iiinova — where your digital self keeps evolving

Joachim Fuchs @joachimfuchs.bsky.social · 31/08/2026

I guess the same applies to the text, but the change in figures just looks so much more dramatic.

Joachim Fuchs @joachimfuchs.bsky.social · 31/08/2026

I always find it amazing how much review figures or graphical abstracts change over time and with more understanding of the subject. If one slow-cooks instead of AI-generates figures that is.. 🧪🎨

1142

Joachim Fuchs @joachimfuchs.bsky.social · 23/08/2026

These neurons don't require their postsynaptic partner to sort correctly. Only through the interactions of their growth cone filopodia in a confined, quasi-2D plane, each growth cone "guides" its neighbours to their final destination. A self-organized visual map without an external organizer. 9/10

continuation of previous figure with top and side views of developing Drosophila visual map in lamina.

Panel B (pre 45h after pupa formation), shows the arrival of each growth cone from a specific angle (note: a very similar angle as it leaves the bundle it arrived from, see previous post). These growth cones re-sort collectively based on interactions with their neighbouring photoreceptor growth cones without requiring guidance signals from their future postsynaptic partner. (see Agi et al. 2024, https://doi.org/10.1126/science.adk3043.
N-Cadherin stabilizes the bond between the growth cone and its postsynaptic partner cells (L-cells).
Panel B shows a later timepoint (after 50h pupal development) where the sorting plane (lamina plexus) disassembles, and growth cones elongate along their postsynaptic partner cell where they form synaptic connections.

102

Joachim Fuchs @joachimfuchs.bsky.social · 23/08/2026

An example occurs in the developing visual map of Drosophila. Here, ca. 5000 growth cones (from the roughly 800 single eyes of a fly's compound eye) re-sort from a retinotopic (one eye = one subregion in the brain) to a visuotopic arrangement (one point in space = one subregion). 8/10

Early phase of Drosophila visual map formation. The image shows a side view and a top view of a developing lamina plexus (around 25h after pupa formation). The side view shows an individual bundle of six photoreceptor axons from one single-eye of the fly's compound eye. The neighbouring photoreceptors are linked by AJ components and sidekick (sdk), the neighbouring bundles by an adhesion-GPCR flamingo (Fmi), thereby preserving the arrangement of cell bodies in the eye to the arrangement of growth cones in the brain (as described by Kehribar et al. 2026, https://doi.org/10.1016/j.cub.2026.01.007). The regularly patterned structure in the fly eye thereby translates into a very similar pattern in the developing brain.

In the lamina plexus, each axon turns by 90° to continue extending in a quasi-2D plane where all neighbouring growth cones of the arriving photoreceptors interact.

112

Joachim Fuchs @joachimfuchs.bsky.social · 23/08/2026

In zebrafish embryos during polster cell migration, AJs coordinate which side of a cell can form filopodia. This way, so called follower cells bias where the cells in front of them grow - basically steering from the backseat. Local interactions enable the coordinated movement of a collective. 6/10

Three panels showing various zoom levels of the "guidance by followers" model as described by Boutillon et al. 2022, https://doi.org/10.1016/j.devcel.2022.05.001
First panel showing the direction of polster cell migration in zebrafish embryo (upwards towards the animal pole, shaped somewhat like a mushroom.
Second panel zooms to the cellular level highlighting filopodia with adherens junctions at their tip predominantly protruding to the direction of migration. The third panel shows the described model by the authors where mechanoresponsive engagement of AJs at the rear of cells prevents the formation of filopodia. This biases filopodia-formation to the front, and thereby steers the cellular collective to migrate forward. Local interactions that seem similar to the movements of a flock of birds

122

Joachim Fuchs @joachimfuchs.bsky.social · 23/08/2026

This enables a tissue-scale function: during wound closure, AJs move into opposite directions on the opposing fronts of epithelia (cells migrate forward by treadmilling F-actin). When the fronts meet, treadmilling strengthens AJ-binding (& seals the wound) and treadmilling & migration stops. 5/10

Reciprocal effects of adherens junctions (AJ) and cortical F-actin flow as described by Noordstra et al. 2023 https://doi.org/10.1016/j.devcel.2023.06.011.

top part shows an AJ coupled to treadmilling F-actin. The AJ moves the same direction as the F-actin it is attached to.

bottom panel: AJs bound transcellularly as well as bound to treadmilling F-actin intracellularly. Given F-actin treadmilling on these cells pulls the AJs into opposite directions, they experience mechanical force, strengthen their binding and accumulate at the interface. This transcellular bond gets so strong, that F-actin can't treadmill any further, and the cells do not move further. In effect the same mechanism that closes the wound tightly (extracellular binding & mechanoresponsive strengthening of AJs), also stops the migration of the cells.

112

Joachim Fuchs @joachimfuchs.bsky.social · 23/08/2026

One cool molecular feature: they adapt their binding strength to mechanical stress they experience. This way, connected cells can buffer acute forces to not tear apart. But at the same time, the reduced adhesion of AJs in the absence of mechanical stress allows for remodelling within tissues. 3/10

Mechanoresponsive strengthening of adherens junctions. With increasing tension, both extracellular binding strength increases (by switching the type of bond from a catch-bond to a slip-bond, and by recruiting more cadherins), as well as the binding to the intracellular actin cytoskeleton (by unfolding of alpha-catenin to link to F-actin, by additionally binding and unfolding Vinculin, and by recruiting afadin as a further linker of AJs to F-actin).
reviewed in much more detail in Campàs et al. 2024 (https://doi.org/10.1038/s41580-023-00688-7) ; Eckert et al. 2025 (https://doi.org/10.1016/j.ceb.2025.102536); James et al. 2025 (https://doi.org/10.1242/jcs.264055); Pinheiro and Bellaïche 2018 (https://doi.org/10.1016/j.devcel.2018.09.014)

112

Joachim Fuchs @joachimfuchs.bsky.social · 23/08/2026

Adherens Junctions (AJs) are cellular contact sites established by adhesive transmembrane proteins of the cadherin and nectin-family, and intracellular scaffold proteins. They link cells to each other not just by gluing their membranes, but also by connecting to the cytoskeleton. 2/10

Core adherens junction. Extracellular domains of Nectins & Cadherins interact transcellularly and bind to intracellular scaffold proteins Afadin, beta-catenin, alpha-catenin and p120-catenin. The intracellular scaffold proteins bind to the actin cytoskeleton.

214

Joachim Fuchs @joachimfuchs.bsky.social · 23/08/2026

How do molecular features of protein complexes allow cells to behave collectively? And what can we learn from collective cell migration in epithelial and mesenchymal cells to understand brain development? These questions motivated our recent review: doi.org/10.1515/hsz-... 🧪🔬🧠 1/10

Adherens junctions balance stability and motility: from cell morphogenesis to neural tissue patterning

Joachim Fuchs, Kawtar Cherkaoui and P. Robin Hiesinger

Published/Copyright: April 8, 2026

1186

Joachim Fuchs @joachimfuchs.bsky.social · 07/11/2024

The workflow could be visualized somewhat like this. Just imagine it as a 2D slice through your image. It is all based on binary operations on the masks (erosion, subtraction, "logical AND" would be the functions in e.g. FIJI). In Imaris (is this the software here?) this might have different names.

$powerpoint sketch illustrating the workflow of generating an outline of a solid object by first eroding one voxel from a copy of the object and subtracting that copy from the original object. using this outline, the area overlapping with a separate object 2 can be calculated as the intersection (logical AND) of the outline object 1 with the full object 2. The fraction of covered surface would be calculated by summing the number of pixels of the intersection and dividing that by the sum of pixels of the outline of object 1.$

110

Joachim Fuchs @joachimfuchs.bsky.social · 12/09/2023

there were some very cool efforts recently by large consortia to map down the connectomes of the complete brains of larva www.biorxiv.org/content/10.1...) and adult fruit flies www.biorxiv.org/content/10.1...). Super insightful and beautiful!

Reconstruction of all cells of the fruit fly larval brain (L1 stage). All cell bodies are found on the surface of two hemispheres with their axons and dendrites projecting inside of the brain. From a distance the image looks a bit like a collection of colorful balloons. Image from the larva connectome paper found here: https://www.science.org/doi/10.1126/science.add9330

132

Joachim Fuchs @joachimfuchs.bsky.social · 11/09/2023

Hi👋 I'm a postdoc in Berlin (#cellbiology, #neuroscience) that tries to understand how a fruit fly wires a brain in only a few days. I'm interested in improving science (culture & methodology), although my contribution mainly is amplifying ideas. Here's a nice hedgehog-cell I imaged a while ago 🤩

The image shows a cultured neuroblastoma cell that looks like a hedgehog I once found in a microscopy session.

14612