Fig 2: Cortical interneuron migration upon loss of PCDH19 in vivo. (A) Representative brain slices for the PCDH19 WT mouse, the PCDH19 HET mouse, and the PCDH19 KO mouse; the dotted line represents the area that was measured. (B) Quantification of GFP positive cells in the measured area shows average of GFP-positive cells percentage in the cortex per mouse genotype. Less migration to the cortex could be detected in the PCDH19 HET mouse compared to the control PCDH19 WT (*p = 0.0135), the Kruskal–Wallis test, and the Dunn's post hoc test.

Fig 3: PCDH19 extracellular domain misexpression significantly decreases cortical IN migration. (A) 72-h post electroporation taken Z-stack example images of brain slices of TdT control and various PCDH19 overexpression constructs for the Dlx5/6-CRE-IRES-eGFP fluorescence, including the dotted line marking the cortical field. (B) Z-stack example images of the brain slices shown in A of TdT control and diverse PCDH19 overexpression constructs for the electroporated neurons (Tomato fluorescence); the dotted line shows the cortical field. (C) A schematic of the ex vivo brain slice electroporation setup to study the overdose of PCDH19 on cortical IN migration from MGE-derived IN. Analysis of 72-h post electroporation in Dlx5/6-CRE-IRES-eGFP brain slices electroporated with bicistronic plasmids co-expressing a PCDH19 subdomain and TdTomato or only TdTomato in the empty bicistronic plasmid that was used as control. (D) After 72 h in culture, IN migration to the cortical field was determined by measuring the intensity of the TdT+ neurons within the cortical field divided by the intensity of the TdT+ neurons spread over the whole brain slice. Upon PCDH19 ECD TM misexpression, significant reduction in cortical IN migration to the cortical field could be measured [the Kruskal–Wallis multicomparison test with the Dunn's post hoc test, *p = 0.0312 (PCDH19 ECDTM) and *p = 0.0122 (PCDH19 FL)]. N number of replicates is indicated within the bars.

Fig 4: Effects of overexpressing PCDH19 subdomains on MGE cell migration and morphology. (A) A schematic of the ex vivo MGE explant electroporation setup to investigate the effect of PCDH19 overexpression in MGE-derived IN. Total minimum migration distance ‘d' was assessed after 48 h of electroporation in Dlx5/6-CRE-IRES-eGFP MGE explants with a PCDH19 construct or the empty TdTomato plasmid. (B–F) Example images of electroporated and cultured explants 48 h post electroporation of the TdTomato control plasmid (B) and PCDH19 constructs (C–F). (G) A dot plot depicting tdTomato+ IN-related minimal distance from the explant edge. Each dot represents one electroporated neuron in the respective condition; colors of the dots relate to different explants. Significantly shorter distance from the explant edge could be measured between PCDH19 FL and TdT control (the Kruskal–Wallis non-parametric test and the Dunn's post hoc test, ****p < 0.0001). (H) Quantification of TdTomato neurons per bin normalized against the total amount of TdT neurons per bin showed non-significant difference per bin (the mixed model ANOVA test). (I) A bar chart showing the percentages of INs that were classified into 4 different categories (insets depicting the diverse neuron morphology categories) from different electroporated explants calculated on the total amount of INs per category within each experimental group. Nonsignificant differences could be identified within the experimental groups per category (two-way repeated measures ANOVA with Holm-Sidak post hoc comparison). (J–L) Boxplots depicting TdTomato+ neuron-associated cable length (J), primary branch length (K), and average branch length (L) measurements in TdT control and PCDH19 constructs-electroporated MGE explants. Significant differences could be measured in the neuron cable length between PCDH19 FL and TdT control (****p < 0.0001), PCDH19 ECDTM, and TdTcontrol (****p < 0.0001) and PCDH19ICD?NLS and TdT control (*p = 0.0026) (the Kruskal–Wallis non-parametric test and the Dunn's post hoc test). The primary branch length (the Kruskal–Wallis non parametric test and the Dunn's post hoc test, *p = 0.0118) and the average branch length (the Kruskal–Wallis non parametric test and the Dunn's post hoc test, ***p = 0.0003) were significantly shorter in PCDH19 ECDTM compared to TdT control. DIV, days in vitro; scale bar: 500 µm.

Fig 5: PCDH19 subdomain constructs target expression to different subcellular locations. (A) NLS prediction within the mouse PCDH19 amino acid sequence with three different programs. Cyan represents the alternate exon and red the consensus sequence. (B) Full-length PCDH19 and overexpression subdomain constructs were produced C-terminally tagged with a GFP tag and inserted into a bicistronic plasmid containing TdTomato. (C) PCDH19 subdomain expression in slice EP neurons 72 h after electroporation. Separate fluorescence channels, as well as the overlay, show a similar subcellular localization to the Neuro2A cells. White arrowheads in PCDH19 FL indicate a GFP signal in the nucleoli.