Fig 2: Evidence for Class I PI3K-Driven Accumulation of PI(3,4)P2 in EGF-Stimulated Mcf10a Cells(A) Measurements by HPLC-MS of PI(3,4,5)P3 (left panel) or PI(3,4)P2 (right panel) in INPP4BA/B-KD,PTEN-KO Mcf10a cells that were starved; pretreated for 20 min with the indicated dilutions of a mixture of PIK75 and TGX115 (1× represents 0.0058 μM PIK75 and 0.13 μM TGX115); and then stimulated with EGF (10 ng/mL) for 0, 1, or 5 min. Data are means ± SD of 3 technical replicates. The table shows IC50s of PIK75 and TGX115 against relevant PI3K activities assayed in vitro.(B) Time courses of PI(3,4,5)P3 (top) and PI(3,4)P2 (bottom) accumulation during EGF stimulation of WT, PTEN-KO, SHIP2-KD, INPP4A/B-KD, SHIP2-KD,PTEN-KO, and INPP4A/B-KD,PTEN-KO cells. Cells were starved and stimulated with EGF (10 ng/mL) at time 0. At 60 s post-EGF stimulation, samples were treated with either vehicle (DMSO; solid circles) or PI-103 (1 μM; crosses), and incubations continued for the times indicated. Data are represented as means ± SD of 3 biological replicates. Lines represent simulations of our mathematical model (see below) in the absence (continuous) or presence (dashed lines) of PI-103. Only INPP4A/B-KD,PTEN-KO cells exhibit a large increase in PI(3,4)P2.(C) List of biochemical reactions represented in the mathematical model, represented using Systems Biology Graphical Notation (SBGN) process descriptions (Le Novère et al., 2009).(D) Simulated maximal fluxes of dephosphorylation of PI(3,4,5)P3 into PI(4,5)P2 by PTEN and into PI(3,4)P2 by SHIP2 and X (the combination of all other relevant 5-phosphatases). In PTEN-KO cells, all PI(3,4,5)P3 is converted into PI(3,4)P2, SHIP2 and X compensating for the absence of PTEN. In SHIP2-KD cells, while X compensates partially for the lack of SHIP2, a larger share of the fluxes is re-routed toward PI(4,5)P2. When both PTEN and SHIP2 are absent, the flux through X cannot fully compensate for the lack of both enzymes.(E) Simulated maximal fluxes of dephosphorylation of PI(3,4)P2 by INPP4B, PTEN, and an unknown phosphatase Y in WT, INPP4A/B-KD, and PTEN-KO cells. In the WT, the fluxes through INPP4B and PTEN are balanced. The fluxes are re-routed in the mutants. While an unknown phosphatase is required to accurately reproduce the experimental results, its activity remains very limited.Supporting information is presented in Figures S4 and S5, the STAR Methods, and Data S1 and S2.

Fig 3: TGFβ−1 Inhibits B-cell proliferation and maintains B-cell checkpoint molecule expression.(A) Histogram and bar plots showing cellular expansion of activated B-cells cultured with exogenous TGFb-1or mock as determined by efluor-450 proliferation dye. B-cells were activated with a B-cell expansion cocktail and TGFb-1 was added every day starting from day 0. Addition of TGFb-1inhibited B-cell proliferation. (B) B-cell proliferation was also quantified based on cell counts. The same number of cells were plated in each group, and the ratio of mock B-cells to TGFb-1 B-cells was determined at days 4 and 7. (C) Bar plot of B-cell expansion showing TGFb-1 induced loss of proliferation could be prevented with a TGFβ-R inhibitor. B-cells were activated with an expansion cocktail and cultured with exogenous TGFβ−1. SB431542 small molecule TGFβ-R inhibitor was also added every day to the experimental group 30 minutes before addition of TGFβ−1. While TGFβ−1 inhibited B-cell proliferation, blockade of the TGFβ receptor was able to maintain normal B-cell proliferation even in the presence of exogenous TGFβ−1. (D-E) Bar plots showing CD22 and CD72 expression on B-cells after culture with TGFβ−1. B-cells were activated with B-cell expansion cocktail and TGFβ−1 was added every day starting from day 0 in the experimental group. While control B-cells decreased expression of CD22 and CD72 as they became more activated, additional of TGFβ−1 maintained elevated levels of CD22 and CD72. (F) Western blot for TRAF6, SHIP1, and SHIP2 in B-cells cultured for 4 days with or without TGFβ−1. Culture with TGFβ−1 led to an increase in TRAF6, SHIP1, and SHIP2 in B-cells. (G) Lyz2-Tgfb1 transgenic mice with Tgfb1 knocked out of myeloid cells (LyzMcre/WT-Tgfb1flox/flox) were challenged with CT2A glioma cells. Animal survival was assessed in these experimental mice compared to wild-type controls (LyzMcre/WT-Tgfb1flox/WT). Knockout of Tgfb1 trended towards improved animal survival. (H) Lyz2-Tgfb1 transgenic mice with Tgfb1 knocked out of myeloid cells (LyzMcre/WT-Tgfb1flox/flox) were challenged with CT2A glioma cells. After 14 days, the tumors were analyzed for B-cell infiltration via flow cytometry. Knockout of Tgfb1 trended towards increased B-cell counts in the tumor. (I) Wild-type C57 mice were challenged with CT2A glioma cells with Tgfb1 knocked out (CT2A-Tgfb1KO). Animal survival was assessed in these experimental tumors compared to wild type tumors (CT2A-Tgfb1VC). Knockout of Tgfb1 in CT2A tumor cells significantly improved animal survival. (J) The tumors of wild-type mice bearing CT2A-Tgfb1KO tumors were analyzed after 14 days for B-cell infiltration compared to mice bearing CT2A-Tgfb1VC controls. Knockout of Tgfb1 in CT2A tumor cells significantly increased B-cell counts in the tumor. ns= p>0.05, * = p<0.05, ** = p<0.01, *** = p<0.001, **** = p<0.0001