A range of phenotypes are observed following loss of cadherin function. Expression of red fluorescent protein (RFP) from RNAi vectors allows cell migration and morphology to be assessed. (A) RNAi-mediated knockdown of Fat2 cadherin causes an accumulation of electroporated cells in the mantle zone (mz) of the neural tube. (B) A similar phenotype is observed after knockdown of Notch1 (Das et al., 2006), whereas cells targeted with luciferase RNAi (C) are located throughout the neural tube in both ventricular zone (vz) and mantle zone. (D) Fat1 RNAi results in a reduction in size and cell number (32% fewer cells) in the electroporated side of the neural tube. (G) Expression domain of Pax6 is reduced relative to the contralateral side. (E,H) Loss of N-cadherin produces a decrease in cell number on the electroporated side, together with disruption of the apical surface of the neural tube, as marked by loss of β-catenin expression (H, arrows). (F,I) Protocadherin 19 RNAi leads to a 50% increase in the number of cells on the electroporated side of the neural tube and an expansion of Pax6 expression compared with the contralateral side (I). Some electroporated cells also migrate into the unelectroporated side of the neural tube (arrows in F), suggesting there may be some loss of roof plate integrity. (J,K) RNAi-mediated knockdown of protocadherin 10 leads to a dorsal shift in expression of Pax6 (J) and Olig2 (K). (L,M) Knockdown of FatJ cadherin leads to an increased number of Lim1⁺/Lim2⁺ cells. (N) This increase is spatially restricted to intermediate regions of the neural tube (bottom panel, arrows); dorsal Lim1/Lim2-expressing cells are unaffected by FatJ RNAi (N, top panel).

Rescue of Fat2 RNAi phenotype by NICD. (A) RNAi-mediated knockdown of Fat2 cadherin causes an accumulation of electroporated cells in the mantle zone of the neural tube. (B,C) A similar phenotype is observed after knockdown of Notch1 (Das et al., 2006) (B), whereas cells targeted with luciferase RNAi (C) are located throughout the neural tube in both ventricular zone (vz) and mantle zone. (D,E) Co-electroporation of the Notch intracellular domain (NICD) rescues the Fat2 knockdown (D), as well as the Notch1 knockdown phenotype (E). The rescue is evident by the location of electroporated (RFP+) cells in both the mantle zone and ventricular zone (compare A with D). This suggests that Fat2 is involved in Notch1 signalling.

Knock down of FatJ mRNA by RNAi vectors and FatJ expression pattern. (A-F) RNAi-mediated knockdown of FatJ was verified by in situ hybridisation. Electroporation with FatJ RNAi vectors A and B led to a marked reduction in FatJ expression, restricted to the electroporated side of the neural tube (A-C), whereas luciferase RNAi vectors had no effect on the expression of FatJ mRNA (D-F). (G-I) Similar levels of FatJ mRNA knockdown were produced by the other FatJ RNAi vectors (data not shown) and all three RNAi vectors produced a similar expansion of the Lim1/Lim2 domain when electroporated individually. Average percentage increase of Lim1⁺/Lim2⁺ cells shown, error bars are s.e.m. (J) At St 10, high levels of FatJ mRNA are detected in the ventral somites with lower levels of expression throughout the neural tube with the exception of ventral and dorsal-most regions. nt, neural tube; scl, sclerotome. (K,L) At St 19 and 22, FatJ is expressed at higher levels in the intermediate region of the neural tube. Expression is restricted to a diamond-shaped domain that appears to correspond to the dp4-vp1 progenitor domains. (M) Quantification of the number of Lim1⁺/Lim2⁺ cells in the neural tube of these embryos reveals all three vectors produced an increase of ~20-30% more Lim1⁺/Lim2⁺ cells compared with the unelectroporated side. Data are mean±s.e.m.

Detailed analysis of increase in interneurons following FatJ knockdown. (A) Following loss of FatJ, there is no change in the number of cells expressing the marker Islet1 (Isl1), which marks the dI3 class of interneurons. (B-F) Loss of FatJ results in an increase in specific interneuron number, relative to the contralateral side: 9.52% more Pax2⁺ cells (B; dI4 and dI6 interneuron); 16.11% more Lmx1b⁺ cells (C; dI5 interneurons); 15.24% more Lim1⁺/Lim2⁺ cells (D; dI2, dI4, dI6, v0 and v1 interneuron classes; 20.00% more Evx1+ cells (E; marking v0 interneurons); 17.91% more En1+ cells (F; v1 interneurons). (G) Loss of FatJ does not increase the number of cells expressing Chx10 (v2 interneuron class). Therefore, the increase in interneuron number is observed from the dI4 to v1 interneurons. Average cell counts for each marker on both sides of the neural tube are shown graphically on the right of the figure (error bars show s.e.m.). Statistical significance of the cell counts was determined using a paired Student's t-test.

Loss of FatJ increases progenitor cell number and does not cause premature differentiation. (A-F) Comparison of FatJ expression (A) with the progenitor markers Pax6 (B), Pax7 (C), Dbx1 (D) and Nkx6.1 (E). A schematic of the defined progenitor and differentiated neuron domains is shown, for reference, in F. (G,H) Loss of FatJ expression does not alter the domain of the neural progenitor markers Dbx1 (G, green cells), Nkx6.1 (G, blue cells) or Pax6 (H). In each case, the domain of expression is similar on the electroporated side and the unelectroporated side of the neural tube. (I) BrdU incorporation is increased in cells targeted by FatJ RNAi, resulting in ~14% more BrdU-positive cells within the normal FatJ expression domain (boxed region). (J) There is a ~20% increase in mitotic cells marked by phospho-histone H3 (pH3) within the FatJ expression domain (boxed region). (K,L) Quantification of BrdU+ (K) and pH3+ (L) cells reveals statistically significant increases in number of mitotically active cells after FatJ RNAi. (M-O) Detailed quantification shows that FatJ knockdown leads to statistically significant increases in the number of Pax6+ (O) and Dbx1+ (N) cells, but does not change the number of Nkx6.1+ cells (M), which lie ventral to the domain of FatJ expression. Cell counts show average cell number±s.e.m.; statistical significance was determined using a paired Student's t-test. (P) Loss of FatJ expression leads to an increased number of En1-expressing cells (left-hand panel, electroporated; right-hand panel, control); however, none of the En1+ cells co-express Pax6 (arrows).

Expression of neural progenitor and apical proteins following FatJ knockdown. (A,B) Loss of FatJ expression does not alter the domain of the neural progenitor markers Pax7 (A) or Nkx2.2 (B). (C,D) In each case, the domain of expression is similar on the electroporated side and the unelectroporated side of the neural tube. Loss of FatJ does not affect the expression of Crumbs2 (C) or N-cadherin (D), as both proteins are markers of the apical cell surface; this suggests that the increase in progenitor cells is not due to disrupted apicobasal polarity. All sections are counterstained with DAPI to show nuclei.

The FatJ knockdown phenotype can be rescued by knockdown of Yap1 and Tead4. (A) Knockdown of Yap1 by RNAi was verified using in situ hybridisation. A significant knockdown of Yap1 mRNA, normally expressed widely in the ventricular zone with the exception of the cells immediately dorsal to the floor plate, was observed following electroporation of Yap1 RNAi vectors (arrow). (B) Luciferase RNAi does not affect Yap1 expression. (C) RNAi-mediated knockdown of Yap1 has no significant effect on the number of Lim1⁺/Lim2⁺ cells in the neural tube. (D,E) Simultaneous knockdown of FatJ and Yap1 rescues the phenotype observed with FatJ knockdown (shown for reference in E). No difference is detected in the number of Lim1⁺/Lim2⁺ cells when compared with the unelectroporated control side. (F) Quantitative analysis showing knockdown of either Yap1 or Tead4 is able to rescue the increase in Lim1⁺/Lim2⁺ cells caused by loss of FatJ expression. FatJ RNAi leads to 16.77% more Lim1⁺/Lim2⁺ cells, whereas loss of Yap1 or Tead4 at the same time as loss of FatJ rescues the phenotype. Error bars are s.e.m., *P<0.05 (paired Student's t-test). (G) Electroporation of dominant negative (DN) Yap1 or Tead4 constructs together with FatJ RNAi vectors results in normal numbers of Lim1⁺/Lim2⁺ interneurons, therefore rescuing the increase in Lim1⁺/Lim2⁺ cells observed with FatJ RNAi alone. Data are mean±s.e.m. (H) Western blot analysis of pYap1 levels 72 hours after electroporation reveals a decrease in pYap1 levels after FatJ RNAi when compared with the Luc RNAi sample. pYap1 and tubulin levels were analysed using Quantity-one software after collection using charge coupled device (CCD) camera detection of chemiluminescent signals and pYap1 values were normalised to tubulin levels. (I) pYap1 levels are reduced within the FatJ expression domain 72 hours after FatJ RNAi, particularly at the lateral edge of the ventricular zone adjacent to the mantle zone (arrows). The lateral boundary of the ventricular zone is indicated by a broken line. (J) pYap1 expression is unchanged 72 hours after electroporation with Luc RNAi vectors.