2017: a moderately baffled review of the year in neuroscience
We came, we saw, we did a little shopping
What the hell. Sorry, I mean: hello, and welcome to The Spike’s 2017 review of the year in neuroscience. Thank any-passing-deity-that-happens-to-be-listening I don’t have to write a review of the year in politics. The combined malevolent incompetence of the US President and UK government will likely exhaust all possible permutations of terms of abuse by June 2018; after that their increasingly demented decisions and pronouncements will pass without response because we’ll have used All The Words. Every single one. Even the French ones. Va te faire foutre Trump, for example.
Brains. That’s what we’re here for. Like zombies, but with an education. There’s been a lot of sound and fury in neuroscience this year. Highbrow publications like Quanta and MIT Technology Review have burst with neuroscience stories. Neuron published so many opinion articles that they had to dedicate an entire bloody issue to them. If there’s any sign that a field of science has reached peak hot air, its that one of its elite journals publishes an issue without science. And Neuron published two in the space of 12 months (I’m aware of the irony of complaining about opinion pieces on a blog, thank you). Sound and fury; and many strong contenders for the Nobel Prize in Stating the Bleedin’ Obvious.
But what did you miss amongst the malestrom? Here for your delectation are four dispatches that arrived in 2017 from the more esoteric reaches of neuroscience. The type of neuroscience that could turn out to be epochal; or just plain old interesting. And what more can one ask?
(1) Sleep little worm, don’t you cry; Manuel’s going to sing you a lullaby
Sleep is a many splendored thing. It revitalises us, and cleans our brains. It seems essential for learning, and for creativity. We’d all like more of it. Without it, we die. Mammals need sleep so badly that, deprived of their slumber, bits of their brain will go to sleep independently of the rest while the animal is doing stuff. With awkward consequences. Horribly jet-lagged attendees at international conferences will recognise this as the bit during the poster session where you pitch head-first into the poster board while apparently still talking, because bits of your motor cortex have gone “god this is dull, I’m going to sleep”.
So we’d all like to know more about how the brain goes to sleep. But brains are big unwieldy beasts — how can we study how a whole brain goes to sleep? Manuel Zimmer and his team hit on a solution: study sleep in the tiny worm C Elegans. Yes, apparently worms with 302 neurons in their entire bodies fall asleep. Technically, it enters a state of lethargus: the worm goes offline, but there’s little evidence that it dreams (other than that mortifying recurring nightmare of being caught eating bacteria with its mouth open).
Zimmer and team showed us the neural activity cycling around the entire brain while the worm turns (and wriggles), then abruptly halting as the worm slept. As many as 75 percent of its neurons shut up entirely; and the rest carried on being active, but in a way that was unrelated to their activity in waking. In other words, they entered a different global state — one that looked very much like a random walk around a fixed point (if you’re excited about such things. Which I am. But then I code dynamical systems models for fun. So, you know, don’t take my word for it). What we gained here is view of sleep not as a thing created by some detailed, discrete sleep “circuit”, but as a switch in the state of the entire brain’s dynamics.
But this paper was also notable for the depths of its content. We often lament that papers in the most elite journals of science, the Natures and Sciences, are all mouth and no trousers: they report amazing results on the flimsiest of evidence. But this paper, in Science, spent the first few pages on a deep and detailed characterisation of how subtly different lineages of C Elegans (all the same species, remember) had different abilities to fall into the lethargic state, and how that changed over their development. It was like a firehose of deep science: no big picture, nothing flashy, but deep careful work to find the precise type of worm most suited to the question of what happens when worms sleep. In Science. Crazy shit. Is there hope yet?
(2) Would you like to hear voices?
Auditory hallucinations. One would think that these were a trifle tricky to objectively research, seeing as they practically define the personal subjective experience. (Pretentious, moi?) Yet as hallucinations are a debilitating hallmark of schizophrenia and other cognitive disorders, knowing something of how the brain creates them would be rather useful. This year Phil Corlett and co hit on a neat idea: why not create auditory hallucinations both in healthy people and in people who spontaneously suffer from them? For if we then measure the difference in how easy it is to create hallucinations in the two groups, we might learn something about how the brains of hallucinators are operating differently.
For it turns out we can make ordinary people have auditory hallucinations. Nothing scary, just hearing something that isn’t there. It’s disarmingly simple: pair the flash of a light with a short sound that you can just about hear. Do that a lot. Then just flash the light on its own. Even ordinary people then report hearing the tone a good percentage of the time. So the crux of this new study was: let’s also do that in people who already hear auditory hallucinations and ask if they hear more hallucinated tones? And if so, why?
The answer was: yes, yes they do. A lot more. And for much quieter sounds too. But the key insight here was that the authors used this data to make a model of “why”. And the answer was, predictably — prediction. Those pairings of light-and-sound set the brain up to make predictions. It predicts that the light means a sound is also present. So when the light flash comes alone, the brain goes right ahead and predicts a sound when none is present. In ordinary people, sometimes this prediction wins out against the sensory evidence that there was no sound, and a sound is heard where none is to be found. But in hallucinators this prediction dominated, so they heard many more fictional sounds.
Which all fits with an interesting theory of hallucinations, that they occur when the the brain starts adding too much credence to the predictions it makes, rather than using the information coming from the outside world. Like football pundits. Basically, the what-the-hell-is-Alan-Shearer-talking-about-now-did-he-even-watch-that-match theory of hallucinations.
(3) A major journal publishes a null result
The title of this bit is not a typographical error. eLife published a paper which was almost entirely about the absence of an effect.
Back in 2004, a highly influential paper in Science suggested that Parkinson’s disease patients on and off their dopamine medication seemed to learn different things about the same task. Those on-medication learnt to pick options that were most likely to be rewarded, suggesting they learnt best from positive feedback. Those off-medication learnt to avoid options that were least likely to be rewarded, suggesting they learnt best from negative feedback (i.e. on missing out on a reward). This finding sparked an entire way of thinking about how dopamine controls learning through its effects on the lumps of neurons we call the basal ganglia.
A paper in 2012 suggested something was amiss. There Shiner and co reported that actually patients on and off-medication could learn the task the same, and learn it just as well as a control group. But what the patients couldn’t do was generalise what they had learnt when faced with new combinations of options — they were unable to take what they’d learnt and use it in new ways. So which was it? Does dopamine medication change the style of learning, or of the ability to adapt learnt information to new choices?
The goal of the eLife study this year from Liz Coulthard, lead-authored by John Grogan, was to get to the bottom of the mystery by repeating the whole study again, and in all possible combinations of whether the patients were on or off medication during learning or during testing. And they found nothing, nada, niente, not a sausage. Bugger all. They found no difference between patients and no difference between patients and controls on anything. And found it three times. A very convincing null result, all told.
A huge amount of work, no “positive” result, and yet published in an elite journal. Slight downside is that our grasp of dopamine is now even more like a panda with its tongue caught in a Rubik’s Cube: painfully baffled.
(4) Turning one nervous system into another.
Many questions about how brains work can’t be answered in complex animals. Things like: can the same behaviour arise from different neural circuits?
But Paul Katz knows how to get around these problems: use dumb-as-a-rock marine invertebrates. In a phenomenal piece of work that mammalian neuroscientists can only dream of, Sakurai and Katz took two closely-related species of swimming sea-slugs and showed that despite having the same swimming movement, and very similar neurons, the wiring between those neurons was different. The same behaviour, the same neurons, but different circuits.
Rhythmic movements like swimming are produced by circuits that self-oscillate, that have neurons bursting with activity at regular intervals, each burst causing a muscle to contract. So S&K said: right, so these two circuits must produce the same bursting but in different ways. To prove this, they blocked the synapses of the same pair of neurons in both species. In one species, this let the other neurons drop into a patterns of slow bursting that didn’t quite line up right. In the other, this blockade totally destroyed bursting. Same neurons, different effects.
These two sea-slugs have a common ancestor. So these two different circuits imply that evolution has driven divergent wiring of this circuit, yet kept the same behaviour. Evolution here was acting on the wires between neurons, not the types of neurons nor the types of transmitters they used. If true, this should mean that we could take one circuit and rewire it to match the other, and end up with the same firing patterns in both. Some science fiction gubbins right there. I mean, imagine if we could that.
We can. This is exactly what S&K did.
To start with, they simplified their task by blocking the same pair of neurons again, taking the circuit down to just four neurons in both species. Again, they got slow, wrong bursting in one; and sod all in the other. All that was different was two connections in the slow bursting circuit that were missing in the sod-all circuit. So they made those connections themselves — they made artificial synapses. They recorded from the source neuron, transformed that neurons’s output into the predicted set of small fluctuations it should cause in a neuron at the other end; and injected the resulting signal as electrical input into the target neuron. Voila, artificial synapse. Rinse and repeat for both missing connections.
And lo, the sod-all circuit started slow, wrong bursting in all the same ways in all the same neurons. They turned one animal’s nervous system into a different animal’s nervous system.
That high-pitched whining sound you can hear is mammalian neuroscientists quietly seething with jealousy. Sea slugs rock (©Angela Bruno’s PhD defense):
And 2017 had so much more. Rafael Yuste finally got to record ALL THE NEURONS (well, Christopher Dupre recorded all the neurons). The game-changing Neuropixels probe, trailed in last year’s review, was finally published. Elsayed and Cunningham asked the awkward question: you know all that exciting stuff we keep finding in large populations of neurons — you know, those Nature papers — what if they’re just artifacts of adding up a lot of single neurons with really dull tuning and nothing to do with the special sauce of the “population’’ at all? They gave us a bunch of ways to test; and found there was some special sauce after all. Phew. We celebrated the 200th anniversary of “An Essay on the Shaking Palsy”, the tract by James Parkinson that launched the systematic study of neurological disorders. And we had the revelation that “scientific papers have decreased in readability over the past few decades” — the authors missing the irony of this sentence. Yes, the words used in scientific papers have become more complex over time. And the worst offender? Genetics. But we knew that already.
One thing we scientists perhaps under-appreciate is that — if you’re paying attention — you get to hear good news all the time. Of advances in our understanding of the natural world; of technological breakthroughs; of medical marvels. And, in these idiotic times of fractured nationalism, we get to hear about countries working together, about co-operation that transcends all differences.
ITER reached the half-way mark. With nuclear fusion perhaps the only viable permanent solution to our simultaneous carbon and energy crises, and ITER being the only attempt on the table to get it working commercially in the near future, having it back on track can only be a wonderful thing. And as a collaboration between 35 countries who historically don’t get along all that well (including the US, Russia, China on the one hand, and China, Japan, and Korea on the other), it is also signifies that we can still pull together.
Better yet, a new synchrotron opened this year. Great, you’re thinking, another source of high intensity light for looking at crystals. Whoop-de-doop. Ah, but this one is the only synchrontron in the Middle East. Based in Jordan, but built by a collaboration between Iran, Jordan, Egypt, Israel, and Palestine. I’ll let you go back and read that list of countries again. Science again transcending conflict and politics.
Oh, and they called it: Synchrotron-light for Experimental Science and Applications in the Middle East. Yes, the official acronym is: SESAME. And the official funding project? Open SESAME. For a laugh. We’re going to be OK.
Want more? Follow us at The Spike
Twitter: @markdhumphries
