Thread
The PDF of our recent paper in @Brain1878 with @DrYohanJohn @NeuroYogacara @C__OCallaghan @jaaanaru @ishan_walpola @GabWainstein and @NatashaL_Taylor on making sense of the serotonergic system just dropped (but you'll have to wait til the end of this thread to get the PDF...) 
The paper stemmed from a talk I gave to @DrYohanJohn's discussion group about how amazing the neuromodulatory system is: www.youtube.com/watch?v=GXf6db4t6-g. Given my past work, I mainly just spoke excitedly about how interesting the noradrenergic system is...
Side-note: If the noradrenergic system is your jam, check out @GabWainstein's recent review paper in @TrendsCognSci here: shine-lab.org/wp-content/uploads/2022/04/2022_tics.pdf
Towards the end of my talk, @DrYohanJohn asked me to say a few words about serotonin. I hadn't really ever found a way to nicely summarise the effects of serotonin on the brain. At least, not in the way that we typically do for acetylcholine, dopamine and noradrenaline. 
As is typical, @DrYohanJohn asked a really prescient question: given that ~95% of the serotonin in the body is in the gastrointestinal tract (GIT), could we use the GIT system as an analogy for how serotonin works in the brain. The answer sent me down an Alice-style rabbit hole.. 
the starting point was trying to get a handle on how serotonin works in the GIT, which forced me to go back over things I'd learned in medical school that I had since over-written with far more important information -- like what on earth a graph Laplacian is (spoiler: its cool) 
it was fun to go back over the old ideas I'd learned about, all those years ago. I never really thought they'd be relevant for a neuroscience problem, but this is one of my favourite aspects about blue-sky science -- you never know what aspects of your knowledge will matter... 
the good news is that "we" (as in, brilliant GIT scientists) have figured out how digestion works in exquisite detail. once food hits your stomach, it triggers neural signals that traverse your brainstem and signal the release of serotonin in your GIT, which triggers peristalsis
when you start to digest food, it releases chemicals (glucose, leptin, ghrelin, insulin, etc) that slowly make their way (via the bloodstream) to the hypothalamus, wherein POMC transcription creates ligands that inhibit the descending serotonin pathways, shutting of digestion! 
our starting point: can we figure out if this same idea of fast positive feedback (when food needed to be digested) and slow negative feedback (when food had been digested) was similar to the way that serotonin was utilised in the brain?
we reviewed a lot (a lot (a lot)) of papers from a tonne of different areas to try to sketch out the details of this idea. it's kind of hard to unpack all of them on the twitter, but some cool parts that might whet your appetite... 
actually, another quick side-note: the other thing i loved about this paper was that the analogy lent itself to a pant-load of hilarious food-related puns, many of which we ultimately had to remove from the paper (thanks to the wisdom of an anonymous reviewer...) 
the basic idea: what if cognition is analogous to digestion? if there is cognitive work to be done (e.g., a problem to solve, or a decision to be made), then serotonin could help recruit the circuits (via G-protein-coupled Rs) that facilitate the completion of the cognitive work
when you first tackle a cognitive problem, you typically try something you've done before that worked -- that is, you act relatively automatically. I've argued before that the cerebellum is ideally organised to facilitate these automatised behaviours: macshine.github.io/publications/2014_delegation.pdf
if the problem isn't solved, you need to create flexible ways around the impasse. the heavily inter-connected circuits of the frontal cerebral cortex are much better suited for this kind of capacity, particularly when the circuits can sample a wide range of feedback signals 
we found some really lovely neuroanatomical evidence for this idea: the cerebellum is serotonin-rich, and in the cerebral cortex, low levels of serotonin inhibit the output of major pyramidal neurons -- these features would promote automatic responses with low serotonin
higher levels of serotonin instead promote coupling between feedback and feedforward signals in the cerebral cortex, which we argue fit nicely with the idea of identifying novel strategies to solve otherwise challenging problems 
(see also a beautiful new preprint from @HiroTaiyoHamada that shows the cortical effects in rodents following optogenetic stimulation of the dorsal raphe [www.biorxiv.org/content/10.1101/2022.08.07.503074v1] -- just wish this was published before we finalised our proof!)
in the rest of the paper, we expanded on our analogy, and thought about how this new perspective might help us think about the link between serotonin and a range of different mental health issues, such as anxiety, depression, and compulsive behaviours...
the paper was so much fun to work on -- from the original ideas, through to different collaborators popping up to help us add new, interesting perspectives and refine our thinking, I thoroughly enjoyed the whole process (except having to remove all of hilarious digestive puns) 
we're excited to hear everyone's feedback too. the ideas are more like guideposts for the lovely intersection we see between neurobiology and non-linear dynamical systems, which Walter Freeman called "God's own
firewall" (hat-tip @MolemanPeter)
firewall" (hat-tip @MolemanPeter)
Ok, so the PDF I promised is here: shine-lab.org/wp-content/uploads/2022/08/2022_brain.pdf. I hope you enjoy reading it as much as we enjoyed writing it... 
