Tumor-associated neutrophils suppress CD8+ T cell immunity in urothelial bladder carcinoma through the COX-2/PGE2/IDO1 Axis

Abstract

Background: Many urothelial bladder carcinoma (UBC) patients don't respond to immune checkpoint blockade (ICB) therapy, possibly due to tumor-associated neutrophils (TANs) suppressing lymphocyte immune response.

Methods: We conducted a meta-analysis on the predictive value of neutrophil-lymphocyte ratio (NLR) in ICB response and investigated TANs' role in UBC. We used RNA-sequencing, HALO spatial analysis, single-cell RNA-sequencing, and flow cytometry to study the impacts of TANs and prostaglandin E2 (PGE2) on IDO1 expression. Animal experiments evaluated celecoxib's efficacy in targeting PGE2 synthesis.

Results: Our analysis showed that higher TAN infiltration predicted worse outcomes in UBC patients receiving ICB therapy. Our research revealed that TANs promote IDO1 expression in cancer cells, resulting in immunosuppression. We also found that PGE2 synthesized by COX-2 in neutrophils played a key role in upregulating IDO1 in cancer cells. Animal experiments showed that targeting PGE2 synthesis in neutrophils with celecoxib enhanced the efficacy of ICB treatment.

Conclusions: TAN-secreted PGE2 upregulates IDO1, dampening T cell function in UBC. Celecoxib targeting of PGE2 synthesis represents a promising approach to enhance ICB efficacy in UBC.

Fig. 1. Meta-analysis of the association between NLR and prognosis in UBC patients receiving immune checkpoint blockade therapies.

a, b Forest plots of association between neutrophils to lymphocytes ratio (NLR) and clinical response to immune checkpoint blockade therapies (a) and overall survival (b) in studies involving ICB treatment on UBC patients.

Fig. 2. Neutrophil and CD8⁺ T cell infiltration predicts the response to immunotherapy in UBCs.

a–d Neoadjuvant ICB combined with chemotherapy cohort. e–h Adjuvant ICB cohort. i–l Reproduced results from IMVigor 210 database with CC 3.0 licence. a, b, c Dotplot showing the expression in IHC of stromal CD66b (a), intratumoral CD8 (b), or stromal CD66b to intratumoral CD8 ratio (c) in pCR or non-pCR patients. n = 8 in the pCR groups and n = 5 in the non-pCR group. d ROC curves represent the predictive value of stromal CD66b to intratumor CD8 score in pCR to neoadjuvant therapy. e, f, g Dotplot showing the expression in IHC of stromal CD66b (e), intratumoral CD8 (f), or stromal CD66b to intratumoral CD8 ratio (g) in patients CR/PR or SD/PD to ICBs. n = 9 in the CR/PR groups and n = 7 in the SD/PD group. h ROC curves represent the predictive value of stromal CD66b to intratumor CD8 score in ICB response. i, j, k Stacked bar plots showing the proportion of CR/PR or SD/PD patients in the IMVigor 210 cohort based on the score of neutrophils (i), CD8⁺ T cells (j), and neutrophils to CD8⁺ T cells ratio (k). For neutrophils or CD8⁺ T cells group classification, patients are divided into two groups according to the two-thirds number of neutrophils score (cut-off value = 2.442) and the one-thirds CD8⁺ T cells score (cut-off value = 0.1881). (i: n = 99 in the high group and 199 in the low group; j: n = 98 in the low group and 200 in the high group; k: n = 24 in the high group and 274 in the low group). l Kaplan-Meier plot displays the estimated overall survival probabilities of groups according to neutrophils to CD8⁺ T cells ratio (high: n = 24, Low: n = 274). Error bars represent the mean ± SEM. p < 0.05 was considered a significant difference, (Mann-Whitney rank test, Chi-square test, fisher exact test, or logrank test). *p < 0.05, **p < 0.01. pCR pathological complete response, CR complete response, PR partial response, SD stable disease, PD progressive disease.

Fig. 3. IDO1 mediates the immunosuppressive effects of the interaction between TANs and cancer cells.

a Flow cytometry analysis displays the proportion of IFN-γ⁺ T cells in the indicated groups, n = 4. b Representative images of intratumoral CD8⁺ T cells infiltration in the tumor core with or without abundant TANs surrounded. c Spearman correlation analysis of stroma CD66b count with intratumoral CD8 count. d Differential gene ranking plot displaying gene expression changes between tumors with high or low neutrophil infiltration, n = 25 in each group. e GSEA analysis showing the tryptophan metabolism pathway is enriched in tumors with high neutrophil infiltration. f Representative images illustrating methods of HALO immune infiltration analysis and analysis results of spatial distribution and expression intensity of IDO1 in relation to tumor margins with or without massive neutrophils surrounded, n = 5–6. g Representative images and statistical plot illustrating the expression of IDO1 in the indicated groups in the neutrophil depletion animal experiments, n = 4. h, i Flow cytometry (h), and IHC analysis (i) indicating CD8 infiltration in the indicated groups. j qRT-PCR result shows relative fold expression of IDO1 on T24 cells in the indicated groups, n = 4. k, Flow cytometry analysis shows IDO1 expression in tumor cells in the indicated groups. l qRT-PCR result shows relative fold expression of IDO1 on the WT and IDO1-OE T24 cells. m Cell viability results show the relative cell growth in the indicated groups, n = 4. n Representation of the 3D cell culture in 48 well plates and the infiltration of the PBMCs. o Representative images of the 3D cell culture models in the indicated groups. p Representative images of EdU assays in the 3D cell culture models models. Scale bar: black, 100 μm. Error bars represent the mean ± SEM. p < 0.05 was considered a significant difference, ns was no significance (one-way ANOVA with Holm-Sidak multiple comparison tests, unpaired Student’s t-test). *p < 0.05, **p < 0.01, ***p < 0.001, NC negative control.

Fig. 4. Combination treatment of IDO1 inhibition and anti-PD1 delays tumor growth.

a, b Tumor growth curves (a) and tumor weights (b) of wild type and IDO1-KO groups with or without anti-PD1 therapy were measured, n = 5. c, d Tumor growth curves (c) and tumor weights (d) of each group were measured following anti-PD1, indoximod, or a combination of indoximod and anti-PD1 treatment, n = 5. e Flow cytometry analysis of the proportion of Granzyme B⁺ cells within CD8⁺ T cells from the indicated groups, n = 3. f Representative images and statistical plot illustrating the expression of CD8 in the indicated groups, n = 4. Scale bar: black, 100 μm. Error bars represent the mean ± SEM. p < 0.05 was considered a significant difference, ns was no significance (one-way ANOVA with Holm-Sidak multiple comparison test, unpaired parametric Student’s t-test). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. IDO1i: IDO1 inhibitor, indoximod; WT: wild type.

Fig. 5. PGE2 secreted by neutrophils upregulated IDO1 via the PKC-GSK pathway.

a Multiple cytokines assay results indicate the relative concentration of cytokines and chemokines in the supernatant of peripheral neutrophils from UBC patients. b Flow cytometry analysis reveals that IDO1 expression in cancer cells cocultured with neutrophils following the indicated cytokines are targeted with corresponding antibodies. c Flow cytometry analysis displays IDO1 expression in cancer cells cocultured with unboiled or boiled neutrophils supernatant, n = 4-5. d tSNE plot of myeloid single cells from scRNA-seq colored by major cell types. e Volcano plot displays the differential expressed genes of neutrophils. f Ridgeplot shows expression levels of PTGS2 in the indicated clusters. g Histograms of ELISA results show the PGE2 concentration in the indicated groups, n = 6. h Flow cytometry analysis and statistical plot of IDO1 expression in cancer cells cocultured with neutrophils after celecoxib was treated, n = 3-5. i Flow cytometry analysis and statistical plot of IDO1 expression in cancer cells cocultured with neutrophils after Ro-31-8425, a PKC inhibitor, or LY294002 a pan-PI3K inhibitor, was treated, n = 5. Error bars represent the mean ± SEM. p < 0.05 was considered a significant difference, ns was no significance (one-way ANOVA with Holm-Sidak multiple comparison test). *p < 0.05, **p < 0.01, ***p < 0.001, **** p < 0.0001. NC negative control.

Fig. 6. Celecoxib enhances the therapeutic activity of PD-1 blockade in combination therapy.

a Schematic representation of vehilce, celecoxib, anti-PD1, and combination treatment schedule. b, c Tumor growth curves (b) and tumor weights (c) of each group were measured after treatment with celecoxib, anti-PD1, or a combination of celecoxib and anti-PD1. d Representative images and statistic analysis illustrating the expression of IDO1 in the indicated groups. e Flow cytometry analysis of the proportion of Granzyme B⁺ cells within CD8⁺ T cells from the indicated groups. f Representative images and statistic analysis illustrating the expression of CD8 in the indicated groups. Scale bar: white, 100μm. Error bars represent the mean ± SEM. n = 6 in (b, c); n = 3 in (d); n = 4–5 in (e); n = 5 in (f). p < 0.05 was considered a significant difference, (one-way ANOVA with Holm-Sidak multiple comparison tests, unpaired Student’s t-test). *p < 0.01, ***p < 0.001, **** p < 0.0001.