nhsl1b is expressed in mesodermal cells during gastrulation.

A Whole-mount in situ hybridisation with nhsl1b or control probes, on embryos fixed at the 1-cell, shield (onset of gastrulation), 75% epiboly (mid-gastrulation) and 12-somite stages. B, C Whole mount in situ hybridisation with nhsl1b and sox17 probes and subsequent GFP immunostaining. Bnhsl1b is expressed in induced mesendodermal cells. N = 4 experiments. C No expression of nhsl1b is observed in induced endodermal cells. N = 3 experiments. D Whole-mount double in situ hybridisation with sox17 (red) and nhsl1b (green) probes at 75% epiboly (end of gastrulation) and bud stages. nhsl1b is not expressed in the sox17 expressing fore-runner cells. Scale bar 100 µm unless specified.

nhsl1b knockdown affects lateral mesoderm migration.

A Representative lateral views of Tg(tbx16:EGFP) embryos, at early gastrulation (60% epiboly; t = 0) and 1 hour later, in different conditions. White dashed lines indicate the position of the margin of the blastoderm, yellow dashed lines indicate the position of the front of the migrating mesoderm. Mesoderm progression was measured as the variation in distance between these two lines (see also Supplementary Fig. S1). Scale bar 100 µm. B Quantification of the lateral mesoderm progression in MO Control injected embryos (n = 11 embryos), MO Nhsl1b injected embryos (n = 12 embryos), MO-2 Nhsl1b injected embryos (n = 21 embryos), MO-2 Nhsl1b and nhsl1b mRNA co-injected embryos (n = 15 embryos). N = 16 experiments in total with more than 4 experiments per conditions. Kruskal–Wallis test followed by Dunn’s test. Adjusted p-values: MO Control vs MO Nhsl1b: 0.0093 **; MO Control vs MO-2 Nhsl1b 0.0002 ***; MO-2 Nhsl1b vs MO-2 Nhsl1b + Nhsl1b 0.0218 *; MO Control vs MO-2 Nhsl1b + Nhsl1b 0.6481 ns.

nhsl1b knockdown reduces cell speed and persistence.

A Representative image of mesodermal cell nuclei tracking, in a Tg(tbx16:EGFP) embryo expressing H2B-mCherry. B, C Instant cell speed and directionality ratio. Circles on the violin plot indicate mean per embryo. Likelihood ratio test of a linear mixed effects model with treatment as a fixed effect and embryos as a random effect against a model without the fixed effect. Adjusted p-values: (B) MO Control vs MO Nhsl1b: 0.0100 *; MO Control vs MO-2 Nhsl1b: 0.0131 *; (C) MO Control vs MO Nhsl1b: 0.0301 *; MO Control vs MO-2 Nhsl1b: 0.0301 *. D Directional autocorrelation. Likelihood ratio test of a non-linear mixed effects model (see methods) with treatment as a fixed effect and embryos as a random effect against a model without the fixed effect. Adjusted p-values: MO Control vs MO Nhsl1b: 0.045 *; MO Control vs MO-2 Nhsl1b: 0.0054 *. MO control (n = 5), MO Nhsl1b (n = 8) and MO-2 Nhsl1b (n = 6) injected embryos. E, F Representative cell trajectories of mesodermal cells in MO control and MO Nhsl1b injected embryos, colour-coded for cell migration persistence (directionality ratio). x-axis represents distance along the dorsal-ventral axis, y-axis represents distance along the animal-vegetal axis.

nhsl1b overexpression affects lateral mesodermal migration, reducing cell speed and persistence.

A Representative lateral views of Tg(tbx16:EGFP) embryos, at early gastrulation (60% epiboly; t = 0) and 1 hour later, in embryos injected with lacZ (control, n = 6) or nhsl1b (n = 10) mRNAs. Dashed lines indicate the positions of the front of the migrating mesoderm on the first (yellow) and last (white) frames. Mesoderm progression was measured as the distance between these two lines (see also Supplementary Fig. S1). Scale bar 100 µm. B Quantification of the lateral mesoderm progression in control and nhsl1b mRNA injected embryos. N = 6 experiments. Mann–Whitney. p-value: Control vs Nhsl1b: 0.0075*, C, D Instant cell speed and directionality ratio. Circles on the violon plot indicate mean per embryo. Likelihood ratio test of a linear mixed effects model with treatment as a fixed effect and embryos as a random effect against a model without the fixed effect. p-values: 0.0029 ** and 0.0111 *. E Directional autocorrelation. Likelihood ratio test of a non-linear mixed effects model (see methods) with treatment as a fixed effect and embryos as a random effect against a model without the fixed effect. p-value: 0.03589 *. Control (n = 4) and nhsl1b (n = 10) mRNA injected embryos.

Nhsl1b localises at cell-cell contacts and at the tip of actin-rich protrusions.

Nhsl1b-mNeongreen and Lifeact-mCherry expressing mesodermal cells plated on a coverslip (A–C) or in vivo, transplanted in a non-labelled host embryo (D, E). A Nhsl1b localises at the very edge of the lamellipodium (arrowhead), where it stands during lamellipodium progression, as revealed by the kymograph (performed over the region delineated by the dashed yellow box). B, C In cultured cell clusters, Nhsl1b localises at cell-cell contacts (arrow) and at the very tip of actin-rich protrusions (arrowhead). D, E Similar localisations are observed in mesodermal cells in vivo. C, E Quantification of the max-normalised intensity of Nhsl1b-mNeongreen and Lifeact-mCherry along the segments indicated in the inset. Scale bars: (A) 10 µm; 2 µm, (B) 20 µm; 10 µm, (D) 40 µm; 20 µm.

Nhsl1b regulates protrusion dynamics.

A Protrusions of mesodermal cells injected with Lifeact-mCherry mRNAs and with a MO control or a MO Nhsl1b, transplanted in the mesoderm of a non-labelled embryo. Selected time points showing the protrusion elongation. B, C Quantification of the lifetime and maximum length of protrusions. Likelihood ratio test of a linear mixed effects model with treatment as a fixed effect and cells as a random effect against a model without the fixed effect. Adjusted p-values: (B) MO Control vs MO Nhsl1b: 0.0309*; MO Control vs MO-2 Nhsl1b: 0.0011 ***; (C) MO Control vs MO Nhsl1b: 0.0001***; MO Control vs MO-2 Nhsl1b: 4,89E−06****. MO Control (n = 5 embryos; n = 12 cells), MO Nhsl1b (n = 6 embryos; n = 14 cells) and MO-2 Nhsl1b (n = 7 embryos, 49 cells). D Protrusions of mesodermal cells injected with Lifeact-mCherry mRNAs and with lacZ (control) or Nhsl1b mRNAs, transplanted in the mesoderm of a non-labelled embryo. Selected time points showing the protrusion elongation. E, F Quantification of the lifetime and maximum length of protrusions. Likelihood ratio test of a linear mixed effects model with treatment as a fixed effect and cells as a random effect against a model without the fixed effect. p-values: E Control vs Nhsl1b: 0.0094**; (F) Control vs Nhsl1b: 0.00012***. Control (n = 5 embryos, 30 cells) or Nhsl1b (n = 6 embryos, 28 cells). Scale bars 20 µm.

nhsl1b knockdown increases F-actin retrograde flow and assembly rate.

A Selected time points from a high temporal resolution time-lapse showing growing protrusions in MO Control and MO Nhsl1b injected mesodermal cells expressing Lifeact-mNeongreen. Dashed yellow box indicate the regions used to generate the kymographs in (B). B Kymographs colour-coded to highlight the forward and backward moving pixels. The yellow dashed line corresponds to the extending front of the protrusion, the white dashed line corresponds to the actin retrograde flow. Quantification of the protrusion extension speed (C), retrograde flow speed (D), and F-actin assembly rate (E). Likelihood ratio test of a linear mixed effects model with treatment as a fixed effect and cells as a random effect against a model without the fixed effect. p-values: (C) MO Control vs MO Nhsl1b: 0,8362 ns.; (D) MO Control vs MO Nhsl1b: 0.0051**; (E) MO Control vs MO Nhsl1b: 0.0248*. MO Control (n = 84 cells in 14 embryos) and MO Nhsl1b (n = 89 cells in 17 embryos). Scale bars 2 µm.