Fig 2: Prethalamic pioneer axons belong to the Gsx2 lineage and express Pax6. (A) EGFP reporter of Cre recombinase activity and L1 staining in Gsx2Cre embryos at E14.5. (B) Immunohistochemistry showing Pax6 expression. (C) EGFP reporter of Cre activity in Zic4Cre embryos at E13.5 and L1 staining. (D) High-power image showing the absence of axonal projections from Zic4-lineage prethalamic cells towards the thalamus. (E-E?) Neurons belonging to the Gsx2 lineage extend parallel projections across the thalamus. E' and E? are a higher magnification of the area framed in E. (F-H') Injection of the tracer neurobiotin into the thalamus at E13.5 in the retrogradely labelled neurons in the prethalamus (arrow in F). (F-G) Double labelling of GFP and neurobiotin shows that prethalamic labelled neurons belong to the Gsx2 lineage. (H,H') Parallel section of F? processed for Pax6 immunohistochemistry combined with neurobiotin visualization showing that most prethalamic neurobiotin-positive neurons are also Pax6 positive (arrows). (I) Schematic summary. Scale bars: 500 µm in A-C,E; 250 µm in F-F?; 100 µm in D,E,G-H'. Ctx, cortex; Nb, neurobiotin; PTh, prethalamus; Th, thalamus; vLG, ventral lateral geniculate nucleus; vTel, ventral telencephalon. All descriptions were observed in at least three embryos from three different litters.

Fig 3: Efficient derivation of NPCs and neurons from wild-type and MT5-MMP-deficient hiPSCs. (A,B) Wild-type and MT5-MMP-deficient (KO) hiPSCs were efficiently differentiated into NPCs co-expressing the paired box 6 (PAX6, red) and nestin (green) (A,A’), as well as the SRY (sex-determining region Y)-box 2 (SOX2, red) and Nestin (green) markers (B,B’). Scale bar 20 µm. (C,D) Representative bright-field images of wild-type (C) and MT5-MMP-deficient NPCs (D) undergoing neuronal differentiation at DIV3, scale bar 50 µm. Note that multiple neurite-like protrusions were already observed at this early stage of differentiation, with no clearly distinct morphology between neurons of both genotypes at this stage. (E–H) Representative images of immunolabeled hiPSC-derived neuronal cultures immunolabelled with antibodies against the neuron-specific marker NeuN (green) and the astrocyte-specific marker S100ß (red) (E,F, scale bar 50 µm) after 60 days of differentiation (DIV60). In addition, cells expressed neuron-specific cytoskeletal protein MAP2B (G,H, scale bar 20 µm).

Fig 4: Ubiquitous conditional Pax6 deletion at E9.5 causes mapping errors in the thalamocortical connection. (A) Thalamic axons map to specific areas of the cortex via the thalamocortical pathway. (B) Schematic showing the cortical location where the tracers were placed. (C-H') Transverse sections showing axons retrogradely labelled with DiI and DiA. Although in controls (C,C',C1,C2,D,D',E,E') the areas labelled in the thalamus were clearly separated (dotted lines), in CAGCreER Pax6 cKOs (F-H') they overlapped, indicating the existence of mapping errors in these mutants. Arrows in F-H indicate the thalamocortical bundle segregating in visual (green) and somatosensory (red) halves at the level of the exit of the bundle from the diencephalon. (C',D',E',F',G',H') High-power images of the boxed areas in C,D,E,F,G,H, respectively. Arrows in G' and H' indicate DiI-labelled axons with abnormal trajectories towards medial thalamic regions. (I) Box plots showing the area occupied by each tracer (DiI or DiA) in each defined area (dLGN or VP) in controls versus Pax6 cKOs. Seven E15.5 embryos (four controls and three Pax6 cKOs), from three different litters were analysed. For each embryo, at least five rostrocaudal sections were quantified. In the box plots, the ends of the whiskers represent minimum and maximum values; the bottom of the box is the first quartile, the middle horizontal line is the median and the top of the box is the third quartile. Each dot represents an area value of a single section. P values were calculated by fitting the data to a generalized mixed linear model and the nested factors litter:embryo were considered as biological replicates (n=3). *P<0.001, ***P<1.0 e-15. (J) Summary of the tract tracing results. Scale bars: 200 µm in C,C1,C2,D,E,F,F1,F2,G,G1,G2,H,H1,H2; 50 µm in C',D',E',F',G',H'. dLGN, dorsal lateral geniculate nucleus; FC, fold change; OB, olfactory bulb; VL, ventral lateral nucleus; vLG, ventral lateral geniculate nucleus; VM, ventral medial nucleus; VMP, ventral medial posterior nucleus; VP, ventral posterior; vTel, ventral telencephalon; WT, wild type.

Fig 5: Axon guidance molecules and their receptors are altered in the diencephalon of CAGCreER Pax6 cKOs. (A-J) Expression pattern of Ntn1 and Sema3a mRNA and their receptors Unc5c and Plxna1 in the E13.5 diencephalon of controls (A-E) and CAGCreER Pax6 cKOs (F-J). C and H show two views of 3D reconstructions of the Ntn1 and Sema3a expression pattern from transverse in situ hybridization sections. (K) Schematics of the changes in Ntn1 and Sema3a expression in the thalamus of controls (left) versus CAGCreER Pax6 cKOs (right) and how we hypothesize this might affect the guidance of different subsets of thalamic axons. In controls (left), lateral thalamic neurons express Unc5c and Plxna1, being repelled by both guidance cues, which direct their axons out from the thalamus and towards the prethalamus in a straight trajectory. In Pax6 cKOs (right), some lateral thalamic neurons lose their Unc5c expression but maintain their Plxna1 expression, losing their repulsion to Ntn1 but maintaining it for Sema3a, presumably making their axons deviate towards medial thalamic regions. Scale bars: 500 µm in A,B,D,E-G,I,J; 100 µm in D',I'. Ctx, cortex; PTh, prethalamus; Th, thalamus; VTel, ventral telencephalon. All described patterns were observed at least in three embryos belonging to three different litters for each genotype, unless otherwise stated.