Nature Genet. 3, 241– 246 (1993).

Originally, we reported that the 3′ end of Gli3 is not deleted in the XtJ allele (Fig. 4). U. Rüther brought to our attention that the XtJ allele indeed contains a deletion including the 3′ UTR of Gli3. Subsequently, we repeated the experiments and found that the data described in Fig. 4 using probe d was due to probe d containing sequences other than Gli3. These findings indicate that the deletion in the XtJ allele might not be intragenic; however, it does not affect our main conclusion that XtJ is a mouse model of GCPS. We regret this error.