DEV001255 Supplementary Material

Files in this Data Supplement:

• Supplemental Figure 1 -

  Fig. S1. Spatio-temporal dynamics of combinatorial expression of patterning and proneural HLH factors during early neurogenesis. All panels show double immunostaining of transverse sections of the rat embryo spinal cord at E11.5 and E12.5 (the dorsal side is top). Molecules stained in red and green are shown on the top, and the embryonic stages and axial level of the sections are shown on the left. Horizontal lines indicate the boundaries between the progenitor domains p2, pMN and p3. A circle in Ca indicates migrating Nkx2.2⁺ OLPs that began to emerge at E12.5 at the brachial level. Notice that these cells were negative for Olig2. Scale bars: 50 μm (shown in Cg).

• Supplemental Figure 2 -

  Fig. S2. Expression of patterning factors in the VZ at early (E12.5) and late (E16.5) stages. (A-F) Transverse sections of the brachial spinal cord were stained for Pax6 (A,D), Olig2 (B,E) and Nkx2.2 (C, F) (shown in green), and cell nuclei were stained with propiodium iodide (red). At the early neurogenic period, Pax6, Olig2 and Nkx2.2 are homogenously expressed in the VZ. By contrast, from E16.5 onwards, their expression is gradually downregulated in a subset of VZ cells (indicated by arrowheads). Scale bars: 50 μm (shown in F).

• Supplemental Figure 3 -

  Fig. S3. Overlapping expression of proneural HLH factors during early neurogenesis. Insets are higher-magnification views of the areas indicated by thick arrows. Mash1 and Ngn2 were expressed in separate domains between E11.5 and E12.5, the stage when motoneurons are generated in the Olig2⁺ domain (Aa-Da). However, their co-expression was detected in a subset of cells in the Pax6⁺ domain at E14.5 (Da, inset) and E16.5 (data not shown). The co-expression of Mash1 and Ngn1 in the Pax6⁺ domain began at around E12.5 at the brachial level (Cb), continued at E14.5 (Db) and disappeared at around E16.5 (data not shown). These Ngn1⁺, Ngn2⁺ and Mash1⁺ cells partly overlaps (30-50%) with each other in the Pax6⁺ domain between E12.5 and E16.5 (insets in Cb and Db). As a whole, however, they comprised only a fraction (<50%) of VZ cells at all stages examined. Likewise, Ngn2⁺ cells were only a part (<50%) of the VZ cells of the Olig2⁺ domain at the peak of somatic motoneuron generation (Bc). In the Nkx2.2⁺ domain, Ngn3 and Mash1 were co-expressed in a subset of progenitors between E12.5 and E14.5. Approximately 70% of Ngn3⁺ cells co-expressed Mash1, and 30% of Mash⁺ cells were Ngn3⁺ at E12.5 (inset in Cc). These cells comprised less than 40% of the total VZ cells at both E12.5 and E14.5 (data not shown). Scale bars: 50 μm in Aa, Ba-Bc, Ca-Cc (shown in Cc); 100 μm in Da, Db (shown in Db).

• Supplemental Figure 4 -

  Fig. S4. Neurons and oligodendrocytes among Nkx2.2⁺ and Olig2⁺ cells. All panels show transverse sections of the spinal cord at the brachial level. (A,B) Expression of the oligodendrocyte marker O4 (green) in Nkx2.2⁺ (A) and Olig2⁺ (B) cells (red) in the ventral MZ of E16.5 rat embryos. The majority of OLPs detected in the MZ at this stage express either Nkx2.2 or Olig2, but not both. Both populations express the early OLP marker O4 (arrowheads). (C,D) Expression of the neuronal marker TuJ1 (red) in Nkx2.2⁺ and Olig2⁺ cells (green) in the ventral MZ (C) and lateral MZ (D), respectively, in E14.5 rat embryos. The images on the right and bottom of C are three-dimensional images of a double-positive cell, indicated by arrow. D shows a cluster of Olig2⁺/TuJ1⁺ neurons (arrowheads). (E) Olig2⁺/Islet1⁺ cells detected in the lateral MZ in E14.5 rat embryos. Unlike somatic motoneurons generated earlier, a cluster of Olig2⁺ cells detected in the lateral MZ co-expresses Islet1 (arrowheads). (F,G) Expression of the oligodendrocyte marker O4 in Nkx2.2⁺ (A) and Olig2⁺ (B) cells detected in Pax6^−/− mice at E13.5. Nkx2.2⁺ and Olig2⁺ cells detected at ectopic dorsal positions in the mutant mouse are O4⁺ oligodendrocytes (arrowheads). Scale bars: 25 μm in A-C,F,G (shown in B, C, G); 50 μm in D, E (shown in E).

• Supplemental Figure 5 -

  Fig. S5. Expression patterns of inhibitory HLH factors early in development. Horizontal lines indicate the boundaries between the progenitor domains p2, pMN (p* at E14.5 and E16.5) and p3. The expression of Id1, Id2 and Hes1 began in the ventral spinal cord at around E10.5, closely following the expression of Pax6, Olig2 and Nkx2.2 (data not shown). At E11.5, Id1 expression mostly overlapped with that of Olig2 (Aa, Ac), whereas Id2 and Hes1 were expressed in more dorsal and ventral domains, respectively (Ab,Ad,Ae). Their expression domains gradually expanded from E12.5 onwards (Ba-B,Ca-Cg) and eventually overlapped with each other by E14.5 (Da-Dg). Their expression always overlapped with patterning factors until E14.5 (insets in Da-Dg). Within the same progenitor domains, however, cells expressing proneural and inhibitory HLH factors were mutually exclusive (Af-Ef,Ag-Eg). At E16.5, some Id1⁺ and Hes1⁺ cells from the VZ began to delaminate from the VZ, which were labeled with the astrocyte marker S100β (see Fig. 2). Insets in Ea and Ec are higher-magnification views of the areas indicated by thick arrows showing no overlap between Id1⁺ cells and Olig2⁺ and Nkx2.2⁺ OLPs in the MZ. Scale bars: 50 μm in Aa-Cg (shown in Cg); 100 μm in Da-Eg (shown in Eg).

• Supplemental Figure 6 -

  Fig. S6. Cell fate-specification activities of patterning and HLH factors in vitro. The population assay was extended to day 10 after plating, and the percentages of O4⁺ oligodendrocyte (A) and GFAP⁺ astrocyte (B) cells among total GFP⁺ cells were quantified (mean±s.d. from four independent culture experiments). Viruses used for infection are shown at the bottom. Dashed horizontal lines indicate the values in the control virus-infected culture. *, p<0.05 and **, p<0.01, compared to the control virus-infected culture; $, p<0.01 compared to cultures infected with Mash1 viruses.

• Supplemental Figure 7 -

  Fig. S7. Oligodendrocyte development in Mash1 and Ngn2 mutant mice at E14.5. (Aa-De) Distribution patterns of OLPs in the wild-type (Wt; Aa-Ae), Mash1^−/− (Ba-Be), Mash1^{KI Ngn2} (Ca-Ce) and Ngn2^{KI Mash1} (Da-De) mice at E14.5. Aa-Da and Ac-Dc show Olig2⁺ and Olig1⁺ cells (arrowheads), respectively, in the Olig2⁺ domain (p*) (indicated by arrows) and their migratory descendants around the VZ. Ab-Db and Ad-Dd are lower-magnification views of the spinal cord sections showing the distribution of Olig2⁺ and Olig1⁺ OLPs in the MZ. The dashed lines indicate the outline of the section and ventricular lumen in the midline. Ae-De show NG2⁺ OLPs (arrowheads) in the dorsal MZ. Intense tubular staining derives from blood vessels. Notice that, in Mash1^−/− and Mash1^{KI Ngn2} mice, more Olig2⁺ cells remained in the VZ, and, conversely, much less Olig1⁺, Olig2⁺, NG2⁺ OLPs were detected in the MZ compared with the wild type. (E-G) The numbers of Olig2⁺ (E), Olig1⁺ (F) and NG2⁺ (G) OLPs detected per section of the spinal cord at E14.5. Percentages shown in the graphs are values relative to wild type (indicated as 100%). Data are mean±s.d., obtained from the staining of 6-8 sections derived from 2-3 embryos for each genotype. *, p< 0.05 and **, p<0.01, compared to wild type; $, p<0.01 compared to Mash1^−/− mice. Notice that the numbers of Olig2⁺, Olig1⁺ and NG2⁺ OLPs are significantly smaller in Mash1^−/− mice than those in wild type at E14.5, and even larger reductions were observed in Mash1^{KI Ngn2} mice. In particular, few OLPs were found in the dorsal aspect of the spinal cord, as shown in the NG2 staining (Be, Ce). On the other hand, in Ngn2^{KI Mash1} mice, slightly larger numbers of Olig1⁺ and Olig2⁺ cells, and a comparable number of NG2⁺ cells, were detected compared with wild type despite a dramatic reduction of the overall size of the spinal cord, due to profound defects in neurogenesis (see Parras et al., 2002). Scale bars: 100 μm in Aa-Dd (shown in Dd); 50 μm in Ae-De (shown in De).