Fig 2: Analysis schematicRadical cystoprostatectomy and extended pelvic lymph node dissection were performed, and fresh tissue from four regions of the bladder tumor and representative left and right lymph nodes were dissociated into single cell suspensions and fluorescence activated cell sorting isolated the EPCAM+CD44+CD49f+ cancer cell subpopulation. DNA and RNA were isolated from the bulk tumor regions (samples A-D) and lymph nodes (samples L and R) and the corresponding cancer cell subpopulations (denoted with') and utilized for whole-exome sequencing and RNA-sequencing (A). Representative histology of primary tumor, left and right sampled lymph nodes (B).

Fig 3: Characterization of Troy-expressing cells using organoid technology(A) Experimental setup.(B) Representative flow cytometry scatterplots of viable cells isolated from Troy-EGFP knockin mice stained for ITGA6 and SCA1 (n = 4).(C) Representative bright-field images of sorted cell populations grown as organoids for 7 days.(D–F) Quantification of the cell viability per well (D), number of organoids formed per well (E), and size of organoids formed (F) after 7 days of culture. The data are presented as mean ± SEM (two to four replicates/wells per mouse, n = 4 mice). Student's t test, **p < 0.01, ***p < 0.001, ****p < 0.0001. A.U., arbitrary units; L.U., light units.

Fig 4: Characterization of progeny of Troy-expressing cells(A) Schematic overview of experimental outline.(B) Representative flow cytometry scatterplot of viable cells isolated from TroyEGFP-IRES-CreERT2 × LSL-tdTomato mice 7 days post induction of lineage tracing stained against ITGA6. Sorting gates are indicated in colors (for color coding, see (G)).(C–E) t-SNE map of single-cell sequencing data generated from 7 day induced mice (n = 2 mice) projected onto the dataset generated for Figure 4 C, with assigned clusters (D) and all sorting gates indicated (E).(F) Stacked column chart indicating cluster breakdown by gate.(G) Tail epidermal whole mount of a TroyEGFP-IRES-CreERT2 × LSL-tdTomato mouse 7 days post tamoxifen injection stained for tdTomato and ITGA6. Zoom-ins and arrows indicate suprabasal tracing (yellow arrows) or basal tracing (white arrows).See also Figure S6.

Fig 5: Characterization of Troy-expressing cells in adult murine telogen skin(A) Experimental setup.(B and C) Representative flow cytometry scatterplots of viable cells isolated from Troy-EGFP knockin mice stained for ITGA6 (CD49f) (B) and SCA1 (C). Histogram of Troy-EGFP expression normalized to mode (B). Dark green, ITGA6bright cells; green, ITGA6mid cells; gray, ITGA6dim cells (B). Column chart indicating the percentage of SCA1– and SCA1+ cells within the Troy-EGFP+ population (C). The data are presented as mean ± SEM (n = 3 mice). Student's t test, ****p < 0.0001.(D) Heatmap of the bulk mRNA sequencing showing the differentially expressed genes comparing Troy-EGFP+ and Troy-EGFP– ITGA6bright cells.(E–H) Representative image of paraffin sections of telogen back skin stained RNAscope probes against Troy/Tnfrsf19 and Krt10. Yellow arrows in (G) and (H) indicate basal layer cells positive for Krt10 mRNA and Troy mRNA.(I) Tail epidermal whole mount of Troy-EGFP mice stained for EGFP and KI67.(J) Schematic representation of the genetic constructs.(K) Representative flow cytometry scatterplot of viable SCA1+ ITGA6+ cells assessed for expression of Troy-EGFP and KI67-tagRFP. Column chart indicating the percentage of KI67-tagRFP– and KI67-tagRFP+ cells within the Troy-EGFP+ population. The data are presented as mean ± SEM (n = 3 mice). Student's t test, ***p < 0.001.See also Figure S3 and Table S1.