Figure 3. Gradients of Fibronectin matrix and F-actin correlate with in silico gradients of tension.

(A) Heat map of the tension distribution along tissue surfaces in silico. Warmer colors represent higher tension. Inset: magnification of the PSM|NT interface showing a medial-lateral tension gradient on the neural tube side of the interface. (B–D) Transverse sections taken 60 μm away from last somite boundary on a 12–14 s stage embryo co-labeled for Fibronectin (FN, red) and F-actin (green). There are colocalized medial-lateral gradients of both F-actin and Fibronectin matrix along the PSM|NT interface (dashed box). Star denotes differentiating myofibers that show high F-actin signal. Scale bars = 20 μm. (E) 3D schematic of the PSM|NT interface (grey) and the PSM|E interface (orange). To quantify matrix assembly along these interfaces, FN matrix signal is sorted into two categories based on size: small matrix elements (cyan) and large matrix elements (magenta). See SI for details. The F-actin signal (F, I), FN signal (G, J), and processed images of small and large matrix elements (H, K) are shown. The PSM|NT interface exhibits medial-lateral gradients of F-actin and Fibronectin (F–H), whereas the PSM|E interface shows an increase in F-actin and Fibronectin matrix assembly along the posterior to anterior axis (I–K). All images are projected dorsal views. A = anterior, P=posterior, M = medial, L = lateral. Scale bar = 10 μm. Quantification of the medial-lateral distributions of F-actin (L) and small and large FN matrix elements (M) along the PSM|NT interface. The bracket in (L) denotes the differentiating myofibers rich in F-actin. Quantification of the medial-lateral distributions of F-actin (N) and small and large FN matrix elements (O) along the PSM|E interface. Quantification the density of F-actin (P) and small and large matrix elements (Q) along the anterior-posterior axis of the PSM|E interface. Data represent means and SD. Sample sizes: L, n = 8 PSMs from six embryos; N, P, n = 7 PSMs from five embryos; M, O, Q, n = 10 PSMs from six embryos. (R) The NT-PSM interfaces represented as a lap-joint with a single sided strap. Neural tube (magenta) acts as a strap that is adhered to the two PSMs (cyan) via a graded adhesive (red) made of Fibronectin. Medial and ventral edges of PSM are attached to the notochord and yolk surface respectively (dashed region). Black arrows denote neural tube convergence and green arrows denote resistance to this convergence via the adhesive. Neural tube convergence with respect to the adhesive produces shear stress. Established theories of lap-joint predict a stress gradient with higher stress at the lateral edge of the PSM|NT interface. Extra adhesive, called a ‘spew fillet,’ in an arced shape at the lateral sides of the strap strengths the joint. (S) Transverse section on a 12–14 s stage embryo labeled for Fibronectin (FN, red) illustrating the spew fillet of FN matrix (arrowhead). Scale bar = 20 μm.

Figure 3—figure supplement 1. Medial-lateral gradients of Myosin II and Fibronectin matrix tension at the PSM|NT interface.

(A–C) Embryos were injected with myl12.1-EGFP mRNA to visualize distribution of Myosin-II (Araya et al., 2019; Behrndt et al., 2012). (A) A dorsal view of the neural tube of a live 12-somite stage embryo. The image shows the first 100 μm posterior of the last somite boundary in a single confocal plane roughly 10 μm underneath the dorsal surface of the neural tube. Local enrichment of Myosin-II can be seen in the lateral surface of the neural tube (white arrowheads). Anterior is is up. The dashed line indicates the position of the transverse section in B. Scale bar = 25 μm. (B) A transverse section 60 μm away from last somite where enrichment of Myosin-II can be seen both along the lateral side of the neural tube (yellow arrowhead) and the lateral part of the PSM|NT interface (white arrowhead). Scale bar = 25 μm. (C) Quantification of the relative Myosin II levels reveals a medial to lateral increase of Myosin-II levels from the medial part of PSM|NT interface to the lateral side of the neural tube. Transverse sections were generated every 5 μm along the first 100 μm posterior of the last somite boundary. For each section, the neural tube border was divided into three regions as indicated on the transverse section (top panel): the medial half of the PSM|NT interface (purple), the lateral half of the PSM|NT interface (blue), and the adjacent lateral side of the NT (orange). The mean myl12.1-EGFP intensity was measured in each region, and the relative intensity ratios (Lateral PSM|NT/Medial PSM|NT; NT side/Lateral PSM|NT; NT side/Medial PSM|NT) were calculated on each slice and averaged for each embryo. Average ratios across all embryos were then plotted (bottom panel). All ratios are above 1 according to the 95% confidence intervals (error bars). The difference between the ratios was evaluated via T-test. Sample size: n = 25 data points (each point is the average of 21 slices per embryo), from 13 embryos (two sides per embryo) from four experiments. (D–I) Immunostaining with the H5 antibody on 12 s stage embryos injected with FN1a-mKIKGR13.2-hsFNIII10-11 mRNA. The H5 signal (D, G), mKIKGR signal (E, H), and processed heatmaps of the H5/mKIKGR ratio (F, I) are shown. All images are projected dorsal views with anterior to the top and medial to the left. Scale bars = 10 μm. The PSM|NT interface (D–F) exhibits an increasing medial-lateral gradient of H5. By contrast, the H5/mKIK ratio shows an opposite gradient as the tension on individual fibers is lower laterally where stresses are distributed over more Fibronectin molecules. (G–I) The PSM|E interface does not exhibit any medial-lateral gradient. (J–M) Quantification of the medial-lateral distributions of H5 and mKIKGR levels (J, L) and the H5/mKIKGR ratio (K, M) along the PSM|NT (J, K) or PSM|E (L, M) interface. Sample size: n = 12 PSM|NT and 14 PSM|E interfaces from 10 and 9 embryos, respectively from three experiments. The average slopes of the H5/mKIKGR ratios for the PSM|NT and PSM|E interfaces differ (**, p=0.0046, T-test).

Figure 3—figure supplement 2. Epithelialization of PSM surface cells and increases in F-actin intensity in PSM cells as the PSM matures from posterior to anterior.

(A, B) Transverse sections taken in the anterior PSM (A, 0–100 μm away from last somite boundary) or posterior PSM (B, 200–300 μm away from last somite boundary) on 12–14 s stage embryos co-stained for Fibronectin (FN, red) and F-actin (green). Asterisks denote PSM surface cells. Scale bars = 10 μm. (C) Quantification of PSM surface cell aspect ratio in the anterior and posterior portions of the PSM. Sample size: n = 93 cells (anterior PSM) and 92 cells (posterior PSM) from three embryos. Anterior vs posterior, p=7.46e-5. (D–E) Quantification of the mean F-actin intensity of the PSM surface cells region (D) or PSM internal cells region (E) in the anterior and posterior PSM from three embryos. Each dot represents a transverse section for which the mean F-actin signal within the surface cells or internal cells was measured and normalized by the average intensity of the signal in the posterior sections of the same embryo. Sample size: n = 24 sections (anterior PSM) and 25 sections (posterior PSM). ***p<0.0005 using a T-test. (D) Anterior vs posterior, p=5.62e-7. (E) (D) Anterior vs posterior, p=7.7e-4.