Fig 2: Gpr125+ cells have a MaSC/MRU profile and encompass distinct progenitor populations.a Representative FACS dot plots of total MECs (left) isolated from 6-week Adgra3lz/+ pubertal mice stained with antibodies against CD24 and CD49f. n = 4 mice/ experiment and repeated five times. The basal population is depleted in FDG? populations (center). Gpr125+/FDG+ cells gated within the CD24med/low/CD49f+/hi basal population (right). b Histograms showing expression of CD49f, CD29, Sca1 and CD61 in Gpr125+/FDG+ (red line), Gpr125?/FDG?(black line) and total MECs (gray line) n = 4 mice/experiment repeated three times. c Detection of reporters in the cap cell layer of TEBs: left panel: co-expression of s-SHIP-EGFP detected by immunohistochemistry (brown), on top of Gpr125-ß-gal detected by X-gal (dark blue) counterstained with hematoxylin (purple); and individual expression in center panel: Gpr125-ß-gal detected by X-gal (blue); right panel: s-SHIP-EGFP fluorescence (green). Scale bar = 50 µm (n = 2 mice/panel). d Adgra3 and Lgr5 mRNA expression in the basal cell cluster, visualized by t-SNE plots extracted from the Tabula Muris dataset42. e–g RNA-seq based transcriptomic comparison of FACS-sorted Gpr125+ with Gpr125? cells isolated from TEB-distal (purple) and nipple-proximal regions (yellow) of pubertal Adgra3lz/+ mammary glands. e Venn diagram shows 140 differentially expressed genes (gray) are shared by TEB-distal and nipple-proximal Gpr125+ cells (twofold change, FDR = < 0.1). Biological significance was evaluated by (f) gene ontology (GO) and (g) gene set enrichment analysis (GSEA) analysis of TEB-distal and nipple-proximal Gpr125+ cells compared to their Gpr-negative basal cell counterparts. GSEA shows a correlation between genes highly expressed in TEB-distal (blue) or nipple-proximal (yellow) Gpr125+ cells compared to the ‘Lim_mammary_stem_cell_UP’ gene signature. h Volcano plot comparing gene expression of nipple-proximal Gpr125+ cells and TEB-distal Gpr125+ including highlighted candidates in nipple Gpr125+(blue dots) and TEB cells (red dots) (FDR = 0.1). i GO analysis yielded gene enrichment in biological processes (red bars) and molecular function (green bars) that characterize Gpr125+ cells (FDR B&H). j Differently expressed genes between TEB-distal Gpr125+ and nipple-proximal Gpr125+ cells were confirmed and mapped to distinct basal cell clusters by scRNAseq analysis of the pubertal dataset of Pal et al.28. KNetL plots are shown for cell populations expressing Adgra3, Krt14, Tcf7, Sox11, Fbln1, Lrfn5, Lgr5, Slpi Wnt10a, Wnt5a, Wnt6. k Distal Sox11 expression in TEBs is detected in Tcf1+ cap and Tcf1? body cells and some adjacent mesenchymal cells. (n = 4 mice). Scale bar = 50 µm. Source data are provided as a Source Data file.

Fig 3: High Gpr125 expression is predictive of poor outcome.a X-gal/NFR stained sections showing higher Gpr125-ß-gal expression in a short (7w) latency MMTV-Wnt1;Adgra3lz/+ tumor (Wnt1-early) versus a long (36w) latency (Wnt1-late) tumor. Scale bar 100 µm and 50 µm. (n = 5 tumors/subtype). b qRT-PCR analysis of relative Adgra3 mRNA expression in early and long latency tumors MMTV-Wnt1;Adgra3lz/+ tumors (n = 3 tumors/subtype, **p = 0.0052). c Representative FACS dot plots of total MECs isolated from MMTV-Wnt1;Adgra3lz/+ hyperplastic uninvolved glands and associated short and long latency tumors stained with CD24 and CD49f. FDG+ cells in: 36w tumor = 13331; 36w hyperplasia = 4048; 7w tumor = 64889; 7w hyperplasia = 13156. Plots are representative of three independent experiments. d ADGRA3 expression in (left) ER- (n = 445) and ER+ (n = 1459) and in (right) PR- (n-895) and PR+ (n = 1009) BC patient samples from the METABRIC dataset **p = 0.0042, ****p < 0.0001. e, f ADGRA3 expression in BC patient samples from TGCA PanCancer Atlas (n = 1084) (e) and METABRIC (n = 2509) (f) datasets classified using the PAM50 geneset; ****p < 0.0001. g Integrative clusters within the METABRIC dataset showing high ADGRA3 expression levels in cluster 10; ****p < 0.0001, Mann–Whitney test two tailed. h ADGRA3 expression in human breast cancer cell line from Depmap database (https://depmap.org/portal/) **p = 0.0068, ***p = 0.0002, Unpaired t test two-tailed. i–k Kaplan–Meyer curves depicting ‘Relapse-free survival’ and ‘Distant Metastasis-Free Survival’ in months for high (red lines) and low (black lines) ADGRA3 mRNA expressing human basal-type (h, i) and basal-like-1 TNBC (j) breast cancer subtypes. Auto best fit cutoff was used to divide patients into high and low expression (indicated in beehive plots) sourced from (https://kmplot.com/analysis/index.php?p=service&cancer=breast) KMplotter75. logrank p-value and Hazard Ratio (HR) indicate a significant association between high expression and poor prognosis. Each scatter dot plot represents the mean ± SEM. Source data for this figure are provided as a Source Data file.

Fig 4: Obesity enhances luminal and decreases basal/myoepithelial cell populations. Mammary epithelial cells isolated from control (Con) and obese (high-fat diet (HFD)) C57Bl/6 mice were stained with EpCAM and CD49f and quantified using flow cytometry (a, b). a Representative contour plots depicting gates for luminal cells (EpCAMhiCD49flo) and basal cells (EpCAMloCD49fhi). The percentage of luminal and basal cells were quantified from total mammary-lineage-positive cells in three experiments (n = 9 mice/group). b Results were determined as a ratio of luminal to basal cells from obese mice compared with control mice and represented as mean ± s.d. Human breast epithelial cells were isolated from reduction mammoplasty samples, stained with EpCAM and CD49f, and quantified using flow cytometry (n = 11 patient samples). c Representative contour plot depicting gates for mature luminal (ML, EpCAMhiCD49f-), luminal progenitor (LP, EpCAMhiCD49f+), mature basal (MB, EpCAMloCD49f+), and basal progenitor cells (BP, EpCAM-CD49f+). d Correlation scatter plot with best fit line representing the ratio of the percentage of luminal to basal/myoepithelial cells. The percentage of positive cells was the number of positive cells in the gate compared to the total number of mammary-lineage-positive cells. Correlation scatter plot with best fit line representing the percentage of MB (e), BP (f), ML (g), and LP (h) against body mass index (BMI)