Inflammation after solid organ transplantation is a key contributor to post-transplant complications. Post-transplant inflammation is a unique form of tissue injury, marked by a rapid and robust inflammatory response with contributions from donor and recipient derived innate and adaptive immune cells. Early detection and mitigation of the inflammatory cascade improves allograft outcomes. Currently, common immunosuppressive regimens largely target recipient T cells; however, these medications are wrought with untoward effects such as end-organ injury, infection, and malignancy.

Recently, studies have shown that targeting innate immune cells can mitigate allograft rejection by modulating antigen presenting cells and downstream adaptive immune responses. To this end, our group has recently demonstrated that targeting donor macrophages prior to transplantation improves transplant outcomes and modifies the recipient innate and adaptive immune landscape [1], [2], [3], [4], [5], [6]. Strategies to capitalize on this mechanism include targeting immune cell activation, effector cell responses, and/or cell survival. Targeting upstream inflammatory cell death pathways that result in the activation of macrophages and other antigen presenting cells represents an attractive option. In this review, we will focus on the role of ferroptosis in the post-transplant inflammatory response.

Regulated cell death pathways include apoptosis and necrotic cell death pathways. Unlike apoptosis, necrotic cell death such as necroptosis, ferroptosis, and pyroptosis result in uncontrolled release of intracellular contents leading to necroinflammation [7]. Ferroptosis is an iron-dependent non-apoptotic cell death pathway characterized by oxidative stress and lipid peroxidation. First described in 2012 [8], ferroptosis was found to be regulated by iron, glutathione peroxidase 4 (GPX4), and the cystine/glutamate antiporter system Xc- [9]. Polyunsaturated fatty acids (PUFAs) in the phospholipid bilayers of cells can undergo lipid peroxidation. Free iron, via the Fenton reaction, promotes lipid peroxidation while GPX4 converts these toxic lipid peroxides into non-toxic lipid alcohols [10]. A small molecule called erastin was found to induce ferroptosis via inhibition of system Xc- (which is upstream of GPX4), while Ras-selective lethal small molecule 3 (RSL3) induces ferroptosis via GPX4 inhibition [9]. GPX4 inhibition or inactivation then results in excessive lipid peroxidation, accumulation of reactive oxygen species, and subsequent cell death [10], [11]. Ferroptotic cell death is highly immunogenic due to the release of damage-associated molecular patterns (DAMPs) and subsequent production of inflammatory cytokines and chemokines [12].

Ferroptosis is an essential mediator and therapeutic target in heart [13], lung [14], kidney [15], and liver [16] diseases, among many others [17]. From these studies in other disease processes, it appears that targeting ferroptosis reduces inflammation and improves organ function. This is clinically relevant given the availability of therapeutic agents that target components of the ferroptosis pathway.

Primary graft dysfunction (PGD) is a clinical syndrome resulting in allograft dysfunction immediately after and up to several days after transplantation [18]. PGD is common, occurring in 30% of heart [19], 36% of lung [20], [21], [22], 38% of liver [23], and 21% of kidney [24] transplants. Donor and recipient factors as well as peri-operative organ management contribute to PGD. PGD can be associated with poor short- and long-term outcomes and can predispose to future rejection episodes. Ischemia reperfusion injury (IRI) is a major factor contributing to PGD [25], [26]. IRI results from the composite of ischemic conditions associated with donor death, organ procurement and transportation, and organ reperfusion. Hypoxic tissue injury and metabolic stress during this time result in various forms of regulated cell death, which subsequently leads to the release of cell contents, including DAMPs and intracellular stores of iron. Together, these effects initiate an inflammatory cascade and ultimately PGD. While PGD has different clinical definitions across different solid organ transplants [19], [21], it is characterized by tissue damage, increased vascular permeability, interstitial edema, endothelial cell dysfunction, and infiltration of neutrophils, monocytes, and macrophages [27], [28], [29], [30], [31], [32], [33].

Ferroptosis has been extensively studied in the setting of IRI. Ferroptosis can mediate or potentiate IRI; as such, a number of pharmacologic inhibitors of ferroptosis have been investigated in IRI models [34], [35], [36], [37], [38], [39]. In a murine model of liver ischemia, inflicted by clamping across the portal triad for 60 minutes, there was a significant increase in liver transaminases (markers of liver injury), which was abrogated with administration of ferrostatin-1 (Fer-1), a ferroptosis inhibitor that scavenges peroxidized lipids [40]. In this model, other markers of ferroptosis, including prostaglandin endoperoxide synthase 2 (Ptgs2) and 4-hydroxynonenal (4-HNE), were also reduced with Fer-1 administration. Inflammatory cytokines and chemokines (e.g., IL-1β, IL-6, TNFα, CCL2) induced with liver ischemic injury were also reduced with Fer-1 treatment [40]. As ferroptosis is an iron-dependent form of cell death, this group also found that iron overload exacerbated ischemic injury. To this end, mice fed a high iron diet prior to undergoing liver ischemia experienced worsened injury [40]. These results are supported by clinical findings as a retrospective review of 202 liver transplants demonstrated that high donor serum ferritin levels are an independent risk factor for liver injury after transplant [40]. In another liver IRI model, worse liver injury correlated with increased expression of transmembrane member 16A (TMEM16A), which is a chloride channel thought to promote ferroptosis via ubiquitination-induced degradation of GPX4 [41]. Additionally, cold storage-induced stress during liver ischemia has been specifically shown to induce ferroptosis-mediated ASK1/p-38 kinase pathway activation [42]. Mitochondrial calcium uptake 1 (MICU1) is necessary for cold stress-induced ferroptosis, and MICU1 deficiency results in decreased cold stress-induced ASK1 activation [43]. Finally, ferroptosis inhibitors used during cold storage have been shown to decrease liver IRI [44], [45]. These studies provide a basis for further investigation into the use of anti-ferroptosis agents during donor organ cold preservation.

Iron metabolism is a particular problem in the setting of tissue injury, as cell death leads to the release of intracellular iron stores, which can then further aggravate ferroptotic cell death. Phagocytic cells, such as macrophages, are recruited to areas of injury, take up this newly released iron, and perpetuate this cycle of injury. To build upon this cycle of inflammation, iron overload can polarize macrophages toward an inflammatory phenotype, leading to fibrogenesis and steatohepatitis in a study of nonalcoholic fatty liver disease [46]. In the setting of IRI, iron overload in macrophages can exacerbate tissue injury, and ferrostatins and iron chelators have been shown to decrease this injury [47].

It is important to note that iron overload and ferroptosis may not be uniformly detrimental. There are promising early findings in the cancer literature that ferroptosis may mediate macrophage-induced anti-tumor effects. Specifically, nanoclusters loaded with iron oxides and TGF-β/PD-1 inhibitors were injected into the tumor microenvironment and induced pro-inflammatory macrophages, generation of reactive oxygen species, and ultimately ferroptosis of tumor cells specifically [48]. This introduces a novel therapeutic avenue that synergizes the effects of immunomodulatory drugs and ferroptosis. While ferroptosis of tumor cells is certainly desirable in the cancer setting, the underlying mechanisms of this pathway could be inhibited to reduce injury of the allograft in the setting of transplantation. Indeed, a better understanding of the regulation of iron metabolism by macrophages in the transplant setting is necessary.

Cells and tissues that die via an inflammatory form of death during IRI release DAMPs – such as ATP, HMGB1, S100s, HSP, Histone, F-actin [49] – that are bound by pattern recognition receptors (PRRs) – such as P2X7, RAGE, TLRs, DNGR1 – on innate and adaptive immune cells and/or stromal cells to activate inflammatory signaling [50]. Notably, inhibitors of ferroptosis have led to decreased HMGB1 release and improvements after tissue IRI [51]. Additionally, lipid metabolites from ferroptotic lipid peroxidation, such as prostaglandin E (PGE) and hydroxydodecanoid acids (HETE) can themselves serve as DAMPs and activate the immune response. DAMPs also bind complement (C3), leading to the formation of the membrane attack complex and perpetuating tissue injury. Longer ischemic times and donation after circulatory death (DCD) have been linked with higher complement activation and an increase in rates of PGD [26], [52]. Cells have multiple defense mechanisms against lipid peroxidation and ferroptosis, including GPX4, ferroptosis suppression protein 1 (FSP1), GTP cyclohydrolase (GCH1)/tetrahydrobiopterin (BH4), dihydroorotate dehydrogenase (DHODH), and amino acid oxidase interleukin-4-induced-1 (IL4i1) [11], [43], [53], [54], [55]. In severe IRI, some of these protective factors may be overcome and DAMP release could lead to downstream immune activation, resulting in a relentless cycle of injury (Figure 1).

Prior to discussing the early immune response following organ transplantation, it is worth highlighting the distinction and overlap between necroptosis and ferroptosis [56] (see Figure 2). Both necroptosis and ferroptosis are pro-inflammatory and immunogenic forms of necrotic cell death, resulting in DAMP release. Unlike ferroptosis, which is mediated by iron-dependent lipid peroxidation, necroptosis is induced by cell death receptors including TNF receptor 1 (TNFR1) and Toll-like receptor 4 (TLR4). TNFR1 signaling induces the formation of a necrosome comprised of RIPK1 and RIPK3, which phosphorylates MLKL and results in necroptosis of the cell. TLR4 signaling, in a TRIF-dependent manner, activates RIPK3 which subsequently phosphorylates MLKL and triggers necroptosis. The RIPK1 inhibitor necrostatin-1 (Nec-1) has been developed to inhibit necroptosis. In liver IRI models, Nec-1 protects Kupffer cells from necrotic cell death and attenuates tissue injury [57]. In heart, lung, and kidney transplant models, Nec-1 treatment reduces early neutrophil accumulation and allograft injury [58], [59]. Interestingly, Nec-1 also has an off-target effect of ferroptosis inhibition in a RIPK1- and IDO-dependent fashion [16], [59], [60]. Notably, Nec-1 administration reduced ferroptosis-induced tubular necrosis and improved survival in murine models of kidney IRI [59].

Our group has shown in a murine model of cardiac transplantation that ferroptotic cell death results in the release of DAMPs that signal through TLR4 on graft vascular endothelial cells to promote the recruitment of neutrophils [34]. When we administered Nec-1 to the recipient mice, there was decreased neutrophil infiltration into the cardiac graft. We then used RIPK3-deficient donors and found that neutrophil infiltration was not affected, suggesting that necroptosis of graft cells is not responsible for this phenotype [34]. Since Nec-1 has dual inhibitory effects on necroptosis and ferroptosis, we used pharmacological inhibition of ferroptosis with Fer-1, which resulted in reduced neutrophil infiltration [34]. Together, these results suggest that ferroptosis rather than necroptosis may be responsible for neutrophil infiltration following murine heart transplantation. We acknowledge that the phenotypic difference between genetic RIPK3 deletion and pharmacologic RIPK1 inhibition with Nec-1 may not be completely attributable to ferroptosis. Nevertheless, ferroptotic cell death plays a unique and critical role in neutrophil recruitment following cardiac transplantation and should be further studied as a means of decreasing the post-transplant immune response.

It is clear that early immune activation and recruitment to site of injury are critical mediators of alloimmune activation and long-term allograft outcomes. Neutrophils mediate ongoing tissue injury in part by regulating neutrophil extracellular traps (NETs) and amplifying the subsequent alloimmune response by promoting the recruitment of T cells [61], [62]. Specifically, blockade of neutrophil-mediated tissue damage during reperfusion via neutrophil depletion has been associated with decreased T cell-mediated rejection and improved survival of cardiac allografts [62]. As such, inhibition of ferroptosis and resultant decreased neutrophil infiltration in the early reperfusion period may have long-term beneficial effects of prolonging graft survival. Beyond neutrophils, recruitment of other myeloid cells such as macrophages and dendritic cells may also be affected by targeting ferroptosis. Further studies are needed to investigate the use of ferroptosis inhibitors for short- and long-term graft protection.

DAMP release from ferroptotic cell death also has downstream effects beyond the early post-reperfusion timepoint. The binding of DAMPs to PRRs on myeloid cells, such as macrophages, is key to the concept of trained immunity in organ transplantation, which can provide a link between early innate immune activation and later rejection episodes. Trained immunity can occur through metabolic and epigenetic changes of myeloid cells that cause them to become hyper-responsive upon re-stimulation [26], [63], [64]. HMGB1 is implicated in the long-term reprogramming of myeloid cells as it remains elevated for up to 8 weeks after cardiac transplantation and has been associated with the presence of inflammatory dendritic cells at this late timepoint [65].

Acute rejection is an early complication that affects a significant proportion of patients following solid organ transplantation, with an incidence of around 3-12% after kidney [66], 30% after liver [67], 5-20% after heart [68] and 50% after lung transplantation [69]. Episodes of acute rejection can predispose to future episodes of chronic rejection, which further threatens long-term allograft and patient survival. Acute rejection can be divided into two subtypes: cellular rejection and antibody-mediated rejection. Both forms of rejection result in allograft dysfunction that can vary in severity. Diagnosis generally involves obtaining a biopsy of the allograft with the subsequent classifications and gradations varying by organ. Diagnosis of antibody-mediated rejection can be additionally supported by complement (C3d/C4d) deposition. Treatment of acute rejection involves corticosteroid and cytolytic therapies that have many undesirable adverse effects including infection, metabolic derangements, and future malignancy. Furthermore, recurrent rejection episodes are common. Thus, the prevention of acute rejection is paramount and requires a better understanding of underlying mechanisms.

Among the different types of regulated cell death, necroptosis is known to mediate allograft rejection. Inhibition of necroptosis with Nec-1 and/or genetic RIPK3 deletion decreases rejection in murine models of kidney and heart transplantation [70], [71]. As previously discussed, Nec-1 also has off-target effects of ferroptosis inhibition [60], so the role of ferroptosis in these rejection models must be considered. Recent work has suggested that ferroptosis may be involved in acute T cell-mediated rejection. In early acute tubular injury after kidney transplantation, the pro-ferroptosis gene ACSL4 was robustly expressed, raising the possibility that ferroptosis plays a role in early allograft injury [72]. Furthermore, ferroptosis-related gene expression signatures have been developed and linked to T cell-mediated rejection in kidney transplantation. One such gene signature developed from human kidney transplant biopsies identified a few key genes including TLR4, CD44, and IFNγ as potential mediators of ferroptosis in early T cell-mediated rejection [73]. Another group developed a similar model of ferroptosis-related genes, including TLR4, to predict graft loss following kidney transplantation [74]. While these studies highlight the diagnostic and predictive utility of ferroptosis-related gene signatures, specific mechanistic studies into whether and how ferroptosis contributes to acute rejection are largely lacking at the present time.

As previously discussed, ferroptosis-triggered neutrophil infiltration at the time of graft reperfusion may predispose to subsequent T cell-mediated rejection [62]. Vice versa, activated T cells during episodes of acute cellular rejection may reciprocally induce ferroptosis of graft cells. One group has developed a model of acute rejection after murine allogeneic cardiac transplantation without immunosuppression, and found that the rejection phenotype is attenuated when recipient mice are deficient in tumor necrosis factor-α-induced protein-8 like 2 (TIPE2) [75]. TIPE2 is a regulatory molecule that is critical for the maintenance of immune homeostasis [76]. Mechanistically, these TIPE2-deficient mice had decreased alloreactive T cell activation and production of IFNγ, which resulted in decreased ferroptosis of graft cells and reduced acute allograft rejection [75]. IFNγ has also been shown to increase sensitivity to ferroptosis via inhibition of GPX4 [11], [77]. Thus, IFNγ release from activated T cells may amplify tissue injury during acute rejection by promoting ferroptosis in the allograft. Ferroptosis inhibition at the time of acute rejection episodes may help ameliorate graft injury by interrupting this cycle of amplification.

Adaptive immune cells themselves are also susceptible to ferroptosis. In the cancer setting, the tumor microenvironment is enriched in lipids, which are taken up by tumor-infiltrating CD8+ T cells. These CD8+ T cells undergo excessive lipid peroxidation and ferroptosis, and thus collectively have decreased effector function [78]. Just as GPX4 inhibits ferroptosis in cancer cells [79], [80], it has the same regulatory function in T cells. GPX4 overexpression in T cells decreases lipid peroxidation and restores anti-tumor CD8+ T cell effector function [78]. Further supporting the importance of GPX4 in T cell function, GPX4-deficient T cells had decreased abilities to protect from viral and parasitic infections and were more likely to undergo ferroptosis [81]. Similarly, within the B cell compartment, GPX4 has been shown to protect against ferroptosis by inhibiting lipid peroxidation specifically in B1 and marginal zone B cells [82]. Taken together, these studies show that B and T cell function and survival are regulated by components of the ferroptosis pathway.

Peripheral graft tolerance, the “holy grail” of transplant immunology, refers to allograft survival in the absence of ongoing immunosuppression, which can be reliably achieved in murine models [83]. Tolerance induction is dependent on apoptosis of activated T cells, as mice with defective T cell apoptosis were unable to develop graft tolerance following transplantation [84], [85]. Since T cell death appears important for allograft survival, the induction of ferroptosis in T cells could also be harnessed as a method of immunosuppression. Many innovations in the field of transplant immunology borrow from the cancer literature, and the development of T cell ferroptosis-directed immunosuppression is no exception. Capecitabine, a chemotherapeutic drug, exerts an immunosuppressive effect in a pre-clinical transplantation model [86]. Continuous, low-dose administration of capecitabine reduced acute rejection following liver transplantation in rats. It is believed that 5-fluorouracil (the end-metabolite of capecitabine) inhibits the Nrf2-HO-1/GPX4 antioxidant system, resulting in T cell ferroptosis [86].

When considering inducing T cell ferroptosis as a method of immunosuppression, an important caveat is that regulatory T cells (Tregs) should be excluded from the target population. For example, in murine heart and lung transplantation models, Tregs have been shown to be critical for inducing and maintaining allograft tolerance [87], [88]. Yet, Tregs are susceptible to ferroptosis – a recent report has demonstrated that GPX4 is critical for suppressing lipid peroxidation and ferroptosis in activated Tregs so that they can exert antitumor immunity [89]. Thus, for eventual clinical translation of ferroptosis-targeting therapeutics to organ transplantation, it would be essential to understand whether different T cell subtypes have differential susceptibility to ferroptosis. Furthermore, as ferroptosis is an inherently inflammatory process, its induction in T cells may result in additional inflammation. Thus, any potential ferroptosis-induced inflammation must be balanced against the benefit of targeting alloreactive T and B cells for acute rejection.

In addition to the induction of T and B cell ferroptosis directly, ferroptosis of other immune cells can be targeted for immunosuppression. Polymorphonuclear myeloid-derived suppressor cells (PMN-MDSCs) have suppressive functions in anti-tumor immunity [90]. One mediator of their immunosuppressive property is via ferroptosis. In tumors, PMN-MDSCs undergo ferroptosis which, despite reducing their abundance, generates an inflammatory milieu with release of oxygenated lipids that subsequently inhibit T cell activity [91]. Adapting this to organ transplantation, the induction of PMN-MDSC ferroptosis could potentially decrease allograft rejection through T cell inhibition. Although very much still preliminary, these studies provide the basis for further exploration of ferroptosis-targeting therapies that can be applied to organ transplantation. It is important to acknowledge that the induction or inhibition of ferroptosis to reduce allograft rejection is dependent on many variables, including cell type, time point, and environmental context. Certainly, more work needs to be done to better understand how innate and adaptive immune cell susceptibility to ferroptosis can be harnessed to prevent and/or treat allograft rejection.

Chronic rejection encompasses various phenotypes and affects all types of solid organ transplants. Transplant vasculopathy is marked by endothelial dysfunction and obliteration of the vascular lumen, the most well-described of which is cardiac allograft vasculopathy (CAV). The best studied forms of chronic rejection characterized by graft fibrosis include chronic renal allograft interstitial fibrosis and chronic lung allograft dysfunction, the latter of which consists of two distinct fibrotic phenotypes called bronchiolitis obliterans syndrome (BOS) and restrictive allograft syndrome (RAS) [92]. BOS is characterized by submucosal fibrosis of the respiratory bronchioles, resulting in complete obliteration of airway lumina [93], while RAS is characterized by pleuroparenchymal fibroelastosis [94].

Ferroptosis is associated with endothelial cell dysfunction and neointima formation [95], [96]. In a carotid ligation model, RSL3-activated ferroptosis promoted conversion of vascular smooth muscle cells from a contractive to a synthetic phenotype, resulting in neointima formation [96]. This process was inhibited by administration of Fer-1 [96]. Ferroptosis-induced endothelial dysfunction could potentially induce or exacerbate CAV following heart transplantation. Furthermore, HMGB1 release promotes the production of the proinflammatory cytokines TNFα and IL-6 [97], both of which have been implicated in the development of CAV [98], [99].

Chronic renal allograft interstitial fibrosis is a leading cause of late graft failure following kidney transplantation and ferroptosis has been implicated in its pathogenesis. Proinflammatory cytokines such as TNFα are released during episodes of chronic rejection, which has been shown to contribute to epithelial-to-mesenchymal transition of cells in kidney transplant recipients with chronic allograft dysfunction [100]. TNFα has also been shown to induce ferroptosis of renal tubular epithelial cells via suppression of GPX4. Ferroptosis of these renal tubular epithelial cells can trigger the release of pro-fibrotic cytokines such as PDGF-BB and IL-6, thereby contributing to allograft interstitial fibrosis [101].

As previously discussed, ferroptosis is only one mechanism of cell death that occurs following organ transplantation. In a heart chronic rejection model, inhibition of graft necroptosis with RIPK3-deficient donors attenuated CD4+ T cell-mediated graft injury and vasculopathy, but did not completely eliminate graft rejection [102]. It is likely that multiple forms of regulated cell death contribute to tissue injury in chronic rejection. When necroptosis is inhibited with RIPK3 deletion, other forms of cell death may partially account for the chronic rejection phenotype in this heart transplant model [102]. In fact, following ischemic kidney injury, inhibition of necroptosis has been shown to contribute to ferroptosis-mediated kidney injury [103]. It is possible that ferroptosis, among other pathways of regulated cell death, may compensate for necroptosis when the latter is compromised. Indeed, ferroptosis alone and the interplay between different types of necrotic cell death needs to be further investigated in the context of chronic allograft rejection.

Finally, as in the acute post-transplant setting, regulated cell death pathways are also not uniformly detrimental in the chronic setting. Another cause of poor outcomes after transplantation is the accumulation of senescent cells in the donor graft. Senolytic compounds, such as the combination of dasatinib and quercetin, have been tested in pre-clinical models of transplantation. Pre-transplant treatment of donor mice with senolytics appeared to prolong cardiac allograft survival [104]. This effect is thought to be mediated by elimination of senescent cells via apoptosis; however, some studies show that ferroptosis may be an alternative pathway [105]. In a murine model of kidney transplant, treatment with the ferroptosis-inducer RSL3 selectively eliminated senescent cells while leaving healthy cells unaffected, and was shown to reduce kidney damage and inflammatory cell infiltration [105]. Thus, ferroptosis plays a multifaceted role in long-term allograft health, and a better understanding of its regulation will be instrumental to prolonging graft survival.

Ferroptosis plays an important role in ischemia reperfusion injury and organ transplantation. There are exciting ways to translate the above studies clinically, in a diagnostic and/or therapeutic capacity. For example, in a study of patients in the intensive care unit, plasma levels of catalytic iron (Fec) and malondialdehyde (MDA), a lipid peroxidation degradation product, positively correlated with the median sequential organ failure assessment (SOFA) score [106]. This study demonstrates the feasibility of using these plasma markers as biomarkers of ferroptosis in transplant recipients. From a therapeutic perspective, diacetyl-bis(4-methyl-3-thiosemicarbazonato)copperII (CuII (atsm)) is a drug that has shown promise in the treatment of amyotrophic lateral sclerosis and Parkinson’s disease in Phase 1 studies [107], [108], possibly by inhibition of ferroptosis. In fact, one group has demonstrated that CuII (atsm) is comparable to liproxstatin-1 (a potent ferroptosis inhibitor [109]) in neuronal cell culture [110]. Notably, CuII (atsm) has good oral bioavailability and is currently undergoing Phase 2/3 clinical trials for use in amyotrophic lateral sclerosis [111]. Accordingly, CuII (atsm) could theoretically be translated for use in transplant recipients in the future.