Fig 2: Smad7 Limits PD1 on CD4+ T Cells and PD1-Mediated Treg Differentiation and Promotes T Cell-Mediated Colitis(A) Representative FACS histograms (left) and MFIs (right) of phospho-Smad2/3 in naive CD4+ T cells from Smad7fl/fl and T-Smad7−/− mice stimulated with or without TGF-β (2.5 ng/mL), for 1 h (n = 4).(B) Representative FACS histograms (left) and MFIs (right) of PD1 in naive CD4+ T cells from these mice stimulated with low-dose plate-bound anti-CD3/CD28 (1 μg/mL) with or without TGF-β (2.5 ng/mL), for 18 h (n = 8).(C) Representative FACS plots (top), frequencies (bottom left), and cell counts per well (bottom right) of Foxp3+ populations in naive CD4+ T cells from these mice stimulated with or without TGF-β (0.5 ng/mL) and recombinant plate-bound PDL1-Fc or PDL2-Fc (30 μg/mL), for 4 days (n = 5).(D–H) Colitis was induced by adoptive transfer of 6 × 105 CD4+CD45RBhi T cells i.p. into Rag-1 mice, followed by monitoring for 7 weeks. (D) Percentage body weight changes (left) and linear regression analysis (right) of Rag-1−/− mice injected with CD4+CD45RBhi T cells from Smad7fl/fl or T-Smad7−/− mice (n = 8). (E) Representative histological sections stained with H&E of distal colons (left) and histological scores (n = 8) (right) from recipient Rag-1−/− mice, based on the degree of epithelial damage, as scored blinded by a pathologist at HRHCF. Scale bars represent ~1 mm (top) and ~100 μm (bottom). (F) qRT-PCR of IL-12, IL-1β, IL-6, IL-23, and TNF-α in colitic tissue from recipient mice (n = 8). (G) Representative FACS plots (left) and frequencies (right) of Foxp3+ populations in MLN (top) and splenic (bottom) CD4+ T cells from recipient mice at 7 weeks (n = 5–8). (H) Representative FACS histograms (left) and frequencies (right) of PD1 in MLN (top) and splenic (bottom) CD4+ T cells from recipient mice at 1 week (n = 6).Data representative of ≥2 independent experiments. qRT-PCR and MFI (reflective of CD4+ population) data expressed as FC from Smad7fl/fl media condition. Statistical comparisons to Smad7fl/fl condition. Means ± SEMs. *p < 0.05, **p < 0.01, and ***p < 0.001 by unpaired t test. See also Figures S4 and S5.

Fig 3: PD-L1 and PD-L2 Expression Levels and Clinical Outcomes in Glioma. (A) Kaplan-Meier disease free survival estimates for GBM patients with high expression of PD-L2 (left) and either high expression of both PD-L1/L2, high expression of either PD-L1/L2, or low expression of both (right)(*p = 0.0072; #, ns = not significant). (B) Kaplan-Meier estimates of OS (left panels) and DFS (right panels) based on high (dark) or low (grey) PD-L2 expression. (C) Kaplan-Meier estimates of OS (left) and DFS (right) in LGG patients with either high expression of both PD-L1/L2, high expression of either PD-L1/L2, or low expression of both (** =p < 0.0001, * = p = 0.009).(D) Multivariate analysis for DFS shows PD-L2 is an independent factor in LGG patients after adjusting by covariates of age, gender and IDH1 mutation (p = 0.0062) (left); Multivariate analysis shows PD-L1/2 both high is independently associated with shorter DFS in LGG patients compared to “both low” (p = 0.0055) or “one high” (p = 0.0452) after adjusting by covariates age, gender and IDH1 mutation (right).

Fig 4: NextA improves the anti-tumor activity of anti-PD-1 immune checkpoint blockade. (A) C57BL/6 mice were subcutaneously injected with 1 × 106 SM1 murine melanoma tumor cells. Mice were treated with a vehicle control, 3 mg/kg anti-PD-1, 15 mg/kg NextA, or a combination of both agents for 25 days. (B) Kaplan-Meier survival plot of the previous study. (C) Individual group plots representation for the previous study. (D–H) Tumors were collected at the end point from mice treated with a vehicle control, anti-PD-1, NextA, or combination of both agents and the presence of the costimulatory markers PD-L1, PD-L2, OX40L, MHC class I, and MHC class II was evaluated by flow cytometry. (I) The expression of PD-L1, PD-L2 and GAPDH was evaluated by immunoblot. (J) Tumors were collected at the end point from mice treated with vehicle control, anti-PD-1, NextA, or a combination of both agents and the presence of M1 and M2 surface markers were evaluated by flow cytometry. M1/M2 ratios were calculated from three independent populations from M1 and M2. (K–O) Total RNA was isolated from the same aforementioned sample conditions and qRT-PCR was done to evaluate the expression of IFN?, IL-2, IL-12, IL-10 and TGFß.

Fig 5: Identification of Novel CD274 and PDCD1LG2 cis-Regulatory Elements. (A) Cell surface expression of PD-L2 in brain tumor cells (grey line: isotype control; black line: PD-L2; dotted line: PD-L2 with IFN-γ stimulation). (B) CD274 promoter (Pr1) and enhancer (Pr2 and Pr3) regions cloned and assessed. (C) Normalized luciferase activity in cells with high and low endogenous PD-L1 expression using PD-L1.Pr1,.Pr2 and.Pr3 elements. (D) Schematic of PD-L1 enhancer region (Pr3) and truncated minimal region (Pr3M). (E) Luciferase assay of PD-L1.Pr3 (black) and.Pr3M (white) in PD-L1 high- and low-expressing cell lines. (F) PDCD1LG2 element (.Pr1 + 2) and truncated.Pr1 and.Pr2. (G) Luciferase assay of PD-L2.Pr1 + 2 (left) and PD-L2.Pr1 and.Pr2 (right) in PD-L2 high- and low-expressing cells. (H) PD-L2.Pr1 mutants lacking predicted STAT1 (ΔSTAT1), GATA2/3 (ΔGATA2/3), or both binding sites (ΔSTAT1 & ΔGATA2/3). (I) Luciferase assay of PD-L2.Pr1 truncations lacking STAT1 (ΔSTAT1) or GATA2/3 (ΔGATA2/3)(top) or also both (ΔSTAT1 & ΔGATA2/3)(bottom) in high and low expressing cell lines.