Fig 2: Adipose eosinophil and macrophage populations are influenced by sex, diet, and RELMα.(A) t-Stochastic neighbor embedding (tSNE) analysis to identify stromal vascular fraction (SVF) leukocyte populations. (B) tSNE analyses of SVF from the different groups (male ♂, female ♀, wild-type [WT], or RELMα knockout [KO]) after 12 weeks of diet exposure (Ctr or high-fat diet [HFD]) revealed changes in eosinophil RELMα-dependent and diet-induced changes in eosinophil (red outline) and macrophage (green outline) subsets. (C, D) CD11c surface expression in CD45+MerTK+CD64+ macrophages was analyzed by tSNE, where dashed red outline shows CD11cHi cells, and quantified. (E) CD301b surface expression on SVF macrophages was examined by tSNE and quantified by mean fluorescent intensity (MFI). (F) Siglec-F surface expression on CD45+SiglecF+CD11b+ SVF eosinophils was examined by tSNE and quantified by mean fluorescent intensity (MFI). tSNE data are one representative animal per group. All other data are presented as individual points for each animal, where lines represent group means ± standard error of the mean (SEM). Statistical significance was determined by three-way analysis of variance (ANOVA) Sidak’s with multiple comparisons test (ns, no significant; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001). Data are representative of 3 experiments with 4–6 mice per group.

Fig 3: COS inhibited the DDC diet feeding-induced cholestatic liver disease and liver injury. Cholestatic liver disease was induced by DDC diet feeding for 28 days as described in the Materials and methods section. Two dosages of COS (10 mg/kg, and 30 mg/kg) were used to treat cholestatic liver disease for 14 days. (A) The histogram showed liver weight and liver weight to body weight ratio on day 29. Statistical analysis was done with one-way ANOVA and Tukey as an additional multiple comparison test. The values of ∗P < 0.032, ∗∗P < 0.0021, ∗∗∗P < 0 0.0002, and ∗∗∗P < 0 0.0001 were marked in the results. Histogram showed serum AST, and ALT levels in DDC diet feeding mice and COS treated mice. (B) Histogram showed serum AST, and ALT levels in DDC diet feeding mice and COS treated mice. Statistical method as above. (C) Histogram showed serum ALP level in DDC diet feeding mice and COS treated mice. Statistical method as above. (D) Histogram showed serum total and direct bilirubin levels in DDC diet feeding mice and COS treated mice. Statistical method as above. (E) The hepatic histopathological results, including HE, MASSON, and immunohistochemistry of CK19 and CD45, showed a typical ductular reaction, fibrosis, and inflammation infiltration features of DDC diet feeding induce cholestatic liver disease model. Two dosages of COS treatment resulted in improvement of histopathological features.

Fig 4: TNFR2 blockade and PTX combination therapy. A Schematic representation of 4T1-transplanted murine model and therapeutic procedure. B Tumor growth curves. C Primary tumors harvested from mice. D Representative histogram (left) and quantification (right) of PD-L1 expression on tumor ECs (CD45−CD31+PD-L1+) isolated from mice. E, F Gating scheme for (E) CD45+CD3+CD8+ T cells and (F) PD-1+ T cells (CD45+CD3+CD8+PD-1+) from tumor-bearing mice. G Quantification of the proportion of PD-1+CD8+ T cells. H Schematic illustration showing immune features of CD8+ T cells and ECs from responders and non-responders following PTX therapy. The expression of the ligands PD-L1, PVR, and MHC class II molecules on ECs and the corresponding receptors PD-1 and LAG3 on CD8+ T cells increase in non-responders but decrease in responders. Endothelial TNF–TNFR2 signaling enables inhibition of CD8+ T cell through PD-L1/PD-1 axis via regulating cellular NF-κB–glycolysis pathway on the basis of IFNγ-mediated immune feedback. Schematic illustration was created on the website of “Home for Researchers”. *P < 0.05, ***P < 0.001, ****P < 0.0001

Fig 5: Validation of Y chromosome loss in MB49 tumor models and immune cells.a-b. Genomic DNA (10 ng per reaction) was isolated from sorted epithelial cells, non–T cells, and T cells derived from CRISPR‐engineered (Y-Scr or Y-KO) MB49 tumors. Wild‐type (WT) C57Bl/6 N genomic DNA (lacking Cas9) served as the reference in the ΔΔCt analysis. (a) Ct values were normalized to Gapdh and plotted as fold change relative to WT. (b) Ct values were normalized to Actb and similarly plotted as fold change over WT. Error bars represent mean ± s.e.m. from 4 biological replicates in each group. c. DNA qPCR analysis of Y chromosome genes in TdTomato+ CD45- epithelial cells (left) and CD3+ T cells (right) from CRISPR Y-Scr and CRISPR Y-KO groups. Fold changes normalized to Gapdh (purple dots) and Actb (yellow dots). P-values from two-way ANOVA with Fisher’s LSD test and Geisser-Greenhouse correction. Error bars represent mean ± s.e.m.; Each dot denotes pooled samples from six independent tumors. Six independent tumors were mashed and pooled together for substantial number of T cells isolation and then randomly aliquoted to 4 tubes to increase the efficacy of the result and reduce human error. Each dot of one color represent data from one such tube. d. DNA qPCR analysis of Y chromosome genes (Kdm5d, Uty, Eif2s3y, and Ddx3y) in PBMCs extracted from BBN-treated mice. Each dot represents the pooled DNA sample from 3-4 mice. On X axis: the time points of samples collection, along with a timeline of BBN administration and bladder tumor onset; on the Y axis the fold change expressed as relative gene expression level compared to untreated ctrl at initiation of BBN. e. Changes in DNA levels of Y chromosome genes (Ddx3y, Kdm5d, Zfy1/2, Ssty1, Ssty2) upon long term activation of primary mouse T cells in vitro. qPCR data normalized to housekeeping genes Actin (left) or B2m (right) relative to day 1 after mouse T cell isolation. Mean ± SD, n = 4. SD, Standard Deviation. f-g. Expression changes of T cell exhaustion markers (Ctla4, Pdcd1, Tcf7, Tim3) during long term activation of primary mouse T cells in vitro. Expression normalized to housekeeping genes Actin (f) or Gapdh (g) are shown relative to day 1 post T cell isolation. Mean ± SD, n = 4. h-i. Mean expression of exhaustion (h) and Y chromosome genes (i) in Naïve P14 CD8+ T cells at baseline (Day 0) and compared to at Days 8, 15, and 30 post-infection under acute (LCMV Armstrong; solid lines) or chronic (LCMV Clone 13; dashed lines) conditions. Data were extracted from publicly available transcriptomic datasets39. Source Data