Fig. 6. Commissural axons are dorsally mispolarized in vivo in BmprIb^-/- single mutants and Math1::cre;BmprIa^flox/flox; BmprIb^-/- double mutants. (A,B) In either wild-type (A) or BMPRIA-deficient (B) fillets of the spinal cord, very few Tag1⁺ axons extend towards the RP (A, open arrowhead) and no axons cross the RP. (C-F') By contrast, many Tag1⁺ axons extend into the RP in either BmprIb^-/- single mutants or Math1::cre;BmprIa^flox/flox; BmprIb^-/- double mutants (open arrowheads, C,F), with commissural axons (closed arrowheads, D',F') now observed to cross the RP (outlined in D,F). (G) There is no significant difference (P>0.27) between the percentage of mispolarized axons in BmprIa^flox/flox control fillets (0.75%±0.19 s.e.m., n=8213 pLh2⁺ neurons from 9 embryos) and the BMPRIA-deficient (Math1::cre;BmprIa^flox/flox) fillets (0.95%±0.20 s.e.m., n=9262 pLh2⁺ neurons, 9 embryos). By contrast, a significant increase (P<0.002) is observed in BmprIb^-/- mutants (1.32%±0.14 s.e.m., n=7940 pLh2⁺ neurons from 10 embryos) compared with wild-type litter-mates (0.73%±0.10 s.e.m., n=10379 pLh2⁺ neurons, 12 embryos). The percentage of mispolarized commissural axons seen in fillets from the Math1::cre;BmprIa^flox/flox; BmprIb^-/- double mutant embryos (3.5%±0.35 s.e.m., n=2114 pLh2⁺ neurons, 2 embryos) is statistically identical (P>0.4) to that seen in fillets from Gdf7^-/- embryos (Butler and Dodd, 2003). Scale bar in B: 10 µm for A-F'.