Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

TNF-α-producing macrophages determine subtype identity and prognosis via AP1 enhancer reprogramming in pancreatic cancer

Abstract

Large-scale genomic profiling of pancreatic cancer (PDAC) has revealed two distinct subtypes: ‘classical’ and ‘basal-like’. Their variable coexistence within the stromal immune microenvironment is linked to differential prognosis; however, the extent to which these neoplastic subtypes shape the stromal immune landscape and impact clinical outcome remains unclear. By combining preclinical models, patient-derived xenografts, as well as FACS-sorted PDAC patient biopsies, we show that the basal-like neoplastic state is sustained via BRD4-mediated cJUN/AP1 expression, which induces CCL2 to recruit tumor necrosis factor (TNF)-α-secreting macrophages. TNF-α+ macrophages force classical neoplastic cells into an aggressive phenotypic state via lineage reprogramming. Integration of ATAC-, ChIP- and RNA-seq data revealed distinct JUNB/AP1 (classical) and cJUN/AP1 (basal-like)-driven regulation of PDAC subtype identity. Pharmacological inhibition of BRD4 led to suppression of the BRD4–cJUN–CCL2–TNF-α axis, restoration of classical subtype identity and a favorable prognosis. Hence, patient-tailored therapy for a cJUNhigh/TNF-αhigh subtype is paramount in overcoming highly inflamed and aggressive PDAC states.

Your institute does not have access to this article

Access options

Buy article

Get time limited or full article access on ReadCube.

$32.00

All prices are NET prices.

Fig. 1: BL/poorly differentiated PDAC tumors associated with high TNF-α expression.
Fig. 2: TNF-α switches CLA to BL phenotypic state via lineage reprograming.
Fig. 3: BL/poorly differentiated tumors are enriched with TNF-α+ macrophage infiltrations.
Fig. 4: Subtype-specific CCL2 expression and JUN/AP1 transcriptional program.
Fig. 5: cJUN controls immune surveillance and JUNB regulates differentiation.
Fig. 6: BRD4 controls cJUN for CCL2-mediated macrophages infiltration.
Fig. 7: cJUN–CCL2 axis links to TNF-α+ immune cells and confers poor prognosis.
Fig. 8: Pharmacological inhibition of the BRD4–JUN–CCL2 axis improves survival and restores favorable outcome.

Data availability

For this study, the Molecular Signatures Database (https://www.gsea-msigdb.org/gsea/msigdb/) and HOCOMOCO v.11 (https://hocomoco11.autosome.ru/) databases were used. Previously published microarray, single-cell and bulk RNA-seq, as well as ChIP-seq data that were re-analyzed here are available in ArrayExpress under accession code E-MTAB-6134 and in Gene Expression Omnibus under accession codes GSE125588, GSE64560 and GSE74651. Microarray as well as RNA-, ChIP-, ATAC- and i4C-seq data generated for this study have been deposited in Gene Expression Omnibus with the accession codes GSE173159 and GSE172103. FACS-sorted epithelial patient tumor RNA-seq data are available at EGA under accession code EGAS00001004660. Access is restricted due to patient data protection but can be made available through the DKFZ-HIPO Data Access Committee of Heidelberg Center for Personalized Oncology (hipo_daco@dkfz-heidelberg.de) upon reasonable request. Source data are provided with this paper. All other data supporting the findings of this study are available from the corresponding author on reasonable request.

Code availability

No custom code was used in this study.

References

  1. Siegel, R. L., Miller, K. D. & Jemal, A. Cancer statistics, 2020. CA Cancer J. Clin. 70, 7–30 (2020).

    PubMed  Google Scholar 

  2. Dreyer, S. B., Chang, D. K., Bailey, P. & Biankin, A. V. Pancreatic cancer genomes: implications for clinical management and therapeutic development. Clin. Cancer Res. 23, 1638–1646 (2017).

    PubMed  Google Scholar 

  3. Singh, S., Hasselluhn, M. C. & Neesse, A. A tangled tale of molecular subtypes in pancreatic cancer. Gut 68, 953–954 (2019).

    CAS  PubMed  Google Scholar 

  4. Hessmann, E. et al. Microenvironmental determinants of pancreatic cancer. Physiol. Rev. 100, 1707–1751 (2020).

    PubMed  Google Scholar 

  5. Witkiewicz, A. K. et al. Whole-exome sequencing of pancreatic cancer defines genetic diversity and therapeutic targets. Nat. Commun. 6, 6744 (2015).

    CAS  PubMed  Google Scholar 

  6. Waddell, N. et al. Whole genomes redefine the mutational landscape of pancreatic cancer. Nature 518, 495–501 (2015).

    CAS  PubMed  PubMed Central  Google Scholar 

  7. Bailey, P. et al. Genomic analyses identify molecular subtypes of pancreatic cancer. Nature 531, 47–52 (2016).

    CAS  PubMed  Google Scholar 

  8. Collisson, E. A. et al. Subtypes of pancreatic ductal adenocarcinoma and their differing responses to therapy. Nat. Med. 17, 500–503 (2011).

    CAS  PubMed  PubMed Central  Google Scholar 

  9. Moffitt, R. A. et al. Virtual microdissection identifies distinct tumor- and stroma-specific subtypes of pancreatic ductal adenocarcinoma. Nat. Genet. 47, 1168–1178 (2015).

    CAS  PubMed  PubMed Central  Google Scholar 

  10. Cancer Genome Atlas Research Network. Integrated genomic characterization of pancreatic ductal adenocarcinoma. Cancer Cell https://doi.org/10.1016/j.ccell.2017.07.007 (2017).

  11. Aung, K. L. et al. Genomics-driven precision medicine for advanced pancreatic cancer: early results from the COMPASS trial. Clin. Cancer Res. 24, 1344–1354 (2018).

    CAS  PubMed  Google Scholar 

  12. Puleo, F. et al. Stratification of pancreatic ductal adenocarcinomas based on tumor and microenvironment features. Gastroenterology 155, 1999–2013 (2018).

    PubMed  Google Scholar 

  13. Maurer, C. et al. Experimental microdissection enables functional harmonisation of pancreatic cancer subtypes. Gut 68, 1034–1043 (2019).

    CAS  PubMed  Google Scholar 

  14. Diaferia, G. R. et al. Dissection of transcriptional and cis-regulatory control of differentiation in human pancreatic cancer. EMBO J. 35, 595–617 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

  15. Roe, J. S. et al. Enhancer reprogramming promotes pancreatic cancer metastasis. Cell 170, 875–888 (2017).

    CAS  PubMed  PubMed Central  Google Scholar 

  16. Kalisz, M. et al. HNF1A recruits KDM6A to activate differentiated acinar cell programs that suppress pancreatic cancer. EMBO J. 39, e102808 (2020).

    CAS  PubMed  PubMed Central  Google Scholar 

  17. Andricovich, J. et al. Loss of KDM6A activates super-enhancers to induce gender-specific squamous-like pancreatic cancer and confers sensitivity to BET inhibitors. Cancer Cell 33, 512–526 (2018).

    CAS  PubMed  PubMed Central  Google Scholar 

  18. Adams, C.R. et al. Transcriptional control of subtype switching ensures adaptation and growth of pancreatic cancer. eLife https://doi.org/10.7554/eLife.45313 (2019).

  19. Hayashi, A. et al. A unifying paradigm for transcriptional heterogeneity and squamous features in pancreatic ductal adenocarcinoma. Nat. Cancer 1, 59–74 (2020).

    PubMed  PubMed Central  Google Scholar 

  20. Milan, M. et al. FOXA2 controls the cis-regulatory networks of pancreatic cancer cells in a differentiation grade-specific manner. EMBO J. 38, e102161 (2019).

    CAS  PubMed  PubMed Central  Google Scholar 

  21. Todoric, J. & Karin, M. The fire within: cell-autonomous mechanisms in inflammation-driven cancer. Cancer Cell 35, 714–720 (2019).

    CAS  PubMed  Google Scholar 

  22. Cobo, I. et al. Transcriptional regulation by NR5A2 links differentiation and inflammation in the pancreas. Nature 554, 533–537 (2018).

    CAS  PubMed  PubMed Central  Google Scholar 

  23. LaFave, L. M. & Buenrostro, J. D. Unlocking PDAC initiation with AP-1. Nat. Cancer 2, 14–15 (2021).

    PubMed  Google Scholar 

  24. Crawford, H. C., Pasca di Magliano, M. & Banerjee, S. Signaling networks that control cellular plasticity in pancreatic tumorigenesis, progression, and metastasis. Gastroenterology 156, 2073–2084 (2019).

    CAS  PubMed  Google Scholar 

  25. Hosein, A. N. et al. Cellular heterogeneity during mouse pancreatic ductal adenocarcinoma progression at single-cell resolution. JCI Insight 5, e129212 (2019).

    Google Scholar 

  26. Aiello, N. M. et al. Metastatic progression is associated with dynamic changes in the local microenvironment. Nat. Commun. 7, 12819 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

  27. Rhim, A. D. et al. EMT and dissemination precede pancreatic tumor formation. Cell 148, 349–361 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  28. Singh, S. K. et al. Antithetical NFATc1-Sox2 and p53-miR200 signaling networks govern pancreatic cancer cell plasticity. EMBO J. 34, 517–530 (2015).

    CAS  PubMed  PubMed Central  Google Scholar 

  29. Das, S., Shapiro, B., Vucic, E. A., Vogt, S. & Bar-Sagi, D. Tumor cell-derived IL1β promotes desmoplasia and immune suppression in pancreatic cancer. Cancer Res. 80, 1088–1101 (2020).

    CAS  PubMed  PubMed Central  Google Scholar 

  30. Biffi, G. et al. IL1-induced JAK/STAT signaling is antagonized by TGFβ to shape CAF heterogeneity in pancreatic ductal adenocarcinoma. Cancer Discov. 9, 282–301 (2019).

    PubMed  Google Scholar 

  31. Huang, H. et al. Targeting TGFβR2-mutant tumors exposes vulnerabilities to stromal TGFβ blockade in pancreatic cancer. EMBO Mol. Med. 11, e10515 (2019).

    CAS  PubMed  PubMed Central  Google Scholar 

  32. Zhao, X. et al. Inhibiting tumor necrosis factor-α diminishes desmoplasia and inflammation to overcome chemoresistance in pancreatic ductal adenocarcinoma. Oncotarget 7, 81110–81122 (2016).

    PubMed  PubMed Central  Google Scholar 

  33. Alam, M. S. et al. Selective inhibition of the p38 alternative activation pathway in infiltrating T cells inhibits pancreatic cancer progression. Nat. Med. 21, 1337–1343 (2015).

    CAS  PubMed  PubMed Central  Google Scholar 

  34. Hingorani, S. R. et al. Trp53R172H and KrasG12D cooperate to promote chromosomal instability and widely metastatic pancreatic ductal adenocarcinoma in mice. Cancer Cell 7, 469–483 (2005).

    CAS  PubMed  Google Scholar 

  35. Noll, E. M. et al. CYP3A5 mediates basal and acquired therapy resistance in different subtypes of pancreatic ductal adenocarcinoma. Nat. Med. 22, 278–287 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

  36. Muckenhuber, A. et al. Pancreatic ductal adenocarcinoma subtyping using the biomarkers hepatocyte nuclear factor-1A and cytokeratin-81 correlates with outcome and treatment response. Clin. Cancer Res. 24, 351–359 (2018).

    CAS  PubMed  Google Scholar 

  37. Miyabayashi, K. et al. Intraductal transplantation models of human pancreatic ductal adenocarcinoma reveal progressive transition of molecular subtypes. Cancer Discov. 10, 1566–1589 (2020).

    CAS  PubMed  PubMed Central  Google Scholar 

  38. Martinelli, P. et al. GATA6 regulates EMT and tumour dissemination, and is a marker of response to adjuvant chemotherapy in pancreatic cancer. Gut 66, 1665–1676 (2017).

    CAS  PubMed  Google Scholar 

  39. Chen, E. Y. et al. Enrichr: interactive and collaborative HTML5 gene list enrichment analysis tool. BMC Bioinf. 14, 128 (2013).

    Google Scholar 

  40. Kuleshov, M. V. et al. Enrichr: a comprehensive gene set enrichment analysis web server 2016 update. Nucleic Acids Res. 44, W90–W97 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

  41. Bhagwat, A. S. et al. BET bromodomain inhibition releases the mediator complex from select cis-regulatory elements. Cell Rep. 15, 519–530 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

  42. Chen, N. M. et al. NFATc1 Links EGFR signaling to induction of sox9 transcription and acinar-ductal transdifferentiation in the pancreas. Gastroenterology 148, 1024–1034 (2015).

    CAS  PubMed  Google Scholar 

  43. Lipner, M. B. et al. Irreversible JNK1-JUN inhibition by JNK-IN-8 sensitizes pancreatic cancer to 5-FU/FOLFOX chemotherapy. JCI Insight 5, e129905 (2020).

    PubMed Central  Google Scholar 

  44. Calcinotto, A. et al. Targeting TNF-α to neoangiogenic vessels enhances lymphocyte infiltration in tumors and increases the therapeutic potential of immunotherapy. J. Immunol. 188, 2687–2694 (2012).

    CAS  PubMed  Google Scholar 

  45. Ikemoto, S. et al. TNFα, IL-1β and IL-6 production by peripheral blood monocytes in patients with renal cell carcinoma. Anticancer Res. 20, 317–321 (2000).

    CAS  PubMed  Google Scholar 

  46. Chan-Seng-Yue, M. et al. Transcription phenotypes of pancreatic cancer are driven by genomic events during tumor evolution. Nat. Genet. 52, 231–240 (2020).

    CAS  PubMed  Google Scholar 

  47. Chiu, R., Angel, P. & Karin, M. Jun-B differs in its biological properties from, and is a negative regulator of, c-Jun. Cell 59, 979–986 (1989).

    CAS  PubMed  Google Scholar 

  48. Deng, T. & Karin, M. JunB differs from c-Jun in its DNA-binding and dimerization domains, and represses c-Jun by formation of inactive heterodimers. Genes Dev. 7, 479–490 (1993).

    CAS  PubMed  Google Scholar 

  49. Szabowski, A. et al. c-Jun and JunB antagonistically control cytokine-regulated mesenchymal-epidermal interaction in skin. Cell 103, 745–755 (2000).

    CAS  PubMed  Google Scholar 

  50. Somerville, T. D. et al. Squamous trans-differentiation of pancreatic cancer cells promotes stromal inflammation. eLife 9, e53381 (2020).

    PubMed  PubMed Central  Google Scholar 

  51. Sanford, D. E. et al. Inflammatory monocyte mobilization decreases patient survival in pancreatic cancer: a role for targeting the CCL2/CCR2 axis. Clin. Cancer Res. 19, 3404–3415 (2013).

    CAS  PubMed  PubMed Central  Google Scholar 

  52. Nywening, T. M. et al. Targeting both tumour-associated CXCR2(+) neutrophils and CCR2(+) macrophages disrupts myeloid recruitment and improves chemotherapeutic responses in pancreatic ductal adenocarcinoma. Gut 67, 1112–1123 (2018).

    CAS  PubMed  Google Scholar 

  53. Sherman, M. H. et al. Stromal cues regulate the pancreatic cancer epigenome and metabolome. Proc. Natl Acad. Sci. USA 114, 1129–1134 (2017).

    CAS  PubMed  PubMed Central  Google Scholar 

  54. Shu, S. et al. Synthetic lethal and resistance interactions with BET bromodomain inhibitors in triple-negative breast cancer. Mol. Cell 78, 1096–1113 (2020).

    CAS  PubMed  PubMed Central  Google Scholar 

  55. Mazur, P. K. et al. Combined inhibition of BET family proteins and histone deacetylases as a potential epigenetics-based therapy for pancreatic ductal adenocarcinoma. Nat. Med. 21, 1163–1171 (2015).

    CAS  PubMed  PubMed Central  Google Scholar 

  56. Patzak, M. S. et al. Cytosolic 5′-nucleotidase 1A is overexpressed in pancreatic cancer and mediates gemcitabine resistance by reducing intracellular gemcitabine metabolites. EBioMedicine 40, 394–405 (2019).

    PubMed  PubMed Central  Google Scholar 

  57. Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013).

    CAS  PubMed  Google Scholar 

  58. Afgan, E. et al. The Galaxy platform for accessible, reproducible and collaborative biomedical analyses: 2018 update. Nucleic Acids Res. 46, W537–W544 (2018).

    CAS  PubMed  PubMed Central  Google Scholar 

  59. Anders, S., Pyl, P. T. & Huber, W. HTSeq–a Python framework to work with high-throughput sequencing data. Bioinformatics 31, 166–169 (2015).

    CAS  PubMed  Google Scholar 

  60. Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550 (2014).

    PubMed  PubMed Central  Google Scholar 

  61. Subramanian, A. et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl Acad. Sci. USA 102, 15545–15550 (2005).

    CAS  PubMed  PubMed Central  Google Scholar 

  62. Yu, G., Wang, L. G., Han, Y. & He, Q. Y. clusterProfiler: an R package for comparing biological themes among gene clusters. Omics 16, 284–287 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  63. Liberzon, A. et al. The Molecular Signatures Database (MSigDB) hallmark gene set collection. Cell Syst. 1, 417–425 (2015).

    CAS  PubMed  PubMed Central  Google Scholar 

  64. Buenrostro, J. D., Giresi, P. G., Zaba, L. C., Chang, H. Y. & Greenleaf, W. J. Transposition of native chromatin for fast and sensitive epigenomic profiling of open chromatin, DNA-binding proteins and nucleosome position. Nat. Methods 10, 1213–1218 (2013).

    CAS  PubMed  PubMed Central  Google Scholar 

  65. Langmead, B. & Salzberg, S. L. Fast gapped-read alignment with Bowtie 2. Nat. Methods 9, 357–359 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  66. Yu, G., Wang, L. G. & He, Q. Y. ChIPseeker: an R/Bioconductor package for ChIP peak annotation, comparison and visualization. Bioinformatics 31, 2382–2383 (2015).

    CAS  PubMed  Google Scholar 

  67. Zhou, Y. et al. Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nat. Commun. 10, 1523 (2019).

    PubMed  PubMed Central  Google Scholar 

  68. Li, H. & Durbin, R. Fast and accurate long-read alignment with Burrows–Wheeler transform. Bioinformatics 26, 589–595 (2010).

    PubMed  PubMed Central  Google Scholar 

  69. Zhang, Y. et al. Model-based analysis of ChIP-Seq (MACS). Genome Biol. 9, R137 (2008).

    PubMed  PubMed Central  Google Scholar 

  70. Zhu, L. J. et al. ChIPpeakAnno: a Bioconductor package to annotate ChIP-seq and ChIP-chip data. BMC Bioinf. 11, 237 (2010).

    Google Scholar 

  71. Ramirez, F. et al. deepTools2: a next generation web server for deep-sequencing data analysis. Nucleic Acids Res. 44, W160–W165 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

  72. McLeay, R. C. & Bailey, T. L. Motif enrichment analysis: a unified framework and an evaluation on ChIP data. BMC Bioinf. 11, 165 (2010).

    Google Scholar 

  73. Kulakovskiy, I. V. et al. HOCOMOCO: towards a complete collection of transcription factor binding models for human and mouse via large-scale ChIP-Seq analysis. Nucleic Acids Res. 46, D252–D259 (2018).

    CAS  PubMed  Google Scholar 

  74. Brant, L. et al. Exploiting native forces to capture chromosome conformation in mammalian cell nuclei. Mol. Syst. Biol. 12, 891 (2016).

    PubMed  PubMed Central  Google Scholar 

  75. Stadhouders, R. et al. Multiplexed chromosome conformation capture sequencing for rapid genome-scale high-resolution detection of long-range chromatin interactions. Nat. Protoc. 8, 509–524 (2013).

    CAS  PubMed  Google Scholar 

  76. Patil, S. et al. EZH2 regulates pancreatic cancer subtype identity and tumor progression via transcriptional repression of GATA6. Cancer Res. 80, 4620–4632 (2020).

    CAS  PubMed  Google Scholar 

  77. Singh, S. K. et al. Post-translational modifications of OLIG2 regulate glioma invasion through the TGF-β pathway. Cell Rep. 16, 950–966 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

  78. Schindelin, J. et al. Fiji: an open-source platform for biological-image analysis. Nat. Methods 9, 676–682 (2012).

    CAS  PubMed  Google Scholar 

  79. Espinet, E. et al. Aggressive PDACs show hypomethylation of repetitive elements and the execution of an intrinsic IFN program linked to a ductal cell of origin. Cancer Discov. 11, 638–659 (2021).

    CAS  PubMed  Google Scholar 

  80. Stuart, T. et al. Comprehensive integration of single-cell data. Cell 177, 1888–1902 e1821 (2019).

    CAS  PubMed  PubMed Central  Google Scholar 

  81. Tan, Y. & Cahan, P. SingleCellNet: a computational tool to classify single cell RNA-seq data across platforms and across species. Cell Syst. 9, 207–213 (2019).

    CAS  PubMed  PubMed Central  Google Scholar 

  82. Baron, M. et al. A single-cell transcriptomic map of the human and mouse pancreas reveals inter- and intra-cell population structure. Cell Syst. 3, 346–360 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

The study was supported by the German Cancer Aid to S.K.S. (70112999; Max-Eder group), to A.N. (70113213; Max-Eder group) and to E.H. (70112108). The DFG grant to S.K.S., A.P. and E.H. (all KFO 5002), the Fritz-Thyssen Stiftung (1842610) to S.K.S. and the Volkswagen-Stiftung/Ministry for Culture and Science in Lower Saxony to V.E. (11-25 76251-12-3/16). M.T. was supported by the China Scholarship Council. The processing of human samples was supported by the Dietmar-Hopp Foundation and the BioRNSpitzen cluster ‘Molecular- and Cell-based Medicine’; the German Ministry of Science and Education e:Med program for systems biology (PANC-STRAT consortium, grant nos. 01ZX1305C and 01ZX1605C). The collection and processing of specimens via PancoBank was supported by Heidelberger Stiftung Chirurgie. We thank D. Spyropoulou, E. Sidhu, M. Romey, F. Penz, K. Reutlinger and W. Kopp for their technical assistance. We also thank the EPZ-Biobank and the Department of General and Visceral Surgery of the University Hospital Heidelberg, especially N. A. Giese, T. Hackert, O. Strobel and M. Büchler for the collaboration involving fresh primary PDAC specimens.

Author information

Authors and Affiliations

Authors

Contributions

M.T., L.K. and S.K.S. designed the overall study. S.K.S. wrote the manuscript with contribution of L.K. U.K. and A.P. edited the manuscript. M.T. and L.K. performed preclinical studies. E.E. performed FACS-sorted RNA-seq analysis in patient biopsies, Z.G. performed bioinformatics analyses and A.T. helped in data interpretation. L.U. and S.K. performed tumor histology, RNAscope and related quantification studies. M.T., L.K., L.X. and T.G. conducted RNA-, ATAC-, ChIP-seq analyses and A.P. helped in data interpretation. L.K., F.W. and U.G. helped with the bioinformatics analyses. A.D.M. analyzed scRNA-seq and helped with bioinformatics analysis. M.T. and A.M. conducted i4C assays and A.P. helped in bioinformatics analysis. K.B. and B.C. helped and provided reagents for lenti- and retroviral experiments. S.H., P.S. and E.H. provided human PDX specimens. P.S., A.N., S.K.S. and V.E. critically analyzed TMA, murine and human histopathological examination. P.S., U.K., A.N. and E.H., provided human PDAC specimens and were involved in data analysis. A.P. and V.E. helped with experimental design and data interpretation. S.K.S. supervised the project, interpreted data and wrote the manuscript. All authors read and approved the manuscript.

Corresponding author

Correspondence to Shiv K. Singh.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Peer review information Nature Cancer thanks Francisco X. Real and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended Data Fig. 1 Genetic and orthotopic PDAC models recapitulate human pancreatic tumors.

a, Gene set over-representation analysis of genes enriched in squamous (n = 25 patients) versus pancreatic progenitor tumors (n = 30 patients) of the Bailey dataset7. Data was retrieved and differential analysis performed using R2 (Methods). b, Representative H&E staining of KPC mouse tumors. Scale bar: 20 µm. G1, n = 3; G2, n = 3; G3, n = 3; G4, n = 2 mice. c,d, Representative H&E staining of orthotopically implanted classical (CAPAN1, 2) (c) and basal-like (PANC1, MiaPaCa2) (d) cell lines in NMRI-Foxn1nu/nu mice. Images are shown for the primary tumor (top) and liver and/or liver metastases. Scale bar: 20 µm. c, CAPAN1, n = 6; CAPAN2, n = 6 mice. d, PANC1, n = 8; MiaPaCa2, n = 8 mice. e-h, Representative tile image scans of H&E staining (e,f,h) as well as IF staining for TNFα (g) in primary resected human PDAC tissue of well (e), moderately (f,g) and poorly (h) differentiated tumors, graded by expert pathologists. e, n = 2. f,g, n = 3. h, n = 6. e-h, n refers to number of patients. g, I-III designates respective higher-magnification images. Scale bar: representative area (e,f,h, top g), 200 µm; higher magnification (bottom g), 10 µm.

Extended Data Fig. 2 Exogenous TNFα activates EMT and inflammatory lineage programing.

a-c, Representative IF staining for keratin 81 (KRT81) and α-amylase (AMYL) (a), CD45 (b), as well as KRT81, AMYL and CD45 together (c) in TNFα- and vehicle control (VC)-treated mice bearing orthotopic CAPAN1 tumors in NMRI-Foxn1nu/nu mice. Scale bar: 50 µm. a, TNFα, n = 5; VC, n = 7. b, TNFα, n = 6; VC, n = 5. c, TNFα, n = 3; VC, n = 3. a-c, n refers to number of mice. d, PCA plot of RNA-seq performed on CAPAN1 cells treated for 18 h with TNFα (T1-3) and or vehicle control (C1-3). n = 3 independent experiments. e, Clustered heatmap of top 500 differentially expressed genes, based on RNA-seq performed on CAPAN1 cells treated for 18 h with TNFα (T1-3) or VC (C1-3). Cell color indicates z-score. f, Normalized enrichment scores of selected Hallmark gene sets of the Molecular Signatures Database between CAPAN1 cells treated for 18 h with TNFα or VC. g, Heatmap of genes previously linked to an activated stroma phenotype in PDAC13. Cell color indicates z-score. h, Network plot for the gene sets “epithelial-mesenchymal transition” and “inflammatory response”, indicating the involved genes, their fold change following TNFα treatment, and the size of the gene set. i, WB for indicated targets in CAPAN2 cells transduced with either empty vector (EV) or TNFα overexpression (TNFα-OE) constructs. Representative of n = 3 independent experiments.

Source data

Extended Data Fig. 3 Immune cell recruitment in orthotopic PDAC models.

a, Flow cytometry analysis of CLA (CAPAN1, 2) and BL (PANC1, MiaPaCa2) cells for TNFα expression. Flow cytometry was performed using an anti-TNFα antibody, without any staining (UN) as well as with isotype control (IgG1κ). Counts of gated cells are shown against fluorescence intensity, which correlates to TNFα expression. n = 3 independent experiments. b, Concentration of TNFα in the supernatant of cultured CLA (CAPAN1, 2) and BL (PANC1, MiaPaCa2) cells, measured by human-specific cytokine ELISA. n = 2 independent experiments. No TNFα above the detection limit could be measured in all cell lines. c, RNAscope for murine-specific and human-specific TNFα in orthotopically implanted BL (PANC1) and CLA (CAPAN1) cell lines in NMRI-Foxn1nu/nu mice. Arrows indicate exemplary positive cells. Scale bar: large area (left), 40 µm; higher-magnification insert (right), 10 µm. n = 3 mice per group. d, Representative IF staining for CD45 in orthotopically implanted CLA (CAPAN1, 2) and BL (PANC1, MiaPaCa2) cell lines in NMRI-Foxn1nu/nu mice. Scale bar: 50 µm. CAPAN1, n = 3; CAPAN2, n = 3; PANC1, n = 6; MiaPaCa2, n = 3 mice. e, Representative H&E as well as IF staining for CD68 in TNFα- and vehicle control (VC)-treated mice bearing orthotopic CAPAN1 tumors in NMRI-Foxn1nu/nu mice. Scale bar: 50 µm. For H&E, TNFα, n = 8; VC, n = 5 mice. For IF, TNFα, n = 3; VC, n = 3 mice.

Extended Data Fig. 4 Single-cell RNA-seq reveals macrophages as source of TNFα in PDAC mouse models.

a, Heatmap of the classification of late KIC tumor single cell data as epithelial or macrophage identity. Column color indicates the classification score per cell. b-d, Violin plots showing cluster-specific expression of TNFα (b), CD68 (c), and KRT19 (d) in clusters of late KIC tumors. Overlay boxplots show 25th to 75th percentile with median as box and highest and lowest data point in 1.5 times interquartile range as whiskers. Number of cells in each cluster are indicated. e, tSNE plot visualization of clusters of late KPC tumor sample. Clusters identified as macrophages (clusters 0, 2, 5) or epithelial cells (cluster 3), along with the fraction of total cells identified as either group are indicated. f-h, Feature plots of late KPC tumor sample with expression of TNFα (f), CD68 (g) and KRT19 (h). Macrophage and epithelial cell clusters are indicated as in e. As in a, for late KPC tumor sample. j-l, As in b-d, for the late KPC tumor sample.

Extended Data Fig. 5 Mutually exclusive expressions of cJUN and JUNB in PDAC tumors.

a, Cytokine array performed on supernatant of CLA (CAPAN2) and BL (PT45P1) cell lines. b, Representative bright-field images of cultured PDX-derived cell lines GCDX62 and GCDX103. Morphology was monitored over several passages. c,d, Trans-well invasion assay of GCDX62 and GCDX103, showing representative DAPI staining of invaded cells (c) as well as quantification thereof (d). b,c, Scale bar: 100 µm. d, Data given as average counts per F.o.V., with means ± s.d. Unpaired, two-tailed Student’s t-test. n = 3 independent experiments. e,f, Representative IF staining showing mutually exclusive expression of cJUN and JUNB in orthotopically implanted CLA (CAPAN2) and BL (PANC1) tumors (e), as well as in well or moderately (W/M) and poorly differentiated KPC tumors (f), quantified in g,h. Scale bar: 50 µm. e, CLA, n = 3; BL, n = 3 mice. g,h, Quantification for cJUN+ (g) or JUNB+ (h) cells of f. Data given as count per F.o.V., with means ± s.d. Two-tailed Mann-Whitney test. g, 50 F.o.V.’s were evaluated from n = 5 mice each. h, 62 (W/M) and 48 (poorly) F.o.V.’s were evaluated from n = 5 (W/M) and n = 4 (poorly) mice. i, Representative IF staining for ECAD with either cJUN or JUNB in moderately or poorly differentiated areas of primary PDAC tissue derived from 46 patients. Scale bar: 50 µm.

Source data

Extended Data Fig. 6 cJUN and JUNB regulate lineage-specific programs to maintain subtype identity.

a, Representative ATAC-seq and ChIP-seq coverage plots from CLA (CAPAN1, 2) and BL (PANC1, MiaPaCa2) cell lines in JUNB and CXCL5 loci. b, Correlation between ATAC-seq replicates. c, Venn diagrams showing replicate overlap of ATAC-seq peaks from CLA (CAPAN1, 2) and BL (PANC1, MiaPaCa2) cell lines. d, Annotation of ATAC-seq peaks in CLA (CAPAN1, 2) and BL (PANC1, MiaPaCa2) cell lines. UTR: untranslated region. e, Correlation matrix of ChIP-seq data for cJUN and JUNB in PANC1 and CAPAN1, respectively. Boxplots show 25th to 75th percentile with median as box and highest and lowest data point in 1.5 times interquartile range as whiskers. f, As in b, but for cJUN/JUNB replicates. g, As in c, but for cJUN/JUNB ChIP-seq peaks. h, As in d, but for ChIP-seq peaks from all replicates. i,j, Coverage plots of ATAC-seq and ChIP-seq for H3K27ac and H3K4me1 in CAPAN2 (blue) and PANC1 (black), as well as JUNB ChIP-seq in CAPAN1 (i) or cJUN in PANC1 (j), at the CDH1 (i) and VIM (j) loci. For H3K27ac and H3K4me1 public datasets were used14 (GSE64560; Methods). Promoter (Pro) and enhancer (En) regions used for ChIP-qPCR (k-p) are indicated. k-p, ChIP-qPCR in CAPAN1 (k,l), PANC1 (m,n) and CAPAN1 cJUN-overexpressing cells (OE; o,p) for indicated regions in i,j, for JUNB (k) or cJUN (m,o) and H3K27ac (l,n,p) pulldown. Data given as percentage of input, with means ± s.d and average isotype control in red. n = 3 independent experiments. q, Reporter assay for a fused VIM promoter-enhancer luciferase construct in CAPAN2 cJUN-OE and empty vector (EV) control cells. The En2 and promoter regions were used (Methods). Firefly luciferase luminescence is standardized to Renilla control. Data given as relative changes to EV, with means ± s.d. Two-tailed Mann-Whitney test. n = 5 independent experiments.

Source data

Extended Data Fig. 7 Overexpression of cJUN induces BL phenotype in CLA cells.

a,b, Expression correlation analysis in 78 PDAC patient tumors12 (E-MTAB-6134) with high tumor cellularity between cJUN and VIM (a) as well as GATA6 and VIM (b). RMA-normalized probe intensity values were plotted. A linear regression with 95% CI is shown in orange. Pearson’s correlation (R) and corresponding two-tailed P value are indicated. c, Representative bright-field images of GCDX62 cells transduced with empty vector (EV) or cJUN overexpression (cJUN-OE) constructs. Morphology was monitored over several passages. d-f, RNA-seq analysis was performed on GCDX62 cells transduced with EV or cJUN-OE. n = 3 independent cultures. d, PCA plot. e,f, Enrichment plots for gene set enrichment analysis between cJUN-OE and EV samples for ‘classical’ and ‘quasi-mesenchymal’ PDAC13 (e), as well as the top genes up- and downregulated following TNFα treatment in CLA (CAPAN1) cells (f). g, WB for indicated targets in CAPAN1 cells transduced with EV or cJUN-OE. Representative of n = 3 independent experiments. h, Representative bright-field images of CAPAN1 cells transduced with EV or cJUN-OE. Morphology was monitored over several passages. c,h, Scale bar: 200 µm. i,j, Trans-well invasion assay of CAPAN1 cells transduced with EV or cJUN-OE, showing representative DAPI staining of invaded cells (i) as well as quantification thereof (j). i, Scale bar: 100 µm. j, Data given as average counts per F.o.V., with means ± s.d. Unpaired, two-tailed Student’s t-test. n = 3 independent experiments. k-m, Mean cell viability ± s.d. at different concentrations of gemcitabine (k), oxaliplatin (l) and SN38 (m) in CAPAN1 cells transduced with EV or cJUN-OE. IC50 values for each drug are indicated. n = 3 independent experiments.

Source data

Extended Data Fig. 8 TNFα-mediated BL tumor immune plasticity is driven by cJUN.

a,b, Expression correlation analysis in 78 PDAC patient tumors with high tumor cellularity between cJUN and CCL2 (a), 309 patients12 (E-MTAB-6134) between cJUN and CD68 (b). RMA-normalized probe intensity values plotted. Pearson’s correlation (R) and corresponding two-tailed P value indicated. Linear regression with 95% CI shown in orange. c, WB using PANC1 cells transfected with siCtrl or cJUN-siRNAs (sicJUN #1/#2). sicJUN #1 used for all other experiments. d, qRT-PCR for cJUN and CCL2 in PANC1 cells transfected with siCtrl/sicJUN. Means ± s.d. Unpaired, two-tailed Student’s t-test. n = 3. e-g, WB using GCDX57 transfected with siCtrl/sicJUN (e), CAPAN2 EV or cJUN-OE cells (f), and PDX-derived primary PDAC cells (g). h, IF staining in orthotopically implanted CLA (CAPAN1, 2) and BL (PANC1, MiaPaCa2) tumors, quantified for double-positive cells in i. Scale bar: 50 µm. i, Data as average counts per F.o.V. per animal, with means ± s.d. CLA, n = 5; BL, n = 6 mice. j, IF staining in W/M and poorly differentiated KPC tumors, quantified for cJUN-CCL2 (k), and CD45-CCL2 (l) double-positive cells. Inserts, higher magnification. Scale bar: 50 µm. k,l, Data as counts per F.o.V., with means ± s.d. 50 F.o.V.’s evaluated from n = 5 mice each. For i,k,l, two-tailed Mann-Whitney test. m, WB using CAPAN1, CAPAN2 and GCDX62 cells treated with TNFα (10 ng/μL). n, Representative IF staining as indicated in TNFα- and vehicle control (VC)-treated mice bearing orthotopic CAPAN1 tumors. Scale bar: 50 µm. TNFα, n = 3; VC, n = 3 mice. o, Flow cytometry analysis of CAPAN1 and CAPAN2 EV or TNFα-overexpressing (TNFα-OE) cells. Counts of gated cells shown against fluorescence intensity, which correlates to TNFα expression. n = 3. p,q, WB using CAPAN1 EV or TNFα-OE cells (p) or additionally transfected with siCtrl or sicJUN (q). For all WB analyses, representative of n = 3. d,o-q, n, independent experiments.

Source data

Extended Data Fig. 9 Pharmacological inhibition of the BRD4-cJUN axis reduces macrophage invasion.

a, Treatment model and Kaplan-Meier survival plot of orthotopically implanted KPC cells into syngeneic C57Bl6 mice (KPCbl6). After tumor development, mice were treated with anti-TNFα antibody (n = 7 mice) or IgG (n = 6 mice). Median survival (ms) was not significantly (ns) altered (log-rank test). b, H&E staining of a, quantified for well differentiated area in c. Scale bar: 200 µm. c, Data as percentage of total tumor area/animal, means ± s.d. IgG, n = 5; anti-TNFα n = 4 mice. d, IF staining of a for CD45, quantified in e. Scale bar: 50 µm. e, Data as average values/animal of counts/F.o.V., means ± s.d. IgG, n = 5; anti-TNFα, n = 4 mice. For c,e, two-tailed Mann-Whitney test. f, Expression correlation analysis in 78 PDAC patient tumors12 with high tumor cellularity between cJUN and BRD4. RMA-normalized probe intensity values plotted. Pearson’s correlation (R) and corresponding two-tailed P value indicated. Linear regression with 95% CI shown in orange. g, WB using PANC1 cells, transfected with siCtrl or BRD4-targeting siRNAs (siBRD4 #1/#2). siBRD4 #1 used for all other experiments. h, WB using nuclear (NL) and cytosolic (Cyt) fractions of PANC1 cells transfected with siCtrl/siBRD4. g,h, Representative of n = 3. i, qRT-PCR for cJUN and CCL2 in PANC1 and MiaPaCa2 cells, treated with 500 nM JQ1 or VC for 24 h. Means indicated. n = 2. j, Macrophage (MΦ) differentiation and co-culture invasion assay. k-n, Trans-well co-culture invasion assay of M1/M2-differentiated THP-1 macrophages with PANC1 (k,l) or TNFα-OE CAPAN2 cells (m,n) seeded into wells, treated as indicated. k,m, DAPI staining of invading macrophages, quantified in l and n, respectively. Scale bar: 100 µm. l,n, Data as average counts per F.o.V., means ± s.d. k,l, n = 3. m,n, n = 4. For l,n, unpaired, two-tailed Student’s t-test. o-r, qRT-PCR for macrophage differentiation markers in undifferentiated (M0), M1 and M2-differentiated THP-1 macrophages. Means indicated. n = 2. g-i,k-r, n, independent experiments.

Source data

Supplementary information

Reporting Summary

Supplementary Tables 1–4

Supplementary Data 1

Flow cytometry gating strategy applied for TNF-α expression in cell lines. Supplementary Data 2 Gating strategy used for cell sorting of PDAC patient biopsies.

Source data

Source Data Fig. 1

Statistical source data for Fig. 1.

Source Data Fig. 2

Statistical source data for Fig. 2.

Source Data Fig. 2

Unprocessed western blots for Fig. 2.

Source Data Fig. 3

Statistical source data for Fig. 3.

Source Data Fig. 4

Statistical source data for Fig. 4.

Source Data Fig. 4

Unprocessed western blots for Fig. 4.

Source Data Fig. 5

Unprocessed western blots for Fig. 5.

Source Data Fig. 6

Statistical source data for Fig. 6.

Source Data Fig. 6

Unprocessed western blots for Fig. 6.

Source Data Fig. 7

Statistical source data for Fig. 7.

Source Data Fig. 8

Statistical source data for Fig. 8.

Source Data Fig. 8

Unprocessed western blots for Fig. 8.

Source Data Extended Data Fig. 2

Unprocessed western blots for Extended Data Fig. 2.

Source Data Extended Data Fig. 5

Statistical source data for Extended Data Fig. 5.

Source Data Extended Data Fig. 5

Unprocessed western blots for Extended Data Fig. 5.

Source Data Extended Data Fig. 6

Statistical source data for Extended Data Fig. 6.

Source Data Extended Data Fig. 7

Statistical source data for Extended Data Fig. 7.

Source Data Extended Data Fig. 7

Unprocessed western blots for Extended Data Fig. 7.

Source Data Extended Data Fig. 8

Statistical source data for Extended Data Fig. 8.

Source Data Extended Data Fig. 8

Unprocessed western blots for Extended Data Fig. 8.

Source Data Extended Data Fig. 9

Statistical source data for Extended Data Fig. 9.

Source Data Extended Data Fig. 9

Unprocessed western blots for Extended Data Fig. 9.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Tu, M., Klein, L., Espinet, E. et al. TNF-α-producing macrophages determine subtype identity and prognosis via AP1 enhancer reprogramming in pancreatic cancer. Nat Cancer 2, 1185–1203 (2021). https://doi.org/10.1038/s43018-021-00258-w

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/s43018-021-00258-w

Search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing