Mutant p53 Exploits Enhancers to Elevate Immunosuppressive Chemokine Expression and Impair Immune Checkpoint Inhibitors in Pancreatic Cancer

Abstract

Pancreatic ductal adenocarcinoma (PDAC) is an aggressive cancer without effective treatments. It is characterized by activating KRAS mutations and p53 alterations. However, how these mutations dysregulate cancer-cell-intrinsic gene programs to influence the immune landscape of the tumor microenvironment (TME) remains poorly understood. Here, we show that p53^R172H establishes an immunosuppressive TME, diminishes the efficacy of immune checkpoint inhibitors (ICIs), and enhances tumor growth. Our findings reveal that the upregulation of the immunosuppressive chemokine Cxcl1 mediates these pro-tumorigenic functions of p53^R172H. Mechanistically, we show that p53^R172H associates with the distal enhancers of the Cxcl1 gene, increasing enhancer activity and Cxcl1 expression. p53^R172H occupies these enhancers in an NF-κB-pathway-dependent manner, suggesting NF-κB's role in recruiting p53^R172H to the Cxcl1 enhancers. Our work uncovers how a common mutation in a tumor-suppressor transcription factor appropriates enhancers, stimulating chemokine expression and establishing an immunosuppressive TME that diminishes ICI efficacy in PDAC.

Keywords: Cxcl1; PDAC; chemokine; eRNA; enhancer; immune checkpoint inhibitors; immunosuppression; immunotherapy; mutant p53; p53R172H; p53R175H.

Figure 1.. p53^R172H elevates the expression of a subset of chemokine genes.

A, Generation of Trp53^−/− isogenic cells from the parental Trp53^R172H/− cells. Trp53 gene with R172H mutation in exon-5 is deleted from intron-1 to intron-10 using CRISPR/Cas9. B, Minus-average (MA) plot of RNA-seq transcripts showing the differentially expressed genes between Trp53^R172H/− and Trp53^−/− isogenic cells. Significantly upregulated and downregulated genes (adjusted p-value < 0.001 and four-fold change in normalized counts) are shown in blue and red, respectively. The wild-type p53-regulated genes are shown in yellow. C, Pathways (left) and transcription factor targets (right) enriched in p53^R172H-upregulated genes. Gene Ontology analysis was performed using Enrichr against the KEGG pathway database (left) and TRRUST database (right). Blue dots represent significant gene sets (p-value < 0.05), and the darker color represents higher significance. D, Generation of Trp53^R172H-restored isogenic cells in Trp53^−/− cells using a Trp53^R172H cDNA expression cassette in piggyback vector (Trp53^−/− + pTrp53^R172H). An empty vector without the Trp53^R172H cDNA expression cassette (Trp53^−/− + pEV) was inserted in Trp53^−/− as a control. E, mRNA levels of the expressed chemokine genes are shown as the z-score heatmap of RNA-seq transcripts per million (TPM) in the four isogenic cells. F, Quantification of the three chemokine genes under p53^R172H control by ELISA in the tissue culture media of the four isogenic cells. P-values are calculated from a one-way ANOVA test followed by a post hoc test with Benjamini-Hochberg correction. Panels B, C, E, and F use Trp53^−/− clone-1 isogenic cells.

Figure 2.. p53^R172H creates immunosuppressive TME and abrogates ICIs efficacy.

A, Schematics and the experimental design of Trp53^R172H/− and Trp53^−/− isogenic cells’ orthotopic implantation in mouse pancreas, immune profiling of the tumors, ICIs treatment, and survival analysis. B, Weights of the Trp53^R172H/− and Trp53^−/− tumors. The p-value is calculated using a two-tailed t-test. C, Effect of the Trp53 status on T cell infiltration in PDAC tumors. The p-values are calculated using a two-tailed t-test. D, Effect of the Trp53 status on MDSCs infiltration in PDAC tumors. The p-value is calculated using a two-tailed t-test. E, Representative multiplex immunofluorescence staining of Trp53^R172H/− and Trp53^−/− tumors (left), and quantification of immune cells density from randomly selected 500 × 500 μm regions (right). The scale bars correspond to 200 μm and 50 μm in the main view and the magnified view, respectively. P-values are calculated using a two-tailed t-test. F, Effect of the Trp53 status and ICIs on PDAC tumor growth. Control mice were treated with IgG. Tumor volumes from the last measurement with at least three mice left in the cohort were used to calculate p-values using a two-tailed t-test. G, Kaplan-Meier survival curves showing the effect of Trp53 status and ICIs on the survival of mice implanted with either Trp53^R172H/− or Trp53^−/− cells. Control mice were treated with IgG. P-values are calculated using a log-rank (Mantel-Cox) test. H, Experimental design and timeline of tumor challenge experiment in the long-term survivor mice implanted with Trp53^−/− PDAC tumors and treated with ICIs.

Figure 3.. Cxcl1 mediates the immunosuppressive role of p53^R172H and abrogates ICIs efficacy.

A, Generation of Trp53^R172H/−;Cxcl1^−/− and Trp53^R172H/−;Cxcl5^−/− isogenic cells from the parental Trp53^R172H/− cells. A single guide-RNA-mediated genome editing using CRISPR/Cas9 resulted in a frameshift mutation. B, Weights of the Trp53^R172H/−;Cxcl1^−/− and Trp53^R172H/−;Cxcl5^−/− tumors compared with the Trp53^R172H/− tumors. P-values are calculated from a one-way ANOVA test followed by a post hoc test with Benjamini-Hochberg correction. C, Effect of the Cxcl1 status in T cell infiltration in PDAC tumors. P-values are calculated using a two-tailed t-test. D, Effect of the Cxcl1 status in leukocyte (CD45⁺) and neutrophil (CD45⁺CD11b⁺MHCII^-Ly-6G⁺) infiltration in PDAC tumors. P-values are calculated using a two-tailed t-test. E, Effect of the Cxcl1 status in lymphocyte (CD45⁺CD11b^-CD90⁺NK1.1^-) and NK cell (CD45⁺CD11b^-CD90⁺NK1.1⁺) infiltration in PDAC tumors. P-values are calculated using a two-tailed t-test. F, Effect of the Cxcl1 status in macrophage (CD45⁺CD11b⁺CD64⁺) infiltration in PDAC tumors. P-values are calculated using a two-tailed t-test. G, Representative multiplex immunofluorescence staining of Trp53^R172H/− and Trp53^R172H/−;Cxcl1^−/− tumors (left), and quantification of immune cells density from randomly selected 500 × 500 μm regions (right). The scale bars correspond to 200 μm and 50 μm in the main view and the magnified view, respectively. P-values are calculated using a two-tailed t-test. H, Effect of the Cxcl1 status and ICIs in PDAC tumor growth. Control mice were treated with IgG. Trp53^R172H/− tumor growth cohort is the same as in Figure S4C. Tumor volumes from the last measurement with at least three mice left in the cohort were used to calculate p-values using a two-tailed t-test. I, Kaplan-Meier survival curves showing the effect of Cxcl1 status and ICIs in the survival of mice implanted with parental Trp53^R172H/− or Trp53^R172H/−;Cxcl1^−/− cells. Control mice were treated with IgG. Trp53^R172H/− mice survival cohort is the same as in Figure S4D. P-values are calculated using a log-rank (Mantel-Cox) test.

Figure 4.. p53^R172H occupies and modulates Cxcl1 enhancers.

A, p53^R172H occupancy at and around the Cxcl1 gene in Trp53^R172H/− and Trp53^−/− cells using CUT&RUN. Non-specific IgG in Trp53^R172H/− cells is used as a control. B, p300, H3K27Ac, and H3K4me3 occupancy at and around the Cxcl1 gene in Trp53^R172H/− and Trp53^−/− cells using CUT&RUN. C, Nascent RNA profiles at and around the Cxcl1 gene in Trp53^R172H/−, Trp53^−/−, or Trp53^−/− + pTrp53^R172H cells using PRO-seq. Nascent RNA profiles are used to call de novo enhancers using dREG, and the annotated enhancers are shown at the bottom (names begin with “e”). D, Quantification of enhancer RNA (eRNA) from PRO-seq in the three enhancers around the Cxcl1 gene. E, Nascent RNA levels of the expressed Chemokine genes shown as the z-cores heatmap of PRO-seq RPM in the three isogenic cells. F, Quantification of p300 and histone modification levels (left) and p53^R172H occupancy (right) at the promoter and enhancers of the Cxcl1 gene using CUT&RUN. P-values are calculated excluding e8697 using a paired t-test.

Figure 5.. p53^R172H-occupied Enhancers regulate Cxcl1 expression and dictate Cxcl1-mediated immunosuppression.

A, Generation of enhancer-deleted isogenic cells from the parental Trp53^R172H/− cells. Two enhancers upstream of the Cxcl1 gene and one downstream of the Cxcl1 gene are deleted individually or in a pair using CRISPR/Cas9. The size of deleted regions is indicated in parentheses. B, Quantification of the Cxcl1 chemokine level in enhancer-deleted isogenic cells and comparison with the Trp53^R172H/− cells. P-values are calculated from a one-way ANOVA test followed by a post hoc test with Benjamini-Hochberg correction. C, Weights of the enhancer-deleted isogenic tumors compared with the Trp53^R172H/− tumors. Trp53^R172H/−;Δe8695;Δe8696 cells were not implanted in mice due to the minimal effect over either Trp53^R172H/−;Δe8695 or Trp53^R172H/−;Δe8696 cells. P-values are calculated from a one-way ANOVA test followed by a post hoc test with Benjamini-Hochberg correction. D, Effect of the e8696 and e8697 status in leukocyte and neutrophil infiltration in PDAC tumors. P-values are calculated from a one-way ANOVA test followed by a post hoc test with Benjamini-Hochberg correction. E, Representative multiplex immunofluorescence staining of Trp53^R172H/−;Δe8695 and Trp53^R172H/−;Δe8696 tumors. The scale bars correspond to 200 μm and 50 μm in the main view and the magnified view, respectively. F, Quantification of immune cell density from randomly selected 500 × 500 μm regions. P-values are calculated from a one-way ANOVA test followed by a post hoc test with Benjamini-Hochberg correction. G, Kaplan-Meier survival curves showing the effects of e8696 or e8697 status and ICIs in the survival of mice implanted with parental Trp53^R172H/− or Trp53^R172H/−;Δe8696, or Trp53^R172H/−;Δe8697 cells. Control mice were treated with IgG. Trp53^R172H/− mice survival cohort is the same as in Figure S4D & Figure 3I. P-values are calculated using a log-rank (Mantel-Cox) test.

Figure 6.. p53^R172H occupancy in Cxcl1 Enhancers is dependent on NF-κB.

A, Enriched transcription factor motifs at the promoters and enhancers of the p53^R172H-dependent chemokine genes identified using HOMER. B, NF-κB occupancy at the Cxcl1 gene and enhancers in Trp53^R172H/− and Trp53^−/− cells using CUT&RUN. Non-specific IgG is used as a control. The HOMER-identified NF-κB motifs are shown at the bottom (red bars). C, Quantification of NF-κB occupancy at the promoter and enhancers of the Cxcl1 gene using CUT&RUN. P-values are calculated excluding e8697 using a paired t-test. D, Quantification of NF-κB occupancy at the NF-κB CUT&RUN peaks using CUT&RUN. E, Quantification of NF-κB phosphorylation in the Trp53^R172H/− and Trp53^−/− cells using ELISA on cell extracts. The antibody targets S536 in the p65 subunit of NF-κB. F, Quantification of NF-κB localization in the cytoplasm vs nucleus using western blot (serial dilution of protein lysates). Vinculin and histone H3 were used as the markers of cytoplasmic and nuclear fractions, respectively. P-values are calculated from a two-way ANOVA test.

Figure 7.. NF-κB inhibition abrogates p53^R172H occupancy.

A, Effect of inhibiting NF-κB activation (5 uM TPCA-1) in NF-κB and p53^R172H occupancy at the Cxcl1 gene and enhancers. DMSO treatment serves as a control. B, Change in p53^R172H occupancy (left) and NF-κB occupancy (right) by NF-κB inhibition in the p53^R172H and NF-κB CUT&RUN peaks, respectively. C. Effects of TPCA-1 treatment in NF-κB (left) and p53^R172H (right) occupancy in the two detectable enhancers of the Cxcl1 gene. D, Effects of TPCA-1 treatment in Trp53^R172H (left) and Cxcl1 (right) mRNA levels using RT-qPCR.