Immunogenic cell death in colon cancer prevention and therapy

Abstract

Colorectal cancer (CRC) is a leading cause of cancer-related death worldwide. The colonic mucosa constitutes a critical barrier and a major site of immune regulation. The immune system plays important roles in cancer development and treatment, and immune activation caused by chronic infection or inflammation is well-known to increase cancer risk. During tumor development, neoplastic cells continuously interact with and shape the tumor microenvironment (TME) which becomes progressively immunosuppressive. The clinical success of immune checkpoint blockade therapies (ICBs) is limited to a small set of CRCs with high tumor mutational load and tumor infiltrating T cells. Induction of immunogenic cell death (ICD), a type of cell death eliciting an immune response, can therefore help break the immunosuppressive TME, engage the innate components and prime T cell-mediated adaptive immunity for long-term tumor control. In this review, we discuss the current understanding of ICD induced by antineoplastic agents, the influence of driver mutations, and recent developments to harness ICD in colon cancer. Mechanism-guided combinations of ICD-inducing agents with immunotherapy and actionable biomarkers will likely offer more tailored and durable benefits to colon cancer patients.

1. Introduction

Colorectal cancer (CRC) is a major cancer-related killer worldwide, with an increased incidence projected in many developing countries. The survival rate of metastatic CRC remains low at around 11% [1]. Most sporadic CRCs develop from benign pre-neoplastic polyp-like lesions following somatic inactivation of the APC tumor suppressor and progress with a series of genetic and epigenetic alterations accumulated over years or decades in KRAS/BRAF, SMAD/TGFBR2, p53, and PIKCA3 among others [2]. Existing research has highlighted critical interactions between the immune system and emerging tumor cells, and the complex changes in the tumor microenvironment (TME) during cancer progression towards immunosuppression and loss of immunosurveillance [3]. Cancer cell-intrinsic mechanisms are believed to play a key role in shaping local immune landscapes and therapeutic responses [4]. The breakthrough in immune checkpoint inhibitors (ICBs), such as antibodies against Programmed Cell Death-1 (PD-1), Programmed Death Ligand 1(PD-L1), and Cytotoxic T Lymphocyte associated Antigen 4 (CTLA-4) [5,6], supports immune normalization or restoration in tumor control [7]. However, T-cell activating therapies have limited success in solid tumors. Induction of immunogenic cell death (ICD), a type of cell death eliciting an immune response, is, therefore, an attractive approach to break the immunosuppressive TME and re-establish immune surveillance by engaging both the innate and adaptive components [8,9]. ICD can be induced by a variety of chemopreventive agents, radiation, chemotherapeutics, and targeted therapies, and is shown to be critical for the antitumor effects in preclinical models through a systemic antitumor immunity [10]. A better mechanistic understanding of drug-induced cell death and immunologic consequences, particularly the influence of oncogenic signals, will likely hold the key to help devise more efficacious and precise strategies to prevent and treat colon cancer. This review will discuss recent findings that can help advance the research and clinical development of ICD in this direction.

2. Changing immune landscapes during CRC development

Immunogenomic analysis of over 10,000 tumors across 33 diverse cancer types using data compiled in The Cancer Genome Atlas (TCGA) revealed that tumor immune landscapes differ greatly between and within cancer types [4,11]. Driver mutations can dictate the immune contexture of tumors, and their bidirectional interactions shape tumor evolution and therapeutic response [4,11]. The accumulation of genetic and epigenetic changes in cancer cells reflects clonal expansion [12] driven by selective pressure including that from immune cells and neutral drift intrinsic to normal intestinal stem cell replacement [13,14]. CRC progression is associated with changes in the composition, density, and location of tumor-infiltrating lymphocytes (TILs) that predict clinical outcomes. CRCs heavily infiltrated with TILs and natural killer cells (NKs) have a more favorable prognosis, and the infiltration of CD3+, CD8+, or CD45RO memory T cells is a strong prognostic factor [15,16]. Tumor-suppressive role of immunosurveillance has also been established in mice. Immunodeficient IFN-γ−/−, RAG2−/− or Prkdc^scid (better known as severe combined immune deficiency, or SCID) mice are more susceptible to chemically-induced and spontaneous tumorigenesis [17], or intestinal polyposis driven by APC loss [18].

Immunosurveillance is gradually diminished during malignant transformation, from tumor cell elimination, equilibrium, and finally to escape accompanied by ongoing immune editing [3,19,20]. The escape phase is associated with poor prognosis and characterized by diminished tumor-infiltrating cytotoxic T lymphocytes (CTLs), type I helper (Th1) CD4+ T lymphocytes, and NKs, and an increased presence of immunosuppressive regulatory T (T_reg) and myeloid cells producing TGF-β and interleukin-10 in the TME [21]. Escaping from immunosurveillance occurs on tumor cells due to the loss of tumor antigens, the major histocompatibility complex (MHC) proteins, sensitivity to the complement system, T cell or NK cell lysis, and production of immunosuppressive molecules such as TGF-β, indolamine-2,3-dioxygenase (IDO), PD-L1, PD-L2 or CTLA-4 [5,6], making tumor cells a poor target for an immune attack [3].

3. Immunogenic cell death of cancer cells

ICD is a type of cell death that primes a systemic immune response [8]. ICD is essential for the host defense against viral and bacterial infections and can be induced by various anticancer agents in a context-dependent manner [9]. The role and regulation of ICD in cancer therapy remains an active area of investigation.

Death by many ways

Cell death is an irreversible fate, highly regulated and stimulus-specific [22]. Apoptosis is an evolutionarily conserved form of cell death important for normal development in the animal kingdom. Apoptosis is regulated by mitochondria-dependent intrinsic and death receptor-dependent extrinsic pathways, converging on the activation of executioner caspases-3 and −7 [23,24]. During transformation, neoplastic cells frequently become resistant to apoptosis via genetic and epigenetic mechanisms, driving the accumulation of additional oncogenic events and therapeutic resistance [25,26].

Several forms of nonapoptotic cell death have been discovered, including necroptosis, the most extensively studied in inflammation and cancer [22,27]. Necroptosis can be initiated upon activation of the extended TNF-α receptor family on the cell surface, and propagated through the receptor-interacting serine/threonine kinases, RIP1 and/or RIP3 [28–30], particularly when apoptosis is compromised due to blocked activation of caspase-8 or other caspases [31,32]. The cell death is then executed by plasma membrane pores formed by Mixed Lineage Kinase Domain-Like protein (MLKL) [33], and amplified by autocrine TNF-α production [34,35]. Intracellular pathogens, lipid oxidation, loss of attachment, increased reactive oxygen species, PARP-1 hyperactivation mediated-bioenergetic failure, or prolonged mitotic arrest can trigger pyroptosis, ferroptosis, anoikis, autophagic cell death, parthanatos, and mitotic death, respectively [22].

Immunogenic cell death

ICD is associated with several hallmarks in dying cells, such as calreticulin (CRT or CLAR) cell-surface translocation, extracellular release of damage-associated molecular patterns (DAMPs) such as HMGB1, heat shock proteins, or ATP, and the production of Type I interferons (IFNs) among others [36–38]. DAMPs attract innate immune cells such as neutrophils, macrophages, dendritic cells (DCs), and NKs through various pattern recognition receptors (PRRs), and promote their maturation and/or activation, including dead cell clearance, optimal antigen uptake, processing and presentation, cytokine production, and cell killing within days [9]. Cross-priming of CD8+ CTLs is triggered by mature DCs and γδT cells in an IL-1β- and IL-17-dependent manner. Primed CTLs then elicit a direct cytotoxic response to kill tumor cells through the generation of IFN-γ, perforin-1, and granzyme B, followed by memory (Figure 1). T-cell mediated immunity is regulated by a constant and complex interplay between stimulatory and inhibitory signals to promote antigen-specific adaptive response while avoiding autoimmunity [9].

Figure 1. Heating up “cold” tumors with the right kill.

MSI and MSS CRCs display different immune TME and response to ICBs while sharing mutations in several major drivers. Anticancer drugs often induce mixed types of cell death and immunological outcomes, which are strongly influenced by tumor intrinsic factors (genetic and epigenetic) such as MSI, mutant APC, KRAS/BRAF or p53, or silenced RIP1/RIP3 as detailed in the text. ICD breaks the “cold’ TME, engages innate response to prime adaptive response required for long-term tumor control. The “right” kill is expected to be dominated by pro-inflammatory (red) effectors (ICD markers and cytokines), and cells (i.e., CD8+ T cells) and over suppressive (blue) effectors and cells (i.e., T_reg and MDSCs).

Calreticulin plasma membrane exposure

CRT is a 46-kDa, Ca²⁺-binding, endoplasmic reticulum (ER) chaperone. CRT translocation to the outer leaflet of the plasma membrane is considered an “eat-me” signal, occurs during ER stress-associated cell death and is mediated by endocytosis and the protein complex SNAP Receptor (SNARE) [38–40]. ER stress is an arm of the evolutionarily conserved and multipronged integrated stress response (ISR) [41], characterized by the phosphorylation of eukaryotic translation initiation factor 2α (eIF2α) by PKR-like ER kinase (PERK) in response to unfolded proteins. Exposed CRT is recognized and engulfed by cells expressing CD91, also known as LDL Receptor Related Protein 1 (LRP1) (e.g., DCs and macrophages). CRT binding to CD91 triggers a cascade of events to facilitate the recruitment of antigen-presenting cells (e.g., DCs), optimal antigen presentation, release of pro-inflammatory cytokines (e.g., TNF-α and IL-6) and activation of type 17 helper T (Th17) cells [42].

Extra Cellular DAMP Release

HMGB1 is an abundant DNA minor groove binding protein, and its release from dying cells triggers a strong inflammatory response [43,44]. HMGB1 binds to several PRRs including Toll-like receptor 4, (TLR4), receptor for advanced glycation end products (RAGE) to activate MAPKs (p38 and ERK1/2), and NF-κB in DCs. HMGB1 also facilitates DC maturation, migration, presentation of tumor-associated antigens to T cells, and production of pro-inflammatory cytokines. Release of chaperones such as HSP70 and HSP90 from dying cells can also stimulate tumor antigen uptake, DC maturation, and function [44].

ATP secretion from dying tumor cells triggers an immune response and requires autophagy [45,46]. Extracellular ATP from dying tumor cells generates a strong “find-me” signal for DCs and macrophages, upon its binding to purinergic receptor P2Y2 on the target cells [47], to promote DC maturation and macrophage expansion [47]. In fact, increased ATP hydrolysis by ectonucleotidases (e.g., CD39 and CD73) in the TME [48,49] blunted TILs and tumor response to chemotherapy [50].

Type 1 IFN response

Virtually all cells can secrete Type 1 interferons (IFNs) upon viral or bacterial infections and the activation of nucleic acid (RNA and DNA) sensors such as TLR3, TLR7, TLR9, RIG1/MDA5-MAVs (Retinoic Acid Inducible Gene 1 Protein/Melanoma Differentiation-Associated protein 5 Mitochondrial antiviral-signaling protein), and cGAS-STING (cyclic GMP-AMP Synthase -Stimulator of interferon genes) [51]. The binding of Type 1 IFNs (alpha and beta) to IFNAR1–IFNAR2 heterodimers activates JAK/STAT signaling to establish an antiviral state, from inhibiting viral production, activating antigen presentation and innate immunity to kill and remove infected cells, and ultimately to mounting adaptive immunity against future infections [51,52]. Type 1 IFN response regulates highly cell-type specific expression of numerous Interferon Stimulated Genes (ISGs), and controls the production and function of numerous chemokines, cytokines, and immune cells via crosstalk with TNF-α/NF-κB and IL-18/IL-1β pathways. The host and immune cell-dependent roles of Type 1 IFN response in controlling infection and tumor growth are well-established [51,53], while its tumor-intrinsic role is less understood.

ICD defined in vivo

The type of cell death can influence immunologic outcomes in animals. Apoptosis is considered to be less immunogenic due to the lack of ICD markers and direct suppression of necroptosis, TNF-α/NFκB, or type 1 IFN signaling as a result of caspase-mediated cleavage of RIP1 [31,32] or cytoplasmic RNA/DNA sensors [54]. However, no single type of cell death or marker defines ICD, and mixed types of cell death and shared regulators or effectors are often observed [22]. As such, tumor cell ICD is defined by two major criteria in immunocompetent mice [55,56]. First, tumor cells upon ICD induction in vitro without any adjuvant function as a vaccine to protect mice against a subsequent challenge with live tumor cells of the same type. Second, antitumor immunity in vivo is associated with the infiltration of immune effector cells into the TME and is dependent on the host immune system and CD8+ cells.

4. ICD-inducing agents in CRC prevention and therapy

Emerging evidence supports that the clinical activity of most, if not all, conventional and targeted antineoplastic agents currently used in humans can be attributed to the re-establishment of immune surveillance and reactivation in the TME [10]. These agents can induce ICD and alter TILs abundance and composition, which is associated with more favorable therapeutic responses or prognosis in cancer patients [9]. However, a detailed mechanistic understanding of ICD in relation to often highly variable in vivo efficacy is lacking. The use of more physiological and immune-competent models, such as APC^Min mice and MC-38 and CT-26 syngeneic models, will help explain how preventive and therapeutic agents work under substantially different oncogenic signals and the TME [57,58]. We will focus our discussion on CRC relevant agents (Table 1).

Table 1.

Drug-induced ICD in colon cancer.

ICD inducer	Model	Mechanism and *Biomarker	References
Prevention
NSAIDs (Sulindac, Aspirin)	Human, CRC cells, APCMin mice	apoptosis, ER stress, IFN and IL-8 signaling, increased TILs, decreased Tregs and PEG2	[62–70]
Metformin	Human, CT-26 syngeneic model	apoptosis, increased MHC-1 expression and T-cell functions	[74–77]
Radiation and chemotherapy
Radiation	CRC patients, CT-26 and MC-38 syngeneic models	cell death, the abscopal effect, modulated peptide repertoire, enhanced MHC-1 expression and host type 1 IFN response	[78–80]
Chemotherapy (5-FU, Irinotecan, Oxaliplatin, Doxorubicin)	CRC cells, CT-26 and MC-38 syngeneic models	cell death, increased surface CRT, MHC-1 expression, and TILs, especially in combination	[37,40,86,87]
Targeted therapies
VEGF antibodies (Bevacizumab)	mCRC patients	Increased B-cell and T-cell compartments	[92]
EGFR antibody (Cetuximab)	mCRC patients, CRC cells, and CT-26	Th1-cytotoxic phenotype, CTL-dependent, DC phagocytosis, *mutant KRAS/BRAF for exclusion	[93–95]
Regorafenib	mCRC patients, CRC cells, xenografts and CT-26 syngeneic model	apoptosis, enhanced anti-tumor immunity via macrophage modulation	[96–98]
PI3K inhibitor	CT-26 syngeneic model	enhanced activity of effector CD8+ T-cells	[99]
Emerging agents
Inhibitors of proteasome, HSP, GRP78, mTOR and BET	CRC cells, xenografts, 3D spheroids	cell death, ER stress, single agent or in combination	[101–108]
Natural compounds, epigenetic drugs and oncolytic viruses	CRC cells, CT-26 syngeneic	cell death, increased CD8+ T cells, CRT exposure and HMGB1 release, and IFN signaling	[87,109–114]
Immune checkpoint inhibitor
Anti-PD1	mCRC patients	T-cell activation in the TME, *MSI, high TMB, pretreatment TILs for inclusion	[118,119]

CRC preventive agents

A variety of pharmacological agents, natural products, and dietary components have been shown to have CRC preventive activity in human studies and animal models. The most potent and best studied are non-steroidal anti-inflammatory drugs (NSAIDs) [59–61]. NSAIDs induce apoptosis selectively in APC-deficient and Myc-high preneoplastic stem and cancer cells via ER stress, involving the crosstalk of the death receptor and mitochondrial apoptotic pathways [62–65]. The efficacy of NSAID in human CRC patients is associated with single nucleotide polymorphisms (SNPs) in interferon and interleukin-8 (IL-8) signaling [66–68], as well as increased active TILs and reduced immunosuppressive T_reg cells [69] or cytokines [70] in the TME, and increased MHC expression in tumor cells [71]. In APC^Min mice, NSAIDs promoted the antitumor M1 polarization of macrophages [72] and inhibited COX2-mediated production of prostaglandin E2(PGE2) and immunosuppressive Myeloid-Derived Suppressor Cells (MDSCs), T_reg and TH2 cells [73]. The anti-diabetic drug metformin reduced CRC risk in epidemiological and preclinical studies [74], which is associated with growth suppression, apoptosis induction, and improved MHC1 expression and T-cell functions [75–77].

Radiation and chemotherapy

Radiation therapy and most conventional chemotherapeutic agents trigger DNA damage ultimately apoptotic and non-apoptotic cell death [23]. Besides direct cytotoxicity to tumor cells caused by non-reparable double-stranded DNA breaks, the abscopal effect of radiation therapy has been well-recognized with systemic antitumor immunity capable of shrinking established tumors that have not been subjected to treatment. Radiation can modulate the peptide repertoire, enhance MHC class I expression and response to immunotherapy in syngeneic tumor models such as MC-38 and CT-26 [78–80]. cGAS-STING-dependent Type 1 IFN response in the host is important in these preclinical models. IFN therapy (Type 1 and 2), however, has not been efficacious in solid tumors.

5-fluorouracil (5-FU) is a pyrimidine analog that inhibits thymidylate synthase required for nucleotide synthesis [81,82], and is standard of care for CRC patients. The platinum drug Oxaliplatin or the topoisomerase I inhibitor irinotecan is often given in combination with 5-FU as FOLFOX (fluorouracil, leucovorin, and oxaliplatin) or FOLFIRI (fluorouracil, leucovorin, and irinotecan) for improved efficacy [83]. 5-FU or irinotecan treatment induced ICD hallmarks CRT exposure and MHC-1 expression in the mouse colon cancer cell line CT-26 and human cancer cells [40]. Genetic ablation of Caspase-3 enhanced RIP1 dependent necroptosis and response to 5-FU in HCT 116 cells and xenografts [84]. Interestingly, agents with similar chemical structures may have very different immunological consequences [85]. Oxaliplatin, but not cisplatin, acts as a potent ICD inducer in CRC syngeneic murine models by triggering ER stress and CRT exposure [37,86]. A combination of cytotoxic agents such as doxorubicin [87] or 5-FU and radiation [88] can potentiate ICD, CD8+ CTLs and therapeutic responses in mice.

Targeted therapy

Several targeted drugs are approved by the US Food and Drug Administration (FDA) for CRC treatment, including the anti-vascular endothelial growth factor (VEGF) antibody bevacizumab, the anti-EGFR antibodies cetuximab and panitumumab, and the multi-kinase inhibitor regorafenib [89,90]. These agents inhibit RAS/RAF/MEK/ERK and PI3K/AKT signaling and can lead to transient remissions but rarely cures, as feedback activation of survival pathways and suppression of apoptosis leads to a drug-tolerant state and resistance [91]. Bevacizumab-based first-line therapy increased T and B cell compartments in metastatic CRC (mCRC) patients [83,92]. Cetuximab with chemotherapy increased circulating DCs, NKs, memory T cells, and cancer-specific CTLs in mCRC patients, and promoted ER stress and phagocytosis by DCs in vitro [93–95]. Regorafenib promoted apoptosis and NF-κB activation in human CRC models [96,97], and enhanced anti-tumor immunity in mCRC patients [98]. PI3Kα/δ inhibition was shown to promote anti-tumor immunity through direct enhancement of effector CD8(+) T-cell activity in the CT-26 syngeneic model [99].

Emerging therapeutic agents

ER stress inducers including FDA-approved cancer drugs can trigger ICD. Bortezomib is a specific inhibitor of the 26S proteasome subunit and induces apoptosis of various cancer cells, associated with enhanced phagocytosis and DC cross-presentation of tumor antigens to T cells [100]. In preclinical CRC models, bortezomib, HSP inhibitors, mTOR/PI3K inhibitors, BET inhibitors or a GRP78 inhibitor induced ER stress-associated cell death [101–106]. Interestingly, several drug combinations were particularly effective in killing CRCs harboring mutant KRAS/BRAF or SPOP through elevated ER stress and DR5 expression [107,108]. Various natural compounds, epigenetic drugs, and oncolytic viruses can also induce or enhance ICD [109–111]. Anticancer efficacy observed was associated with increased CD8+ T cells [87], tumor cell CRT exposure and HMGB1 release [112], or activation of endogenous retroviruses and IFN signaling [113,114].

5. ICD regulation by cancer intrinsic mechanisms in CRC

Therapeutic responses to anticancer agents and induction of ICD hallmarks are highly variable, lacking clinically actionable biomarkers (Table 1). Studies on cancer-immune interactions reveal that virtually every step from innate to adaptive immunity can be altered in the TME, which differs significantly among cancers and cancer types [3]. A transcriptomic meta-analysis defined five major CRC subtypes with distinct mutational, growth, stroma, and immune profiles [85,115], supporting tumor-intrinsic mechanisms in shaping disease heterogeneity, local immune landscapes and therapeutic responses [4]. We will focus on CRC-relevant ICD mechanisms in the context of differential therapeutic responses and “Hot” and “Cold” tumors (Figure 1).

Mismatch Repair Deficiency

Approximately 15% of CRCs display microsatellite instability (MSI) due to mutation or promoter methylation in DNA mismatch repair (MMR) genes MSH2, MLH1, MSH6, and PMS2. MSI is characterized by highly elevated DNA replication errors in microsatellite repeats in both noncoding and coding sequences [2]. MSI CRCs have high tumor mutational burden (TMB), more T-cell inflamed or “Hot” TME [116,117], and respond favorably to ICBs such as anti-PD-1 [118,119]. However, MSI CRCs respond poorly to 5-FU-based chemotherapy [120]. Nearly all microsatellite stable (MSS) CRCs and approximately 30–50% of MSI CRCs do not respond to ICBs, and relatively few tumor-specific antigens have been confirmed [121,122]. Resistance mechanisms in MSI tumors appear to be complex, most notably mutational inactivation of MHC-1 and IFN regulators such as B2M, NLRC5, JAK1/2 (over 60%) [123–125], Transforming Growth Factor β Receptor 2 (TGFBR2), PTEN, Asteroid Homolog 1 (ASTE1), Caspase 5 [117], and BAX [62]. ICB resistance is associated with the lack of pretreatment T cells or “inflamed” CD8+ T cell and DC gene signature [126,127].

Wnt/Myc

Hyperactivation of Wnt/β-catenin and oncogenic Myc is an early and near-universal event in CRCs [2] and is found to be strongly associated with immunosuppressive TME, absence of TILs, and ICB resistance across cancer types [128,129]. Wnt/β-catenin/Myc signaling impairs a multitude of immune functions in the TME by reducing MHC-1 expression [130] and NKG2D ligand in tumor cells [131], as well as activation and expansion of cytotoxic T-cells [132] through inhibitory signals and metabolic deprivation [133,134]. Wnt/Myc also upregulates T cell inhibitory PD-L1, as well as “don’t-eat-me” signals CD47 [135] and CD24 [136], whose binding to their respective receptors signal-regulatory protein α (SIRPα), and Siglec-10 on immune cells triggers phagocytosis checkpoints and protects cancer cells from immunosurveillance [137]. CD47 was also upregulated by TNF/NF-κB [138] and hypoxia-inducible factor 1 (HIF-1) [139]. High mRNA expression of CD47 [140] and CD24 [136] is associated with poor survival in cancer patients. Myc can inhibit IFNβ production via direct binding to the STAT1 promoter [141].

Mutant KRAS/BRAF

KRAS and BRAF mutations are mutually exclusive and found in over 50% of CRCs. Hot spots include codons 12, 13, 61, 146 in KRAS, and codon 600 in BRAF, leading to constitutive RAF/MAPK and PI3K signaling [2]. Mutant KRAS/BRAF is a significant resistance mechanism in CRCs [142] in response to front-line therapy, targeted therapies against EGFR [143–146], mutant BRAF [147], and mTOR [102], which activates compensatory survival pathways and blocks apoptosis. Mutant KRAS was recently shown to drive resistance to immune therapy through IRF2 loss- and CXCL3/CXCR2-dependent recruitment of MDSCs [148]. Mutant KRAS also cooperated with Myc to promote CLL9 and IL-23-dependent inflammation and stroma immunosuppression [149] and is associated with more frequent JAK1/2 or B2M mutations [124]. Mutant BRAF was reported to inhibit antigen presentation and T cell infiltration in melanoma [150]. These findings suggest defective ICD as a potential therapeutic target in CRCs harboring mutant KRAS/BRAF.

Compromised cell death, genome integrity, and immune recognition

Widespread mutational inactivation of p53 and activation of RAS/RAF and PI3K/AKT pathways in CRC impair DNA damage repair [151], removal of cells with chromosomal instability (CIN) [152], and apoptosis, and leads to therapeutic resistance (reviewed by [91]). Mutant FBW7 was recently shown to cause resistance to targeted therapies in CRC preclinical models by blocking drug-induced Mcl-1 phosphorylation and its subsequent degradation or dissociation from BH3-only proteins [97,153,154]. Epigenetic silencing of RIP1 or RIP3 blocks necroptosis induced by anticancer agents including 5-FU in CRC cells (reviewed by [155]).

Nearly all MSS CRCs display CIN at early stages [2,156,157], with little or no response to ICBs even in combination trials with conventional or targeted therapies [158]. CIN has been shown to activate the cGAS-STING pathway to promote genomic heterogeneity and inflammatory signaling [159]. Compromised Type 1 IFN response can lead to resistance to radiation, anthracycline-based chemotherapy and immune therapy [121,160–162]. Besides rare JAK1/2 mutations [123–125], the cause appears to be mostly non-genetic, such as promoter methylation and reduced expression of cGAS or STING [163] and loss of INFAR1 protein expression [164]. Elevated T-cell inhibitory signals such as CTLA-4, T cell immunoreceptor with Ig and ITIM domains (TIGIT), PD-1, TIM-3 are also attributable to epigenetic mechanisms such as loss of DNA methylation or suppressive histone marks in CRC [165].

Harnessing ICD via combination therapies

Cancer immunotherapy comes in different forms, including immune effectors such as antibodies, cytokines, activated T cells, NK cells that act directly on tumor cells, or those designed to educate or activate (e.g. vaccines) the immune response in the patient [3]. ICD bridges innate and adaptive immunity and therefore can potentiate ICBs and chimeric antigen receptor T cell therapy [166]. Anti-CD47 in combination with other therapeutic agents enhanced cell-surface CRT and reduced T_reg and MDSCs [167]. STING [162] and TLR [168] agonists have entered early phase clinical testing. The intestinal barrier is a major site of epithelial, immune, and microbiota interactions. The gut microbiota regulates mucosal immunity, colon cancer development, therapeutic responses, and likely influences ICD through tumor and immune interactions including heterogeneity [169,170]. The use of syngeneic models, human-derived models, and ongoing clinical trials will likely help guide the development of biomarkers to facilitate patient-tailored therapy.

6. Conclusions

Emerging evidence supports the establishment of antitumor immunity in the clinical success of antineoplastic agents. Tumor intrinsic mechanisms play a critical role in shaping the immune TME, tumor development and therapeutic responses. Drug-triggered ICD represents a promising approach to restore the immunogenicity in “cold” tumors by engaging both the innate and adaptive components to prime the TME for a greater antitumor response (Figure 1). A better understanding of this process will likely help translate ICD into the clinic. Some critical questions await answers. Cell death, stress response, and immunological outcomes are interconnected and highly cell type- and stimulus-specific. It remains to be determined if and how specific types of cell death, ER stress or other arms of ISR regulate immunogenicity in the context of CRC driver mutations and subtypes. The hallmarks of ICD and immune effectors seem numerous. However, key molecular determinants in the activation of innate immunity upon cell death required to prime adaptive anticancer immunity remain to be elucidated. Lastly, actionable ICD biomarkers will be critical to help guide treatment decisions.