Title: A Circuit for Gradient Climbing in C elegans Chemotaxis
In this paper, their goal is to understand how C. elegans can perceive scents over at least four orders of magnitude concentration range. Worms can sense and respond to extremely small amounts (10 nM) of diacetyl (an 'intensely buttery-smelling' compound, and not surprisingly, naturally found in butter), all the way up to massive concentrations of diacetyl (100 uM). How do they avoid sensor saturation? If a sensor saturates, it's no longer able to detect a gradient, and if a worm can't detect a gradient, it cant do chemotaxis!
There are two sensory neurons that respond to diacetyl - AWA and AWC. They focus on AWA, and disregard AWC for reasons I'm not clear on. Additionally, they later focus on one downstream interneuron - AIA - which has a gap junction with AWA (and receives a chemical synapse from AWC).
They first clearly show chemotaxis in spatial gradients as small as 0.8 nM/mm.
They next consider calcium dynamics of AWA for various odor concentrations. At low concentrations (10nM), AWA behaves like a high-gain lowpass filter. At intermediate concentrations (100-1000nM) it behaves much more like a rectifying differentiator, spiking sharply when the concentration goes up, and then rapidly diminishing even though the input remains (they refer to this as 'desensitization'). At high concentrations (10-100uM), it again goes back to acting like a low-pass filter, but with much lower gain!
This is complicated. Where does this highly nonlinear behavior come from? How does nothing saturate?
To make matters more complicated, they note that the sensory response habituates over repeated stimuli, at a rate that depends on the stimulus concentration, and also on the stimulus number. Specifically, they show rapid habituation for the first 6 or 7 stimuli (minute spacing between stimuli), followed by very gradual habituation over the next 300 stimuli.
Even more complicated! Many distinct timescales, from one second to five hours! They use Chrimson (a red-shifted channelrhodopsin enabling them to do simultaneous GCaMP calcium imaging) to depolarize AWA, and ask if the same desensitization and/or habituation process occurs. Unfortunately, it's hard to match the Chrimson stimulus intensity to the diacetyl concentration to make sure you're comparing apples and apples (remember that the desensitization and habituation rates depend on the diacetyl concentration!). They assume that since they match the GCaMP fluorescence levels that they're close, but I don't know if that's fair or not.
Long story short, they conclude that "AWA desensitizes to intermediate odor concentrations at a step in olfactory transduction before cell-body depolarization." I don't know if I 100% buy this because I'm not sure Chrimson activation is really identical to a 'later' step in the sequence of events that occurs following odor stimulation. It's certainly a plausible explanation, I'm just not convinced it's the only one.
They also looked at a panel of mutants to see whether there were mutants that differed in their desensitization or habituation kinetics. They found a number of mutants in interflagellar transport, and note that all the mutants had 'abnormal' cilia, which in the one micrograph they show, looks like the cilia is (are?) just plain missing. Strikingly, these mutants aren't impaired in their sensory abilities - just the kinetics of desensitization and habituation. They also note a mutant in phosphoinositol-5-phosphatase that habituates faster, and note that its expression of the diacetyl receptor in the cilia was 'subtly altered.'
They move onto one of AWA's downstream interneurons, AIA. AIA appears to somehow interpret its inputs (which again, include AWA, AWC, and ASE) to yield a function as a very well-behaved fold-gradient-increase detector. It doesnt respond to steps down, but for a fixed fold increase in diacetyl, it briefly spikes in a manner proportional to the fold change. There are some mutants discussed as well.
They finally turn to AWA and AIA's effects on chemotaxis. Strikingly, Chrimson activation of AWA suppresses turning, but when the light goes off, there is massive 'rebound' turning! A strange and interesting result, that more or less goes unexplained.
Questions that came to mind:
The rebound effect after they stop depolarizing AWA is striking. Why does this happen??
Why didn't they ever use a continuous gradient instead of a step? It seems like they would predict that would yield a DC output from AIA that would be directly interpretable as the gradient magnitude!
I had a lot of questions about methods in this paper. Two I made note of are here:
- Does the blue light from GCaMP imaging really not drive the Chrimson? In the maintext at least, there is no control to show this.
- They show a single ephys trace from a patch experiment to prove that Chrimson activation can provide something akin to constant current current clamp. This is super important, but begs the question why the rest of the experiments weren't done with patch. I presume it's because patch is just too hard, but these experiments would be so much cleaner if one could directly record voltage during odor presentation, and if one could then compare stimulus presentation with current injection.