Fig 2: Single-Cell Proteomic Analysis of Cellular Subpopulations in EME6/7t EECs and Primary Tumors(A and B) tSNE plots of mass cytometry (CyTOF) analysis visualizing 74,048 epithelial cells, which were divided into clusters (n = 27) based on their proteomic profiles (A). These cells included immortalized EEC line EME6/7t (exposed to ASC-CM ± Tip) and primary endometrial tumors of patients with known BMIs (B).(C) tSNE-based phenographs showing abundance and cellular distributions of JAC proteins.(D) tSNE-based phenographs showing EpCAM and CK8/18 as epithelial cell markers and TGFR-2, SMAD2, and SMAD4 as TGF-ß factors.See also Figure S6 and Tables S2 and S3.

Fig 3: Pontin interacts with LEF1 to activate TGFBR2 gene transcription in GBM cells.a Venn plot showed the overlapping of Pontin interacting proteins with possible transcription factors regulating TGFBR2 gene expression. SF splicing factors, TF transcription factors. b Representative images showing that Pontin (red) colocalized with LEF1 (green) in U87MG and U251 cell nuclei. Scale bar, 10 μm. c Co-IP confirmation of the interaction between Pontin (flag-tagged) and endogenous LEF1 (upper) or HA-tagged LEF1 (bottom) in U87MG cells. d Effect of Pontin or LEF1 on the luciferase transcriptional response of the reporter driven by TGFBR2 promoter (−2000 to 0). e LEF1-binding sites predicted in the TGFBR2 promoter (left) and ChIP-qPCR analyses of the capacities for them to bind with LEF1 (right). Data in d, e are expressed as means ± SDs, n = 3.

Fig 4: TGFßR2 restoration recovers xenograft growth and its higher levels indicate the worse outcomes in glioma patients.a Representative bioluminescent images reveal the xenografts in mice brain formed by indicated U87MG cells. b Representative H&E staining images of the indicated xenografts. Red dotted lines outline the junctions between xenografts and surrounding non-tumoral brain tissues. Invasion distances was indicated by red double-sided arrows. Scale bar, 50 µm (×100). c Survival curve of the glioma-bearing nude mice. d Box plots comparing TGFBR2 mRNA expression between normal brains (N) and tumours (T) in GBM and LGG (lower-grade glioma) using the data from GEPIA dataset. e Kaplan–Meier analyses the OS of all the glioma patients (GBM and LGG) from the GEPIA dataset. Patients were divided into high and low expression groups according to TGFBR2 expression medians of the corresponding cohorts. P-values calculated by the log-rank (Mantel–Cox) test are shown.