The immune checkpoint B7x expands tumor-infiltrating Tregs and promotes resistance to anti-CTLA-4 therapy

Abstract

Immune checkpoint molecules play critical roles in regulating the anti-tumor immune response, and tumor cells often exploit these pathways to inhibit and evade the immune system. The B7-family immune checkpoint B7x is widely expressed in a broad variety of cancer types, and is generally associated with advanced disease progression and poorer clinical outcomes, but the underlying mechanisms are unclear. Here, we show that transduction and stable expression of B7x in multiple syngeneic tumor models leads to the expansion of immunosuppressive regulatory T cells (Tregs). Mechanistically, B7x does not cause increased proliferation of Tregs in tumors, but instead promotes the conversion of conventional CD4⁺ T cells into Tregs. Further, we find that B7x induces global transcriptomic changes in Tregs, driving these cells to adopt an activated and suppressive phenotype. B7x increases the expression of the Treg-specific transcription factor Foxp3 in CD4⁺ T cells by modulating the Akt/Foxo pathway. B7x-mediated regulation of Tregs reduces the efficacy of anti-CTLA-4 treatment, a therapeutic that partially relies on Treg-depletion. However, combination treatment of anti-B7x and anti-CTLA-4 leads to synergistic therapeutic efficacy and overcomes the B7x-mediated resistance to anti-CTLA-4. Altogether, B7x mediates an immunosuppressive Treg-promoting pathway within tumors and is a promising candidate for combination immunotherapy.

Fig. 1. Tumor-expressed B7x promotes infiltrating Treg populations.

a Experimental scheme of MC38 tumor engraftment. b Representative flow cytometry analysis of B7x expression in MC38-B7x and MC38-Control cells. c Representative gating strategy for Tregs in CD45⁺ cells. d, e Populations of CD4⁺ T cells and CD8⁺ T cells were analyzed within dissociated tumors (d) and draining lymph nodes (e). f, g Gated on Tregs, surface TGF-LAP and intracellular Ki67 (f), and surface Nrp1 and intracellular Helios (g) was determined. n = 9 mice in control group and 10 mice in B7x group, representative of three independent experiments. For (d–g), error bars represent SEM and P values were calculated by two-tailed Student’s T-test, unadjusted for multiple comparisons.

Fig. 2. B7x promotes the expression of Foxp3 in CD4⁺ T cells.

a, b Splenic CD4⁺ T cells were isolated from wildtype mice and were cultured with iTreg-inducing conditions: anti-CD3, anti-CD28, IL-2, 10μg/mL recombinant B7x-Ig fusion protein or Control-Ig, and varying concentrations of TGFβ1, and expression of Foxp3 was measured by flow cytometry after 4 days. c CD4⁺ T cells were cultured as described in (a), except with fixed 1 ng/mL TGFβ1 and varying concentrations of B7x-Ig or Control-Ig, and expression of Foxp3 was assayed after 4 days. d CD4⁺ T cells from Foxp3-GFP/DTR mice were cultured in iTreg-inducing conditions with either B7x-Ig or Control-Ig and 2 ng/mL TGFβ1, and expression of GFP was measured after 4 days. e CD4⁺ T cells were cultured with anti-CD3, anti-CD28, and either B7x-transduced or control MC38 or Hepa1-6 cells, and expression of Foxp3 in the T cells was analyzed after 4 days. f iTregs were induced with B7x-Ig or Control-Ig as described in (a), and expression of intracellular IL-10 and surface TGF-LAP was measured after 4 days. Error bars represent SEM, P values were calculated by two-tailed Student’s T-test. Results are representative of three independent experiments.

Fig. 3. B7x promotes immunosuppressive function in Tregs.

a–d MC38-B7x and MC38-Control tumors were engrafted into Foxp3-GFP/DTR mice, and after 7 days, tumors were dissociated and GFP⁺ CD4⁺ T cells were flow-sorted for RNA extraction and whole transcriptome RNA-seq analysis, n = 3 samples per group. a Principle component analysis (PCA) plots based on all expressed genes in MC38-B7x and MC38-Control tumors. b Microarray (MA) plot showing gene expression changes between the two Treg groups, red dots indicate significantly differentially expressed genes based on log₂fold change $\ge$1 or $\le$−1 and adjusted p-value < 0.05 as calculated by the Benjamini and Hochberg method. c Gene ontology (GO) analysis on molecular functions is shown, with GO terms enriched in Tregs from B7x⁺ tumors on the right and Tregs from B7x⁻ control tumors on left. d Gene set enrichment analysis was performed on DEGs with gene sets for Treg Suppressive and Effector phenotypes, normalized enrichment score (NES) and familywise error rate (FWER) p values based on modified Kolmogorov Smirnov test are shown in top-right of each plot. e–g CD4⁺ T cells from Foxp3-GFP/DTR mice were induced into iTregs in vitro in the presence of either Control-Ig or B7x-Ig, and subsequently GFP⁺ CD4⁺ T cells were flow-sorted for RNA extraction and RNA-seq (n = 3 samples per group). e PCA plots based on all expressed genes in the Control-Ig and B7x-Ig Tregs. f Volcano plot of differential gene expression comparing Control-Ig Tregs to B7x-Ig Tregs, genes in red are upregulated in either group at log₂ fold change ≥1 or ≤−1 and adjusted p-value < 0.05 as calculated by the Benjamini and Hochberg method. g Double-gradient heatmap depicting fold change of selected DEGs relevant to Treg differentiation for each sample in the Control-Ig and B7x-Ig Treg groups. h All genes from the datasets of (b) and (f) were plotted by magnitude of fold change, genes appearing as DEGs in both data sets are displayed in red (left). Venn diagrams listing numbers of DEGs in each data set are shown (right). i, j iTregs were generated with B7x-Ig or Control-Ig as described in (d), and resulting GFP⁺ iTregs were flow-sorted and co-cultured with CellTrace Violet (CTV)-labeled responder T cells and anti-CD3/CD28 Dynabeads, and proliferation of responder T cells was analyzed by CTV dilution after 3 days. Error bars represent SEM, P values were calculated by two-tailed Student’s T-test. Results are representative of three independent experiments.

Fig. 4. B7x regulates the Akt-Foxo1 pathway to enhance Foxp3 expression.

a–f CD4⁺ T cells were cultured in iTreg-inducing conditions with either B7x-Ig or Control-Ig for up to 48 h, and phosphorylation status of c-Jun (a), p65, b, Akt, c, total Foxo1 (d), and phospho-Foxo1 (e) was determined by phospho-flow cytometry. f The percentage of cells with dephosphorylated Foxo1 (% total Foxo1⁺ – % p-Foxo1⁺) and fraction of dephosphorylated Foxo1 of Foxo1⁺ cells are shown. Error bars represent SEM, P values were calculated by two-tailed Student’s T-test, unadjusted for multiple comparison. Results are representative of 4 independent experiments. g–i iTregs were generated as described in (a) and were subsequently stained with nuclear marker DAPI, cytoplasmic marker phalloidin, and anti-Foxo1, and were analyzed by confocal microscopy. g Representative images are shown, arrowheads represent nuclear localization of Foxo1. h Colocalization analysis for overlap of Foxo1 and DAPI signal was performed to calculate Pearson’s R coefficient and Intensity Correlation Coefficient, each point represents a unique field-of-view from separate experiments (n = 6 per group). Error bars represent SEM. Representative of three independent experiments. i Percent of cells with nuclear localization of Foxo1 per field-of-view. j CD4⁺ T cells were induced into iTregs with either B7x-Ig or Control-Ig and were treated with inhibitors against Akt (GSK690693), Foxo1/3 (AS1842856, carbenoxolone), and Shp1 (TPI-1), and Foxp3 expression was analyzed after 4 days. Error bars represent SEM, P values were calculated by two-tailed Student’s T-test. Results are representative of three experiments.

Fig. 5. Expression of B7x in tumors reduces efficacy of anti-CTLA-4 treatment.

a Mice were engrafted with MC38-Control or MC38-B7x tumors, sacrificed after 7 days, and tumors were dissociated to analyze expression of CTLA-4 on Foxp3⁺ Tregs. n = 10 mice per group, representative of three independent experiments. b Splenic CD4⁺ T cells were differentiated into Tregs with either Control-Ig or B7x-Ig and varying amounts of TGFβ1, and expression of CTLA-4 was analyzed after 4 days. Results are representative of 3 independent experiments. Representative flow cytometry plots (left), and graphs quantifying percent CTLA-4⁺ of total Tregs and CTLA-4 MFI of Tregs are shown (right). c Experimental scheme of MC38 tumor engraftment and anti-CTLA-4 (+) or IgG isotype (−) treatment. d Tumor volumes were tracked, individual tumor volumes (left) were tracked and numbers designate tumor clearance (final tumor volume <50 mm³), and mean tumor volumes were calculated (right). There were 14 mice in the MC38-Control isotype-treated group, 15 mice in the MC38-Control anti-CTLA-4-treated group, 15 mice in the MC38-B7x isotype-treated group, and 15 mice in the MC38-B7x anti-CTLA-4 treated group. Representative of five independent experiments. e Mice were engrafted with tumors and treated as described in (c), sacrificed at day 17, and tumors were extracted and weighed. There were seven mice in the MC38-Control isotype-treated group, eight mice in the MC38-Control anti-CTLA-4-treated group, five mice in the MC38-B7x isotype-treated group, and six mice in the MC38-B7x anti-CTLA-4 treated group. Representative of three independent experiments. f, g Mice were engrafted with tumors and treated as in (c), mice were sacrificed for tumor dissociation at day 17, and cells were analyzed by flow cytometry. CD45⁺ immune cells were quantified and calculated as proportion of total immune cell population (f), and fold change of the anti-CTLA-4 treated groups normalized to each group’s Isotype-treated population (g). h Proportions of Foxp3⁺ cells of CD4⁺ T cells were determined. i Cell suspensions were stimulated with PMA/ionomycin and were stained intracellularly for IFN-γ. For (f–i), there were seven mice in the MC38-Control isotype-treated group, nine mice in the MC38-Control anti-CTLA-4-treated group, seven mice in the MC38-B7x isotype-treated group, and 9 mice in the MC38-B7x anti-CTLA-4 treated group. Representative of three independent experiments. Error bars represent SEM, P values were calculated by two-tailed Student’s T-test.

Fig. 6. Combination anti-B7x and anti-CTLA-4 therapy has synergistic therapeutic effect.

a Representative staining of MC38-B7x and MC38-Control cells stained with anti-B7x clone 1H3. b Experimental scheme of MC38-B7x engraftment and anti-B7x or anti-CTLA-4 treatment. c Tumor volumes were tracked and shown as individual tracings (top), numbers designate tumor clearance (final tumor volume < 50 mm³), and mean tumor volumes were calculated (bottom). d, e CD45⁺ immune cells were quantified and calculated as proportion of total immune cell population (d), and fold change of treated groups were normalized to isotype-treated population (e). f Tregs were quantified. g Cell suspensions were stimulated with PMA/ionomycin and were stained intracellularly for IFN-γ expression. n = 12 mice per group, representative of three independent experiments. Error bars represent SEM, P values were calculated by two-tailed Student’s T-test.