CG individually encase clonally related cells regardless of their identity.

(A, B) Adaptation of the CG network to NSC tumours. Type I pros (Nrv2::GFP, wor-GAL4; tub-GAL80ts > pros RNAi) and Type II brat (Nrv2::GFP, wor-GAL4; tub-GAL80ts > brat RNAi) tumours were induced, and the organisation of CG membrane was monitored by Nrv2::GFP (green). Control line for both (Nrv2::GFP, wor-GAL4; tub-GAL80ts > w1118). For pros, larvae were dissected after 48 h at 18°C, followed by 2 h heatshock at 37°C and 48–52 h at 29°C. For brat, larvae were dissected after 72 h at 29°C. See S1 Table for detailed genetics, timing, and conditions of larval rearing. (A) Left panel shows a representative whole thoracic VNC for control (n = 10 VNCs) and pros tumour (n = 10 VNCs), and (A’) middle and right panels show a close-up of the yellow boxes in (A). NSCs/tumour NSCs are labelled with anti-Dpn (grey), and neuron with anti-ElaV (magenta). (B) Left panel shows a whole representative CB for control (n = 10 CBs) and brat tumour (n = 10 CBs), and (B’) middle and right panels show a close-up of the yellow boxes in (B). NSCs/tumour NSCs are labelled with anti-Dpn (grey), Type I NSC and mINPs with anti-Asense (Ase, magenta), and GMCs plus newborn neurons with anti-Prospero (Pros, blue). For both, dashed white lines delineate one CG chamber. (C, D) Relationship between individual encasing and cell identity. Type I pros (wor > raeppli-NLS, pros RNAi) and Type II brat (wor > raeppli-NLS, brat RNAi) tumours were induced together with the multicolour lineage tracing Raeppli-NLS. One out of 4 colours (blue, white, orange, and red) is stochastically induced in the transformed NSCs using hs-Flp. For pros, larvae were dissected after 24 h at 18°C, followed by 2 h heatshock at 37°C and 48–52 h at 29°C. For brat, larvae were dissected after 48 h at 18°C, followed by 2 h heatshock at 37°C and 48–52 h at 29°C. See S1 Table for detailed genetics, timing, and conditions of larval rearing. (C) Top and orthogonal close-up views of representative control (n = 10 VNCs) and pros (n = 9 VNCs) Type I lineages marked with Raeppli-NLS. (D) Top and orthogonal close-up views of representative control (n = 8 CBs) and brat (n = 7 CBs) Type II lineages marked with Raeppli-NLS. For both, CG membrane is visualised with Nrv2::GFP (green). Dashed white lines delineate one CG chamber. White stars indicate Type II lineages. (E) brat RNAi tumours are individually encased by the CG. Representative confocal pictures of the CG network in control Type II NSC clone and brat RNAi Type II NSC clone. Top view and orthogonal views are shown. The CoinFLP system was used to generate wild-type and brat RNAi clones in the Type II NSCs, using grhD4-FLP for induction. The grhD4 enhancer drives in most NSCs from late embryonic stage. As such, clones in Type I NSCs can also be induced (marked by a cyan star) yet do not lead to any tumour. Larvae were dissected after 72–96 h at 25°C. See S1 Table for detailed genetics, timing, and conditions of larval rearing. The membrane of the clone is marked by mCD8::RFP (magenta), and CG membrane is visualised with Nrv2::GFP (green). Dashed white lines delineate one Type II NSC chamber. n(brat RNAi clones) = 20 clones in 15 CBs and n(control clones) = 15 clones in 11 CBs. All (100%) of the Type II NSC clones in both control and brat RNAi conditions were found individually encased by the CG. (F) Schematic of the different hypotheses explaining individual encasing of NSC lineages by CG. Panel I depicts the NSC-driven timing of NSC encapsulation, prior to lineage generation, as the instructive cue. Panel II describes the use of adhesion mechanisms: (1) specific CG to NSC lineage and (2) generic, based on difference in strength between intralineage adhesion (A(LNSC)) and adhesion linking NSC lineages and CG (A(LNSC-CG)); in this case, the prediction is that A(LNSC-CG) NSC). CB, central brain; CG, cortex glia; GMC, ganglion mother cell; NSC, neural stem cell; VNC, ventral nerve cord.