Channelrhodopsin2 application: the race is on!!

So channelrhodopsin2 (ChR2) is the hottest thing in (optical) neuroscience just now. As noted in an earlier entry on this blog, it was cloned by the group of Nagel & Bamberg in Germany in 2002/03 (ChR1: Science (2002) 296:2395-2398; ChR2: PNAS 2003) 100:13940-13945). On both of these papers Peter Hagemann was a collaborator. He had been studying these proteins in green algae since 1991 (at least), but had not cloned them. Bamberg had been interested in the partial reactions (rate constants for the many kinetic steps) of many integral membrane proteins for 30 years. In particular, he had studied bacteriorhodopsin in great detail (I know Bamberg from his work of the Na,K-ATPase). The Science paper gives the sequence of the core amino acids for ChR1, which is a light-driven proton channel. It was natural for Bamberg and Hagemann to collaborate here, as both sytems use cis-trans isomerisation of retinal to driven proton fluxes. Though there is an important difference between ChR1 and bR, in that that latter is an ion pump, and the former an ion channel. From the biology of green algae they realised that another channel existed, so they cloned and expressed ChR2 in Xenopus oocytes and showed that ChR2 is highly permeable to sodium and calcium ions.

At this point I am not quite sure what happened, but Karl Deisseroth saw the potential of ChR2, got the DNA, put it in neurons, and reported that with bluish light you could depolarised neurons (Neuron (2005) 8:1263-1268). What is remarkable about this is that one does not need to add any co-factor! What do I mean? Neurons have enough retinal to condense with the haloprotein to make the photoresponsive channel. Why, we don't know.

With this report, the race was on to do something sexy (sexier?) with ChR2. Recently there have been a few reports showing that ChR2 expressed in mice can be used to depolarise neurons in situ (ex or in vivo): Nature Neuroscience (2007) 10:663-668; PNAS (2007) 104:8134-8138; Neuron (2007) 54:205-218. The last of these is the coolest as it is the first report of activation of neurons in the living mouse (in the olfactory bulb).

During this flurry of work, Nagel, Bamberg, Deisseroth, and collbaorators had been working on what must now be called "obvious": photo-inhibition of neuronal circuits (Nature (2007) 446:633-641). For anyone who had studied those "boring" light-driven pumps this would be obvious, as halorhodopsin (the light-driven chloride pump) had been known for many years (JBC (1982) 257:10306-10313). Deisseroth and collaborators used chloride pumps from archaea. They found one of these (called NpHR) worked well as a counterpart to ChR2. Remarkably, ChR2 and NpHR have different absorption spectra (Figure (a) above is from the Nature paper, a beautifully written article I would add). Thus, "NpHR and ChR2 form a complete system for multimodal, high-speed, genetically targeted, all-optical interrogation of living neural circuits".

The only comfort I can derive from this is that I am told that ChR2 is not 2-photon sensitive, leaving room for caged compounds (whew!). I would suggest that caged calcium or glutamate could be a useful compliment to the photoproteins, as these photoprobes are 2P active, and absorb light at a shorter wavelength. For example, NDBF-EGTA (our new caged Ca, Nature Methods (2006) 4:35-41) has an absorption maximum at 330 nm, where the photoproteins do not absorb (Figure (b) above).

It is going to be interesting to see where the revolution takes us next.