A differential developmental pattern of spinal interneuron apoptosis during synaptogenesis: insights from genetic analyses of the protocadherin- gene cluster

DEV026807 Supplementary Material

Files in this Data Supplement:

• Supplemental Figure S1 -

  Fig. S1. A subset of FoxP2-positive interneurons cells are dI2-derived. In spinal cord sections from E16 Atoh1^+/− mice (left panel), immunostaining with an antibody against β-galactosidase (red) labels Atoh1-expressing dI1 interneurons; none of these coexpress FoxP2 (green). In sections from Atoh1^−/− mice, some β-galactosidase-positive dI1 neurons adopt a dI2 fate (Gowan et al., 2001; Helms and Johnson, 1998), and these are observed to coexpress FoxP2 (inset, arrowheads).

• Supplemental Figure S2 -

  Fig. S2. Synapse loss and spinal interneuron apoptosis are spatially correlated in Pcdh-γ null mutant neonates. Quantification of the number of immunostained synaptic puncta in Pcdh-γ^del/del; Bax^−/− double-mutant spinal cords at P0, as compared with controls. Synapse loss is restricted to the ventral horn, which is where interneuron apoptosis is primarily occurring prenatally in both wild-type (see Fig. 6) and Pcdh-γ^del/del (Figs 2 and 3) mice. Data are expressed for mutants as a percentage of control values. **P<0.001; *P<0.05.

• Supplemental Figure S3 -

  Fig. S3. Characterization of a novel conditional mutant allele of Pcdh-γ. (A) Simplified schematics of the wild-type (Wt) Pcdh-γ locus and of the Pcdh-γ^fcon3 allele, showing loxP sites (orange triangles) flanking constant exon (ce) 3 fused in-frame to GFP. (B) Western blot of brain lysates from P0 Actin-Cre; Pcdh-γ^+/fcon3 and Actin-Cre; Pcdh-γ^fcon3/fcon3 probed with an antiserum recognizing ce1/2 demonstrates appropriately sized bands in heterozygote controls but none in mutants; blots were probed with antibodies to γ-tubulin as a loading control. (C) RT-PCR analysis of RNA from P0 Actin-Cre; Pcdh-γ^+/fcon3 and Actin-Cre; Pcdh-γ^fcon3/fcon3 brains demonstrates the expected loss of spliced transcripts containing ce3, but the continued presence of those containing ce1 and 2, in mutants. Spliced transcripts containing variable exons (A8 and B1 are shown) and ce1 and 2 are present but reduced in Actin-Cre; Pcdh-γ^fcon3/fcon3 brain as compared with control levels. (D,E) Analysis of ventral interneuron populations in P0 Actin-Cre; Pcdh-γ^fcon3/fcon3 spinal cords demonstrates that these mutants phenocopy Pcdh-γ^del/del null mutants. Data are expressed as the percentage survival of each population in Actin-Cre; Pcdh-γ^fcon3/fcon3 spinal cords as compared with control numbers. Scale bar: 100 µm.

• Supplemental Figure S4 -

  Fig. S4. Cre-mediated recombination of the Pcdh-γ^fcon3 allele in restricted interneuron populations. (A-C) Sections of E12 spinal cord from Wnt1-Cre; Pcdh-γ^fcon3/fcon3 mice immunostained with antibodies against GFP (green, A,C), Cre (red, A,B) and/or FoxP2 (green, B; red, C) demonstrate the loss of the GFP-fused constant exon 3 from dorsal interneuron populations, including FoxP2-positive dI2 interneurons that are migrating into the ventral horn. (D) Sections of P0 spinal cord from Wnt1-Cre; Pcdh-γ^fcon3/fcon3 mice immunostained with an antibody against GFP demonstrate excision of the floxed allele throughout most of the dorsal horn, but not in the ventral horn regions into which FoxP2-positive, Pax2-negative dI2 interneurons migrate. Although Wnt1-Cre; Pcdh-γ^fcon3/fcon3 mutants were recovered in the expected numbers at P0, none were identified from litters genotyped at P5, suggesting neonatal lethality. (E-E′′) Immunostaining using antibodies against Pax2 (green) and Cre (red) in sections of E11 spinal cord from Pax2-Cre; Pcdh-γ^fcon3/fcon3 mice. Nearly all Pax2-positive cells are also Cre-positive (yellow, E′′). Neither Pax2-Cre; Pcdh-γ^fcon3/+ heterozygotes nor Pax2-Cre; Pcdh-γ^fcon3/fcon3 homozygotes were recovered in the expected Mendelian ratios; although we have not ascertained the reason for this, we suspect that the Pax2-Cre BAC insertion site might be within chromosome 18, where the clustered Pcdh loci reside. Nevertheless, we were able to recover several litters containing Pax2-Cre; Pcdh-γ^fcon3/fcon3 neonates from one breeding pair. (F) Using the Z/EG reporter line (Novak et al., 2000), we confirmed that the Hb9-Cre line was active in many (although not all) motoneurons (arrowheads); a small number of other ventral cells, which might be the previously described Hb9-positive interneurons (Wilson et al., 2005), also expressed the GFP reporter in Hb9-Cre; Z/EG neonates, but these were neither FoxP2-positive nor Pax2-positive. Scale bars: 100 µm in A-E′; 60 µm in E′′; 40 µm in F.