Every nerve cell of a zebrafish’s brain monitored at the same time to identify alertness neurons

Stanford researchers were able to simultaneously monitor activity in every nerve cell of a zebrafish’s brain and determine which types of neurons were tied to alertness.

Previously it was believed that there was a single alertness on/off switch but they have shown it is many cell types throughout the brain together produce the state of alertness.

Problems tied to alertness deficits range from sleep deprivation to depression to brain-trauma-induced somnolence, while conditions such as anxiety, mania and post-traumatic stress disorder are often characterized by excessive alertness.

The researchers employed a cutting-edge technology that allowed them to monitor vast numbers of nerve cells’ activity in the brain simultaneously and to characterize the cells of interest in molecular detail. This new method lets scientists record the activity of any cell type, without needing to produce special, genetically modified animals for each new experiment.

The techniques for decoding the how molecular and cellular functions in the nervous system and brain could be applied to determining other brain functions. This could eventually help with Singularity related goals.

Extending findings to mice

They extended their findings to mice confirming that the evolution left this system similar in mammals whose ancestors split from those of zebrafish hundreds of millions of years ago. The tight conservation of alertness-promoting circuits over hundreds of millions of years strongly implies it is the same in humans.

The Stanford scientists call the new method they developed MultiMAP, which is short for Multiplexed-alignment of Molecular and Activity Phenotypes. It let them track the activity of nearly every neuron in the zebrafish brain and then identify the cell type of every neuron of interest — the crucial step in determining which neuronal circuits are participating in the induction of a brain state such as alertness.

“We looked at every neuron in the fish’s brain during life, when those cells were actively firing, and learned which cells were most active at moments when we knew that the fish was most alert,” said Deisseroth. “Then, after the fish’s brain tissue was preserved with a fixative without altering relative positions of cells within the fish’s head, we could target those neurons with molecular probes and determine their cell types.”

In particular, the investigators wanted to explore the activity of the brain’s neuromodulatory circuitry. Distinct from their more binary cousins — some 98 percent of neurons in the brain act by either exciting or inhibiting downstream nerve cells — clusters of neuromodulatory neurons send out projections that branch throughout the brain and act less like on/off switches than like colors on an artist’s palette. Rather than excite or inhibit, they instead add nuance by secreting substances that render excitatory and inhibitory neurons more or less likely to fire under various circumstances.

Different types of neuromodulatory neurons can be distinguished not only by the different kinds of substances they secrete, such as dopamine, acetylcholine and serotonin, but also by various biological markers, such as proteins that appear in or on one but not another neuromodulatory cell type.

The zebrafish Deisseroth’s group looked at were only 7-10 days old — a time when their bodies are only a few millimeters in length and still relatively transparent, yet are able to execute complex behavior. This feature, along with a few bioengineered tricks permitted direct brain-wide optical inspection of neural activity.

Bioengineered fish

The fish had been bioengineered so that calcium flux within a neuron — an excellent proxy for impulse conduction within that cell — triggered a fluorescent signal that could be picked up and recorded via high-powered microscopy. The researchers constrained the fish’s heads by embedding each tiny animal in an agarose gel. They carved the gel away from the fish’s tail, whose movements would be key to their experiments.

To test alertness, the investigators presented the fish with a visual stimulus suggestive of an approaching predator. This evokes an instinctive tail-swishing response, as the fish attempts to veer away from impending danger. Each of 34 separate larval zebrafish was subjected to repeated exposures to the looming-predator stimulus. The time elapsed between each exposure and the consequent swishing of the fish’s tail — the animal’s reaction time — was measured. In each instance, fluorescence representing activity within tens of thousands of nerve cells just prior to the stimulus was captured and recorded.

After humane euthanasia and preserving the fish tissue in fixative, the team directed fluorescence-labeled molecular probes at the animals’ neurons and imaged the results. Using algorithms, they were able to align the neural-activity data from their experiments on live zebrafish with cell-identity data extracted afterward from those same fish.

Deisseroth and his colleagues identified multiple anatomically and biochemically distinct populations of neurons that were most active just before the fishes’ fastest responses to the looming-predator stimulus. One of these was the norepinephrine-secreting population originating in the locus coeruleus, already well-known to be involved in alertness. This was reassuring, Deisseroth said, because it indicated that the approach was valid.

But the dragnet snagged several other neuromodulatory populations of neurons as well, including cells that release acetylcholine, serotonin, dopamine and the peptides CART, somatostatin and neuropeptide-Y. While all these populations’ activity levels correlated with alertness, they didn’t correlate in perfect lockstep with one another, suggesting that each population has its own slightly different job to do.

Guided by their findings in zebrafish, the researchers then targeted the equivalent neuronal populations in the much more complex brains of mice. To test mice’s reaction times, they trained the mice to lick in response to a particular tonal cue; when the mice did so, they were rewarded with a drop of drinking water.

Deisseroth’s team observed almost precisely the same activity patterns in these equivalent circuits in mice that they had seen in those of the fish.

The scientists went one step further in the mice: They used optogenetics — a technology, developed in Deisseroth’s lab, that allows neurons to be excited or inhibited by a pulse of light at the flick of a switch — to prove, by variously activating or inhibiting neurons in one or another population, that only a subset of the circuits in question were individually capable of increasing alertness. One was the set of neurons coming from the locus coeruleus; this set operates by secreting the substance norepinephrine. Another was an understudied group of so-called CART neurons, previously known to be responsive to cocaine and amphetamines, in a brain region called the Edinger-Westphal nucleus; and the third was a group of acetylcholine-secreting neurons in a deep-brain structure called the lateral dorsal tegmentum. Other cell types correlated with alertness did not directly influence the behavior; Deisseroth said he thinks these neuromodulatory circuits, rather than directly triggering alertness, report on its status to other brain circuits.

The new findings open the door to a whole new route of further exploration, Deisseroth said. “The more we understand the landscape of neurons that underlie a brain state like alertness, the more we understand the brain state concept itself — and we may even be able to help design brain-state targeted clinical interventions.”

10 thoughts on “Every nerve cell of a zebrafish’s brain monitored at the same time to identify alertness neurons”

  1. humane euthanasia
    quite a bit a difference just adding an e.
    seems sort of a leap of faith to conclude with surety they haven’t caught more cells than they wished with that net.

  2. “This could eventually help with Singularity related goals.”
    I was wondering if this would be brought up. The Singularitians are big on the idea of mind uploads into a computer, the first step will be knowing exactly what all the brain does, so it can be simulated on an appropriate hardware platform.

    It seems we are still quite a ways from knowing how the human brain functions in any detail, but this is a good baby step.

    • No mind uploads. The present State of the Art is – in many ways – about the same level as Neolithic Man, having invented the hammer, hitting her partner’s knee, watching the leg spring forth and declaring, “Why look, Snorzl! We’ve figured out the nervous system!”

      Seriously. I’m not even trying to overpun the point.

      And if you’re inclined (for good discussion) to cite the Magnetarians (those who attach dozens-to-thousands of esquisitely sensitive superconducting SQUID devices to the outsides of brains, or even deeper invasive studies involving dozens-to-thousands of little nanometer-wide glass-insulated pure gold wire probes and even intimately integrated chips which digitize things on the fly…), this too is STILL “hammer-shaped-rocks-attached-to-sticks” technology.

      Insofar as I am aware, we have yet to measure more than a few thousand neurons firing in situ, in realtime.

      The mind uploaders often zing into SciFi – presupposing error-free freeze-and-slice tech yielding maybe a million brain slices of YOU, for the upload portion of things. Then – one continues to imagine – that the millions of slices are somehow accurately mapped. Neurons, synapses, state-of-life (dead, dying, alive, ), and that with ginormous computer algorithms that ALL the original connections can be derived and “reconnected” digitally into a 3-D database model that THEN might go on to act as the data-store on which a universal neural network simulation might run.

      Which in the end is supposed to be EXACTLY you. ‘cept for the box in which ‘you’ resides. You, in real time. You in past memories. You in psychotic disorders, you in tenderness, compassion, creativity, curiosity; you in your slothfulness, your lustiness, your urges of hunger, smells, sounds, appreciation of music, hobbies, habits, and political leaning.

      On sufficiently powerful hardware. With sufficient memory.

      Um… OK.
      Um… I don’t think so.


      • I’ll agree with that. Mind uploads are too far away to predict when (or if) they will happen. But you know people see an article like this and think about it, which is why I was a tiny bit surprised to see it mentioned in the article itself. (I didn’t look at the original, that might be Brian’s doing).

        Something I am confident won’t happen is quantum archeology (for resurrecting dead people). I’ve tried arguing with some of the adherents, but they haven’t even taken college-level physics and don’t grasp things like Heisenberg’s uncertainty principle, chaos theory, etc.

      • I can imagine Snorzl grabbing her own stone hammer and trying some retrophrenology.

        (Retrophrenology: Following the discovery of phrenology, that a careful mapping of the bumps on the head correlate to the personality and intellectual capabilities of the head’s owner, it becomes obvious that the way to improve said capabilities and personality is to adjust said bumps. So pass me my hammer and hold still….)

  3. So… I found there are something like 10,000,000 active neurons in a Zebrafish. Thanks Wikipedia. But nowhere did I find the number in the brain of the zebrafish. Nor in this article (at least superficially scanned.)

    So… I wonder how many neurons they (Stanfurd researchers (I’m a Cal guy)) were actually looking at.

    The technique utilized is interesting. Calcium-pathway fluorescence genetic engineering. Essentially, they cobbled together zebrafish who express both a fluorescent tag enzyme (from some sort of jellyfish, one supposes), and an attachment of that fluorescent tag to calcium-metabolic pathways. So, when neurons are firing, they glow.

    But note – there is a fair amount of misleading hubris here, too.

    I reasonably(?) expected that the ‘monitoring’ would have been the firings of millions of neurons in real-time. All those pulses. Fantastic science, if that were the case. However, ‘calcium monitoring’ is a very slow proxy to just firing rate.

    Kind of like (by poor analogy) noting the miles-per-hour of race-cars zinging around the track. Every couple of minutes. Hardly tells how they’re interacting with each other.


    So they ‘monitored’ the brain neurons. Maybe thousands of them? Maybe more. They digitized microscope images with a hi-res (4K ) video camera, at some frame rate that corresponded more or less to the calcium-channel fluorescence lapse rate. Then after, the digitized images were aligned in non-real-time to account for frame jitter, for fish-wiggle and all that. Then after the excitation, they froze the fish and sliced it into wafers; then from those, they figured out what kind of brain neurons were active. Mapping the video onto the slices.

    Do not perceive that I’m underselling this achievement. It is very powerful, and truly an awesome computational/biological science result. Truly. But it totally failed to fit my expectation of what ‘neuron monitoring in real-time’ was going to be.

    Just saying.

    • That’s kind of at odds with their statement earlier : ” This new method lets scientists record the activity of any cell type, without needing to produce special, genetically modified animals for each new experiment.”

      I wonder how they reconcile that with their actual approach, which does use bio-engineered animals.

      Regarding “real-time” analysis, they were apparently able to measure the reaction time of the fish to a predatory alarm, so I would think that would have to be something in at least tens of milliseconds response time. But you’re right, I don’t see how the glow from calcium channels could react anywhere near as fast as nerve cells do.

    • Re-reading that line I quoted (geez, I wish there was a way to edit comments here!), it seems I might have been mislead. Perhaps what they mean is that they can use a “standard” GMO animal for multiple test scenarios, instead of making a new GMO type animal for each test.


Leave a Comment