Ionizing radiation (IR) is a prevalent non-surgical intervention for achieving cures in cancer
[1]. Approximately 50% of cancers are treated with some form of IR
[2][3], and it is estimated that this contributes to 40% of cures
[4][5]. While the cytotoxic effects of IR on cancer were first appreciated in the early 20th century, following Röntgen’s discovery of X-rays in 1895
[6], the complex mechanisms by which the tumor microenvironment (TME) is altered following IR are still being elucidated today
[7][8][9].
Notably, the immune modulatory potential of IR has gained particular interest. Cancers exploit naturally occurring immunosuppressive mechanisms to facilitate their growth and metastasis, and many reports show that IR is capable of reprogramming the TME to alleviate this suppression and even achieve abscopal effects in rare cases
[10]. Following IR exposure, cancer cells undergo an immunogenic form of cell death which results in antigen presenting cell (APC) activation and maturation, leading to eventual cytotoxic T cell responses specific to the tumor
[11]. Interestingly, multiple reports now describe that an intact immune system is essential to the efficacy of IR
[12][13][14][15].
While the large body of evidence for the immune stimulatory effects of IR is certainly convincing, numerous reports also show that IR can induce a wound healing response characterized by tumor-associated macrophage (TAM), myeloid-derived suppressor cell (MDSC) and regulatory T cell (Treg) recruitment to tumors
[16][17]. These cells dampen anti-tumor immune responses and initiate angiogenic processes, augmenting tumor growth and expansion
[18][19]. Moreover, a higher number of these cells correlate with poorer prognosis in many cancers
[20][21]. Hence, a dichotomy exists with this therapy where both immune stimulatory and immunosuppressive effects occur, often simultaneously
[16].
2. Radiation Induced Immunogenic Cell Death
Immediately following radiation, DNA is damaged via direct ionization or indirectly through the action of free radicals (such as reactive oxygen species; ROS) generated when radiation ionizes water molecules or biomolecules in cells. Cells are composed of 80% water, and it has been estimated that the indirect effects of radiation account for 60% of the total cellular damage accumulated
[27]. This DNA damage occurs in the form of single-strand and double-strand breaks that are sensed by ataxia telangiectasia mutated (ATM) and ATM- and RAD3-related (ATR) kinases, which in turn activate DNA repair mechanisms in the cells. However, if the damage accumulated exceeds cellular repair capabilities, the cells will undergo cell death or senescence
[28].
2.1. Senescence and Cell Death
In the case of senescence, the cell terminates division processes through p53/p21 and p16/RB1 signaling
[29]. The cell remains viable but undergoes arrest to prevent further damage and outgrowth of mutated cells. For cell death, on the other hand, several death pathways can occur depending on the extent of the damage. Lower, recoverable levels of damage are associated with programmed mechanisms of death such as apoptosis and autophagy, whereas higher levels evoke irreparable damage resulting in necrosis
[30]. Typically, necrosis has been defined as immunogenic and apoptosis has been defined as immune silent; however, it is now known that all forms of cell death possess some level of immunogenicity
[31]. ICD broadly refers to a form of cell death that involves the release of damage-associated molecular patterns (DAMPs), chemokines, cytokines and tumor antigens that prime APCs and instigate adaptive immune responses (
Figure 1)
[11]. IR has been shown to induce these markers of ICD (calreticulin, high mobility group box 1 (HMGB1) and adenosine triphosphate (ATP)) in a dose-dependent manner (ranging from 0 to 100 Gy)
[32][33].
2.2. Damage-Associated Molecular Patterns
Following IR-induced cellular damage, DAMPs such as calreticulin, heat shock protein-70 (HSP70), HMGB1 and ATP, which are normally endogenous in nature, get released by dying cells to facilitate phagocytosis
[31][34]. Calreticulin and HSP70 are endoplasmic reticulum (ER) molecular chaperones that translocate to the cell membrane following cellular damage. Once expressed on the surface, they can bind to CD91 expressed on APCs and act as an ‘eat me’ signal for the dying cell
[35][36]. Additionally, the HSP family of proteins may also present carried tumor antigens to APCs directly
[37]. HMGB1 is a nuclear protein that governs chromosomal structure and function. Although its main function is to act as a DNA chaperone, it can be passively released during cell death and bind to toll-like receptors (mainly TLR4) on APCs to augment phagocytosis and antigen processing via MyD88 signaling
[38][39]. ATP is somewhat different, as it requires autophagy for its active release from dying cells
[40]. Autophagy is a stress response that initiates the molecular recycling of cellular organelles and cytoplasmic proteins under hostile conditions. Following extracellular release, ATP functions as a chemoattractant for dendritic cells (DC) by binding to purinergic receptors
[41][42]. Conversely, although DAMPs generation in the tumor is an important immune activator, studies assessing correlations between the levels and overall outcome have generated contested results
[43][44]. It is likely that elevating DAMPs alone is not enough to elicit measurable benefits to radiation, but many factors of the tumor microenvironment (TME) are critical when considering this response.
2.3. Cytokine Release
The release of cytokines following IR is a massive field of research, with a large volume of work being published on various cytokines including IFN-γ
[45], IL-1β
[46], TNF-α
[47], IL-6
[48] and TGF-β
[49]. Among these, the Type-I interferons (IFN) have proven particularly prominent in recent times. Type-I IFN is a family of pro-inflammatory cytokines heavily involved in the response to viral infection. They are also pivotal for anti-tumor immune responses following IR, primarily through DC cross-priming
[50][51][52]. IR induces Type-I IFN expression (IFN-α and IFN-β, in particular) through multiple endogenous nucleic acid sensing pathways following the release of double-stranded DNA and RNA into the cytosol of damaged cells
[53]. Cyclic GMP-AMP synthase (cGAS) is an enzyme that catalyzes the formation of cyclic GMP-AMP (cGAMP) when it recognizes cytosolic DNA. cGAMP then binds to and activates STING (stimulator of interferon genes) in the ER
[54][55]. STING is an adaptor molecule, and its activation leads to downstream Type-I IFN expression
[56]. Notably, STING also activates NF-κB signaling, inducing a vast array of proinflammatory cytokines such as TNF-α, IL-1β and IL-6
[57][58]. DNA and RNA are also capable of triggering Type-I IFN release through endosomal TLRs (TLR9 for DNA; TLR3, 7 and 8 for RNA) and their adaptor molecules MyD88 or TRIF
[59][60]. Cytosolic RNA can additionally be detected by RIG-I-like receptors (RLR) and induce IFN-β expression via the mitochondrial adaptor protein MAVS
[61]. Regardless of the pathway involved, however, multiple reports have demonstrated that Type-I IFN production is essential to IR efficacy. When this pathway is inhibited, tumor burden is similar to unirradiated controls, and it has been shown that CD8
+ T cell cytotoxic function in these tumors is impaired
[52][62][63][64].
2.4. Chemokine Release
IR leads to an influx of immune cells into the TME through chemokine gradients as a form of damage control. Chemokines are produced by multiple cells in the TME and bind to their cognate receptors in an autocrine or paracrine fashion
[65]. CCL2 is upregulated in tumors following IR, and this facilitates CCR2
+ monocyte and Treg recruitment to tumors
[66][67]. Recent research points towards STING-dependent expression of CCL2, suggesting this pathway is a double-edged sword in IR treatment
[68]. On one hand, there is Type-I IFN expression that fosters anti-tumor immunity; on the other, there is an influx of immunosuppressive monocytes and Tregs that evoke radioresistance and tumor growth
[69]. Another chemokine ligand produced downstream of the cGAS-STING pathway following IR is CCL5
[70]. CCL5 is also involved in trafficking monocytes to the tumor utilizing the CCR5 receptor, in particular
[70][71]. While the receptors for these chemokines are also expressed on T cells, higher levels of CCL5 expression are typically associated with poorer outcomes in cancer
[72]. This suggests that the immunosuppressive infiltrate dominates any potential anti-tumor immune responses driven by chemokine signaling following IR. Overall, the broad topic of chemokine expression following radiation is highly complex, influencing both pro- and anti-tumor immune cell types and varying based on the tumor, IR dose and time-point measured
[73][74][75]. It has been shown, however, that inhibiting CCR2/CCR5 signaling can sensitize tumors to IR
[71].
Figure 1. Immunogenic cell death elicited by IR. ICD induced by IR is marked by the release of DAMPs (HMGB1, double-stranded DNA and RNA, HSP70, ATP and calreticulin (CRT)), Type I IFN (IFN-α and IFN-β) through cGAS-STING signaling and other cytokines (TNF-α, IL-1ß and IL-6) and chemokines (CCL2 and CCL5) through STING-induced NF-κB activation. Broken lines indicate movement and continuous lines represent the production of the given target.