A Musing on Frequencies

This month a paper was published by one of my former colleagues, in Frontiers1. Using acute brain slices on a multi-electrode array (MEA), they investigate neurone burst firing frequency.

The MEA is a piece of recording equipment that I have never used myself, but I have seen its use. And as someone used to doing single-cell patch clamping, I am interested by the type of data you can get from it, in particular the high quantities of recordings in a short space of time.

In case anyone doesn’t know, the MEA uses an array of recording electrodes in 2D matrix, and the brain slice is set on top. This lets you record action potentials from a number of points across a region of interest, which in this paper was the NTS, PVN and SON.

The benefits of the MEA is that lets you record natural firing dynamics from many neurones simultaneously, and this enable Chrobok et al. to identify some neurones in the NTS that exhibit phasic firing behaviours (Figure 1). This was very interesting behaviour, with 5-10 Hz neurone burst firing frequency interspersed with long periods of complete silence.

My particular interest in these results is how it pertains to in vivo optogenetics, in particular trying to mimic natural neuronal firing behaviours with the stimulation pattern. I usually try to keep the experimental paradigm as straightforward as possible, so I go with 5, 10 or 20 Hz continuous stimulation.

However, it is worth noticing that Betley et al. used a phasic stimulation pattern when investigating AgRP neurone-driven behaviours (they did 20 Hz stim for 1 second, then off for 3 seconds)2. The reason they picked this stimulation paradigm is to mimic AgRP firing behaviours as seen by Van den Top et al.3 I have found this to be an interesting approach, to hopefully improve your in vivo behavioural data by trying to closely mimic the natural firing dynamics.

The importance of matching firing dynamics in an experimental setting extends beyond simply trying to mimic any information that might be encoded in such pattern. In fact, it has been known for over 40 years that phasic firing can enhance the release of neuropeptides from the nerve terminal4. This is an important aspect of optogenetics experiments that is often ignored – when stimulating neurones you are likely to be getting fast neurotransmitter release (ie. glutamate and GABA), but depending on your stimulation paradigm you may not be getting commensurate release of neuropeptide.

We have seen in our lab how this can change the animal behaviour, beyond simply increasing the degree of any behavioural response, eg. going from changes to stress hormone levels at low frequency stimulation to freezing and escape behaviour at high frequency stimulation.

So, back to Chrobok et al., who saw much slower phasic behaviour: approx. 2-4 Hz firing for 4-8 seconds, repeated every 10-100 seconds. Next time I (or someone in the lab) want to optogenetically stimulate NTS circuits in vivo, I will point them towards this paper and suggest they try phasic stimulation to see if that produces better behaviour resoponses.

There is one final point I want to take from the Chrobok paper, which is that the phasic behaviour they see in the NTS is not in sync. This is in contrast to nuclei such as the SCN which has strong synchronicity. Anyway, this got me thinking, if we intend to mimic natural firing dynamics as closely as possible, shouldn’t we try to perform unsynchronised phasic stimulation of the NTS neurones?

I envisage an AAV that has multiple opsins, responsive to different colours of light, whose activity is randomly chosen by mixed LoxP sites, similar to the Brainbow construct. For example, if we had an AAV with ChR2(h134r), C1V1TT, and Chrimson (Figure 2), after a stop codon so you get no expression, but under cross-reactive LoxP sites such that cre will randomly switch on one of the variants only, you could produce a selected population of neurones (eg. TH neurones in the NTS) expressing a variety of opsins.

Then you would need to set up an in vivo optogenetics stimulation system that could switch between 450 nm, 530 nm, and 620 nm, allowing you to produce desynchronised phasic behaviour among that single population. You could easily set it up to cycle through flashing blue light for 5 seconds, then green light for 5 seconds, then red light for 5 seconds to produce desynchronised phasic behaviour.

Furthermore, given that you can easily get multiple insertions in a single neurone, you might well have some members of the population expressing more than one opsin variant, which would give you further variety in phasic behaviour, and even some that simply fire continuously. This is unfortunately A LOT of work simply to see what happens to the behaviour when you try to closely mimic natural phasic firing dynamics, so, while someone out there might be brave enough to do something like this, I don’t think I or anyone in my lab is likely to.

1. Chrobok et al., Front Phys 12, 638695 (2021) Phasic neuronal firing in the rodent nucleus of the solitary tract ex vivo.

2. Betley et al., Cell 155, 1337-1350 (2013) Parallel, redundant circuit organization for homeostatic control of feeding behavior.

3. Van den Top et al., Nat Neurosci 7, 493-494 (2004) Orexigen-sensitive NPY/AgRP pacemaker neurons in the hypothalamic arcuate nucleus.

4. Dutton and Dyball, J Physiol 290, 433-440 (1979) Phasic firing enhances vasopressin release from the rat neurohypophysis.

Leave a Reply

%d bloggers like this: