Mitochondria as mediators of systemic inflammation and organ cross talk in acute kidney injury

Abstract

Acute kidney injury (AKI) is a systemic inflammatory disease that contributes to remote organ failures. Multiple organ failure is the leading cause of death due to AKI, and lack of understanding of the mechanisms involved has precluded the development of novel therapies. Mitochondrial injury in AKI leads to mitochondrial fragmentation and release of damage-associated molecular patterns, which are known to active innate immune pathways and systemic inflammation. This review presents current evidence suggesting that extracellular mitochondrial damage-associated molecular patterns are mediators of remote organ failures during AKI that have the potential to be modifiable.

INTRODUCTION

Acute kidney injury (AKI), as defined by the Kidney Disease Improving Global Outcome guidelines, occurs in one in five hospitalized patients worldwide (1). The 90-day mortality rate of AKI is 25–50% depending on AKI severity and recovery (2–5), and there are no proven treatments for AKI beyond renal replacement therapy. Despite improvements in renal replacement therapy technology and implementation, the mortality of AKI remains unacceptably high. Interestingly, multiple organ failure is a leading cause of death due to AKI (2), and recent data have demonstrated that AKI is a systemic disease that contributes directly to remote organ injuries (6–10). Investigations into the mechanisms of multiple organ failure during AKI have great potential to lead to novel life-saving therapies.

Several inflammatory mediators have been implicated in remote organ injuries due to AKI. Hoke et al. (9) used the ischemia-reperfusion (IR) and bilateral nephrectomy models of AKI to show that circulating levels of multiple cytokines, including IL-6 and IL-8, increase in the circulation just hours after AKI and are associated with neutrophil infiltration in the lung. Klein et al. (11) showed later that treatment with antibodies directed against IL-6 prevented lung injury in these same AKI models, and similar findings have been demonstrated for TNF-α (9, 12). Liu et al. (10) found increased levels of IL-1β, KC, granulocyte colony-stimulating factor, and other chemokines in the brain in mice with IR-AKI, which was associated with increased neurovascular permeability and reduced locomotor activity. This type of inflammatory cross talk has also been found to occur between the kidney and many other remote organs, including the gut (13) and liver (6). Release of cytokines from the injured kidney and reduced kidney function are implicated in the systemic accumulation of inflammatory mediators. However, markers of microvascular leak in the lung and liver have been found to be increased in models of IR-AKI compared with bilateral nephrectomy (6, 7). This suggests that direct release of inflammatory mediators from the injured kidney may be the more dominant mechanism. Unfortunately, treatments targeting specific mediators have not been proven to be beneficial in human studies, and research into novel mechanisms that may be modifiable is imperative.

Mitochondria have emerged as a promising target to prevent the direct complications of kidney injury, as mitochondrial injury is a key pathological feature of AKI (14–16). We have recently shown the temporal changes in mitochondrial structure and function in sepsis-associated AKI (16). Mitochondrial injury is also implicated in multiple organ failure syndromes (17–19), yet the role of mitochondria in remote organ failure due to AKI has not received much attention. Mitochondrial damage leads to the release of mitochondrial DNA (mtDNA) and mitochondrial proteins, such as transcription factor A, mitochondria (TFAM), that act as damage-associated molecular patterns (DAMPs) outside the cell (18, 20). In this review, we discuss how mitochondrial injury in AKI may contribute to the systemic release of mitochondrial DAMPs (mtDAMPs) and how therapies targeting mitochondrial health and mtDAMPs have the potential to be lifesaving in patients suffering from AKI.

ALTERED MITOCHONDRIAL DYNAMICS AND mtDAMPs IN AKI

The Kidney Is Highly Susceptible to Mitochondrial Injury

The kidney is one of the most metabolically active organs in the body due to the high energy required for fluid and electrolyte transport (21). Therefore, a large pool of healthy mitochondria is necessary to produce the ATP required to support these vital processes. For this reason, the kidney has a higher mitochondria content than every other organ in the body except the heart (22). Unlike the heart, the baseline partial pressure of oxygen in the kidney is relatively low despite high renal blood flow and oxygen delivery due to high rates of oxygen consumption (23). The majority of kidney perfusion is directed to the cortex, with decreasing tissue oxygenation from the outer to inner medulla (24). The outer medullary zone is particularly vulnerable to hypoxia/ischemia due to high oxygen demand to support tubular transport with relatively low perfusion and oxygen delivery, resulting in mitochondrial injury. Therefore, it is not surprising that mitochondrial dysfunction is one of the earliest manifestations of cellular injury in AKI (14).

Altered Mitochondrial Dynamics in AKI

Renal tubular cells have thousands of mitochondria that are continuously fusing, dividing, and regenerating to maintain optimal mitochondrial energetics (21). These processes are referred to as mitochondrial dynamics. Mitochondrial fission (division) results in short mitochondrial spheres as visualized by electron microscopy, and fusion (union) results in elongated, filamentous structures (25). The process of mitochondrial fission is largely controlled by the GTPase dynamin-related protein 1 (Drp1) (26). Drp1 resides in the cytoplasm but localizes to the mitochondria after injury (15), and Drp1-mediated fission is believed to be triggered by decreased ATP production and increased reactive oxygen species (ROS) (27). Mitochondrial fusion is also controlled by GTPases with mitofusin 1 (Mfn1), mitofusin 2 (Mfn2), and optic atrophy 1 (OPA1) being the major proteins involved (21). Mfn1 and Mfn2 are responsible for tethering of outer membranes during fusion, and Opa1 is responsible for inner membrane fusion and maintaining cristae structural integrity (28). In in vivo and in vitro models of ischemic and nephrotoxic AKI, Brooks et al. (14) demonstrated that Drp1 activation and fission precede kidney tubular injury. This group has also shown that apoptotic proteins during cellular injury interact with Mfn1 and Mfn2 to suppress fusion (29). Overall, AKI has been characterized as having an increase in fission and a decrease in fusion in tubular mitochondria (21). Interestingly, similar alterations in mitochondrial dynamics have been found to develop in remote organs during AKI. Mice subjected to renal ischemia perfusion demonstrated significant increase in Drp1 and mitochondrial fission in cardiac myocytes along with cardiac dysfunction, mitigated by pretreatment with Drp1 inhibitor (30).

Mitochondrial Fragmentation Into mtDAMPs

Altered mitochondrial dynamics in AKI, mediated by fission and fusion proteins, not only disrupt mitochondrial function and ATP production but also lead to mitochondrial fragmentation. Knockout of Mfn2 and increase in Drp1 expression are both associated with mitochondrial fragmentation, and pharmacological inhibition of Drp1 or siRNA targeting Drp1 has been shown to ameliorate it (14, 31). Electron microscopy of kidney cells has revealed that mitochondrial fragmentation is one of the earliest events postinjury, occurring in nearly 40% of proximal tubule cells immediately following an ischemic event (14). Many mitochondrial fragments function as DAMPs that activate innate immune pathways in the cytoplasm and outside the cell (19). The best-described extracellular mtDAMP is mtDNA, which has been implicated in inflammation in the heart (32, 33), lung (8, 19, 34), and liver (35, 36) via Toll-like receptor 9 (TLR9). For example, Oka et al. (32) showed that mtDNA accumulation in DNase II-deficient mice activated TLR9 and induced dilated cardiomyopathy, and injection of mtDNA into mice and rats leads to increases in serum cytokines, pulmonary edema, and neutrophil infiltration in the lung (19, 34). Similarly, mtDNA caused increases in pro-IL-1β transcript in the liver along with an increase in serum alanine aminotransferase in vivo (36). Interestingly, mtDNA is much more sensitive to damage compared with nuclear DNA. This suggests that there may be a biological role for mtDNA in alarming the cell of an injurious process before irreversible nuclear DNA damage (18, 37). Other extracellular mtDAMPs include TFAM (38, 39), ROS (40), cytochrome c (41), and N-formyl peptide (42), which have all been implicated in cell death or organ failure syndromes.

mtDAMPs Enter the Cytoplasm Through the Permeability Transition Pore

For mtDAMPs to engage innate immune receptors, they need to escape the mitochondrial membrane and enter the cytoplasm. This is believed to be accomplished through the permeability transition pore (PTP) of the inner and outer mitochondrial membrane. Opening of the PTP allows the passage of molecules up to 1,500 Da to escape the mitochondrial membrane (43), and mitochondrial proteins have been shown to accumulate outside the mitochondrial membrane during PTP opening (44). The PTP is activated by increases in ROS and calcium influx (45), which are known to occur in multiple forms of AKI.

Cell Death Leads to the Extracellular Release of mtDAMPs

mtDAMPs need to enter the extracellular space to potentiate systemic inflammation directly. Necrotic cell death, as seen in many forms of AKI, can release cytoplasmic mtDAMPs into the extracellular space. Recently, a type of programmed necrosis referred to as “necroptosis” has become an intense area of interest in AKI and has been implicated in extracellular mtDAMP accumulation (46, 47). Necroptosis is triggered by receptor-interacting protein kinase 1 and 3 upregulation, which activates the executor protein mixed lineage kinase 1, leading to a rapid increase in cellular permeability and release of cellular contents (48). Linkermann et al. (49, 50) demonstrated that these kinases are activated after IR-AKI, and similar findings were shown in a mouse model of contrast-induced injury. Especially relevant to this review are the findings by Zhao et al. (51), who discovered that necroptosis is activated in the kidney and remotely in the lung in a renal allograft model of AKI in rats. Importantly, in vitro and in vivo studies have shown that extracellular mitochondria and mtDAMPs are increased with necroptosis (46, 52).

Clearance of mtDAMPs

mtDAMPs activate multiple innate immune pathways within the cytoplasm, such as cGAS/STING (37) and inflammasomes (36, 53, 54). These are critical aspects of the innate immune response to pathogens, but a sophisticated mechanism to clear these components is necessary to prevent overwhelming inflammation and cell death. mtDAMP clearance is accomplished by a process referred to as mitophagy, a type of selective autophagy that sequesters whole mitochondria and mitochondrial components into autophagasomes (55). These autophagosomes eventually fuse with lysosomes to allow degradation, preventing the release of mtDAMPs into the extracellular space. During kidney injury, ROS, hypoxia, ATP depletion, and the energy-sensing kinases mammalian target of rapamycin (mTOR) and AMP-associated protein kinase are believed to promote mitophagy by promoting expression of the mitophagy proteins phosphatase and tensin homolog deleted on chromosome 10 (PTEN)-induced putative kinase protein 1 and Parkin (56–58). Mitophagy is considered an important component of mitochondrial quality control to remove damaged and dysfunctional mitochondria from the cell (21, 59, 60). There is well-established literature on the role of autophagy in in vitro and in vivo models of IR-AKI and cisplatin-induced AKI, and mice with impaired autophagy demonstrate more profound kidney injury (61–64). Evidence suggests that increased injury in the setting of impaired autophagy is due to the accumulation of mtDAMPs. Nakahira et al. (53) discovered that autophagy inhibition led to an accumulation of mtDNA in the cytoplasm with subsequent inflammasome activation, and Oka and colleagues (32) demonstrated that mtDNA fragments that escape autophagy increase inflammation via TLR9. In addition to these, studies specifically assessing mitophagy and demonstrating a protective role in AKI are growing (21, 59, 60, 65–67).

EXTRACELLULAR mtDAMPs CONTRIBUTING TO ORGAN FAILURES DURING AKI

The Kidney as a Source of Extracellular mtDAMPs

Strong evidence exists for mtDAMPs accumulating in the extracellular space during AKI. Whitaker et al. (68) showed that mtDNA accumulates in the urine just 10 min after kidney IR, and the amount of urinary mtDNA was predictive of the degree of kidney injury. This same group of investigators showed similar results for urinary levels of the mitochondrial protein ATP synthase-β (69). Tsuji and colleagues (17) discovered that mtDNA is increased in the circulation 4 h after septic AKI, and we (8) have recently shown that mtDNA and TFAM are increased in the plasma and lung following IR-AKI. However, there is not definite evidence yet to indicate that kidney cells are the source of these mtDAMPs. It is possible that they are released by leukocytes or other injured tissues.

Extracellular mtDAMPs Contribute to Systemic Inflammation and Remote Organ Failures

Increases in extracellular mtDAMPs are associated with detrimental changes in almost all organ systems, as shown in Fig. 1. Direct endothelial damage, activation of inflammatory pathways, and noninflammatory mechanisms have all been implicated as potential mechanisms of these changes. In a series of elegant experiments, Sun et al. (70) showed that mtDAMPs isolated from human tissues directly increase human endothelial cell permeability in vitro, which is an important pathological feature of multiple organ failure syndromes (71). In this study, mtDAMPs isolated from the liver were cocultured with various types of endothelial cells, and transendothelial electrical resistance was measured as a marker of permeability. mtDAMPs increased permeability in both EA.hy926 cells and human pulmonary artery endothelial cells. However, there was less of an impact on human pulmonary microvascular endothelial cells.

Figure 1.

Mechanisms of direct organ injury by extracellular mitochondrial damage-associated molecular patterns (DAMPs). TLR, Toll-like receptor. [Image created with Biorender.com.]

Regarding inflammatory pathways, extracellular mtDAMPs are well-described activators of inflammasomes and innate immune receptors, including NF-κB (72); NOD-, LRR- and pyrin domain-containing protein 3 (53, 72); and TLRs (17, 73) that lead to leukocyte activation and cytokine production. In a seminal study, Zhang et al. (19) injected healthy rats with mtDAMPs, which led to neutrophil activation and lung injury as demonstrated by histology and an increase in neutrophils in bronchoalveolar lavage fluid (BALF). Cytokine levels were also increased in BALF, along with an increase in BALF albumin and the lung wet-to-dry ratio, which is consistent with capillary leak. Finally, they showed that mtDAMPs cause neutrophils to release IL-8 in vitro. Intraperitoneal injection of mtDNA in mice has also been shown to cause lung injury by histology and increased BALF protein levels, which were attenuated in TLR9 knockout mice (34). Extracellular TFAM and N-formyl peptide have been implicated as inducers of inflammatory lung injury via dendritic cell activation and MAPK pathways (20, 39, 74). TLR9-mediated inflammation via mtDNA has also been found to contribute to the development of heart failure (32) and liver failure (35), and extracellular mtDAMPs have been shown to contribute to cytokine production in the gut (75) and in neuronal cells (76).

Interestingly, there is also evidence implicating noninflammatory mechanisms in remote organ dysfunction due to mtDAMPs in AKI. We have recently shown that intraperitoneal injection of extracellular mtDAMPs isolated from kidney tissue altered lung metabolism in a pattern consistent with mitochondrial dysfunction (8). Similarly, Tsuji et al. (17) showed that mtDAMP administration altered ATP production in the kidney, and Sumida et al. (30) showed altered mitochondrial dynamics develop in the heart after IR-AKI along with an upregulation of TNF-α. Interestingly, therapies aimed at improving mitochondrial dynamics in the latter study improved cardiac function but had no impact on TNF-α expression, indicating that these mitochondrial changes function independently of inflammatory pathways. Finally, human studies have shown that an increase in circulating mtDAMPs is associated with several multiple organ failure syndromes (77–80). Circulating mtDAMPs have also been found in patients with AKI (46).

THERAPEUTIC OPPORTUNITIES

With the increasing evidence for the role of mitochondrial injury and mtDAMPs in the pathophysiology of AKI and remote organ injury, therapeutic approaches targeting them are an area of active investigation. These are shown in Fig. 2 and further discussed below.

Figure 2.

Therapeutic strategies to target mitochondrial injury and mitochondrial damage-associated molecular patterns (mtDAMPS) in remote organ injury in acute kidney injury (AKI). AICAR, 5-Aminoimidazole-4-carboxamide ribonucleotide; Mdivi-1, mitochondrial division inhibitor; mTOR, mammalian target of rapamycin; NFP, N-formyl peptide; TFAM, transcription factor A, mitochondria. [Image created with Biorender.com.]

Preventing Mitochondrial Fragmentation Through Fission Inhibition

Targeted interventions to prevent fission and fusion imbalance during kidney injury would be predicted to prevent the mitochondrial fragmentation that leads to the release of mtDAMPs. This strategy has been successful in multiple preclinical models using a pharmacological inhibitor of Drp1 called mitochondrial division inhibitor (Mdivi-1). Brooks et al. (14) showed that Mdivi-1, given before injury, prevented mitochondrial fragmentation after cisplatin and IR renal injury with attenuation of cytochrome c release and kidney tubular apoptosis. In a rat model of rhabdomyolysis, Tang et al. (81) showed that Mdivi-1 after the induction of injury did not reduce Drp-1 expression; however, it did ameliorate mitochondrial fragmentation and AKI by preventing Drp-1 translocation to the mitochondria. Finally, Sumida and colleagues (30) showed that Mdivi-1 prevented cardiac injury after kidney IR when given as a pretreatment before IR as well as a delayed treatment after IR, which suggests that Mdivi-1 could be used to prevent other remote organ failures.

Promoting Mitophagy and Inhibiting Necroptosis to Prevent mtDAMP Release

Promoting mitophagy via mTOR inhibitors such as rapamycin, given as a pretreatment, has been shown to mitigate IR and nephrotoxic kidney injury with cisplatin (82, 83). However, mTOR inhibition may prevent cell growth during the recovery phases of AKI, limiting its clinical use (57). Mitophagy induction may also be limited by the recent finding that the mitophagy protein PTEN-induced putative kinase protein 1 promotes necroptosis (84); thus therapeutics promoting mitophagy may need to be combined with inhibitors of necroptosis. Recently, a pharmacological inhibitor of necroptosis, necrostatin-1, has shown promise in preventing kidney injury in IR, nephrotoxic, and septic models when given at the time of or before injury (50, 85–87). Necroptosis has also been observed in remote organs in animal models of AKI. Interestingly, necrostatin-1 prevented lung injury in a kidney allograft model when given at the time of grafting (51). While the exact mechanism of this protective effect is not known, prevention of necroptosis in kidney cells with a reduction in the release of extracellular mtDAMPs is highly plausible.

Therapies Directly Targeting mtDAMPs

Finally, interventions targeting mtDAMPs directly may mitigate downstream inflammatory effects. Pretreatment with systemic DNase 1 targeting extracellulIar DNA, including mtDNA, was shown to mitigate endothelial inflammation in a coronary artery bypass model (88). Systemic DNase also decreased liver injury due to acetaminophen in mice when given after the induction of liver injury (89). In a recent study of ventilator-associated pneumonia in rats, Simmons et al. (90) used inhaled DNase, given before and after injury, to target alveolar mtDNA and prevent lung injury. They also showed that patients with ventilator-associated pneumonia had significantly higher levels of mtDNA DAMPs in serum and BAL. Similar results were demonstrated in a lung allograft model with bronchial DNase instillation at the time of grafting (73). Importantly, inhaled DNase has an excellent safety profile and is already Federal Drug Administration approved in the treatment of cystic fibrosis and, hence, an attractive therapeutic option.

CONCLUSIONS

In conclusion, the kidney has a higher mitochondrial content than all other organs except the heart. The kidney is also uniquely prone to hypoxic injury, which makes it a likely site of mitochondrial damage and mtDAMP formation and release. PTP opening, as seen in AKI, allows mtDAMPs to escape the mitochondrial membrane and enter the cytoplasm. Impaired mitophagy to clear these mtDAMPS and subsequent cell death allow these DAMPs to accumulate in the extracellular space and in the circulation. Circulating mtDAMPs are well-described mediators of systemic inflammation resulting in multiple organ failures, which is the leading cause of death due to AKI. Investigations into novel pharmacological therapies targeting mitochondrial health and mtDAMPs have great potential to improve AKI as well as remote organ injury.

GRANTS

This work was supported by Veterans Affairs Merit Award BX002175 (to P.S.), National Institutes of Health (NIH) Grant R01DK107852 (to P.S.), Veterans Affairs Award CDA-2 IK2BX004338-01 (to M.H.), and the University of Alabama at Birmingham-University of California-San Diego O’Brien Center (NIH Grant P30DK079337).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

P.S. conceived and designed research; M.H. prepared figures; M.H. and P.S. drafted manuscript; M.H. and P.S. edited and revised manuscript; M.H. and P.S. approved final version of manuscript.