A mechanosensitive circuit of FAK, ROCK, and ERK controls biomineral growth and morphology in the sea urchin embryo

Abstract

Biomineralization is the utilization of different minerals by a vast array of organisms to form hard tissues and shape them in various forms. Within this diversity, a common feature of all mineralized tissues is their high stiffness, implying that mechanosensing could be commonly used in biomineralization. Yet, the role of mechanosensing in biomineralization is far from clear. Here, we use the sea urchin larval skeletogenesis to investigate the role of substrate stiffness and focal adhesion kinase (FAK) in biomineralization. We demonstrate that substrate stiffness alters spicule morphology and growth, indicating a mechanosensitive response during skeletogenesis. We show that active FAK, F-actin, and vinculin are enriched around the spicules, indicating the formation of focal adhesion complexes and suggesting that the cells sense the mechanical properties of the biomineral. Furthermore, we find that FAK activity is regulated by Rho-associated protein kinase (ROCK) and is crucial for skeletal growth and normal branching. FAK and ROCK activate extracellular signal-regulated kinase (ERK), which regulates skeletogenic gene expression at the tips of the spicules. Thus, the FAK-ROCK-ERK circuit seems to provide essential mechanical feedback on spicule elongation to the skeletogenic gene regulatory network, enabling skeletal growth. Remarkably, the same factors govern mammalian osteoblast differentiation in vitro and pathological calcification in vivo. Thus, this study highlights a common mechanotransduction pathway in biomineralization that was probably independently co-opted across different organisms to shape mineralized structures in metazoans.

Keywords: biomineralization; focal adhesion kinase; gene regulatory networks; mechanotransduction; sea urchin.

Fig. 1.

Substrate stiffness affects spicule morphology and length in vitro. (A) Spicule morphology on 0.25 kPa, 25 kPa, 250 kPa, and glass (>GPa) at 72 hpf. (B) Quantification of skeletogenic phenotypes in different substrate stiffness. (C) Length of spicule rods at 72 hpf. Each box plot shows an average, the first and the third quartiles (edges of boxes), and outliers (dots). ***P < 0.001, Kruskal–Wallis nonparametric test. Experiments were performed in three independent biological replicates, the total number of scored spicules is n = 169 in 0.25 kPa, n = 145 in 25 kPa, n = 137 in 250 kPa, and n = 137 in control.

Fig. 2.

FAK is activated around the spicules during sea urchin skeletogenesis. Representative images showing immunostaining of control embryos at different developmental stages. Phalloidin (green) marks F-actin, pFAK (red) marks phosphorylated FAK, and 6a9 (blue) marks the skeletogenic cell membrane. The Right panels show enlargements of rectangle sections marked in the leftmost panels. (A–E) At 20 hpf FAK is activated around the initial skeletal grain (arrows in B and C) and F-actin accumulates around it (arrow in D). (F–J) At 24 hpf FAK is activated around the triradiate skeleton, surrounded by F-actin (arrows in G–J). (K–O) At 48 hpf FAK is activated around the spicule rods (arrows in L) surrounded by F-actin which is enriched at the tips of the rods (arrows in M–O). (P–T) pFAK and Phalloidin staining in skeletogenic cell culture showing enrichment at the tips (arrows in R–T). Experiments were performed in three independent biological replicates, the numbers at the bottom left of the left column indicate the number of embryos that show this phenotype out of all embryos scored. The scale bar in the whole embryo is 50 μm and in cell culture 20 μm.

Fig. 3.

Vinculin is enriched around the tips of the spicule rods together with active FAK. A representative embryo showing immunostaining of Vinculin (green) and phosphorylated FAK (red) at 48 hpf. (A–C) Whole embryo, (D–F) enlargement of the body rod region marked by the rectangle in A. (G–I) Enlargement of the anterolateral (AL) rod region marked by the rectangle in A. (J–L) Enlargement of the postoral rod region that is out of focus in A. Left panels show Vinculin antibody, Middle panels show pFAK antibody, and Right panels show an overlay of the two antibodies. Arrows point to regions at the tips of the rods where Vinculin and pFAK are enriched. The experiment was performed in three biological replicates and in 42 embryos out of the 44 embryos scored, the reported pattern was observed. The scale bar in C is 50 μm and in F, I, and L is 10 μm.

Fig. 4.

FAK activation depends on ROCK activity at 48 hpf. Immunostaining of phosphorylated FAK (red) and the skeletogenic marker 6a9 (green) of control sea urchin embryo (A–I) and ROCK inhibited embryos (30 μM Y27632, J–R) at 48 hpf. Upper panels are images of whole embryos. Middle panels show enlargement of the spicule region of the body rod (SR rectangle in A and J). Lower panels show enlargement of the tip cells of the body rod (TC in A and J). (M–O) Late ROCK inhibition reduces FAK phosphorylation at the spicule as indicated by the white arrow, while ROCK inhibition does not affect FAK activity in skeletogenic cells in the body rods (P–R) as indicated by the arrowheads. The scale bar is 50 μm. Results are based on three biological repeats, the numbers at the bottom left of A and J indicate the number of embryos that show this phenotype out of all embryos scored.

Fig. 5.

FAK regulates skeletal growth and inhibits branching. (A–H) FAK inhibition (10 µM PF573228 and 10 µM, 20 µM, or 40 µM GSK2256098) in whole embryos before (>20 hpf) and after skeletal initiation (>25 hpf) decreased skeletal growth and increased ectopic branching at 48 hpf. (A–G) representative embryos demonstrating the various phenotypes of FAK inhibition. (Scale bars are 50 µm.) (A) normal embryo, (B and C) short skeleton, (D and E) very short skeleton, (F and G) short and branched skeleton. (H) Statistics of FAK inhibition phenotypes. The graph shows the percentage of embryos showing each of the four phenotypes for the different treatments. Experiments were performed in three biological replicates for all treatments but PF573228 added at 20 hpf that was performed in two biological replicates. The number of embryos scored in each replicate for each treatment is provided in SI Appendix, Table S2. (I–K) FAK inhibition in skeletogenic cell culture after skeletal initiation (>48 hpf) decreased skeletal growth, observed at 72 hpf. (I) representative spicule in control culture and (J) under FAK inhibition. (K) Measurement of the length of the spicules in control and FAK-inhibited cell cultures (PF573228, 10 µM). Each box plot shows an average, the first and the third quartiles (edges of boxes), and all individual measurements (dots). Experiments were conducted in three biological replicates where the total number of spicules measured was n = 97 for control and n = 100 for FAK inhibition. A one-way ANOVA with Bonferroni post hoc tests was performed, *** indicates P < 0.0001. The scale bar is 50 μm.

Fig. 6.

FAK and ROCK activate ERK. (A–C) Immunostaining of dually phosphorylated ERK (red) and 6a9 (green) in control, (D–F) FAK-inhibited embryos (10 µM PF573228, >33 hpf), and (G–I) ROCK-inhibited embryos (30 μM Y27632, >25 hpf). The middle row is the enlargement of body rods, and the lower row is the enlargement of the anterolateral rod. FAK and ROCK inhibition eliminate ERK activation in the skeletogenic cells but not in neighboring cells. Red arrows with a green outline indicate skeletogenic cells with active ERK, red arrows indicate nonskeletogenic neighboring cells with active ERK, and empty arrows with green outlines indicate skeletogenic cells with no ERK activity. (J) A graph of the average ratio of skeletogenic cells with dpERK signals to the total number of skeletogenic cells seen in each rod. Error bars represent SD, and circle diameter represents the number of occurrences observed at that value as indicated by the scale at the top right of the graph. Statistical significance was measured using nonparametric one-way ANOVA along with pairwise tests between groups where ** indicates P < 0.01, and ***P < 0.0001. Results are based on three biological repeats where the total number of scored embryos is n = 28 for control, n = 29 for FAK inhibition, and n = 27 for ROCK inhibition. The scale bar in the whole embryos is 50 μm. SCs – skeletogenic cells, PO – postoral, AL – anterolateral.

Fig. 7.

FAK and ERK signaling are required for normal expression of skeletogenic genes at the tips of postoral and anterolateral rods. (A–C) representative images of skeletal morphology as observed in DIC images for live embryos at 48hpf, in control (A), FAK (PF573228, 10 µM, (B) and ERK (U0126, 10µM, (C) inhibited embryos. The inhibitors were added after skeletal initiation (at 25 hpf). The scale bar is 50 µm. (D–I) Spatial expression of skeletogenic genes in control (D and G), FAK inhibition (E and H), and ERK inhibition (F and I). Gene names are indicated at the top of each panel. (J) Relative gene expression levels of treated embryos vs. control embryos measured by qPCR at 48 hpf. Bars show averages of over three independent biological replicates. Dark green bars show ratios under ERK inhibition and light green bars show ratios under FAK inhibition. Ratio of one indicates that the expression of the gene is unaffected by inhibition treatment. Asterisks indicate P < 0.001, one-tailed z-test. Error bars indicate SD.

Fig. 8.

Proposed model for the role of FAK-ROCK-ERK circuit in sea urchin skeletogenesis. (A) ROCK is essential for the formation of the biomineral grain and enhances F-actin polymerization (37). The stiffness of the biomineral and ROCK activates FAK around the grain. FAK is activated at the membrane engulfing the initial biomineral grain, and F-actin accumulates around it. (B) FAK is activated around the triradiate skeleton, proximal to F-actin. pFAK is necessary for skeletal elongation and inhibits skeletal branching. (C) During spicule elongation, FAK is activated along the entire spicule rods, surrounded by F-actin and both are enriched at the tips of the rods from where the skeleton elongates, possibly assisting in mineral exocytosis. ROCK and FAK are required for the activation of ERK signaling, that regulates skeletogenic gene expression at the tips of the rods. ROCK, FAK, and ERK activity are essential for mineral deposition at the tips of the rods and skeletal elongation.