[0] https://web.stanford.edu/group/dlab/media/papers/Zhang%20Nat...
The virus is injected, and then infects cells. It re-writes their DNA, like a lot of viruses we are all made sick by, and inserts a DNA code into the cell. In wild viruses, that DNA code would make other viruses and start the process of infection all over again. Here, that DNA code is for a protein called an 'opsin'. Just like in your retina and eyes, this opsin protein is affect by light of certain colors. In the case of exciting the nerve cell to 'fire' (the other general case is inhibiting the cell from 'firing') the opsin protein is in the cell wall and opens up the cell to the extra-cellular environment. Similar to osmosis, this causes some of the ions in the fluids (Sodium, Potassium, Calcium, etc) to rush in or out of the cell. The specifics depend a lot on the particular cell, but mostly on the relative concentration of the ions in the cell and in the extracellular fluids.
Shine light, the protein (delivered via viral infection) opens, ions rush in/out, the cell 'fires'.
Also, in longer axons like in the eye or the spine, you do have cross talk like with copper wires. BUT only if those axons are un-myelinated. That is bio-talk for cells that wrap around the axons of nerve cells that are very fatty in composition. Basically, myelin is like the plastic insulation you have on copper wire. They do a lot of other things too, but in the spine and eye, they act as isolators for the nerves that have to go a long distance. When that myelin 'goes wrong', you can get disorders like ALS and such.
Here are some sources to learn more about optogentics:
- The "invasiveness" of optogenetics is at least as bad as electrical microstimulation, in that you need to implant cannulae to inject the virus, plus some sort of fiber. It's possible to implant very thin metal electrodes that do fairly little damage; the optical stuff is currently much bigger.
- Optogenetics does have a selectivity advantage, but it's a bit different. By finding the right promotor, you can target specific cell types so that the optical protein is only expressed in (say) a specific class of interneurons, or a specific layer, or cells that project to another brain region.
- The virus itself isn't switching things on and off. Instead, the virus infects cells and causes the cells to produce a channel which light can switch on/off.
For areas deeper in the brain, however, you can't exactly feed a wire in there without destroying things, so instead, you use optogenetics: you get a virus to be the "end" of the wire (finding the right area of the brain to live in, and then converting light pulses into electrical pulses) and then you point a laser at the brain and the viruses "go off" in response. The virus is essentially just serving as one half of an electrical optoisolator.
As an added benefit for the optogenetic approach, the electrical pulses released by the viruses are each local to their host cells, so they won't have nearly as much "bleed-over" into surrounding tissue as you'd see with one strong pulse from a wire.