Activation of FAK and Src are receptor-proximal events required for netrin signaling

Abstract

The axon guidance cue netrin is importantly involved in neuronal development. DCC (deleted in colorectal cancer) is a functional receptor for netrin and mediates axon outgrowth and the steering response. Here we show that different regions of the intracellular domain of DCC directly interacted with the tyrosine kinases Src and focal adhesion kinase (FAK). Netrin activated both FAK and Src and stimulated tyrosine phosphorylation of DCC. Inhibition of Src family kinases reduced DCC tyrosine phosphorylation and blocked both axon attraction and outgrowth of neurons in response to netrin. Mutation of the tyrosine phosphorylation residue in DCC abolished its function of mediating netrin-induced axon attraction. On the basis of our observations, we suggest a model in which DCC functions as a kinase-coupled receptor, and FAK and Src act immediately downstream of DCC in netrin signaling.

Figure 1

Netrin-stimulated tyrosine phosphorylation of DCC and FAK. (a) Netrin stimulates tyrosine phosphorylation of DCC and FAK in embryonic spinal neurons. IB, immunoblot; anti-pY, anti-phosphotyrosine antibody. (b) DCC extracellular domain antibody blocks netrin-induced FAK tyrosine phosphorylation. Bar graph shows quantitative data. α-DCC, anti-DCC. (c) Netrin stimulates tyrosine phosphorylation of DCC and FAK in transfected HEK293 cells. HEK293 cells were transfected with the indicated expression vectors. After netrin stimulation, DCC was immunoprecipitated and subjected to the indicated IB analysis. Lanes 3 and 4 are duplicates. (d) Co-IP of FAK with DCC. Left, results of IP with anti-HA followed by anti-DCC IB. Right, the reciprocal IP done. Anti-Flag IP was included as a negative control. (e) DCC and FAK interact at endogenous levels. Endogenous FAK was immunoprecipitated from the membrane fraction isolated from mouse brain lysate. The presence of endogenous DCC was determined by IB. Anti-HA and anti-Myc IgGs were used as negative controls. Lane 4 reflects a shorter exposure than lanes 1–3. (f) The C-terminal domain (FRNK) of FAK interacts with DCC. HEK 293 cells were transfected with the indicated plasmids. The various FAK deletions were immunoprecipitated by anti-HA and the presence of DCC was determined by IB. Expression levels of DCC and different FAK deletions in cell lysates are shown by IB. (g) FRNK inhibits netrin-induced DCC tyrosine phosphorylation. HEK293 cells were transfected with DCC with or without FRNK. Cells were stimulated with netrin for 20 min and tyrosine phosphorylation of DCC was determined by IB. The anti-phosphotyrosine western blot in this panel was exposed longer than that in c to visualize DCC tyrosine phosphorylation induced by netrin.

Figure 2

Src interacts with DCC and is activated by netrin. (a) Src activates DCC tyrosine phosphorylation. HEK293 cells were transfected with DCC (500 ng) and Src (50 ng) and treated with PP2 as indicated. Tyrosine phosphorylation of DCC was determined by anti-DCC IP followed by anti-PY20 IB. (b) Netrin stimulates Src. Cells were transfected with DCC (250 ng) and Src (10 ng) and stimulated with netrin. Tyrosine phosphorylation of DCC and Src were determined using anti-pY20 and Src-Y418, respectively. (c) Synergistic tyrosine phosphorylation of DCC by Src and FAK. DCC was cotransfected with various amounts of FAK and Src as indicated. Tyrosine phosphorylation of DCC was determined. (d) Co-IP of Src with DCC. HEK293 cells were transfected as indicated and subjected to IP with either anti-DCC (left) or anti-Src (right). Co-IP of Src (left) or DCC (right) was determined by IB. (e) Interaction of endogenous Src and DCC. Endogenous Src was immunoprecipitated. The presence of endogenous DCC in the Src IP was determined by IB. IP with anti-GAL-4 and anti-Src incubated with buffer only were used as negative controls. (f) Netrin enhances the interaction between DCC and Src in transfected HEK293 cells. Src was immunoprecipitated by Src antibody and the IP was probed with DCC antibody for DCC. (g) Netrin stimulates Src tyrosine phosphorylation and kinase activity in the E12 mouse brain cortex neurons. Left, Src tyrosine phosphorylation; right, quantification of kinase activity after 10 min and 20 min of netrin application. (h) Netrin stimulates Fyn tyrosine phosphorylation and kinase activity. Experiments are similar to those in g. (i) Netrin did not induce DCC tyrosine phosphorylation in Src/Fyn/Yes (S/F/Y) triple-knockout MEF cells. DCC and UNC-5 were cotransfected into wild-type (WT) and Src/Fyn/Yes triple knockout MEFs, and then netrin-induced DCC tyrosine phosphorylation was determined.

Figure 3

Tyr1420 in DCC is phosphorylated and the P3 domain in DCC interacts with FAK. (a) Schematic of DCC intracellular domain, showing the conserved P1, P2 and P3 domains along with four tyrosine residues. Also shown are some of the constructs used in Figure 3b–h. (b) FAK enhanced DCC Tyr1420 phosphorylation. HEK293 cells were transfected with the indicated plasmids. Y1, Y2, Y3 and Y4 denote mutations in the tyrosines of the intracellular domain of DCC, except for the one indicated in parenthesis. (c) As in b, except with Src coexpressed instead of FAK. (d) In vitro phosphorylation of DCC by Src and not FAK. Immunoprecipitated Src and HA-FAK were incubated with MBP-DCC. Above, ³²P incorporation. Below, total amounts of Src, FAK and MBP-DCC used in the kinase reactions. (e) Intracellular domain of DCC. DCC-ΔP1, ΔP2, ΔP3 and Δ1/2P3 correspond to deletions of residues 1147–1171, 1335–1356, 1412–1447 and 1426–1447, respectively. Shaded box denotes mutation of 22 C-terminal residues in DCC-1420Y/FΔP3*. (f) The C-terminal half of the P3 domain of DCC is required for FAK binding. Experiments were similar to those shown in b. (g) FAK binding is essential for tyrosine phosphorylation of DCC by FAK. DCC and various mutants were coexpressed with HA-FAK in HEK293 cells. Tyrosine phosphorylation (top) and total amount of protein (middle) of immunoprecipitated DCC are shown. (h) The P3 domain of DCC is essential for FAK activation. HEK293 cells were transfected as indicated. The netrin-DCC induced phosphorylation status of FAK at Tyr576 and Tyr577 was determined using phospho-specific antibody. (i) Kinase-inactive FAK cooperates with Src to stimulate DCC phosphorylation. DCC was cotransfected in HEK293 cells with 2 ng of wild-type FAK (WT), kinase-dead FAK (KD) and Src as indicated. Phosphorylation of immunoprecipitated DCC was detected (two different exposures are presented).

Figure 4

The 1400PXXP motif in DCC and the SH3 domain in Src are required for DCC and Src interaction. (a) Mutation of all DCC intracellular tyrosine residues did not affect Src-DCC interaction. HEK293 cells were transfected as indicated. Src was immunoprecipitated and bound DCC or DCC-4Y/F was determined by immunoblotting (IB). (b) Src kinase activity is required for optimal Src-DCC interaction. HEK293 cells were transfected as indicated. Cells were treated with PP2 for 20 min or 100 min followed by Src IP. Bound DCC was determined by IB. (c) Diagram of all nine putative PXXP motifs in the DCC intracellular domain. The amino acid number denotes the first proline in the PXXP motif and the P1400A mutant is also shown. (d) DCC-P1400A is compromised in Src binding. HEK293 cells were transfected with indicated plasmids. Src was immunoprecipitated and the presence of DCC in IP was determined by immunoblot. (e) The Src SH3 domain is important for the interaction with DCC. DCC-P1362A was included as a control. (f) Tyrosine phosphorylation of DCC mutants stimulated by Src. Experiments are similar to those shown in Figure 3g.

Figure 5

Src activity is required for netrin-induced axon growth. (a) PP2 inhibits netrin-stimulated axon growth. Explants from the dorsal half of E12 spinal cords were cultured in collagen gels for 20 h. Netrin-1-induced outgrowth of axon bundles was inhibited by PP2 but not PP3. Scale bar, 100 μm. (b) PP2 inhibits netrin-stimulated axon growth. The netrin (20 h) and PP2 (0.2 μM) treatments are indicated (netrin versus control, P < 0.0001; netrin versus netrin + PP2, P < 0.0001; netrin versus netrin + PP3, P < 0.57. (c) PP2 did not inhibit spontaneous axon growth. The explants were cultured for 40 h. PP2 compared with control, P < 0.33; PP3 compared with control, P < 0.88. Numbers denote explants counted.

Figure 6

Src function or DCC tyrosine phosphorylation is required for DCC-mediated axon attraction. (a) Superimposed neurite trajectories of DCC-expressing neurons in the presence of 0.1 μM PP2 or PP3 during 1-h exposure to a netrin gradient (2.5 μg/ml in the pipette, arrows) for all the neurons examined. Arrows, presence of netrin gradient or buffer control. ‘PP2’ and ‘PP3,’ presence of 0.1 μM PP2 and PP3, respectively, in the presence of a netrin gradient. (b) Cumulative distribution of turning angles. Significant differences from the control groups are marked (*P < 0.05, **P < 0.001; Kolmogorov-Smirnov test). (c) Average neurite extension during the 1-h assay for experiments shown in a. Numbers denote the total number of growth cones examined. Error bars, s.e.m. (d) Traced neurite trajectories of Src-overexpressing mutants in DCC-expressing neurons exposed to a netrin gradient similar to that in a. The netrin gradient (arrow) was present in all panels. ‘Control’ indicates no Src cotransfection. Src-KM, Src-SH3M and Src-SH2M are the kinase-inactive, SH3-domain and SH2-domain mutants, respectively. (e) Distribution of turning angles for the data in d. (f) Average neurite extension during the 1-h assay for experiments shown in d. (g) Expression of DCC-Y1420F blocks netrin-induced attraction. X. laevis spinal neurons with ectopic expression of DCC or DCC-Y1420F mutant were analyzed by in vitro axon turning experiments in the presence of a netrin gradient. Graph show the cumulative distribution of turning angles for all neurons examined. Turning responses of growth cones observed in wild-type DCC (DCCWT; open circle, n = 11) and DCC mutant (DCC-Y1420F; closed circle, n = 11) overexpressing neurons in response to a netrin-1 gradient (2.5 μg/ml) are shown. Significant differences from the control groups are marked (**P < 0.001 compared with WT; Kolmogorov-Smirnov test).