Autocrine signaling of type 1 interferons in successful anticancer chemotherapy

Abstract

Anthracycline-based chemotherapies are particularly effective if they succeed in reinstating immunosurveillance by the induction of immunogenic cell death (ICD) in the tumor. Recently, we discovered that ICD is coupled to the induction of type 1 interferons (IFNs-I) that act in an autocrine fashion on cancer cells, thereby increasing their immunogenic potential.

Abbreviations

ATP

adenosine triphosphate

BC

breast cancer

CCL2

chemokine (C-C motif) ligand 2

CRT

calreticulin

CXCL10

(C-X-C) motif chemokine ligand 10

DC

dendritic cell

ER

endoplasmic reticulum

HMGB1

high mobility group box 1

ICD

immunogenic cell death

IFN

interferon

IFNAR

IFNα/β receptor

ISG

IFN-I-stimulated genes

MX1

myxovirus resistance 1

NLRP3

NLR family pyrin domain containing 3

P2RX7

P2X purinoceptor 7

PRR

pathogen recognition receptor

TLR

toll like receptor

UPR

unfolded protein response.

To achieve a long-term control of tumor growth, cancer therapies must not only reduce the tumor mass but also stimulate a protective anticancer immune response.¹ One way to (re)instate immunosurveillance relies on the capacity of a selected array of anticancer agents (exemplified by anthracyclines, oxaliplatin and cyclophosphamide) to induce immunogenic cell death (ICD) in the tumor.^2,3 The rules governing ICD reside in a compendium of biochemical changes in dying cancer cells, including: (i) the pre-apoptotic unfolded protein response (UPR) leading to the translocation of calreticulin (CRT) from the endoplasmic reticulum (ER) lumen to the plasma membrane, where this protein serves as an eat-me signal facilitating the clearance of apoptotic tumor cells by dendritic cells (DCs); (ii) the pre-mortem autophagy-induced release of adenosine trisphosphate (ATP) into the microenvironment, where it acts as a chemotactic signal for inflammatory monocytes and DCs and, by binding to the purinergic receptor P2X ligand-gated ion channel 7 (P2RX7), triggers the activation of the NLR family, pyrin domain containing 3 (NLRP3) inflammasome platform;⁴ and (iii) the post-mortem exodus of nuclear high mobility group box 1 (HMGB1) in the extracellular space, where it binds to toll-like receptor 4 (TLR4) and hence signals for optimal tumor antigen presentation by DCs.^2,5 The engagement of NLRP3 and TLR4 leads to interleukin (IL)-1β secretion by DCs and T_H1/T_C1 cell polarization, both indispensable for the eradication of the tumor. Each of these 3 sequential events is required for chemotherapy to be perceived as immunogenic and hence to induce anticancer immune responses for durable tumor growth-inhibitory effects (Fig. 1).⁶

Figure 1.

Schematic outline of the characteristics of immunogenic cell death. Immunogenic cell death (ICD) inducers can act on cancer cells through parallel mechanisms: (i) the activation of cell-death associated molecular patterns (CDAMPs), which culminates in the exposure and release of molecular moieties by dying tumor cells and stimulates antigen phagocytosis and cross-presentation by dendritic cells (DCs); (ii) an autocrine damage-associate molecular pattern (DAMP), which is initiated by recognition of self ssRNA, single-stranded RNA (ssRNA) by endosomal pathogen recognition receptors (PRRs) and leads to type I interferon/ IFNα/β receptor (IFN/IFNAR) signaling in tumor cells. Altogether, these processes result in a potent anticancer immune response, which can eventually lead to the eradication of chemotherapy-resistant tumor cells. CTL, cytotoxic T lymphocyte; IKK, inhibitor of kappa B kinase; IPS-1, IFN beta promoter stimulator-1; IRF, IFN regulatory factor; MDA5, melanoma differentiation-associated gene 5; MHC, major hystocompatibility complex; RIG-I, retinoic acid-inducible gene-I; TAA, tumor-associated antigen; TBK, TANK-binding kinase; TCR, T-cell receptor; TRIF, TIR-domain-containing adaptor protein-inducing IFN beta.

Recently, we uncovered the fourth hallmark of ICD. When characterizing the effects of anthracycline-based chemotherapy on gene expression in tumors implanted in immunocompetent syngeneic mice, we observed that type 1 interferons (IFNs-I) as well as IFN-I-stimulated genes (ISGs) were massively transactivated. Comparisons of distinct cell fractions within the tumor revealed that tumoricidal dosages of anthracyclines directly trigger an early transcriptional program of ISGs specifically in cancer cells themselves, rather than in stromal cells or infiltrating leukocytes, both in vitro and in vivo.⁷ By mechanistic studies we found that IFNs-I were induced via a complex pathway that seems to involve the release of nucleic acids (and in particular RNA) from dying cells into the tumor environment, where they are sensed by the endosomal pathogen recognition receptor (PRR) TLR3⁸ on still viable cells. TLR3, in turn, signals the transcription of ISGs. At this stage, the precise molecular nature of the nucleic acid mediating TLR3 activation in this context is unknown. Although TLR3 is strictly required for the anthracycline-dependent induction of IFNs-I, whether other PRRs perceive additional danger-associated molecular patterns emanating from dying cells remains elusive. Once released, IFNs-I act on the common IFN-I receptor, [a heterodimeric (IFNAR1/IFNAR2) surface complex] and trigger an autocrine IFN-I signature. Among the most bioactive ISGs, the chemokine (C-X-C motif) ligand 10 (CXCL10) acts as an essential chemotactic factor for the recruitment of immune effectors that selectively attack the tumor (Fig. 1).⁷

Tumors lacking any of the elements of this cascade (such as virus-relevant PRRs or IFNAR) grew similarly to control wild type (WT) cells upon inoculation into the flank of immunocompetent syngeneic mice. However, these tumors did not respond to anthracyclines unless the missing end-products, IFNs-I or CXCL10, were exogenously supplied. On the contrary, experiments performed on Ifnar^-/- mice indicated a dispensable role of IFN-I signaling for the tumoricidal activity of chemotherapy. Corroborating the cancer autocrine effect of TLR3>IFNs-I>IFNAR signaling, the silencing of this cascade abolished the immunogenicity of anthracycline-mediated tumor cell death in prophylaxis experiments.⁷ Altogether, these results indicate that immunogenic chemotherapy phenocopies viral infection leading to an autocrine molecular pathway, in which TLR3 acts upstream of IFNAR (and hence of IFNs-I) and IFNAR in turn acts upstream of CXCL10 (Fig. 1). Based on these data we surmise that “viral mimicry” by chemotherapy is indispensable to elicit a protective anticancer immune response.⁷

Type I and type II IFNs are both required for the anticancer immune response elicited by anthracyclines in vivo. However, it should be noted that IFNs-I and IFNs-II contribute to the anticancer immune response upon chemotherapy in a rather distinct fashion. Major differences concern the kinetic of type I and type II IFN production, their targets, and the nature of their source, the former being produced in an autocrine feed forward loop by stressed cancer cells, the latter being released by host T cell effectors.^2,7 Of note IFNs-I can induce the production of chemokine (C-C motif) ligand 2 (CCL2), which – together with its receptor chemokine (C-C motif) receptor 2 (CCR2) – we recently reported to be necessary for optimal antineoplastic efficacy of anthracyclines.⁹ Further experiments are required to elucidate the precise link between IFNs-I and CCL2.

Finally, we demonstrated that the IFN-I pathway is clinically relevant for the long-term benefit of breast cancer (BC) patients exposed to anthracyclines. By performing studies on cohorts of advanced BC patients treated with neoadjuvant or adjuvant anthracycline-based chemotherapy, we indeed found that the IFN-I metagene myxovirus resistance 1 (MX1) was a positive predictor of pathological complete response (pCR).¹⁰ Moreover, by immunohistochemistry (IHC) and Allred score calculation, we found that anthracyclines promote the expression of MX1 in BC tumors. Hence, MX1 expression at diagnosis might reflect the capacity of BC to respond to therapy by mounting a IFN-I response.⁷ These results highlight the translational importance of IFNs-I for anticancer chemotherapies. Moreover, they suggest that molecular defects residing in any of the main players of this “viral mimicry” might compromise the efficacy of distinct antineoplasic agents and hence deserve prospective validation.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Funding

GK and LZ are supported by the Ligue Nationale contre le Cancer (Equipes labellisées), SIRIC Socrates, ISREC Foundation, Agence Nationale pour la Recherche (ANR AUTOPH, ANR Emergence), European Commission (ArtForce), European Research Council Advanced Investigator Grant (to GK), Fondation pour la Recherche Médicale (FRM), Institut National du Cancer (INCa), Fondation de France, Cancéropôle Ile-de-France, Fondation Bettencourt-Schueller, the LabEx Immuno-Oncology, and the Paris Alliance of Cancer Research Institutes. AS is supported by the Associazione Italiana Ricerca contro il Cancro (AIRC). IV is supported by AIRC (MFAG 14641), the Italian Ministry of Health (RF_GR 2011-2012) and the Programma per i Giovani Ricercatori “Rita Levi Montalcini” 2011.