Netrin-1 attracts axons through FAK-dependent mechanotransduction

Abstract

The mechanism by which extracellular cues influence intracellular biochemical cascades that guide axons is important, yet poorly understood. Because of the mechanical nature of axon extension, we explored whether the physical interactions of growth cones with their guidance cues might be involved. In the context of mouse spinal commissural neuron axon attraction to netrin-1, we found that mechanical attachment of netrin-1 to the substrate was required for axon outgrowth, growth cone expansion, axon attraction and phosphorylation of focal adhesion kinase (FAK) and Crk-associated substrate (CAS). Myosin II activity was necessary for traction forces >30 pN on netrin-1. Interestingly, while these myosin II-dependent forces on netrin-1 substrates or beads were needed to increase the kinase activity and phosphorylation of FAK, they were not necessary for netrin-1 to increase CAS phosphorylation. When FAK kinase activity was inhibited, the growth cone's ability to recruit additional adhesions and to generate forces >60 pN on netrin-1 was disrupted. Together, these findings demonstrate an important role for mechanotransduction during chemoattraction to netrin-1 and that mechanical activation of FAK reinforces interactions with netrin-1 allowing greater forces to be exerted.

Figure 1.

Heparin or deletion of domain C prevents nonspecific adsorption of netrin-1 but preserves DCC binding. A, Domain structure of netrin-1. Domain V and VI bind to netrin-1's receptor DCC, while domain C contributes to substrate adsorption through its many positively charged amino acids. B–D, HEK293 cells were transfected with GFP-tagged DCC (DCC-GFP) and incubated with either 1 μg/ml full-length mCherry-tagged netrin-1 (B), 1 μg/ml full-length NTN1-mCherry and 2 μg/ml heparin (C), or 1 μg/ml NTN1dC (D) for 90 min. As highlighted by the cells that do not express DCC (white arrows), the presence heparin or deletion of domain C restricts binding to only the cells that express DCC. E–G, Differential interference contrast (DIC) images (bottom) and unmodified fluorescent intensity images (top) of spinal commissural neurons incubated for 15 min with, either 1 μg/ml NTN1-mCherry (E), 1 μg/ml NTN1-mCherry with 2 μg/ml heparin (F), or 1 μg/ml NTNdC-mCherry (G). Relative to full-length netrin applied alone, there is a decrease in background binding in the presence of heparin (37%, p < 0.01, n = 40) or when domain C is deleted (20%, p < 0.01, n = 40) and there is distinct labeling of both axon and growth cone. H, I, ELISA detecting the myc tag within recombinant, full-length netrin-1 when media containing 200 ng/ml are incubated with either a collagen cushion for 16 h (H) or a polylysine-coated tissue culture well (I) for time periods of 15 min or less. Significant binding of full-length netrin is seen to collagen gels (n = 16, p < 0.01) and to polylysine dishes within 5 min (n > 15, p < 0.05) compared with media alone. Inclusion of 2 μg/ml heparin significantly reduced the association of netrin-1 to collagen by 79% (n = 16, p < 0.01) and to a polylysine-coated dish by 90% after 15 min. Fluorescent intensity comparison based on average intensities of 10 μm² areas from images of equal exposure (5 s). Scale bars: B, 50 μm; E, 10 μm.

Figure 2.

Substrate adsorption of netrin-1 is required for outgrowth from dorsal spinal cord explants. A–D, Dorsal spinal cord explants cultured for 16 h in a collagen gel. Netrin-1 (NTN1) elicits maximum outgrowth at 200 ng/ml (B, G). However, when 2 μg/ml heparin was coapplied to prevent substrate adsorption of netrin-1, all detectable outgrowth to netrin-1 was eliminated at 100, 200, and 400 ng/ml netrin-1 (n > 10 explants, D). E, F, H, The presence of heparin in the media had no effect on this netrin-independent outgrowth after 36 h (see white arrows in E and F, n = 4 explants). I–L, Transiently transfected HEK293 aggregates expressing either full-length netrin-1 (HEK FL) or netrin-1 lacking the substrate-binding, C-terminal domain (HEK NTNdC) were cultured in close proximity to two dorsal explants for 16 h. I, Full-length netrin-1 triggered robust outgrowth (3.9 bundles per explants, n = 12). However, the presence of heparin within the media largely eliminated this outgrowth (0.1 bundles per explants, n = 11, K) and expression of domain C lacking netrin-1 severely reduced outgrowth (0.9 bundles per explant, n = 11, L). Outgrowth in G and H was quantified at the total length of bundles from each explant. Outgrowth to in J–L was quantified as the number of bundles per explant. Scale bars, 100 μm.

Figure 3.

Substrate adsorption of netrin-1 is important for attracting spinal commissural neuron axons within the developing spinal cord. A, Schematic of the turning assay whereby mouse E10 dorsal cord were dissected and cultured alongside aggregates of HEK293 cells expression either full-length or NTN1dC. B–D, HEK293-expressing full-length netrin-1 deflected axons over an average distance of 164 μm (n = 18). B, E, F, However, when cellular aggregates expressed NTN1dC, the average distance decreased by 53% (mean of 77 μm, n = 22). Scale bar, 100 μm. **p < 0.01 (least significant difference, LSD).

Figure 4.

Netrin-1 Substrate adsorption is important for the phosphorylation of CAS and FAK. A, B, Diagrams of the domain structure of CAS and FAK showing the examined tyrosine phosphorylation sites within the substrate domain (SD) of CAS and the kinase domain of FAK. C, D, Top panels are immunofluorescent images of phosphorylated tyrosine 410 of CAS (pCASY410) or tyrosine 397 of FAK (pFAKY397) within spinal commissural growth cones. Brightness and contrast values are unmodified, but to highlight intensity differences, gray scale images were converted to the “Fire” LUT (Look Up Table) of ImageJ. Bottom panels show DCC labeling. E, Normally, netrin-1 increases average growth cones area approximately doubles (2.1-fold, n = 60, p < 0.01) after 15 min with 200 ng/ml netrin-1. However, reducing substrate adsorption of netrin with 2 μg/ml heparin disrupts this expansion (n = 60) as measured based on DCC staining. F, Addition of netrin-1 increases the integrated density of pCASY410 (3.1-fold, n = 50) and pFAK397 (2.2-fold n = 50) labeling. Coapplication of heparin significantly reduced the pCASY410 by 54% (1.7-fold of baseline), n = 49) and pFAKY397 by 44% (1.0-fold of baseline, n = 47). G, Western blot images from spinal commissural neuron (SCN) cultures for FAK and CAS phosphorylation following 15 min stimulation with 200 ng/ml netrin-1. H, Quantification of the change in netrin-induced phosphorylation relative to the absence of netrin. The presence of heparin significantly reduced the amount of CAS-Y410 (57%, n = 6), FAK-Y397 (61%, n = 8) and FAK-Y576 (71%, n = 6) phosphorylation. **p < 0.01 (LSD). Scale bar, 5 μm.

Figure 5.

Myosin II is required for traction on Netrin-1. A–C, Immunofluorescent labeling of myosin IIa and IIb shows enrichment in the central region of the growth cone. D–F, Normally microtubule filaments are concentrated in the central region of normal growth cones, while actin filaments are found throughout. G–I, However, inhibition of myosin II with 50 μm blebbistatin for 1 h resulted in growth cones with dramatically longer filopodia and dense microtubule arrays that penetrate deep into the peripheral region of the growth cone (arrow). J, K, Five representative plots of the amount of force over time exerted on netrin-1-coated beads normally (J) and in the presence of 50 μm blebbistatin (K). L, Growth cone pulling forces on netrin-1 within 8 min of initial contact were categorized into three responses: <30 pN, between 30 and 60 pN, and >60 pN. Typically, approximately three quarters of spinal commissural neuron growth cones exert >60 pN of force on netrin-1 within 8 min (74%, n = 50). However, in the presence of the myosin II inhibitor blebbistatin (Blebb) not a single growth cone was able to generate >60 pN. Instead the vast majority generated <30 pN (93%, n = 15). Similarly, significantly less growth cones exerted >60 pN in the presence of Y-27632 (47%, n = 19) and ML7 (44%, n = 18). *p < 0.05, **p < 0.01 relative to Ctrl (LSD). Scale bar, 5 μm.

Figure 6.

Myosin II contractions underlie netrin-1-induced FAK Phosphorylation. A, Western blot images of tyrosine phosphorylation within the substrate domain of CAS (Y165 and Y410) or the kinase domain of FAK (Y397 and Y576) following 15 min netrin stimulations (200 ng/ml) in the absence of inhibitors or following preincubation for 1 h with the Src Family Kinase inhibitor PP2 (10 μm) or the myosin II inhibitor blebbistatin (50 μm, blebb). B, Quantification of netrin-1 stimulated phosphorylation intensity changes relative to control cells (in the absence of both netrin and blebbistatin). Inhibition of myosin II reduced netrin-1-induced phosphorylation of FAK on the Y397 and Y576 sites by 77% (n = 8) and 62% (n = 5), respectively. No significant effects were seen on the netrin-1-induced phosphorylation within the substrate domain of CAS (Y165 and Y410). C, Top panels are immunofluorescent images of phosphorylated tyrosine 410 of CAS (pCASY410) and tyrosine 397 of FAK (pFAKY397) within spinal commissural growth cones. Brightness and contrast values are unmodified, but to highlight intensity differences, gray scale images were converted to the Fire LUT of ImageJ. Bottom panels show DCC labeling. D, Inhibition of myosin prevented an increase in the integrated density of FAK-Y397, but not of CAS-Y410, labeling (n = 50 for each condition) relative to control (in the absence of both netrin and blebbistatin). **p < 0.01(LSD). Scale bar, 5 μm.

Figure 7.

A, B, Mechanical force on netrin-1 triggers FAK autophosphorylation. DIC and immunofluorescent images of spinal commissural neurons that were presented with netrin-1-coated beads, then fixed and immunofluorescently processed for FAK autophosphorylation (pFAKY397). In one scenario the bead was released from the optical trap 15–30 s after its initial contact with the growth cone and then processed 5′ later (A). In the other scenario, the bead was held for the entire 5′ period thus allowing the growth cone to build >30 pN of force on the bead (B). The average immunofluorescent signal within each growth cone was normalized to the average intensity of a 5 μm² area of the axon segment immediately adjacent to the growth cone. When Netrin-1-coated beads were held (B) there was a 63% (n = 13, p < 0.05) increase in FAK Y397 autophosphorylation within the growth cone compared with when it is released (A). Brightness levels of immunofluorescent images are unmodified. Dotted circle denotes the location of the bead. Scale bars are 10 μm (left) and 2 μm (right).

Figure 8.

Inhibition of FAK disrupts strong traction forces on netrin-1. A, Western blot images of the autophosphorylation site within FAK (Y397) and total FAK. One hour preincubations with the FAK kinase inhibitor PF-228 (10 μm) efficiently eliminated both endogenous and netrin-induced (15 min, 200 ng/ml) phosphorylation. B, Inhibition of FAK with 10 μm PF-228 reduced the number of growth cones capable of pulling on netrin-1 with >60 pN (10%, n = 30). C, Diagram of FAK domain structure indicating selected protein association regions. FRNK lacks the FERM, the first proline-rich (PR) and kinase domains. DCC binds to FAK's FAT domain. D, Overexpression of FAK had no significant effect on the ability of growth cones to pull on netrin-1 with >60 pN (70%, n = 23). However, fewer axons generated >60 pN on netrin-1 when expressing FAK-Y397F (22%, n = 18) or FRNK (35%, n = 17). E, F, Inhibition of FAK kinase activity or expression of FAKY397F and FRNK reduced the percentage of growth cones that expanded their initial contacts on the netrin-1 beads. Adhesion expansion was quantified as the number of filopodia whose contacts are stabilized on the bead. Numbers above each bar represent the number of growth cones examined for each condition. G, A possible mechanism whereby traction force on netrin activates FAK's kinase activity through separation of FAK's FERM domain from the kinase domain. The C-terminal FAT domain of FAK associates with the P3 intracellular domain of DCC, while its N-terminal FERM domain indirectly associates with actin filaments. *p < 0.05, **p < 0.01 relative to Ctrl (B, E) or FAK (D, F) (LSD).