The body axis of vertebrate embryos is periodically segmented into bilaterally symmetric pairs of somites1,2. The anteroposterior length of somites, their position and left–right symmetry are thought to be molecularly determined before somite morphogenesis3,4. Here we show that, in zebrafish embryos, initial somite anteroposterior lengths and positions are imprecise and, consequently, many somite pairs form left–right asymmetrically. Notably, these imprecisions are not left unchecked and we find that anteroposterior lengths adjust within an hour after somite formation, thereby increasing morphological symmetry. We find that anteroposterior length adjustments result entirely from changes in somite shape without change in somite volume, with changes in anteroposterior length being compensated by corresponding changes in mediolateral length. The anteroposterior adjustment mechanism is facilitated by somite surface tension, which we show by comparing in vivo experiments and in vitro single-somite explant cultures using a mechanical model. Length adjustment is inhibited by perturbation of molecules involved in surface tension, such as integrin and fibronectin. By contrast, the adjustment mechanism is unaffected by perturbations to the segmentation clock, therefore revealing a distinct process that influences morphological segment lengths. We propose that tissue surface tension provides a general mechanism to adjust shapes and ensure precision and symmetry of tissues in developing embryos.
This is a preview of subscription content
Subscribe to Nature+
Get immediate online access to the entire Nature family of 50+ journals
Subscribe to Journal
Get full journal access for 1 year
only $3.90 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Tax calculation will be finalised during checkout.
Get time limited or full article access on ReadCube.
All prices are NET prices.
Map-projected light sheet data are available at Zenodo (https://doi.org/10.5281/zenodo.4146919). Raw imaging data combining both the multiview and single-view light-sheet datasets are 15 TB in size and are available on request from the corresponding author without any restrictions. Source data are provided with this paper.
All custom-developed image analysis codes are available at GitHub (https://github.com/sundar07/SomSeg).
Oates, A. C., Morelli, L. G. & Ares, S. Patterning embryos with oscillations: structure, function and dynamics of the vertebrate segmentation clock. Development 139, 625–639 (2012).
Naganathan, S. R. & Oates, A. C. Patterning and mechanics of somite boundaries in zebrafsh embryos. Semin. Cell Dev. Biol. 107, 170–178 (2020).
Brend, T. & Holley, S. A. Balancing segmentation and laterality during vertebrate development. Semin. Cell Dev. Biol. 20, 472–478 (2009).
Pourquié, O. Segmentation of the vertebrate spine: From clock to scoliosis. Cell 145, 650–663 (2011).
Wang, W. J. et al. Top theories for the etiopathogenesis of adolescent idiopathic scoliosis. J. Pediatr. Orthop. 31, S14–S27 (2011).
Kusumi, K. & Dunwoodie, S. L. The Genetics and Development of Scoliosis (Springer, 2018).
Goldberg, C. J., Fogarty, E. E., Moore, D. P. & Dowling, F. E. Scoliosis and developmental theory adolescent idiopathic scoliosis. Spine 22, 2228–2238 (1997).
Cooke, J. & Zeeman, E. C. A clock and wavefront model for control of the number of repeated structures during animal morphogenesis. J. Theor. Biol. 58, 455–476 (1976).
Palmeirim, I., Henrique, D., Ish-Horowicz, D. & Pourquié, O. Avian hairy gene expression identifies a molecular clock linked to vertebrate segmentation and somitogenesis. Cell 91, 639–648 (1997).
Vermot, J. et al. Retinoic acid controls the bilateral symmetry of somite formation in the mouse embryo. Science 308, 563–566 (2005).
Vermot, J. & Pourquié, O. Retinoic acid coordinates somitogenesis and left–right patterning in vertebrate embryos. Nature 435, 215–220 (2005).
Kawakami, Y., Raya, A., Raya, R. M., Rodríguez-Esteban, C. & Belmonte, J. C. I. Retinoic acid signalling links left–right asymmetric patterning and bilaterally symmetric somitogenesis in the zebrafish embryo. Nature 435, 165–171 (2005).
Packard, D. S. & Jacobson, A. G. Analysis of the physical forces that influence the shape of chick somites. J. Exp. Zool. 207, 81–92 (1979).
Dias, A. S., de Almeida, I., Belmonte, J. M., Glazier, J. A. & Stern, C. D. Somites without a clock. Science 343, 791–795 (2014).
Bard, J. B. L. A traction-based mechanism for somitogenesis in the chick. Rouxs Arch. Dev. Biol. 197, 513–517 (1988).
Nelemans, B. K., Schmitz, M., Tahir, H., Merks, R. M. & Smit, T. H. Somite division and new boundary formation by mechanical strain. iScience 23, 100976 (2020).
Grima, R. & Schnell, S. Can tissue surface tension drive somite formation? Dev. Biol. 307, 248–257 (2007).
Yabe, T., Hoshijima, K., Yamamoto, T. & Takada, S. Quadruple zebrafish mutant reveals different roles of Mesp genes in somite segmentation between mouse and zebrafish. Development 143, 2842–2852 (2016).
Compagnon, J. et al. The notochord breaks bilateral symmetry by controlling cell shapes in the zebrafish laterality organ. Dev. Cell 31, 774–783 (2014).
Guillon, E. et al. Fibronectin is a smart adhesive that both influences and responds to the mechanics of early spinal column development. eLife 9, e48964 (2020).
Mongera, A. et al. A fluid-to-solid jamming transition underlies vertebrate body axis elongation. Nature 561, 401–405 (2018).
Kim, S., Pochitaloff, M., Stooke-Vaughan, G. A. & Campàs, O. Embryonic tissues as active foams. Nat. Phys. 17, 859–866 (2021).
Shelton, E. R. et al. Stress-driven tissue fluidization physically segments vertebrate somites. Preprint at bioRxiv https://doi.org/10.1101/2021.03.27.437325 (2021).
Robinson, E. E. α5β1 integrin mediates strong tissue cohesion. J. Cell Sci. 116, 377–386 (2003).
Lecuit, T. & Lenne, P.-F. Cell surface mechanics and the control of cell shape, tissue patterns and morphogenesis. Nat. Rev. Mol. Cell Biol. 8, 633–644 (2007).
Ehrig, S. et al. Surface tension determines tissue shape and growth kinetics. Sci. Adv. 5, eaav9394 (2019).
Jülich, D., Geisler, R. & Holley, S. A. Integrinα5 and delta/notch signaling have complementary spatiotemporal requirements during zebrafish somitogenesis. Dev. Cell 8, 575–586 (2005).
Riedel-Kruse, I. H., Müller, C. & Oates, A. C. Synchrony dynamics during initiation, failure, and rescue of the segmentation clock. Science 317, 1911–1915 (2007).
Wang, H., Holland, P. W. H. & Takahashi, T. Gene profiling of head mesoderm in early zebrafish development: insights into the evolution of cranial mesoderm. EvoDevo 10, 14 (2019).
Das, D., Chatti, V., Emonet, T. & Holley, S. A. Patterned disordered cell motion ensures vertebral column symmetry. Dev. Cell 42, 170–180 (2017).
Grimes, D. T. Making and breaking symmetry in development, growth and disease. Development 146, dev170985 (2019).
Stern, C. D. & Bellairs, R. The roles of node regression and elongation of the area pellucida in the formation of somites in avian embryos. J. Embryol. Exp. Morphol. 81, 75–92 (1984).
Green, A. A., Mosaliganti, K. R., Swinburne, I. A., Obholzer, N. D. & Megason, S. G. Recovery of shape and size in a developing organ pair: Shape and Size Recovery in Developing Organs. Dev. Dyn. 246, 451–465 (2017).
Abrams, M. J., Basinger, T., Yuan, W., Guo, C.-L. & Goentoro, L. Self-repairing symmetry in jellyfish through mechanically driven reorganization. Proc. Natl Acad. Sci. USA 112, E3365–E3373 (2015).
Hervieux, N. et al. Mechanical shielding of rapidly growing cells buffers growth heterogeneity and contributes to organ shape reproducibility. Curr. Biol. 27, 3468–3479 (2017).
Eritano, A. S. et al. Tissue-scale mechanical coupling reduces morphogenetic noise to ensure precision during epithelial folding. Dev. Cell 53, 212–228 (2020).
Huch, M., Knoblich, J. A., Lutolf, M. P. & Martinez-Arias, A. The hope and the hype of organoid research. Development 144, 938–941 (2017).
Félix, M.-A. & Barkoulas, M. Pervasive robustness in biological systems. Nat. Rev. Genet. 16, 483–496 (2015).
DiFrisco, J. & Jaeger, J. Genetic causation in complex regulatory systems: an integrative dynamic perspective. BioEssays 42, 1900226 (2020).
Weber, M., Mickoleit, M. & Huisken, J. Multilayer mounting for longterm light sheet microscopy of zebrafish. J. Vis. Exp. 84, e51119 (2014).
Schindelin, J. Fiji: an open-source platform for biological-image analysis. Nat. Methods 9, 676–682 (2012).
Preibisch, S., Saalfeld, S., Schindelin, J. & Tomancak, P. Software for bead-based registration of selective plane illumination microscopy data. Nat. Methods 7, 417–418 (2010).
Sendra, G. H., Hoerth, C. H., Wunder, C. & Lorenz, H. 2D map projections for visualization and quantitative analysis of 3D fluorescence micrographs. Sci. Rep. 5, 12457 (2015).
Frangi, A. F., Niessen, W. J., Vincken, K. L. & Viergever, M. A. Multiscale vessel enhancement filtering, in Proc. Medical Image Computing and Computer-Assisted Intervention—MICCAI’98 (Wells, W. M. et al.) Vol. 1496, 130–137 (Springer, 1998).
Pietzsch, T., Saalfeld, S., Preibisch, S. & Tomancak, P. BigDataViewer: visualization and processing for large image data sets. Nat. Methods 12, 481–483 (2015).
Etournay, R. et al. Interplay of cell dynamics and epithelial tension during morphogenesis of the Drosophila pupal wing. eLife 4, e07090 (2015).
Guirao, B. et al. Unified quantitative characterization of epithelial tissue development. eLife 4, e08519 (2015).
Virtanen, P. et al. SciPy 1.0: fundamental algorithms for scientific computing in Python. Nat. Methods 17, 261–272 (2020).
Sawada, A. et al. Zebrafish Mesp family genes, mesp-a and mesp-b are segmentally expressed in the presomitic mesoderm, and Mesp-b confers the anterior identity to the developing somites. Development 127, 1691–1702 (2000).
Narayanan, R. & Oates, A. C. Detection of mRNA by whole mount in situ hybridization and DNA extraction for genotyping of zebrafish embryos. Bio Protoc. 9, e3193 (2019).
We thank the members of the Oates laboratory, M. Gonzales-Gaitan, S. W. Grill, M. Labouesse, P. Tomancak, G. Salbreux, J. Bois, K. Uriu, V. Krishnamurthy, P. Gross, Z. Hadjivasiliou and P. Chugh for comments on the manuscript; and the staff of the zebrafish facility in the Center for PhenoGenomics and the Bioimaging and OPtics (BiOP) core facility of École polytechnique fédérale de Lausanne. M. Popovic thanks M. Wyart for supporting the collaboration with the Oates laboratory. We acknowledge the vital early input of C. Qian in the project; P. Strnad and A. Boni for building the Viventis LS1 live light-sheet system; C. Remy for building the ablation module in the Viventis LS1 live light-sheet system; L. Rhode for input on explant preparation; J-Y. Tinevez and A. Bercowsky-Rama for building and adapting the Mastodon plugin, respectively, for identifying and tracking cells in zebrafish embryos; C. Jollivet, F. Lang and G. Valentin for assistance with zebrafish husbandry and line maintenance; O. Burri for building the laser ablation module in a spinning disk confocal system, R. Narayanan for optimizing the in situ hybridization protocol; the members of the Heisenberg laboratory for sharing the DN fibronectin constructs; and D. Jülich and the members of the Holley laboratory for sharing the fn1a-mKIK construct. This work was supported by EPFL, Wellcome (WT098025MA); the Francis Crick Institute, receiving its core funding from Cancer Research UK, the Medical Research Council, and Wellcome. S.R.N. was supported by a Long-Term Human Frontier Science Program postdoctoral fellowship (LT000078/2016). M.P. was supported by the Swiss National Science Foundation (grant no. 200021-165509) and the Simons Foundation (grant no. 454953).
The authors declare no competing interests.
Peer review information
Nature thanks M. Lisa Manning and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data figures and tables
The boxes represent the major steps involved and below each box, respective details are included.
(A) Schematic of multiview imaging setup. A zebrafish embryo in its chorion is mounted in a PTFE tube filled with 0.25% agarose and fluorescent beads. An embryo with its AP axis oriented along the circumference of the tube, as indicated, is chosen for multiview imaging. Rotation of the tube allows imaging the embryo from different angles. (B) One of the time points from a histone transgenic line viewed from 6 angles. Images represent respective maximum intensity projections. Scale bar, 50 μm. (C) Representative image of multiview fusion, which was performed using the MultiView Reconstruction FIJI plugin. Image generated using Imaris Viewer x64 (version 9.7.0). Scale bar, 150 μm. (D) Sphere fit (in grey) of a representative fused image. Cyan dots represent individual nuclei. (E) The radius of the sphere obtained from the fit does not change over the analysis time window. Individual lines represent the 6 multiview videos used for analysis. (F) Heat map of change in pixel size in equidistant cylindrical map projection. Note that the change in pixel size is negligible within 15° from the equator, which corresponds to regions of somite formation (compare with Fig. 1a). Further details on statistics and reproducibility can be found in the Methods section.
(A) Raw zoomed in image of a representative region of an embryo imaged using Viventis microscope. Scale bar, 25 μm. (B) Image in (A) overlaid with segmented notochord (in green) and somite boundaries (in red). Segment length (indicated in yellow) is determined with respect to the notochord. Yellow arrow represents local angle of the notochord. Scale bar, 25 μm. (C), (G) Quiver plots of the eigen vectors upon application of Frangi vesselness filter. Orientation of the vectors represent directions of least change in fluorescent intensity. Along the notochord, these directions are horizontal (C), while along somite boundaries, they are vertical (G). (D), (H) Raw image in (A) is subdivided into two parts based on the directions of the eigen vectors, thus leading to images with horizontal (D) or vertical lines (H). (E), (I) Lines that are less than 15 pixels apart are joined together by straight lines following which lines that do not represent notochord or somite boundaries are discarded. (F), (J) Any remaining gaps in boundary detection are then connected by straight lines. Further details on statistics and reproducibility can be found in the Methods section.
(A) Initial AP lengths, lengths at 1 hr and lengths at 2 hr of somites one to eight are shown across the six multiview datasets. Both left and right somites are included in the same plot. Grey dots represent individual somites. Red dots, mean AP length. Somites seven and eight were not segmented at two-hour time points. (B) Change in AP length over time for all eight segmented left–right somites across six embryos. Note the decrease in standard deviation (red) over time. (C) Variance of AP lengths over time in each of the six embryos imaged with Zeiss Z1. (D) Comparison of AP lengths over time across Zeiss and Viventis datasets. (E) Variability of LAP of somites 14 to 17 decreases over time. Right, variance of LAP.
(A) Snapshots from Mastodon, a FIJI plugin, indicating detected nuclei (cyan circles) in corresponding histone and utrophin fluorescent images. A single zoomed in z-slice of an embryo imaged using Viventis microscope is shown. Scale bar, 50 μm. (B) Comparison of initial and one-hour cell numbers of somites (r = 0.96, 95% CI = 0.93–0.98). About 15% of the cells divided within an hour after somite formation. Somites three to five are separately colored. Dashed line, slope = 1 in (B,C); blue, regression line. N = 62 somites; 11 embryos from 4 independent spawnings in this and all panels of this figure. (C) Positive correlation of somite AP lengths obtained from boundary segmentation method and shape analysis (r = 0.8, 95% CI = 0.69–0.87). Because nuclei are used for determining lengths in the shape analysis, AP lengths are consistently smaller in the shape analysis compared to lengths determined from boundary segmentation. (D,E) Comparison of initial and one-hour ML (D) and DV lengths (E) indicate decrease and increase in lengths respectively. (F,G) Comparison of initial and 1 h left–right differences in ML (F) and DV (G) length. Data was normalized with respective average lengths at the different time points. (H) Initial left–right volumes are asymmetric consistent with an imprecise prepattern (Fig. 1d) and the volume asymmetry remains during length adjustment. Further details on statistics and reproducibility can be found in the Methods section.
Left image in each of the five explanted somites (A) to (E) shows initial shape of explants. Right images represent time points when explants acquired spherical shape. Scale bar, 25 μm. (A) is same as shown in Fig. 3c. An exponential fit of explant anisotropy over time is shown below each example. Respective relaxation time scales, τe are as indicated.
(A) (a) Schematic of somite dimensions. (b) Normal stress on somite surface due to surface tension. Surface tension forces Fj (red) and Fk (green) acting on the light blue surface i are proportional to the surface edge lengths Lj and Lk. The normal stress σi is then normal force per surface area Ai from which we obtain Eqs. 9, 10 and 11 (see Supplementary Note). (c) Contacts (red) with PSM and anterior somite result in normal stress σa. (d) Left, constraint l(t) imposed on LDV by neural plate and yolk. Right, snapshot of somite 3 from a 3-somite stage embryo. Scale bar, 100 μm. (B,D) 3D plot of relative cell nucleus movements for a representative explant (B) and in vivo somite (D). (C,E) MSRD measurement for 2 explants (C) and two in vivo somites (somite 3) (E). Red line indicates diffusive behaviour and black line indicates the square of typical cell size. (F) Explanted somites contain a Fibronectin-rich extracellular matrix. Fibronectin-mKIK fusion protein from mRNA injected in embryos localizes to somite and notochord boundaries as expected. Red box points to zoomed-in images (right) of the middle and ventral regions of a somite. Scale bar, 50 μm. (G) Fibronectin-mKIK is observed with a similar distribution around explanted somites. Blue, Fn1a-mKIK; yellow, explant manually outlined from Utr:mCherry channel. Scale bar, 25 μm. (H) Representative data for cumulative shape change of cells and somites along AP, ML and DV dimensions. Further details on statistics and reproducibility can be found in the Methods section.
(A) Ablation of PSM adjacent to most-recently formed somite boundary (N = 6 embryos from 4 independent spawnings). Imaging in Viventis light-sheet system was started 15 min after ablation. Right image in each example shows a prominent bulge of the boundary (arrow). Scale bar, 50 μm. (B) Ablation of LPM (N = 7 embryos from 2 independent spawnings) adjacent to lateral boundary of a somite. Yellow arrow points at the same boundary over time. Ablation of LPM was performed in the light-sheet system.
Extended Data Fig. 9 Length adjustment reduces upon perturbing molecules involved in surface tension.
(A) Comparison of AP length LAP under different perturbations. (B) Injected RGD-FITC localizes to notochord and somite boundaries. Scale bar, 50 μm. (C) Injection of 1.5 ng each of fn1a and fn1b MO or 4 ng of itgα5 MO results in disintegration of anterior somites by the 10th somite stage (zoomed in inset). Injection of 2 ng of cdh2 MO results in defective formation of neuroectoderm and tail. Lesser amounts of each of these MOs as indicated in (A) were used to analyse length adjustments. Scale bar, 100 μm. (D) Estimated relaxation timescale τ for wild type anterior somites (black) and the different perturbations. Error bar, fit uncertainty. Further details on statistics and reproducibility can be found in the Methods section.
Extended Data Fig. 10 Left–right difference in somite length is weakly correlated with anterior somite boundary difference.
(A) Schematic of measurement of boundary position difference. The points of intersection of orthogonal lines (cyan) from somite boundaries on the left (L) and right (R) sides with the middle of the notochord (black dashed line) are determined. A difference in the median position (red dots) of these points of intersection represents boundary position difference (black arrow). Boundary positions increase positively from P to A. (B-C) AP length difference between left–right somite pairs is weakly correlated with anterior boundary position difference both initially (r = −0.24, 95% CI = −0.45 to −0.03) (B) and after 1 h (r = −0.29, 95% CI = −0.48 to −0.10) (C). 76 somite boundaries; 17 embryos from 10 independent spawnings.
The mechanical model of AP length adjustment of a newly formed somite and Supplementary References.
Multiview videos: Zebrafish embryos were imaged from six angles (as indicated) using the Zeiss Z1 system and the maximum intensity projection of nucleus channel (H2B:GFP) in each angle is shown.
Map projection enables the visualization and analysis of left–right somite morphogenesis simultaneously. Multiview acquisitions were fused and map-projected. Map-projected time-lapse in which nuclei (top) were visualized with fluorescently-tagged histone (H2B:GFP) and the cell outlines (bottom) were visualized with fluorescently-tagged utrophin (Utr:mCherry). Anterior is to the left and posterior is to the right. Single map-projected layers from each time point were chosen and are shown here.
Single-somite explants round up over time. Single somites were isolated and cultured in vitro. A single z-slice from each time point was chosen and shown here.
About this article
Cite this article
Naganathan, S.R., Popović, M. & Oates, A.C. Left–right symmetry of zebrafish embryos requires somite surface tension. Nature 605, 516–521 (2022). https://doi.org/10.1038/s41586-022-04646-9