Oncogenic KRAS Drives Immune Suppression in Colorectal Cancer

In this issue of Cancer Cell, Liao et al. demonstrate that oncogenic KRAS drives an immune suppressive program in colorectal cancer by repressing IRF2 expression, which leads to downregulation of interferon responsive genes, enhanced expression of CXCL3 and recruitment of suppressive myeloid cells, and subsequent resistance to immune checkpoint blockade. In this issue of Cancer Cell, Liao et al. demonstrate that oncogenic KRAS drives an immune suppressive program in colorectal cancer by repressing IRF2 expression, which leads to downregulation of interferon responsive genes, enhanced expression of CXCL3 and recruitment of suppressive myeloid cells, and subsequent resistance to immune checkpoint blockade. The 5-year survival rate for patients with metastatic colorectal cancer (CRC) is a mere 12%. Immune checkpoint blockade with anti-programmed death (PD)-1 has recently been approved for use in patients with microsatellite instability-high (MSI-H) CRC, with an objective response rate of around 40%; however, efficacy is absent in microsatellite-stable (MSS) disease, which represents the majority of CRCs (Overman et al., 2018Overman M.J. Ernstoff M.S. Morse M.A. Where we stand with immunotherapy in colorectal cancer: deficient mismatch repair, proficient mismatch repair, and toxicity management.Am. Soc. Clin. Oncol. Educ. Book. 2018; : 239-247Crossref PubMed Scopus (74) Google Scholar). In addition to inactivating mutations in APC and TRP53, oncogenic KRAS mutations (usually G12D) are frequent drivers of aggressive disease in MSS CRC (Overman et al., 2018Overman M.J. Ernstoff M.S. Morse M.A. Where we stand with immunotherapy in colorectal cancer: deficient mismatch repair, proficient mismatch repair, and toxicity management.Am. Soc. Clin. Oncol. Educ. Book. 2018; : 239-247Crossref PubMed Scopus (74) Google Scholar). Although KRAS mutations do not predict response to anti-PD-1, they are associated with reduced T cell infiltration, hinting that downstream pathways could regulate immunotherapy and may become relevant as additional combinations of immune checkpoint inhibitors enter the clinic. In this issue of Cancer Cell, Liao et al., 2019Liao W. Overman M.J. Boutin A.T. Shang X. Zhao D. Dey P. Li J. Wang G. Lan Z. Li J. et al.KRAS-IRF2 axis drives immune suppression and immune therapy resistance in colorectal cancer.Cancer Cell. 2019; 35 (this issue): 559-572Abstract Full Text Full Text PDF PubMed Scopus (250) Google Scholar demonstrate that KrasG12D suppresses expression of interferon regulatory factor 2 (IRF2), a negative regulator of the CXCL3 chemokine expression (Figure 1). This leads to enhanced recruitment of myeloid-derived suppressor cells (MDSCs) and restricted T cell accumulation, a resistance mechanism that limits response to anti-PD-1 therapy. To explore the role of KRASG12D in CRC, the authors utilize an advanced murine model employing tamoxifen-inducible cre recombinase under control of the Villin promoter, floxed alleles of Apc and Trp53, and doxycycline-inducible expression of KrasG12D (designated iKAP). This model has a mesenchymal phenotype and activated TGF-β signaling (Boutin et al., 2017Boutin A.T. Liao W.-T. Wang M. Hwang S.S. Karpinets T.V. Cheung H. Chu G.C. Jiang S. Hu J. Chang K. et al.Oncogenic Kras drives invasion and maintains metastases in colorectal cancer.Genes Dev. 2017; 31: 370-382Crossref PubMed Scopus (101) Google Scholar) and reflects MSS disease of the CMS4 subtype (Guinney et al., 2015Guinney J. Dienstmann R. Wang X. de Reyniès A. Schlicker A. Soneson C. Marisa L. Roepman P. Nyamundanda G. Angelino P. et al.The consensus molecular subtypes of colorectal cancer.Nat. Med. 2015; 21: 1350-1356Crossref PubMed Scopus (2688) Google Scholar). To determine whether the model also recapitulated the poor T cell infiltration seen in KRASG12D human tumors, the authors performed cytometry by time of flight (CyTOF) to characterize the immune landscape. As expected, KrasG12D tumors displayed a reduction in CD4+ and CD8+ T cell infiltration, as compared to tumors harboring only null alleles of Apc and Trp53 (designated iAP). The authors also noted a substantial increase in CD11b+Ly6G+ polymorphonuclear (PMN) cells with the capacity to block T cell proliferation in a classic immunosuppressive assay, marking these as PMN-MDSCs. Critically, this infiltration was not a byproduct of tumor development in the presence of KrasG12D, as removal of doxycycline reversed this immune phenotype. Thus, expression of oncogenic KRAS drives the creation of an immunosuppressive tumor microenvironment in CRC. Ingenuity pathway analysis of RNA-seq profiles revealed a strong correlation between KrasG12D expression and downregulation of an interferon (IFN) response. This was true in both whole tumors and in cell lines derived from the murine model, suggesting an intrinsic ability of KRASG12D expression to suppress expression of IFN responsive genes. The authors then compared significant changes in gene expression in their murine data with genomic deletions in human CRC that were mutually exclusive with KRAS mutations, and integrating these datasets led to a single gene of interest, IRF2. To determine whether reduced IRF2 levels in KrasG12D tumors was a driver of the observed phenotype, IRF2 was overexpressed in iKAP cell lines, which reversed the IFN signature, including genes related to immunity and antigen presentation. While a reduced IFN response might explain lower T cell infiltration, it did not explain increased recruitment of PMN-MDSC in iKAP tumors. The authors therefore evaluated migration of MDSCs in response to conditioned media from iKAP cell lines, compared to those overexpressing IRF2. Conversely, IRF2 was knocked down in the KRAS wild-type MC38 cell line. In each case, expression of IRF2 was inversely associated with MDSC migration. Gene expression for cytokines and chemokines was then compared with IRF2 ChIP-seq, leading to the identification of CXCL3. IRF2 suppression of Cxcl3 was confirmed by real-time RT-PCR, and Cxcl3 was expressed at higher levels than other neutrophil chemoattractants. Significantly decreased in vitro migration was also found following neutralization of CXCL3 or inhibition of its receptor CXCR2 with the inhibitor SX-682 (Lu et al., 2017Lu X. Horner J.W. Paul E. Shang X. Troncoso P. Deng P. Jiang S. Chang Q. Spring D.J. Sharma P. et al.Effective combinatorial immunotherapy for castration-resistant prostate cancer.Nature. 2017; 543: 728-732Crossref PubMed Scopus (333) Google Scholar). Most importantly, administration of SX-682 to iKAP, but not iAP mice, dramatically reduced PMN-MDSC recruitment, while increasing recruitment of T cells into tumors, which was similar to the phenotype observed following MDSC depletion. Does PMN-MDSC recruitment lead to changes in tumor growth? iKAP tumors were resistant to anti-PD-1 therapy, and SX-682 alone was without efficacy; however, the combination more than doubled survival over the course of the experiment. This efficacy correlated with increased CD8+ T cell and reduced regulatory T cell (Treg) infiltration. The authors also found that expression of KRASG12D in the MC38 cell line led to anti-PD-1 resistance, a phenotype that was reversed by IRF2 overexpression or treatment with SX-682. Cumulatively, these data support the authors' model in which oncogenic KRAS inhibits expression of IRF2, leading to production of CXCL3, recruitment of CXCR2+ PMN-MDSCs, and suppression of the anti-tumor T cell response. Finally, the authors provide strong correlative data that the identified KRAS-IRF2-IFN axis is relevant in human CRC. This includes KRAS mutant tumors showing lower expression of IRF2, as well as a significantly diminished IFN gene signature. IRF2 expression in KRAS wild-type and mutant patient samples was heterogeneous, which could be due to spatial or expression level heterogeneity of KRAS or indicative of additional regulatory mechanisms. While this limits the utility of KRAS mutational status as a predictive biomarker in immunotherapy, it hints that IRF2 could prove more informative. Impressively, Liao et al. evaluated IRF2 expression in pre-treatment biopsies of 14 MSI CRCs and found that 5/7 patients with low IRF2-expressing CRC were non-responders to anti-PD-1 while 7/7 patients with high IRF2-expressing CRC had a complete response, partial response, or stable disease. Despite the small sample size, this supports further exploration of IRF2 as a biomarker and lends credence to this pathway being an important regulator of the immune system in CRC. Whether this occurs through CXCL3 expression and recruitment of MDSCs in patients requires further validation, as the correlation of CXCL3 with IRF2 expression or KRAS mutation status was relatively weak. However, putative PMN-MDSCs are observed in human CRC (Condamine et al., 2016Condamine T. Dominguez G.A. Youn J.-I. Kossenkov A.V. Mony S. Alicea-Torres K. Tcyganov E. Hashimoto A. Nefedova Y. Lin C. et al.Lectin-type oxidized LDL receptor-1 distinguishes population of human polymorphonuclear myeloid-derived suppressor cells in cancer patients.Sci. Immunol. 2016; 1: aaf8943Crossref PubMed Scopus (435) Google Scholar) and appear to regulate response to immune checkpoint blockade in other cancers (Davis et al., 2017Davis R.J. Moore E.C. Clavijo P.E. Friedman J. Cash H. Chen Z. Silvin C. Van Waes C. Allen C. Anti-PD-L1 efficacy can be enhanced by inhibition of myeloid-derived suppressor cells with a selective inhibitor of PI3Kδ/γ.Cancer Res. 2017; 77: 2607-2619Crossref PubMed Scopus (140) Google Scholar, Weber and Fottner, 2018Weber M.M. Fottner C. Immune checkpoint inhibitors in the treatment of patients with neuroendocrine neoplasia.Oncol. Res. Treat. 2018; 41: 306-312Crossref PubMed Scopus (63) Google Scholar) With the success of immunotherapy, the fields of cancer biology and cancer immunology are finally becoming integrated, and the impact of proto-oncogene and tumor suppressor mutations on the microenvironment and immune response is now being elucidated (Wellenstein and de Visser, 2018Wellenstein M.D. de Visser K.E. Cancer-cell-intrinsic mechanisms shaping the tumor immune landscape.Immunity. 2018; 48: 399-416Abstract Full Text Full Text PDF PubMed Scopus (322) Google Scholar). Oncogenic KRAS has been found to promote MDSC recruitment via induced expression of granulocyte-macrophage colony stimulating factor (GM-CSF) and CXCL1/2 in murine pancreatic and lung cancer, respectively, and to even promote PD-L1 expression in lung cancer (Wellenstein and de Visser, 2018Wellenstein M.D. de Visser K.E. Cancer-cell-intrinsic mechanisms shaping the tumor immune landscape.Immunity. 2018; 48: 399-416Abstract Full Text Full Text PDF PubMed Scopus (322) Google Scholar). Only increased PD-L1 expression was observed in this study, and whether this heterogeneity in resistance mechanisms is due to tissue- or mutation-specific factors remains unknown. It will certainly be interesting to determine whether the KRAS-IRF2 axis translates into other cancer types driven by oncogenic KRAS. Liao et al., 2019Liao W. Overman M.J. Boutin A.T. Shang X. Zhao D. Dey P. Li J. Wang G. Lan Z. Li J. et al.KRAS-IRF2 axis drives immune suppression and immune therapy resistance in colorectal cancer.Cancer Cell. 2019; 35 (this issue): 559-572Abstract Full Text Full Text PDF PubMed Scopus (250) Google Scholar observe that inhibiting PMN-MDSC recruitment led to an opposing increase in Treg recruitment, identifying an adaptive resistance mechanism that has also been problematic with other myeloid targeted therapies (DeNardo and Ruffell, 2019DeNardo D.G. Ruffell B. Macrophages as regulators of tumour immunity and immunotherapy.Nat. Rev. Immunol. 2019; Crossref PubMed Scopus (837) Google Scholar). Fortunately, Treg recruitment was reversed by anti-PD-1 therapy in this model, raising hope for ongoing clinical trials combining SX-692 and anti-PD-1 in melanoma (NCT03161431). Nevertheless, targeting CXCR2 bypasses activation of the IFN response in tumor cells, which could potentially limit therapeutic efficacy. Determining the mechanism by which oncogenic KRAS suppresses IRF2 expression could therefore open up new avenues for therapeutic intervention. K.H. is supported by a Swiss National Science Foundation postdoc mobility fellowship (P400PM_183881). B.R. has a courtesy faculty appointment at the University of South Florida, Tampa, FL 33620. KRAS-IRF2 Axis Drives Immune Suppression and Immune Therapy Resistance in Colorectal CancerLiao et al.Cancer CellMarch 21, 2019In BriefLiao et al. show that oncogenic KRAS represses IRF2 expression leading to high expression of CXCL3, which binds CXCR2 on MDSCs to promote their migration into the tumor microenvironment. Enforced IRF2 expression or CXCR2 blockade overcomes resistance of tumors expressing oncogenic KRAS to anti-PD-1 therapy. Full-Text PDF Open Archive