Scientists Can Watch Cellular Activity Across a Living Vertebrate

A seven-day-old zebrafish lies on its side beneath a microscope, held in a drop of clear agarose. It looks almost still. Inside, it is anything but.

In recordings of larvae like this one, cells flicker across the brain, muscles, kidney, gut, gills and skin. Across separate animals and experiments, researchers found that cold water activated cartilage cells, waves travelled through the kidney and low oxygen caused blood flow to the gut to nearly stop while circulation to the brain and muscle remained largely steady.

Several of these responses had not been reported before. They emerged because researchers were no longer watching one organ at a time.

The system responsible is called WHOLISTIC, short for WHole-Organism Live-Imaging System for recording Tissue and IntraCellular activity. Described in a peer-reviewed Nature paper, it combines engineered fish, three-dimensional fluorescence microscopy and computational methods that can follow cellular-scale signals even as muscles contract and organs shift.

It is a way to search a living vertebrate for coordination across many organs, cell by cell.

Making cellular activity visible

The first ingredient is a transparent animal. Young zebrafish have many organs with counterparts in other vertebrates, but light can pass through much of their bodies. That makes the larvae unusually useful for observing living tissue without surgery.

The researchers engineered the fish to produce GCaMP7f in almost every cell. GCaMP is a fluorescent protein whose brightness changes with the concentration of calcium inside a cell. Calcium participates in processes ranging from neural activity to muscle contraction, secretion and metabolism, so its fluctuations provide a broad, incomplete measure of cellular activity.

That distinction matters. WHOLISTIC is not reading thoughts, and it is not recording every chemical message in the body. A flash in a neuron does not necessarily mean the same thing as a flash in cartilage or kidney tissue. The system follows one widespread molecular signal, then uses anatomy and targeted experiments to investigate what particular patterns might mean.

The fish were mounted on their right sides in agarose rather than allowed to swim freely. A spinning-disk confocal microscope scanned a typical volume measuring 800 by 600 micrometres and about 130 micrometres deep. A complete volume was recorded approximately once every 3.3 seconds. Most experiments concentrated on the anterior third of the animal, where the major internal organs are packed.

It would therefore be misleading to say that the researchers continuously tracked every cell in a complete fish. The defensible claim is that the sensor was expressed across nearly all cells and that the system recorded cellular activity across most organ systems in a transparent vertebrate.

A body that refuses to hold still

Making cells glow solved only part of the problem. A living body bends, twitches, pumps, squeezes and drifts. Even an immobilized fish contains moving organs. If a cell changes position between frames, software can mistake movement for activity or lose the cell entirely.

The team recorded an anatomical reference channel alongside the calcium signal, then used an optical-flow algorithm to estimate local tissue movement in small patches. After aligning the volumes, another program divided the fluorescence into cellular-scale fragments. The researchers grouped those fragments according to the timing of their activity.

This organized the data into what the team calls functional tissue ensembles. Cells with related rhythms often mapped onto recognizable parts of the kidney, liver, gallbladder, gut or muscle.

In one kidney example, the software divided 1,265 cellular fragments into 11 ensembles. Their calcium activity did not rise together. It travelled along the nephron in bursts, followed by quiet intervals.

That finding raises the possibility that filtration or fluid movement in the larval kidney is more intermittent on short timescales than researchers assumed. It does not prove it. The waves were observed, but their physiological meaning still needs a dedicated experiment.

A separate cold-response screen produced another surprise. After researchers added cold water, osteochondral cells near the skull showed synchronized calcium responses. Expanded-tissue imaging and a cartilage gene marker helped identify them as chondrocytes, the principal cells of cartilage.

The response appeared in six cold-exposed animals and was largely absent from two room-temperature controls. The fish were anesthetized for this experiment, reducing the chance that movement or broad neural activity produced the signal. The paper describes the cold response as previously unknown, but the small sample makes independent confirmation important.

Muscle activity revealed an unexpected arrangement too. A small anterior portion of the upper trunk muscle sometimes coordinated with abdominal muscle instead of the neighboring trunk region. Anatomical tracing showed that this section receives different nerve input, providing a structural explanation for the newly reported functional split.

The system also detected ependymal cells, which line fluid-filled spaces in the brain and spinal cord. Their calcium rose for about 15 seconds after motor events. During motor quiescence, they produced much slower oscillations with periods of roughly three to seven minutes.

Those findings make the cells plausible participants in cerebrospinal-fluid regulation and rest-related physiology. They do not establish a role in sleep.

Watching the hindbrain redirect blood

The clearest demonstration of WHOLISTIC's value came from a set of experiments in which researchers lowered oxygen from 21 percent to 10 percent.

Calcium levels changed across several organs, particularly in gastrointestinal tissues. When the team inspected the images before motion correction, the mesenteric artery supplying the viscera appeared to constrict. The researchers then used fish with fluorescent blood vessels and red blood cells to measure the response more directly.

Under hypoxia, estimated blood flow to the gut nearly stopped. Flow to the brain and muscle remained largely unchanged, and gut circulation recovered when normal oxygen returned.

That pattern suggested active control rather than a general circulatory failure. When researchers anesthetized the nervous system, hypoxia no longer produced the same reduction in visceral blood flow.

They then performed a more targeted test. The team engineered neurons to carry a light-sensitive inhibitory protein. Shining light on the hindbrain during hypoxia restored near-normal blood flow to the gut within seconds. The improvement lasted as long as the hindbrain was inhibited. Illuminating a control location did not produce the same effect.

The experiment supports a strong but specific conclusion: hindbrain activity was necessary for the mesenteric artery constriction in this larval-fish assay. The idea that this circuit exists to preserve oxygen for the brain and muscle is plausible, but the authors identify it as speculation rather than an established purpose.

A separate optogenetic experiment demonstrated another way to move from correlation to causation. Activity in part of the motor vagus was associated with a sphincter near the stomach. Stimulating selected vagal neurons with light caused a dose-dependent response in the sphincter. Moving the stimulation site along the motor vagus activated different tissues along the throat and anterior gut.

A discovery tool, not a miniature human scanner

WHOLISTIC has not solved whole-body physiology. Its immediate value is that researchers can cast a much wider net.

An experiment focused on neurons might never look for cold-sensitive cartilage. A kidney study might not record muscles, skin and brain simultaneously. A conventional drug test might measure whether a compound affects its intended target while missing an unexpected response elsewhere.

The researchers also created a calcium-sensor line in Danionella cerebrum, a tiny fish that remains transparent as an adult. For now, that is a proof of concept rather than evidence that the full platform can monitor every system during complex adult behavior.

Other limits remain. Calcium sensors cannot capture every fast signal. Transgene expression can vary among fish lines, and several findings came from small exploratory samples. The approach depends on transparent animals and offers no obvious route to watching cellular activity throughout a mouse or human. No independent replication of the newly published results has yet been identified.

Two authors are also co-inventors on an HHMI provisional patent application related to the method. The paper nevertheless makes raw data and analysis code publicly available, while the WHOLISTIC project site provides protocols and other resources for researchers who want to test or extend the system.

For more than a century, physiology has often learned about bodies by isolating their parts. WHOLISTIC offers a complementary starting point: observe many tissues together, find patterns nobody thought to search for, and return with sharper experiments.

The fish does not reveal every secret. It does, however, bring an extraordinary number of organs onto the same screen.