December 1, 2008;
regulates planar cell polarity (PCP) signaling during vertebrate neural tube closure and establishment of inner ear
hair cell polarity; however, its signaling mechanism is unknown. Here, we demonstrate a new function for PTK7
in Xenopus neural crest migration and use this system in combination with in vitro assays to define the intersection of PTK7
with the non-canonical Wnt signaling pathway that regulates PCP. In vitro, using Xenopus ectodermal explants, we show that PTK7
) to the plasma membrane
, a function that is dependent on the PDZ domain of dsh
, as well as on the conserved kinase domain of PTK7
. Furthermore, endogenous PTK7
is required for frizzled7
localization. Immunoprecipitation experiments confirm that PTK7
can be found in a complex with dsh
, suggesting that it cooperates with frizzled to localize dsh
. To evaluate the in vivo relevance of the PTK7
localization, we analyzed Xenopus neural crest migration, as loss-of-function of PTK7
inhibits neural crest migration in whole embryos as well as in transplanted neural crest cells. Supporting the in vivo role of PTK7
in the localization of dsh
, a PTK7
deletion construct deficient in dsh
binding inhibits neural crest migration. Furthermore, the PTK7
-mediated membrane localization of a dsh
deletion mutant lacking PCP activity inhibits neural crest migration. Thus, PTK7
regulates neural crest migration by recruiting dsh
, providing molecular evidence of how PTK7
intersects with the PCP signaling pathway to regulate vertebrate cell movements.
[+] show captions
Fig. 4. PTK7 is expressed in cranial neural crest cells and required for their migration. (A-C) PTK7 expression pattern in Xenopus laevis detected by whole-mount in situ hybridization. (A,B) PTK7 is broadly expressed in premigratory neural crest cells at early neurula stages (black arrow). (C) At stage 26, PTK7 expression is detected in migrating cranial neural crest cells (black arrow). (D-M) Embryos injected in one blastomere at the two-cell stage with different constructs in combination with 100 pg GFP RNA as a lineage tracer. Neural crest migration was analyzed at neurula stages using the neural crest marker twist (D,F,H) or the midbrain-hindbrain marker engrailed (E,G,I). The injected side is shown on the right. (D,E) GFP-injected embryos. (F,G) Embryos co-injected with 10 ng control MO and GFP RNA. (H,I) Embryos co-injected with 10 ng PTK7 MO and GFP RNA. (J-M) Tadpole embryos analyzed by in situ hybridization with the neural crest markers twist (J,K) or AP-2 (L,M). The injected side is shown on the right. (J,L) Tadpoles injected with 10 ng control MO und 100 pg GFP RNA. (K,M) Tadpoles injected with 10 ng PTK7 MO and 100 pg GFP RNA. Inhibition of cranial neural crest migration is marked by arrows in H,K,M. (N) Graph summarizing the MO injection experiments. Left graph summarizes five independent experiments (analyzing neurula stages) and six independent experiments (analyzing tadpole stages), respectively. The right graph summarizes three independent rescue experiment. The percentage of migrating neural crest cells was determined by twist in situ hybridization. Columns are labeled with the number of injected embryos.
Fig. 7. Neural-crest-specific expression of PTK7 and dsh constructs. Two-cell stage embryos were injected in one blastomere with plasmids containing a minimal slug promoter driving either the expression of GFP (A,B), wild-type PTK7 (C,D) or δkPTK7 (E,F). (A-F) Tadpole stage embryos analyzed by twist in situ hybridization. The injected side is presented in the right panel (B,D,F). (A,B) Embryo injected with 100 pg slug-GFP plasmid. (C,D) Embryo injected with 100 pg slug-PTK7 plasmid. (E,F) Embryo injected with 100 pg slug-δkPTK7. Arrowhead in F indicates the inhibition of neural crest migration. (G) Graph summarizing the percentage of migrating neural crest cells in five independent injection experiments. The number of injected embryos is indicated on each column. (H,I) Embryo injected with 50 pg slug-δDEP. (J-O) Embryo injected with 50 pg slug-PTK7 together with 50 pg slug-δDEP (J,K) or 50 pg slug-dsh (L,M) or 50 pg slug-δPDZ (N,O). (P) Graph summarizing three dsh and PTK7 co-injection experiments. The number of injected embryos is indicated on each column.
Fig. 5. PTK7 overexpression does not affect neural crest migration. (A-C) Embryos were injected with 1 ng PTK7 RNA and 50 pg lacZ RNA in one blastomere at the two-cell stage and analyzed for twist expression by in situ hybridization at neurula (A) and tadpole stages (B,C). Light blue lacZ staining marks the injected side. Twist-positive neural crest cells show normal migration at neurula (A) and tadpole stages (B,C). (B) Injected side of a tadpole stage embryo. (C) Uninjected side. (D) Graph summarizing the percentage of twist-expressing embryos injected with 50 pg lacZ RNA, 0.5 ng PTK7 RNA, 0.5 ng myc-tagged PTK7 RNA or 1 ng PTK7 RNA at neurula and tadpole stages. Numbers on the columns indicate the number of injected embryos. (E) Confirmation of PTK7 expression using western blotting (co, uninjected controls) or (F) immunostaining on transverse sections of neurula stage embryos injected with 0.5 ng myc-tagged PTK7 RNA. In F, membrane localization can be detected with a Cy3-coupled myc antibody (left panel). Right panel shows an overlay with the DAPI staining.
Fig. 1. PTK7 recruits dsh to the plasma membrane. Animal caps were injected with different tagged RNAs, and protein localization was analyzed by confocal microscopy. The GFP-tagged dsh (green, left), the co-expressed myc-tagged protein (red, middle) and the merged pictures (right) are shown. (A) GFP-tagged dsh is localized to the cytoplasm of animal caps injected with 100 pg dsh-GFP RNA. (B) Myc-tagged fz7 is predominantly membrane localized in animal caps injected with 100 pg fz7-myc RNA. (C) Co-injection of 100 pg dsh-GFP and 100 pg fz7-myc RNA leads to membrane recruitment of dsh. (D) PTK7 is membrane localized in animal caps injected with 500 pg myc-tagged PTK7 RNA. (E) Animal caps co-injected with 100 pg dsh-GFP RNA and 500 pg PTK7-myc RNA show membrane-recruitment of dsh. Cells that do not express PTK7 in the membrane do not show membrane-localization of dsh (white arrowhead). (F) Animal caps injected with 250 pg RNA coding for a PTK7 mutant lacking the conserved kinase domain (δkPTK7) and 100 pg dsh-GFP RNA do not show membrane localization of dsh. (G) Graph summarizing the percentage of cells with membrane-localized dsh. For colocalization assays using PTK7 or fz7, only cells in which these proteins were membrane localized were analyzed. To determine the number of cells with cytoplasmic dsh localization DAPI co-staining was used. The total number of cells is indicated above each column.
Fig. 2. The PDZ domain is necessary for PTK7-dependent membrane translocation of dsh. (A) Dsh protein structure indicating the DIX, PDZ and DEP domain. RNA of GFP-tagged deletion mutants of these domains were expressed alone or in combination with wild-type PTK7 RNA in animal caps. Protein localization was determined by confocal microscopy. (B) Animal cap injected with 100 pg δDIX-GFP RNA showing cytoplasmic protein localization. (C) Co-injection of 500 pg PTK7-myc RNA partially translocates the cytoplasmic δDIX-GFP protein to the membrane. (D) Animal caps injected with 100 pg δPDZ-GFP RNA show only cytoplasmic localization of the protein. (E) The same is seen after co-expression with PTK7-myc RNA. (F) Animal caps injected with 100 pg δDEP-GFP RNA express the protein in the cytoplasm, whereas (G) co-injection of 100 pg PTK7-myc RNA leads to membrane recruitment. (H) Graph summarizing the percentage of cells with membrane-localized dsh. Only cells with membrane expression of PTK7 were analyzed for dsh localization. The total number of cells is indicated on each column.
Fig. 3. PTK7 is required for fz7-mediated dsh recruitment and phosphorylation. (A-D) PTK7 is required for fz7-mediated dsh recruitment to the plasma membrane. GFP-tagged dsh is shown in green, myc-tagged fz7 in red, HA-tagged PTK7 in blue. (A) GFP-tagged dsh is localized to the plasma membrane in animal caps injected with 100 pg dsh-GFP RNA, 100 pg fz7-myc RNA and 20 ng control MO. (B) GFP-tagged dsh is not recruited to the plasma membrane in animal caps injected with 100 pg dsh-GFP and 100 pg fz7-myc RNA and 20 ng PTK7 MO. (C) Co-injection of 100 pg wild-type PTK7 RNA lacking the MO binding site rescues dsh-localization of animal caps injected with 100 pg dsh-GFP RNA, 100 pg fz7-myc RNA and 20 ng PTK7 MO. (D) Graph summarizing the percentage of cells with simultaneously membrane-localized dsh and fz7. The total number of cells is indicated on each column. (E) PTK7 is required for fz7-dependent hyperphosphorylation of dsh. Embryos were injected with 100 pg dsh-myc RNA, 100 pg fz7 RNA, 500 pg PTK7-myc RNA, 20 ng control MO or 20 ng PTK7 MO in the combinations indicated. Animal caps were cut at stage 9 and their lysates were analyzed by western blotting using anti-myc antibodies. Hyperphosphorylated dsh is detected as a second high molecular weight band. One representative experiment of three independent experiments is shown.