The Wnt receptor Frizzled-4 modulates ADAM13 metalloprotease activity

Abstract

Cranial neural crest (CNC) cells are a transient population of stem cells that originate at the border of the neural plate and the epidermis, and migrate ventrally to contribute to most of the facial structures including bones, cartilage, muscles and ganglia. ADAM13 is a cell surface metalloprotease that is essential for CNC cell migration. Here, we show in Xenopus laevis embryos that the Wnt receptor Fz4 binds to the cysteine-rich domain of ADAM13 and negatively regulates its proteolytic activity in vivo. Gain of Fz4 function inhibits CNC cell migration and can be rescued by gain of ADAM13 function. Loss of Fz4 function also inhibits CNC cell migration and induces a reduction of mature ADAM13, together with an increase in the ADAM13 cytoplasmic fragment that is known to translocate into the nucleus to regulate gene expression. We propose that Fz4 associates with ADAM13 during its transport to the plasma membrane to regulate its proteolytic activity.

Keywords: ADAM; Development; Neural crest; Wnt; Xenopus laevis.

Fig. 1.

ADAM13 binds to Fz4 and Fz4-v1. (A) Co-immunoprecipitation experiments of Fz4 and Fz4-v1 with ADAM13. HEK293T cells were transfected with the various constructs cloned into the pCS2+ vector alone or in combination. ADAM13 was immunoprecipitated using the cytoplasmic domain antibody g821, and Myc-tagged (mt) Fz4 constructs were detected with mAb 9E10. Fz4-mt and Fz4-v1-mt co-precipitate with ADAM13 only when both proteins are co-transfected. Bands observed in lanes 2 and 3 of the total extract blotted with ADAM13 are non-specific. (B) Schematic diagram of mature ADAM13 including the metalloprotease (M), disintegrin (D), cysteine-rich (C), EGF repeat (E) and the cytoplasmic (Cy) domain. In co-immunoprecipitation experiments using a rabbit polyclonal antibody to the Fz4-v1 protein, all ADAM13-mt constructs containing the cysteine-rich domain were co-precipitated, while the disintegrin domain alone was not (D). (C) Real-time PCR on dissected CNC and stage-matched whole embryos (Total), detecting the level of Slug, Fz4, or Fz4-v1 expression, represented as a percentage of GAPDH expression. The average cycle threshold (CT) values in the CNC were as follows (Slug 24.7, Fz4 28.3, Fz4-v1 29.7, GAPDH 23.9). No CT was seen for water or no RT control. Error bars represent ±s.e.m. (D) Double in situ hybridization using short probes recognizing Fz4 or Fz4-v1 (blue) as well as probes recognizing the two CNC markers Sox10 and Twist (both red). The arrowheads point to the tip of the CNC segment. Scale bar: 500 µm. In situ hybridization for ADAM13 is presented as a reference marker (A13).

Fig. 2.

Fz4 colocalizes with ADAM13 in CNC cells. ADAM13-GFP, Fz4-RFP and Fz4-v1 RFP were expressed in CNC explants placed on fibronectin substrate. The fluorescence was detected using a spinning disc confocal acquiring every 2 seconds for one minute. Frames from each movie were extracted and annotated. ADAM13-GFP colocalizes with Fz4-RFP in vesicles (boxed areas and arrowhead) 10 µm from the fibronectin substrate (upper panel). At the plasma membrane in contact with the substrate, ADAM13-GFP colocalizes with Fz4-RFP in both vesicles (arrowheads) and at the membrane (arrow, middle panel). ADAM13-GFP is detected in close association with Fz4-v1 in membrane protrusions in direct contact with the fibronectin substrate (boxed areas, lower panel).

Fig. 3.

Fz4 inhibits ADAM13 proteolytic activity. (A) Western blot showing the cleavage of the proto-cadherin PAPC by ADAM13. The medium from transfected cells was collected and glycoproteins purified using concanavalin-A–agarose beads (ConA). The extracellular fragment of PAPC (60 kDa) is present when ADAM13 is co-transfected (lane 1) but absent when the protease dead ADAM13-E/A is used (lane 2). Co-transfection of Fz4-v1 decreases PAPC cleavage by about 50%, whereras Fz4 completely inhibits the cleavage (compare lanes 1, 3 and 4). Neither form inhibits ADAM13 self-shedding. (B) Co-immunoprecipitation of ADAM13 and PAPC. ADAM13 was co-transfected either with PAPC, Fz4-mt or Fz4-v1-mt alone, or in various combinations. ADAM13 was immunoprecipitated with g877 antibody and associated PAPC and Fz4 or Fz4-v1 were detected by western blot. Fz4 and Fz4-v1 do not compete for the binding of PAPC to ADAM13 (compare lanes 1 with 3, 4 and 6).

Fig. 4.

Overexpression of Fz4 inhibits CNC cell migration. (A,B) Histogram representing the percentage of embryos with no fluorescent CNC cells in the migration pathway following targeted injection at the 8-cell stage in one dorsal animal blastomere (% inhibition of CNC cell migration). Injection of RFP alone is used to normalize to zero and account for any mis-targeting. (A) Injection of Fz4 inhibits migration and can be rescued by either ADAM13 or the cadherin-11 extracellular fragment EC1-3. Student's t-test was performed to compare % inhibition to Fz4 alone. (B) Similarly, injection of Fz4-v1 inhibits migration, and can be partially rescued with ADAM13, the extracellular domain of cadherin-11 (EC1-3) and ADAM13 lacking the cytoplasmic domain (Δcyto). Student t-tests were used to compare values to Fz4-v1. (C) Fluorescently labeled CNC cells were grafted to determine whether the inhibition of CNC cell migration was cell autonomous. CNC cells overexpressing Fz4-v1 do not migrate. Wild type CNC cells grafted into embryos overexpressing Fz4-v1 in the pathway migrated similar to control. (D) Western blot from HEK293T cells transfected with ADAM13, or the non-proteolytic mutant of ADAM13 (E/A) and PAPC. These cells were mixed with cells transfected with Fz4-v1 or RFP and incubated together for 24 h. The conditioned supernatant, which contains Fz4-v1, does not inhibit ADAM13 cleavage of PAPC. *P<0.05, ***P<0.001. n = number of embryos scored from three or more independent experiments. Error bars represent ±s.e.m. ns, not significant.

Fig. 5.

Loss of Fz4 inhibits CNC cell migration. (A) Representative embryos grafted with fluorescent CNC cells from control embryos (GFP) or embryos injected with the Fz4 morpholino (MO Fz4). Scale bar: 500 µm (B) Histogram representing the percentage of embryos without CNC cell migration following the grafts. (C) Histogram representing the percentage of embryos without CNC cell migration following targeted injection. Loss of Fz4 inhibits CNC cell migration and is not significantly rescued by the overexpression of ADAM13. n: represents the total number of embryos. Asterisks represent statistical significance obtained by using student's t-test. ***P<0.001, ns: not significant. At least three independent experiments were performed. Error bars represent ±s.e.m.

Fig. 6.

Loss of Fz4 affects ADAM13 processing. (A) Western blot of ADAM13 at various stages of embryo development (stage 19 to 30). Embryos were injected with the Fz4 morpholino (+) or the ADAM13 morpholino (MO13, last lane) to visualize the specific bands. In control embryos the pro-form (P, 120 kDa), the mature form (M, 100 kDa) and the cytoplasmic domain (Cyto, 17 kDa) are visible, whereas these are absent in MO13 embryos. There is no difference of processing at stage 19. At stage 24, the mature form of ADAM13 decreases while the cytoplasmic domain increases in MOFz4 injected embryos (red line). At stage 30 the decrease of the mature form is still obvious, whereas there is no visible difference for the cytoplasmic domain fragment. The β1 integrin subunit was used as a loading control. (B) Western blot for ADAM13 in embryos injected with RFP, Fz4 or Fz4-v1 mRNA. Embryos injected with the morpholino to ADAM13 (MO13) serve as a control for the antibody specificity. At stage 19 embryos expressing Fz4 or Fz4-v1 show less ADAM13 cytoplasmic domain fragments (Cyto, 47% and 23%, respectively). At this stage Fz4 also reduces the mature form of ADAM13 (M, 58%), whereas Fz4-v1 does not. At stage 30, Fz4-v1 injected embryos have almost no cytoplasmic domain fragment, whereas the mature form is not affected. (C) Representative example of dissected CNC expressing ADAM13-GFP and mb-Cherry (total of eight explants for each condition). ADAM13 GFP is clearly present in control CNC, whereas it is undetectable in CNC injected with MO Fz4. The morpholino has no effect on mb-Cherry expression. Scale bar: 50 µm.

Fig. 7.

Overexpression of ADAM13 reduces the severity of the CNC-positioning defect caused by loss of Fz4. (A,B) In situ hybridization with Sox10 and Twist to label neural crest cells in embryos injected with various constructs. Representative examples of the various phenotypes are on the left and are color-coded. Asterisks indicate the injected side. The histogram on the right represents the percentage of phenotypes observed for each injection. The number of embryos (n) is indicated at the top of each bar. Gray, wild type; yellow, a mild inhibition; orange, embryos with a more-dorsal position of the CNC and a less defined segment (fused); red, embryos in which the CNC cells have migrated less than half the distance from the control side (strong inhibition). (A) Embryos were injected at the 4-cell stage. Similar results were obtained by injecting one cell at the 8-cell stage (B), similar to the targeted injections. Scale bars: 500 µm.