Collagen Switch Turns Fish Fins Into Limb-Like Structures

When the first fish lineage crawled onto land some 375 million years ago, its fins had to be remodeled into limbs capable of bearing weight. Evolutionary biologists have long focused on the genes that pattern the skeleton—Hox clusters, Sonic hedgehog, and other signaling pathways—but a new study argues that a critical part of the story may lie in the scaffolding around the cells. Researchers in Japan report that deleting a pair of fish-specific genes flips the orientation of collagen fibers in the zebrafish fin from parallel to perpendicular relative to the epidermis, thickening the tissue and compressing its outgrowth into a configuration that looks strikingly like a developing limb bud.

The study, published in iScience by Rintaro Tanimoto, Junpei Kuroda, and colleagues at Osaka University and JT Biohistory Research Hall, centers on actinodin proteins, called And1 and And2. These proteins are components of actinotrichia, the spear-shaped fibrillar collagen bundles that stiffen the distal tips of fish fins and guide their shape. Curiously, the actinodin genes share no sequence similarity with any known gene family except a unique repeat motif, and they were lost entirely in the tetrapod lineage—the vertebrates with limbs. Earlier work had proposed that losing these genes contributed to the fin-to-limb transition, but the developmental consequences of deleting them had never been examined directly.

To do so, the team used CRISPR-Cas9 to create zebrafish carrying disruptions in both actinodin genes. Antibody staining confirmed that the mutant alleles abolished production of functional And proteins. The researchers then turned to a visualization method developed in their laboratory: a fluorescent probe known as diaminofluorescein-FM diacetate, or DAF, which binds covalently to allysine residues in collagen crosslinking intermediates and lights up collagen fibers in whole, living tissue. Because the staining is non-destructive and the toxicity is low enough for normal tissue development, it allowed the team to image the entire three-dimensional collagen architecture of fin tissue at high resolution for the first time.

What they saw in the double-knockout larvae was dramatic. In wild-type fish, actinotrichia form uniform, radially oriented fibers, roughly two to three micrometers thick, that run parallel to the epidermis and radiate distally from the base of the fin fold. In the mutants, fluorescence was reduced, the parallel fibers were thinner, and—most strikingly—abnormal collagen bundles appeared in the interstitial spaces of the fin mesenchyme. These aberrant fibers were not randomly oriented. Quantitative analysis of fiber angles showed a clear bias toward a perpendicular alignment relative to the fin surface, with bundle diameters and spacing capped at around three micrometers. Meanwhile, the fin itself was measurably smaller along both the head-to-tail and dorsal-ventral axes, and the thickness of the interstitial mesenchyme increased by about 20 percent.

Genetic dissection revealed that either actinodin gene on its own could partially compensate: single knockouts with one functional copy showed few abnormal fibers, whereas loss of both produced the full phenotype. The same vertical collagen architecture appeared in an actinodin mutant of rainbowfish, confirming the effect is not a zebrafish quirk but a shared feature of fish fin biology. Taken together, the results indicate that And1 and And2 act redundantly to direct collagen fibers into a parallel arrangement beneath the fin epidermis, and that in their absence the fibers pivot into a perpendicular orientation that suppresses distal outgrowth while allowing tissue to thicken.

The cellular story behind the switch proved equally revealing. In wild-type fins, two cell types collaborate: basal epidermal cells, identified by the marker p63, secrete actinodin proteins and fibrillar collagens beneath the basement membrane, while mesenchymal cells migrate distally along the inner surface of the actinotrichia, elongating parallel to the epidermis in a process the authors liken to durotaxis—cell movement guided by the stiffness and structure of a scaffold. In the double knockouts, the epidermal layers looked normal, but the mesenchymal cells had abandoned their flattened, stretched morphology. Their nuclei rotated to an orientation perpendicular to the epidermis, and the cells moved deeper into the mesenchyme, surrounding themselves with the new vertically oriented collagen bundles.

Transmission electron microscopy independently confirmed the imaging results. In wild-type larvae, the fin tissue is organized into three crisp layers—epidermis, actinotrichia, and a thin mesenchymal sheet—with the collagen bundles showing the characteristic crystalline internal structure of actinotrichia. In the mutants, the parallel bundles were reduced to scattered fibers several hundred nanometers wide, and the expanded mesenchyme was threaded with vertically oriented collagen bundles displaying the classic 67-nanometer banding periodicity of collagen. Crucially, the phenotype was not confined to larvae. In adult zebrafish more than six months old, where bony fin rays replace actinotrichia as the structural framework, the tips of the fin bones lacked their usual distal actinotrichia and instead harbored abundant collagen running perpendicular to the bone, with thickened mesenchyme between the paired bone layers, shortened fins, and distorted bone segments. No defects were observed outside the fins.

The evolutionary punchline came when the team applied the same DAF imaging to axolotl limb buds. Amphibians, as tetrapods, naturally lack actinodin genes, and their developing limbs showed collagen fibers extending through the mesenchyme in a pattern essentially indistinguishable from that of the actinodin-deficient zebrafish fin buds: in both cases, roughly 80 percent of the fibers were oriented between 70 and 90 degrees relative to the epidermis. The study also found the same aberrant perpendicular fibers in the pectoral fin buds of the double-knockout zebrafish, accompanied by bud shortening and thickening of the mesenchyme around the endoskeletal disc.

The authors propose a stepwise model linking gene loss to morphological change. When actinodin function disappears, basal epidermal cells can no longer build robust parallel actinotrichia; the thin residual fibers fail to provide the scaffold cues that normally recruit mesenchymal cells along the epidermis. Those cells instead relocate into the deeper mesenchyme and deposit collagen around themselves, orienting it perpendicular to the surface. This cascade of altered cell behavior and extracellular matrix distribution ultimately reshapes the appendage into a limb-bud-like configuration—shorter and thicker rather than long and thin.

Beyond its evolutionary implications, the work elevates the extracellular matrix from passive scaffold to active agent of morphological evolution, complementing traditional models centered on transcription factors and signaling molecules. It also leaves open questions: how And proteins biochemically control fiber orientation, whether they interact with collagens, fibronectin, fibulin, fibrillin, or integrin-mediated cell adhesion, and how collagen-modifying enzymes such as lysyl oxidases and matrix metalloproteinases contribute to the remodeling. The authors note that their interpretation of mesenchymal cell dynamics rests on nuclear staining and electron microscopy, and that cell-type-specific reporters will be needed to resolve cell shape and adhesion in finer detail. Even so, the identification of mutants that dramatically reorganize collagen architecture offers a new experimental handle on one of biology’s oldest questions—how fins became limbs.

Subject of Research: The role of actinodin genes in collagen fiber orientation and fin morphogenesis in zebrafish

Article Title: A collagen orientation switch reshapes fin architecture

Article References: Tanimoto, R., Miyamoto, K., Tamura, K., Kondo, S., & Kuroda, J. (2026). A collagen orientation switch reshapes fin architecture. iScience, 29(10), Article 117464. https://doi.org/10.1016/j.isci.2026.117464

Image Credits: AI Generated

DOI: 10.1016/j.isci.2026.117464

Keywords: actinodin, collagen, zebrafish, actinotrichia, fin development, fin-to-limb transition, extracellular matrix, limb bud, CRISPR, evolution, morphogenesis, axolotl