.
