The core of CLOK is a clever pairing of two well-established technologies. The first is HaloTag, a self-labeling protein that covalently binds specific synthetic ligands; the second is the family of Janelia Fluor dyes, exceptionally bright and photostable fluorescent molecules that cross cell membranes readily. In CLOK, developing neurons are genetically engineered to express HaloTag as they differentiate. Researchers then bathe the animal in a chosen JF dye at a defined moment. Every neuron that switches on HaloTag around that time becomes permanently fluorescent in that dye’s color. Because the HaloTag-dye complex is stable, a second dye applied later cannot relabel the earlier cohort. By applying different dyes sequentially, the team can paint successive generations of neurons in distinct hues, effectively producing a time-lapse record of neurogenesis written directly into the brain of a living vertebrate.

The validation experiments were rigorous. In larval zebrafish expressing HaloTag under the elavl3 promoter, a well-known marker of early neuronal differentiation, incubation with the dye JF525 labeled 97.4 percent of HaloTag-expressing neurons, confirming near-complete labeling efficiency. Control experiments showed that the dye-HaloTag complex remains stable over days, so neurons born after a dye pulse stay dark until the next dye is applied. Unbound dye washes out almost completely within four hours, keeping background fluorescence low, and loading single or multiple dyes early in development had no measurable effect on survival, swim bladder development, or body length. When the team compared CLOK against the classic thymidine-analog birthdating method EdU, which labels cells undergoing DNA replication but requires tissue fixation, both approaches revealed similar brain-wide topographic organization. However, quantitative analysis using Gaussian mixture modeling of dye signal ratios showed that CLOK labeled significantly more specific neuronal populations across all age groups, with far less overlap between cohorts. EdU, by contrast, also stained non-neuronal progenitors in the ventricular zone and produced broader, blurrier spatial distributions.

The temporal precision of CLOK proved remarkable. Short-pulse experiments demonstrated that neurons born within a six-hour window could be clearly distinguished, suggesting a resolution of six hours or better for the elavl3 promoter. This is a dramatic improvement over methods that typically resolve cohorts only at the scale of days. The system also scales in color: by applying five different JF dyes at 24-hour intervals, the researchers simultaneously visualized five distinct neuronal age cohorts within a single living fish, reconstructing the chronological architecture of the entire larval brain in one animal rather than averaging across many specimens.

One of CLOK’s most striking capabilities is subcellular resolution. By fusing HaloTag to localization sequences, the team targeted the protein either to axon terminals, using a synaptophysin tag, or to somatodendritic compartments, using the Kv2.1 localization sequence from the Voltron1 construct. This revealed an unexpected level of chronological organization across the brain. In regions including the optic tectum, cerebellum, and hindbrain, axon terminals from progressively later-born populations occupied successively more peripheral layers of the neuropil, while most dendritic domains were dominated by the earliest-born neurons. In other words, the brain not only organizes neuronal cell bodies by age but also arranges their processes according to birthdate, a layer of developmental order that had never been visualized at this scale before.

The labeling also proved remarkably durable. Neurons tagged at larval stages remained clearly fluorescent for up to eight weeks, well into juvenile stages, and dyes could still be successfully incorporated at 14 days post fertilization. Triple-labeling experiments confirmed that distinct birthdate cohorts remained separable weeks after the original dye pulses. This persistence opens the door to longitudinal studies that follow the same neurons from their birth through their integration into circuits and beyond, something previously impossible with fixation-based methods.

Crucially, CLOK is not just an anatomical tool; it is compatible with functional imaging. The team combined birthdate labeling with the calcium indicator GCaMP6s and presented head-fixed larvae with a battery of visual stimuli, including flashes, moving gratings, and looming shadows that mimic approaching predators, while performing two-photon calcium imaging in the optic tectum. Early-born tectal neurons, which occupy central regions, generally responded more strongly than late-born peripheral neurons, particularly to moving gratings and looming stimuli. Bright flashes, by contrast, evoked similar responses in both populations, suggesting comparable baseline visual sensitivity. Single-cell analysis using the ratio of the two dyes as a continuous birthdate proxy confirmed that older neurons were more likely to be activated across all stimulus types. The team also found that direction selectivity, the ability of neurons to respond preferentially to motion in a specific direction, was stronger and more consistent in early-born neurons, consistent with the idea that this property emerges progressively during development.

The approach worked even in brain regions where position does not predict age. Tectal neuropil interneurons, whose somata are spatially intermingled regardless of when they were born, could be sorted by birthdate using the dual-dye ratio, demonstrating that CLOK can establish birthdate-function relationships even where anatomy alone is uninformative. In the hindbrain, the team birthdated neurons born before and after 24 hours post fertilization and tracked their activity during fictive locomotion at two and four days post fertilization. At the earlier stage, fish could escape in response to electrical shocks but rarely swam spontaneously, and only a small proportion of late-born neurons were active during these behaviors. By four days, fish had acquired spontaneous swimming with a much broader repertoire of muscle activity patterns, and the late-born population showed a dramatic increase in both the proportion of active neurons and the diversity of their functional clusters, eventually exceeding the diversity of the early-born group. Notably, early-born neurons were also recruited into behaviors that did not exist at the earlier stage, hinting at functional coordination between age groups during behavioral maturation.

Perhaps most ambitiously, the researchers coupled CLOK to chemigenetic optical actuators and sensors, creating a complete birthdate-resolved toolkit for manipulating and measuring neural activity. They validated the approach with WHaloCaMP, a calcium indicator built on a HaloTag scaffold, showing that spontaneous activity could be recorded separately from early- and late-born neurons, with the early-born population displaying higher event frequencies. They applied the same logic to Voltron2, a rhodopsin-based voltage indicator, recording fast fluorescent transients corresponding to individual action potentials in the olfactory epithelium of birthdate-defined populations. Finally, they fused HaloTag to the CoChR opsin and performed all-optical experiments in which specific neurons were identified by their birthdate colors, selectively stimulated with two-photon holographic illumination, and monitored via GCaMP calcium imaging. Photostimulation-evoked responses were similar whether early- or late-born neurons were targeted, demonstrating that either cohort can be activated at will with equal efficiency. Notably, the JF dyes’ two-photon absorption profile, with minimal absorption near 920 nanometers, makes them ideal partners for GFP-based indicators, and GFP-tagged opsins produced substantially larger imaging-induced photocurrents than the dye-tagged versions, highlighting a real technical advantage of the chemigenetic design.

The authors emphasize that CLOK’s power comes from its flexibility. Because HaloTag itself is not fluorescent, transgenic lines are spectrally neutral, letting experimenters choose dyes that best complement whatever fluorescent proteins or indicators their experiment requires. The method is primarily demonstrated in zebrafish, where dye penetration through the skin is excellent, but the authors note that the underlying mechanism could extend to other models including worms, flies, mice, and even brain organoids, as well as to studies of glial development and hematopoietic lineages. Some caveats remain: promoter choice affects temporal sharpness, dye bioavailability varies with lactone-zwitterion equilibrium, and interpreting dye ratios as continuous age measures requires careful validation. Still, by uniting single-cell birthdating, functional imaging, and all-optical circuit manipulation within one framework, CLOK gives developmental neuroscientists something they have long lacked: the ability to ask not just where a neuron is and what it does, but exactly when it was born, and to watch how that birthdate shapes its journey into a working circuit.

Subject of Research: A chemigenetic multicolor labeling system for visualizing neuronal birthdate and circuit integration in developing zebrafish

Article Title: CLOK: a chemigenetic multicolor labeling system to visualize neuronal birthdate and circuit integration

Article References: Faini, G., Tuffery, M., Saleem, A., Zhang, L., Du, F., Le Bourdelles, G., Duroure, K., Schreiter, E. R., Tanese, D., Emiliani, V., Del Bene, F., & Koyama, M. (2026). CLOK: a chemigenetic multicolor labeling system to visualize neuronal birthdate and circuit integration. Nature Neuroscience. https://doi.org/10.1038/s41593-026-02423-9

Image Credits: AI Generated

DOI: 10.1038/s41593-026-02423-9

Keywords: CLOK, HaloTag, Janelia Fluor dyes, neuronal birthdating, zebrafish, neurogenesis, circuit development, calcium imaging, voltage imaging, optogenetics, optic tectum, chemigenetic indicators