Summary: C. elegans has a ton of neuropeptides (apparently 115 proneuropeptide genes for up to 250 mature peptides, many of which have been observed experimentally by mass spec). These neuropeptides are generally secreted extrasynaptically, and have been shown to modulate choline/acetylcholine release at neuromuscular junctions in C. elegans. In this paper, they want to know how neuropeptides can control synaptic plasticity.
A simple assay to probe cholinergic transmission is aldicarb-induced paralysis. Aldicarb is a cholinesterase inhibitor, and thus aldicarb enhances a single vesicle's effect on the muscle, until eventually (about 2 hours in) the worms are paralyzed because the muscles are all completely activated.
Many neuropeptide mutants have delayed paralysis (aldicarb 'resistance', though I don't like this terminology). While this would suggest that they may also have decreased cholinergic secretion at NMJs, this has not been observed for three out of four neuropeptide-deficient mutants previously studied with electrophysiology (pkc-1, unc-8, ric-19 had normal baseline transmission, whereas unc-31 transmission was reduced).
First, they show aldicarb potentiates the pre-synaptic side. Using whole-cell patch on body wall muscles, they show that spontaneous excitatory post-synaptic currents (EPSCs) increase in frequency almost two-fold within 1h of aldicarb treatment. This suggest that motor neurons are releasing more synaptic vesicles in response to the aldicarb. If they extracellularly stimulate the ventral nerve cord, the muscle response is similar regardless of aldicarb presence (although it has a longer tail in the presence of aldicarb, leading to almost a two-fold increase in the total charge transmitted). Interestingly, if they stimulate with acetylcholine after 1h of aldicarb treatment, the response is reduced nearly two-fold, suggesting the muscle has habituated.
Combining the fact that aldicarb doesnt change EPSC amplitude (only frequency), and halves the muscle acetylcholine response, does this suggest that the vesicular acetylcholine concentration has increased two-fold??? Perhaps I've missed this in the paper, but I'm confused about this.
Next, they observe that the mutant egl-3, with 'severe defects in proneuropeptide processing' does not show aldicarb-induced potentiation, but have similar baseline EPSC frequencies and amplitudes. Wormbase says egl-3 codes for a proprotein convertase, consistent with the paper in suggesting that egl-3 mutants are deficient in probably many neuropeptides. They then used an RNAi screen to knock down individual neuropeptides, and see which ones yield aldicarb resistance (suggesting a failure of aldicarb-induced synaptic potentiation). Four proneuropeptide mutants cause aldicarb resistance (ins-22, ins-31, flp-1, and nlp-12). They only had genetic mutants for flp-1 and nlp-12, so they seemingly don't investigate ins-22 or ins-31 further, and the nlp-12 mutant was more aldicarb resistant than the flp-1 mutant so they focus on the nlp-12 mutant.
Long story short, nlp-12 mutants basically phenocopy egl-3 mutants in terms of spontaneous EPSCs, suggesting that nlp-12 is a crucial peptide for aldicarb-induced potentiation. Similarly, its receptor (ckr-2) is also crucial, and the ckr-2 mutant yields basically the same phenotype as nlp-12 and egl-3. ckr-2 mutant aldicarb-induced potentiation can be rescued by ckr-2 expression in cholinergic motor neurons, suggesting that's where it's really needed.
To figure out where nlp-12 is expressed, they both expressed GFP under the nlp-12 promoter, and also made an NLP-12::YFP fusion protein. Both were found expressed in DVA (a neuron near the tail of the animal), and the fusion protein was found in puncta in both the ventral nerve cord and the nerve ring. According to wormbase, DVA is both an interneuron and a stretch sensory neuron. Based on the punctate expression pattern, that the YFP-tagged NLP-12 protein still rescued aldicarb-induced potentation (suggesting the peptide was still trafficked somewhat okay and still functional) and the fact that the puncta built up in mutants that can't properly secrete dense-core vesicles, they conclude that these puncta are likely dense-core vesicles. Further, when they add aldicarb, these puncta mostly disappear, suggesting they've been secreted. Take-home: aldicarb induces DVA to secrete NLP-12 via dense-core vesicles.
Apparently a paper in 2006 suggested DVA is a stretch receptor and uses the TRP-4 mechanosensitive ion channel. The trp-4 mutant looks identical to the nlp-12, clk-2 and egl-3 mutants, suggesting that aldicarb-induced potentiation required trp-4.
Questions I still have:
- Is there really evidence that this is synaptic plasticity? Or is it that NLP-12 simply slightly depolarizes motor neurons? (given what we know about graded synaptic transmission at neuromuscular junctions in C elegans).
- Can you decouple aldicarb treatment and the worm's motion? (for example, as in this paper the next year?)
- Why does aldicarb induce more EPSCs?
- "This feedback mechanisms consists of a stretch sensitive neuron (DVA), which secretes the neuropeptide NLP-12 in response to muscle contraction. Activation of CKR-2, an NLP-12 receptor, potentiates transmission at cholinergic NMJs. This mechanosensory feedback is employed during spontaneous locomotion behavior to determine locomotion rate." Put another way:
- Motion causes stretch, stretch causes TRP-4 activation and NLP-12 secretion, which activates CKR-2 which increases vesicle secretion rates, which increases motion. In short, this comes across as a positive feedback loop. Is this reasonable? Positive feedback loops tend to drive themselves to the maximum activation level and just stay there.
- Wouldn't this suggest that when aldicarb starts to induce paralysis, the NLP-12 secretion should stop, not increase? The authors claim that aldicarb induces 'muscle contraction' (very generally) which drives NLP-12 secretion, but "stretch" and "muscle contraction" aren't synonymous.
- to answer this question, how does aldicarb affect overall motion and particularly stretch over the hour in which the worm is in drug?
- Where are the TRP-4 receptors on DVA? DVA runs the entire length of the worm on the ventral side. Again, how does aldicarb modulate the 'stretch' signal?