Molecular and functional dissection using CaMPARI-seq reveals the neuronal organization for dissociating optic flow-dependent behaviors

Background:

Early work in zebrafish established that optic flow, the retinal motion pattern generated during self-movement, drives two complementary, conserved behaviors: the optomotor response (OMR) for stabilizing translation and the optokinetic response (OKR) for stabilizing rotation^1-4. But how are translational vs rotational cues segregated and routed from retina to premotor circuits? Method overhauls, particularly real-time calcium imaging and virtual-reality stimulation, made it possible to measure these transforms across brain areas with cellular resolution^5-7.

Anatomically, the zebrafish pretectum emerged as the major optic-flow hub that sits just downstream of direction-selective retinal ganglion cells (DS-RGCs). The retinal projectome revealed precise, brain-area-specific targeting of RGC axons, including dense inputs to arborization fields AF5/AF6 in pretectum^8-9.  Functionally, whole-brain recordings combined with network modeling produced brain-scale circuit descriptions that link pretectal representations to downstream premotor pathways, showing how visual motion maps onto OMR control³. Parallel studies charted parallel channels and population codes for motion in pretectum/tectum, showing that rotation- and translation-preferring neurons occupy partially segregated territories and that binocular integration relies on interhemispheric interactions^10,12,13. Together, previous results support a working model in which DS-RGC streams are integrated in the pretectum into distinct rotation vs. translation channels that feed separate ocular and locomotor controllers¹.

What remained largely unresolved was molecular identity and causal subtype assignment within these functionally defined pretectal populations. In this preprint, Matsuda and colleagues use CaMPARI-seq in an attempt to close this gap by photoconverting pretectal neurons active during defined OMR/OKR stimuli, FACS-isolate the labeled cells, and perform scRNA-seq to directly couple activity with molecular identity. This reveals a tcf7l2-positive, largely GABAergic pretectal population that splits into mafaa⁺ local neurons and nkx1.2lb⁺ commissural neurons, linking specific transcriptomic types to distinct circuit roles in optic-flow-driven behavior (Fig. 1).

 

Fig. 1 – Anatomical and functional characteristics of nkx1.2lb+ pretectal neurons. Fig. 1 of the preprint, made available under a CC-BY-NC-ND 4.0 International license. 

Key Findings:

A new functional transcriptomics pipeline (CaMPARI-seq):

The authors engineered Tg(elavl3:NLS-CaMPARI2) larvae to photoconvert only neurons active during optic-flow stimulation. By coupling this with FACS and scRNA-seq, they profiled ~15,000 neurons and identified 30 transcriptional clusters, of which eight were enriched for optic flow-responsive activity. This represents the first activity-dependent molecular catalog of pretectal neurons in zebrafish.

tcf7l2 identifies a pretectal optic-flow hub:

Among enriched clusters, cluster 5 expressed tcf7l2, pax7a, tal2, and gata3 genes that delineate the pretectum. Importantly, ~80% of these cells were gad1b⁺ GABAergic. Using CRISPR-based knock-ins and Gal4 driver lines, the authors confirmed dense labeling of pretectal neurons by tcf7l2, and in vivo calcium imaging showed they encompass all major monocular and binocular flow-selective response types, positioning them as the core molecular handle for optic-flow circuits

Molecular subtypes with distinct anatomy and function:

Reclustering of tcf7l2⁺ neurons revealed seven molecular subtypes, each defined by distinct markers (e.g., nkx1.2lb, mafaa, npy, penkb). HCR mapping showed that these subclusters occupy non-overlapping anatomical domains of the pretectum. Notably: 1) mafaa⁺ neurons localized to the lateral pretectal migrated area (M1), projecting locally within AF5/AF6 where DS-RGC axons arborize, and functionally tuned to monocular temporal motion. 2) nkx1.2lb⁺ neurons localized dorsomedially, projecting commissurally via the posterior commissure to the contralateral pretectum, and broadly responded to multiple optic-flow types.

Revising circuit models of optic flow:

Earlier models posited that commissural connectivity was excitatory, relayed via interneurons. Here, genetic access shows instead that direct inhibitory commissural cells (nkx1.2lb⁺) mediate interhemispheric suppression, providing a mechanistic basis for how binocular translation signals are selectively computed in the pretectum (Fig. 2). This shifts the field’s understanding of circuit logic from an indirect to a direct inhibitory framework.

Fig. 2 – Proposed circuit model. Model for the optic flow processing circuit for controlling OKR and OMR. Fig. 2 of the preprint, made available under a CC-BY-NC-ND 4.0 International license.

Why I highlight this preprint?

I was particularly struck by how the authors combine functional imaging, intersectional genetics and behavior to demonstrate causal specificity: targeted manipulation shows that nkx1.2lb⁺ neurons are required for OMR but not OKR, linking a transcriptomic subtype to the computational problem of translational-flow integration. This precision exemplifies how zebrafish systems neuroscience can reveal conserved principles of visual processing.

What makes the study transformative is twofold. Methodological ingenuity: CaMPARI-seq elegantly solves the “typing problem” by capturing the transcriptomes of neurons that are actively engaged during defined behaviors, avoiding the sampling biases of anatomy-only scRNA-seq. The nuclear CaMPARI2 design which keeps photoconverted signal in the soma during dissociation greatly improves cell recovery and sets a practical template for activity→genome mapping in behaving animals. Causal precision: by pairing genetic access with behavior, the authors show that commissural inhibition, not excitation, implements interhemispheric crosstalk for translational flow; nkx1.2lb⁺ GABAergic neurons are therefore essential circuit elements for OMR.

In doing so, the paper overturns two common assumptions in the field: 1) “Binocular computation requires excitatory commissural neurons” → False: direct inhibitory (nkx1.2lb⁺) commissural neurons mediate interhemispheric interactions. 2) “Pretectal outputs are primarily excitatory” → False: ~80% of optic-flow responsive pretectal cells are GABAergic.

On a personal note, I recently visited Prof. Kubo’s lab and saw first-hand how meticulous their functional imaging and genetic pipelines are. I admire their lab’s blend of careful physiology, elegant genetic access and rigorous behavioral readouts; that mixture is why I chose to highlight this preprint. The study gives the community robust genetic handles (tcf7l2, mafaa, nkx1.2lb) that I expect will accelerate causal circuit studies across labs.

Questions for the authors:

1. Do nkx1.2lb⁺ commissural neurons synapse directly onto descending OMR pathways (nMLF/reticulospinal neurons) or do they instead shape local computation via interneurons? Would paired optogenetics + EM/volume connectomics resolve this?
2. Given the predominance of GABAergic pretectal cells, how do local glutamatergic interneurons and DS-RGC inputs converge on mafaa⁺ versus nkx1.2lb⁺ neurons to balance excitation and inhibition during OKR?
3. mafaa⁺ local GABAergic cells receive AF5 inputs, do they also sample binocular inputs or are they strictly monocular? Could local inhibition implement receptive-field size tuning?
4. Does experience or habituation modulate the photoconversion profiles of these subtypes, indicating adaptive circuit tuning?

References:

1. Matsuda, K., & Kubo, F. (2021). Circuit organization underlying optic flow processing in zebrafish. Frontiers in Neural Circuits, 15, 709048.
2. Orger, M. B. (2016). The cellular organization of zebrafish visuomotor circuits. Current Biology, 26(9), R377-R385.
3. Naumann, E. A., Fitzgerald, J. E., Dunn, T. W., Rihel, J., Sompolinsky, H., & Engert, F. (2016). From whole-brain data to functional circuit models: the zebrafish optomotor response. Cell, 167(4), 947-960.
4. Gebhardt, C., Baier, H., & Bene, F. D. (2013). Direction selectivity in the visual system of the zebrafish larva. Frontiers in neural circuits, 7, 111.
5. Ahrens, M. B., Orger, M. B., Robson, D. N., Li, J. M., & Keller, P. J. (2013). Whole-brain functional imaging at cellular resolution using light-sheet microscopy. Nature methods, 10(5), 413-420.
6. Duchemin, A., Privat, M., & Sumbre, G. (2022). Fourier motion processing in the optic tectum and pretectum of the zebrafish larva. Frontiers in Neural Circuits, 15, 814128.
7. Muto, A., Ohkura, M., Abe, G., Nakai, J., & Kawakami, K. (2013). Real-time visualization of neuronal activity during perception. Current Biology, 23(4), 307-311.
8. Robles, E., Laurell, E., & Baier, H. (2014). The retinal projectome reveals brain-area-specific visual representations generated by ganglion cell diversity. Current Biology, 24(18), 2085-2096.
9. Kölsch, Y., Hahn, J., Sappington, A., Stemmer, M., Fernandes, A. M., Helmbrecht, T. O., … & Baier, H. (2021). Molecular classification of zebrafish retinal ganglion cells links genes to cell types to behavior. Neuron, 109(4), 645-662.
10. Kramer, A., Wu, Y., Baier, H., & Kubo, F. (2019). Neuronal architecture of a visual center that processes optic flow. Neuron, 103(1), 118-132.
11. Baier, H., & Wullimann, M. F. (2021). Anatomy and function of retinorecipient arborization fields in zebrafish. Journal of Comparative Neurology, 529(15), 3454-3476.
12. Wang, K., Hinz, J., Zhang, Y., Thiele, T. R., & Arrenberg, A. B. (2020). Parallel channels for motion feature extraction in the pretectum and tectum of larval zebrafish. Cell Reports, 30(2), 442-453.
13. Portugues, R., Feierstein, C. E., Engert, F., & Orger, M. B. (2014). Whole-brain activity maps reveal stereotyped, distributed networks for visuomotor behavior. Neuron, 81(6), 1328-1343.

Tags: #systemsneuroscience, zebrafish

doi: https://doi.org/10.1242/prelights.41791

Read preprint (No Ratings Yet)