Fig 2: Defective autophagy and MHC II processing in lamp2 KO cTECs. (A-B) Autophagic flux in cTECs from 12 days old WT and lamp2 KO mice was analyzed by flow cytometry. (A) Histograms show representative Cyto-ID analysis in cTEC and mTEC from WT and lamp2 KO. Graphs represent the MFI of Cyto-ID in the indicated subsets. (B) Scheme (top left) represents autophagic flux using RFP-GFP-LC3 mice. Differential detection of GFP and RFP allows the distinction between AP (GFP+ RFP+) and AL (RFP+). Dot plot (bottom left) represents GFP and RFP expression in cTECs of WT and lamp2 KO RFP-GFP-LC3 mice. Graphs (top right) represent the average frequency of RFP+ GFPhi (yellow) and RFP+ GFPlow (red) cells in WT and lamp2 KO RFP-GFP-LC3 cTECs. Graphs (bottom right) represent the MFI of GFP and RFP expression in WT (gray) and lamp2 KO (blue) RFP-GFP-LC3. (C) WT and lamp2 KO RFP-GFP-LC3 cTECs (EPCAM+ ENPEP/LY51+) were analyzed by imaging flow cytometry. Representative images are shown. Graph show the distribution of the number of RFP puncta in WT (n = 1345 cells) and lamp2 KO (n = 1422 cells) RFP-GFP-LC3 cTECs. Data in A-C include an average of 2–3 independent experiments (n = 4–6 WT and n = 4–6 lamp2 KO. (D) Flow cytometry analysis of 15G4 staining on cell surface of cTECs (ENPEP/LY51+ UEA-1-) and mTECs (ENPEP/LY51-UEA-1+) from 12 days old WT and lamp2 KO mice. Histograms show representative staining and graphs represent the MFI of 15G4 staining in the indicated subsets. Data include an average of 4 independent experiments (n = 6 WT and 7 lamp2 KO). (E) Histograms show representative 15G4 staining on the cell surface of WT and lamp2 KO c/mTECs from lamp2 WT/KO heterozygous mice. Graphs represent the MFI found in indicated subsets in control (gray) and mutant (blue) cells (n = 8 from 4 experiments). (F) Postnatal day 3–5 WT thymus were treated overnight with chloroquine (50 µM) or with Baf-A1 (0.5 µM) for 5 h. Histograms show the representative staining with 15G4 staining in control (light gray) and chloroquine-treated (black) or Baf-A1-treated (black) in WT cTECs. Graphs represent the MFI in control (gray) and treated (dark gray/ black) cTEC (n = 4–6 animals from 3–4 independent experiments). Results in A-F are shown as mean ± SEM. * p < 0.05; ** p < 0.01; *** p < 0.001.

Fig 3: Genome-wide transcriptomic analysis of TF subsets identifies stages with distinctive gene expression profiles. (A) Principal component analysis plot and dendrogram, detailing the hierarchical clustering between the biological samples, performed with data obtained from total RNA-sequencing analysis of sorted TFA (CD45-EpCAM-GP38+SCA-1-) (n=3), TFB (CD45-EpCAM-GP38+SCA-1+) (n=3) and MC (CD45-EpCAM-GP38-SCA-1-Ly51+) (n=3) populations. (B) Heat maps representing the deviation from average expression of the phenotypic markers used to identify TMC populations, of genes previously associated with pericytes and of genes previously associated with thymic fibroblasts. (C) Heat maps representing the deviation from average expression of the uniquely upregulated genes identified for populations TFA and TFB and the associated molecular functions identified by GO analysis. Genes with FDR<10% were considered as differentially expressed. Enriched GO terms (molecular functions) were identified using MGSA. Represented categories had a marginal posterior probability estimate higher than 0.65.

Fig 4: GP38 and SCA1 expression on TMC subsets. (A) Total thymi cells from 1-week-old mice were isolated, and total TMCs (CD45-EpCAM-) were analysed by flow cytometry. t-SNE representation of the expression of CD31, CD140a, CD140ß, GP38, SCA-1, Ly51 and aSMA. (B) Three main clusters were identified: cluster 1 (CD31+), cluster 2 (CD140a+ß+) and cluster 3 (CD140a-ß+). Clusters 2 and 3 were respectively subdivided into cluster 2.1 (CD140a+ß+GP38+SCA1-) and 2.2 (CD140a+ß+GP38+SCA1+); and 3.1 (CD140a-ß+Ly51+aSMA-) and 3.2 (CD140a-ß+Ly51+aSMA+). (C) TMCs (CD45-EpCAM-CD31-) were analysed for the indicated markers, and sub-cluster 2.1 (red gate), 2.2 (green gate), 3.1 and 3.2 (light- and dark-blue gates) were identified. (D) Analysis of GP38, SCA-1 and DPP4 expression in TFA (red gate) and TFB (green gate) populations at E14, E17, 1 week old (W) and 4W. Numbers in plots indicate the frequency of cells found within each gate. Plots are of a representative analysis per time point. (E) Bar graphs showing mean+s.d. of the frequency and cellularity of TFA and TFB subsets, of three independent analyses per time point. Differences in TF subsets, CD140a+ß+GP38+SCA1- (red) and CD140a+ß+GP38+SCA1+ (green), were statistically analysed at different ages: ***P<0.001, **P<0.01, *P<0.05.