On 2016 Jul 25, Maarten Zwart commented:

None

On 2016 Jul 25, Maarten Zwart commented:

Thanks for the comments, John!

I agree the paired recordings you're suggesting would be interesting to do, but it's very difficult (I've not had any success) to separate excitation from inhibition in voltage clamp because of space clamp issues in these neurons. It seems interesting enough to try to get it to work, though. A more general description of the rhythm generator(s) will hopefully also come from further EM-based reconstructions.

Most larval muscles receive input from a MN that only innervates that particular muscle ("unique innervator"), as well as a MN that innervates multiple muscles ("common innervator"). LT1-4 and LO1 are different; they only receive input from the "unique" MN, so that simplifies things. There are no known inhibitory MNs in the larva, which is an interesting quirk if this indeed holds up.

It's not been exhaustively explored how different larval MNs compare in their intrinsic properties, but there is an interesting difference between the "unique" innervators and the "common" ones, with the latter showing a delay-to-first-spike caused by an Ia-type current (Choi, Park, and Griffith, 2004). I looked into intrinsic properties to test whether a similar delay-to-first-spike mediated the sequence. There will certainly be differences in input resistances between some MNs as they are not all the same size, but fast and slow ones have yet to be described.

Thanks for the heads up on the PTX effect. We've seen different effects at different concentrations, with higher concentrations affecting intersegmental coordination in addition to intrasegmental coordination, and we've jotted these down to simply a more effective receptor block, but that's very interesting!

Thanks again John!

On 2016 Jul 25, John Tuthill commented:

Do transverse and longitudinal MNs receive any synchronous or correlated excitatory input? Figure S4 shows paired recordings between aCC and LO1 and they look relatively correlated. Would be interesting to look at LO1 and LT2 pairs to see whether the inputs they share drive synchronous activation at particular phases of the fictive rhythm cycle, which might be suppressed by inhibition (from iINs) at other phases. This would provide some indication of whether there is a single “CPG” that serves as a common clock/oscillator for all the MNs within a segment. It would also have some bearing on your model that intra-segmental timing is generated by selective inhibition, rather than specificity of excitation.

Each larval muscle is controlled by multiple MNs. These different MNs receive many, but not all presynaptic inputs in common (figure 2). How does this affect the phase relationship of MNs that innervate a common muscle? A broader question might be, in an oscillating population of MNs, how well can you predict phase relationships by quantifying the proportion of overlapping presynaptic inputs to those MNs?

Are larval MNs divided into fast/slow neurons, as in the adult? On a related note, do all larval MNs exhibit the vanilla intrinsic properties shown in Fig 1? Do fly larvae have inhibitory MNs like many adult insects? (Interested in these questions since we are working on them in the adult).

Technical note: @ 1 micromolar, picrotoxin only blocks GABAa receptors, not GluCl (at least in the adult CNS, see Wilson and Laurent, 2005 and Liu and Wilson, 2013).

On 2016 Jul 25, John Tuthill commented:

Do transverse and longitudinal MNs receive any synchronous or correlated excitatory input? Figure S4 shows paired recordings between aCC and LO1 and they look relatively correlated. Would be interesting to look at LO1 and LT2 pairs to see whether the inputs they share drive synchronous activation at particular phases of the fictive rhythm cycle, which might be suppressed by inhibition (from iINs) at other phases. This would provide some indication of whether there is a single “CPG” that serves as a common clock/oscillator for all the MNs within a segment. It would also have some bearing on your model that intra-segmental timing is generated by selective inhibition, rather than specificity of excitation.

Each larval muscle is controlled by multiple MNs. These different MNs receive many, but not all presynaptic inputs in common (figure 2). How does this affect the phase relationship of MNs that innervate a common muscle? A broader question might be, in an oscillating population of MNs, how well can you predict phase relationships by quantifying the proportion of overlapping presynaptic inputs to those MNs?

Are larval MNs divided into fast/slow neurons, as in the adult? On a related note, do all larval MNs exhibit the vanilla intrinsic properties shown in Fig 1? Do fly larvae have inhibitory MNs like many adult insects? (Interested in these questions since we are working on them in the adult).

Technical note: @ 1 micromolar, picrotoxin only blocks GABAa receptors, not GluCl (at least in the adult CNS, see Wilson and Laurent, 2005 and Liu and Wilson, 2013).

On 2016 Jul 25, Maarten Zwart commented:

None