The developing Drosophila wing imaginal disc is sculpted not by differences in growth between its two epithelial layers but by a difference in how the extracellular matrix (ECM) enveloping the organ grows: the ECM on the peripodial (top) side grows in the plane, while the ECM on the disc-proper (bottom) side grows in 3D and thickens, so the faster-growing tissue buckles against its constraint into a dome. Harmansa, Erlich, Eloy, Zurlo and Lecuit (Nature Communications 14:1220, 2023) captured this with an axisymmetric neo-Hookean bilayer model and showed that the matrix metalloproteinase MMP2, expressed at high levels in the peripodial epithelium, is required for planar growth of the top ECM, proposing that differential ECM turnover and the resulting viscous dissipation set the difference between layers. Their Discussion poses as open questions: what molecular mechanisms actually establish the difference in ECM growth anisotropy between the top and bottom layers, and how the model must be extended — to the full non-axisymmetric 3D morphology including dorsal-ventral folds, to actomyosin contractility during pupal disc eversion, and to realistic ECM material behavior (strain-stiffening beyond the neo-Hookean law, hyaluronan-mediated osmotic swelling, polar growth anisotropy) — to capture complete organ morphogenesis. No published work determines what establishes the layer-to-layer difference in ECM growth anisotropy between the top and bottom layers, and no non-axisymmetric, contractility-coupled, or non-neo-Hookean model of the disc has been published. The problem remains open.

Background

Imaginal discs are the larval primordia that give rise to adult Drosophila appendages; the wing imaginal disc is a sac-like epithelial organ whose main wall, the disc proper (DP), is a pseudostratified epithelium, while the opposite, much thinner wall is the peripodial epithelium (PPE). Each epithelium secretes and rests on a basement membrane (BM), a thin sheet-like extracellular matrix (ECM) composed mainly of collagen IV, laminins, perlecan and nidogen; in the disc this ECM envelope is called the top (peripodial-side) and bottom (DP-side) ECM layer. During larval development the disc grows about a hundred-fold in volume and progressively bends into a dome, a shape change that is later used when, at metamorphosis, the disc everts (turns inside out through a constricting opening) to form the pupal wing. Harmansa et al. (2023) measured 3D growth of both tissue layers and both ECM layers over time and found that the two tissue layers both grow essentially in the plane, whereas the bottom ECM grows in 3D and increases in thickness while the top ECM grows in the plane. Growth anisotropy here means that the growth tensor — the part of the deformation describing stress-free volume addition — has unequal components along the sheet normal and within the sheet. Because the tissue grows more in magnitude than the constraining bottom ECM expands in area, the layers are geometrically frustrated, elastic stress accumulates, and the composite bends: differential growth anisotropy between tissue and ECM, not differential growth between the two tissue layers, drives disc doming. The authors captured this quantitatively with a mechanical bilayer model, treating the DP plus its ECM as layers of a nearly incompressible neo-Hookean material — the simplest constitutive law of finite-strain (nonlinear) elasticity, whose stress derives from one parameter, the shear modulus — with growth implemented as a stress-free anisotropic expansion, and with axisymmetry and a force-free disc edge assumed. The model reproduced measured dome shapes, predicted the degree of ECM growth anisotropy and the stiffness ratio between layers, and matched perturbation experiments: enzymatically digesting the ECM (collagenase) or decellularizing it abolished doming. Mechanistically, the matrix metalloproteinase MMP2 (an enzyme that degrades ECM components such as collagen IV), which is expressed far more highly in the PPE than in the DP, is required for planar growth of the top ECM: tissue-specific MMP2 knockdown converts top-ECM growth toward the thickness-growing mode of the bottom ECM. This led to the hypothesis that local MMP2 raises ECM turnover, and the associated viscous dissipation (stress relaxation) of elastic energy, favoring planar integration of newly secreted matrix, whereas low turnover in the bottom ECM favors thickness growth. At pupariation, degradation of the ECM controlled by ecdysone (the steroid hormone that triggers insect molting and metamorphosis) and lysis of the PPE release the stored elastic energy and allow eversion, a process in which supracellular actomyosin cables (contractile actin-myosin II bundles spanning many cells) are proposed to form a constrictive ring. The study's Discussion leaves several issues open, anchored by the explicit question of what controls ECM growth anisotropy and what differs between the top and bottom layers, together with the named model limitations: the axisymmetric, force-free-edge idealization fits the anterior-posterior direction better than the dorsal-ventral direction where complex folds form; actomyosin contractility is absent from the model; the polar (radial versus orthoradial) components of the growth tensor were assumed equal, an assumption that perturbations could invalidate; and real ECMs can strain-stiffen (stiffen under large deformation, captured by Fung or Gent constitutive models with a strain-hardening parameter) or swell osmotically via hyaluronan (a large polysaccharide that binds water), effects the neo-Hookean law ignores. Subsequent literature has made the general point vivid: basement membranes are now treated as dynamic, actively regulated materials whose turnover controls their mechanics (e.g., collagen-IV lifetime setting the stress-relaxation time of the network), and protease-dependent BM patterning controls organ shape in other Drosophila organs. What is still missing is the layer-resolved, causal account for the wing disc itself.

Problem Statement

As posed in the Discussion of Harmansa et al. (2023), the problem is to complete the account of how the ECM sculpts the developing wing disc, with one central mechanistic question and linked model-completion questions. Central question: what controls the growth anisotropy of the ECM, and what differs between the top (peripodial-side) and bottom (disc-proper-side) ECM layers during larval growth? In particular, is the planar growth of the top ECM versus the thickness growth of the bottom ECM established by differential ECM turnover rates and the associated viscous dissipation controlled by MMP2 and possibly other enzymes — high turnover favoring planar integration of new matrix, low turnover favoring thickness growth — or by another mechanism, such as layer-specific local production of ECM components (e.g., laminins) or differential availability of ECM components? A causal answer must identify the molecular difference between the two ECM layers and demonstrate, by loss- and gain-of-function in a layer-resolved manner, that it is necessary and sufficient to switch ECM growth between the planar and thickness-growing modes. Model-completion questions, which operationalize the mechanism within the mechanical framework: can the axisymmetric neo-Hookean bilayer model be extended to (a) the complete non-axisymmetric 3D disc morphology, including the complex folds of the dorsal-ventral direction; (b) pupal disc eversion, which requires incorporating actomyosin contractility, including possible supracellular actomyosin cables forming a constrictive ring; and (c) more realistic ECM material behavior — strain-stiffening beyond the incompressible neo-Hookean law (e.g., Fung or Gent constitutive models), osmotic swelling mediated by hyaluronan and contributions of other ECM components — while evaluating whether polar (radial versus orthoradial) growth anisotropy, negligible in the unperturbed disc, becomes significant under perturbation? Endowed with the identified mechanism and the named missing physics, the extended model must quantitatively reproduce the disc's measured 3D morphogenesis — including dome formation, fold geometry, and the response to perturbations such as collagenase digestion and MMP2 knockdown.

The central mechanistic question and the model extensions (a)–(c) are independently attackable but jointly required: a full account joins the layer-resolved causal identification of the mechanism with the model extensions (a), (b) and (c), including the polar-anisotropy evaluation.

References

  1. Harmansa, S., Erlich, A., Eloy, C., Zurlo, G., Lecuit, T. (2023). Growth anisotropy of the extracellular matrix shapes a developing organ. Nature Communications 14:1220. DOI: 10.1038/s41467-023-36739-y. https://doi.org/10.1038/s41467-023-36739-y (source of the open problem)
  2. Harmansa, S., Erlich, A., Lecuit, T. (2026). Mechanical regulation of cuboidal-to-squamous epithelial transition in the Drosophila developing wing. Current Biology 36(6):1525-1540. DOI: 10.1016/j.cub.2026.02.035. https://doi.org/10.1016/j.cub.2026.02.035
  3. Martín, M., Guarch, E., Pastor-Pareja, J. C. (2026). Tissue morphogenesis: Regional differences in basement membrane mechanics influence epithelial geometry. Current Biology 36(6):R249-R252. DOI: 10.1016/j.cub.2026.01.063. https://doi.org/10.1016/j.cub.2026.01.063
  4. Erlich, A., Harmansa, S. (2025). How growth-induced stresses guide shape changes during animal morphogenesis: Mechanisms and implications. Seminars in Cell & Developmental Biology 175:103661. DOI: 10.1016/j.semcdb.2025.103661. https://doi.org/10.1016/j.semcdb.2025.103661
  5. Mai, J., Anil Gayathry, A., Richards, D. M., Harmansa, S. (2026). Multiscale biophysical control of epithelial shape transitions during animal development. Seminars in Cell & Developmental Biology 184:103686. DOI: 10.1016/j.semcdb.2026.103686. https://doi.org/10.1016/j.semcdb.2026.103686
  6. Barrientos, R., Meadowcroft, B., Sánchez-Sánchez, B. J., Stramer, B. M., Paluch, E. K., Charras, G., Banerjee, S., Šarić, A., Mao, Y. (2026). Basement membrane turnover controls cell shape. Cell (online ahead of print, July 2026). DOI: 10.1016/j.cell.2026.07.010. https://doi.org/10.1016/j.cell.2026.07.010
  7. Hinata, C., Nakagawa, H., Nonaka, S., Nozaki, K., Sato, Y., Ohsawa, S. (2026). Extrinsic MMPs drive epithelial shape change via basal extracellular matrix disassembly in the Drosophila wing disc. Communications Biology. DOI: 10.1038/s42003-026-10376-9. https://doi.org/10.1038/s42003-026-10376-9
  8. Song, Y., Martín, P., Sun, T., Fernández-Herrero, J., Sánchez-Herrero, E., Pastor-Pareja, J. C. (2025). Mechanical coupling between dorsal and ventral surfaces shapes the Drosophila haltere. Current Biology 35(13):3090-3105. DOI: 10.1016/j.cub.2025.05.040. https://doi.org/10.1016/j.cub.2025.05.040
  9. Töpfer, U., Ryu, J., Guerra Santillán, K. Y., Schulze, J., Fischer-Friedrich, E., Tanentzapf, G., Dahmann, C. (2024). AdamTS proteases control basement membrane heterogeneity and organ shape in Drosophila. Cell Reports 43(7):114399. DOI: 10.1016/j.celrep.2024.114399. https://doi.org/10.1016/j.celrep.2024.114399
  10. Ku, H.-Y., Bilder, D. (2025). Basement membrane patterning by spatial deployment of a secretion-regulating protease. Proceedings of the National Academy of Sciences 122(20):e2412161122. DOI: 10.1073/pnas.2412161122. https://doi.org/10.1073/pnas.2412161122
  11. Guerra Santillán, K. Y., Dahmann, C., Fischer-Friedrich, E. (2025). ECM proteins shape topographical patterns in the basement membrane of Drosophila wing discs. Matrix Biology 140:78-87. DOI: 10.1016/j.matbio.2025.06.003. https://doi.org/10.1016/j.matbio.2025.06.003
  12. Fuhrmann, J. F., Schimmenti, V. M., Cwikla, G., Lee, S., Yuan, M., Wilsch-Bräuninger, M., Jülicher, F., Popović, M., Dye, N. A. (2025). Apical extracellular matrix regulates fold morphogenesis in the Drosophila wing disc. bioRxiv preprint. DOI: 10.1101/2025.09.06.674631. https://doi.org/10.1101/2025.09.06.674631
  13. Hadjaje, S., Andrade-Silva, I., Dalbe, M.-J., Clément, R., Marthelot, J. (2024). Mechanics of Drosophila wing deployment. Nature Communications 15:10577. DOI: 10.1038/s41467-024-54527-0. https://doi.org/10.1038/s41467-024-54527-0
  14. Kumar, N., Mim, M. S., Dowling, A., Zartman, J. J. (2024). Reverse engineering morphogenesis through Bayesian optimization of physics-based models. npj Systems Biology and Applications 10:49. DOI: 10.1038/s41540-024-00375-z. https://doi.org/10.1038/s41540-024-00375-z
  15. Dilsha, C., Shiju, S., Shah, N. A., Inamdar, M. M., Shashidhara, L. S. (2024). Differences in Cellular mechanics and ECM dynamics shape differential development of wing and haltere in Drosophila. bioRxiv preprint. DOI: 10.1101/2024.04.26.591286. https://doi.org/10.1101/2024.04.26.591286